24281570-The-Crusades-From-the-Prespective-of-Byzantium-and-the-Muslim-World - PDFCOFFEE.COM (2024)

Preface The studies in this volume originated in a symposium organized by Roy P. Mottahedeh and myself and held at Dumbarton Oaks on 2–4 May 1997. Over the two previous years, a considerable number of scholarly conferences had been planned, to commemorate the nine hundredth anniversary of the Council of Clermont and its results. Most of these focused on the Crusades from the viewpoint of Western Europe. Indeed, the considerable and exciting scholarship produced during recent decades has also, to a large extent, been concerned with the internal, Western, aspects of this movement. Notable exceptions do, of course, exist. Still, it seemed to us that there was need of a conference that would look at the crusade from the perspective of those areas to which it was primarily directed, namely, the Eastern Muslim areas and the Byzantine Empire. The Dumbarton Oaks symposium took place on the nine hundredth anniversary of the appearance of the crusading armies outside the city of Nicaea. Our purpose was to examine several important issues that, in one way or another, affected the Byzantine and Muslim worlds at the time of the Crusades or because of them. The movement having been a lengthy and recurrent one, our time frame extended to the late thirteenth century. The first essay published here analyzes the development of the historiography of the Crusades. The other essays discuss various topics ranging from the problem of the holy war in Byzantium and Islam to the question of attitudes and perceptions, the effect on art, and the impact of the Crusades on the economies of the East. We neither expected nor planned a comprehensive examination of the crusading movement seen from Constantinople, Baghdad or Cairo. Rather, we hope that this volume, by contributing to the very lively scholarly discussion on the Crusades, stimulates further research on developments that engaged the eastern Mediterranean, both Christian and Muslim, and the Muslim world generally. Angeliki E. Laiou Harvard University and Academy of Athens

The Historiography of the Crusades Giles Constable I. The Development of Crusading Historiography The crusades were from their inception seen from many different points of view, and every account and reference in the sources must be interpreted in the light of where, when, by whom, and in whose interests it was written.1 Each participant made his— and in few cases her—own crusade, and the leaders had their own interests, motives, and objectives, which often put them at odds with one another. They were all distrusted by the Byzantine emperor Alexios Komnenos, whose point of view is presented in the Alexiad written in the middle of the twelfth century by his daughter Anna Komnene. The Turkish sultan Kilij Arslan naturally saw things from another perspective, as did the indigenous Christian populations in the east, especially the Armenians, and the peoples of the Muslim principalities of the eastern Mediterranean. The rulers of Edessa, Antioch, Aleppo, and Damascus, and beyond them Cairo and Baghdad, each had their own attitudes toward the crusades, which are reflected in the sources. To these must be added the peoples through whose lands the crusaders passed on their way to the east, and in particular the Jews who suffered at the hands of the followers of Peter the Hermit.2 The historiography of the crusades thus begins with the earliest accounts of their origins and history. Aside from some studies of individual sources, however, and a number of bibliographies and bibliographical articles,3 the historiography has received com1 This article is a revised version of the paper presented at the symposium. It concentrates on general problems concerning the crusades to the east. The references to secondary works are illustrative and are not designed to give a bibliography of the crusades. I am indebted to Benjamin Z. Kedar for various suggestions. A shortened version of part I will appear (in Russian) in the forthcoming Festschrift for Aaron Gurevich. 2 The Jews and the Crusaders: The Hebrew Chronicles of the First and Second Crusades, ed. and trans. S. Eidelberg (Madison, Wisc., 1977). Among secondary works, see most recently D. Lohrmann, “Albert von Aachen und die Judenpogrome des Jahres 1096,” Zeitschrift des Aachener Geschichtsvereins 100 (1995–96): 129–51. 3 H. E. Mayer, Bibliographie zur Geschichte der Kreuzzu¨ge (Hannover, 1960), and idem and J. McLellan, “Select Bibliography of the Crusades,” in A History of the Crusades, ed. K. M. Setton (Madison, Wisc., 1955–89), 6:511–664. Other general bibliographies are L. de Germon and L. Polain, Catalogue de la bibliothe`que de feu M. le comte Riant, pt. 2 (Paris, 1899), and A. S. Atiya, The Crusade: Historiography and Bibliography (Bloomington, Ind.London, 1962), which has a section on historiography (17–28). Among the review articles, see G. Schnu¨rer, “Neuere Arbeiten zur Geschichte der Kreuzzu¨ge,” HJ 34 (1914): 848–55; T. S. R. Boase, “Recent Developments in Crusading Historiography,” History, n.s., 22 (1937): 110–25; J. La Monte, “Some Problems in Crusading Historiography,” Speculum 15 (1940): 57–75; J. A. Brundage, “Recent Crusade Historiography: Some Observations and Suggestions,” CHR 49 (1964): 493–507; F. Cardini, “Gli studi sulle crociate dal 1945 ad oggi,” RSI 80 (1968): 79–106; and H. Mo¨hring, “Kreuzzug und Dschihad in der mediaevistischen und orienta-

[ 2 ] Historiography of the Crusades paratively little attention from scholars. The only general works are a long and still useful appendix to the first (but not the second) edition of Heinrich von Sybel’s Geschichte des ersten Kreuzzugs, which appeared at Du¨sseldorf in 1841 and was translated into English in 1861, and the two volumes (in Russian) by M. A. Zaborov entitled Vvedenie v istoriografiju Krestovykh pokhodov (Introduction to the historiography of the crusades), which deals with the medieval sources, and Istoriografija Krestovykh pokhodov (XV–XIX vv.) (Historiography of the crusades [15th–19th century]), which were published in Moscow in 1966 and 1971 respectively.4 To these can be added a long article, partly historiographical and partly bibliographical, by Laetitia Boehm entitled “‘Gesta Dei per Francos’—oder ‘Gesta Francorum’? Die Kreuzzu¨ge als historiographisches Problem” and a chapter by Jonathan Riley-Smith on “The Crusading Movement and Historians” in the Oxford Illustrated History of the Crusades.5 It is interesting, and perhaps significant, that there is no sustained treatment of historiography in the general histories of the crusades by Rene´ Grousset, Steven Runciman, and Hans Eberhard Mayer, nor in the six-volume cooperative History of the Crusades edited by Kenneth Setton. The historiography of the crusades as seen from the west, with which this article is concerned, can be divided into three periods, of which the first, and longest, went from 1095 until the end of the sixteenth century; the second covered the seventeenth and eighteenth centuries; and the third began in the early nineteenth century and comes down to the present. There was some overlap between the periods, but broadly speaking, during the first, the Muslims were a continuing threat to Western Europe and the defense of Christendom was seen as a pressing concern. In the second period, the crusades moved increasingly into the past, but a past that was colored by confessional or rationalist values, which changed in the third period, when the crusades were subjected to serious, though not always impartial, scholarly investigation. This third period breaks down into the nineteenth century, when the crusades were generally well regarded, and the twentieth century, when there has been a rising tide of criticism and, more recently, a growing division between scholarly and popular views of the crusades. Interest in the crusades today is still influenced by political and ideological interests, including the consequences of European colonialism, the tensions between western and non-western societies, especially in the Middle East, and, more broadly, the legitimacy of using force to promote even worthy and legitimate causes.6 These concerns contriblistischen Forschung 1965–1985,” Innsbrucker historische Studien 10–11 (1988): 361–86. See also H. E. Mayer, “America and the Crusades,” PAPS 125 (1981): 38–45; C. R. Young, “The Crusades: A Tragic Episode in East-West Relations,” South Atlantic Quarterly 55 (1956): 87–97; and the collection of reprints and excerpts in The Crusades: Motives and Achievements, ed. J. A. Brundage, Problems in European Civilization (Boston, 1964). 4 H. von Sybel, Geschichte des ersten Kreuzzugs (Du¨sseldorf, 1841), 148–80, trans. L. D. Gordon, The History and Literature of the Crusades (London, 1861), 311–56; and M. A. Zaborov, Vvedenie v istoriografiju Krestovykh pokhodov (Moscow, 1966) and Istoriografija Krestovykh pokhodov (XV–XIX vv.) (Moscow, 1971). For these references I am indebted to Alexander Kazhdan, who also summarized the contents for me. 5 L. Boehm, “‘Gesta Dei per Francos’—oder ‘Gesta Francorum’? Die Kreuzzu¨ge als historiographisches Problem,” Saeculum 8 (1957): 43–81, and J. Riley-Smith, “The Crusading Movement and Historians,” in The Oxford Illustrated History of the Crusades, ed. J. Riley-Smith (Oxford, 1995), 1–12. See also L. Boehm, “Die Kreuzzu¨ge in bibliographischer und historiographischer Sicht,” HJ 81 (1962): 223–37. 6 P. Rousset, Histoire d’une ide´ologie: La croisade (Lausanne, 1983), 206–8; K. Armstrong, Holy War (London, 1988), xiii–xiv; J. Riley-Smith, “History, the Crusades and the Latin East, 1095–1204: A Personal View,” in

Giles Constable [ 3 ] uted to the change from the comparatively favorable attitude toward the crusades that prevailed in the nineteenth and early twentieth centuries into a more critical, and even hostile, view. Steven Runciman, in the conclusion to his History of the Crusades, called the crusades “a tragic and destructive episode” and said that “the Holy War itself was nothing more than a long act of intolerance in the name of God, which is the sin against the Holy Ghost.” 7 Geoffrey Barraclough echoed this view in 1970: “We no longer regard the crusades . . . as a great movement in defense of Western Christendom, but rather as the manifestation of a new, driving, aggressive spirit which now became the mark of Western civilization. We no longer regard the Latin states of Asia Minor as outposts of civilization in a world of unbelievers, but rather as radically unstable centers of colonial exploitation.” He attributed this change in “our verdict on the Crusades” to “our experience of total war and the hazards of living in a thermonuclear age. War is always evil, if sometimes an inescapable evil; Holy War is the evil of evils.” 8 And John Ward described the crusades in 1995 as “a movement of violent white supremacist colonialism.” 9 This view is now common in works addressed to the general public, including popular presentations and movies. A leaflet distributed in Clermont during the conference held in 1995 to commemorate the summons to the First Crusade was headed “The Crusades—did God will it?” echoing the crusading cry of “Deus le volt.” It went on to ask “Can the Church memorialize the Crusades without asking forgiveness?” and called on the pope to deny that any war can be holy and that sins can be forgiven by killing pagans. According to this view, the crusaders were inspired by greed and religious fanaticism and the Muslims were the innocent victims of expansionist aggression. Many scholars today, however, reject this hostile judgment and emphasize the defensive character of the crusades as they were seen by contemporaries, who believed that Christianity was endangered by enemies who had already overrun much of the traditional Christian world, including Jerusalem and the Holy Land, and who threatened to take over the remainder. Almost all the historians and chroniclers of the expeditions that were later called the First Crusade considered them a response to the Muslim threats to Christian holy places and peoples in the east.10 They wrote from different points of view, however, and used Crusaders and Muslims in Twelfth-Century Syria, ed. M. Shatzmiller, The Medieval Mediterranean 1 (LeidenNew York-Cologne, 1993), 7–8; and idem, “Revival and Survival,” in Oxford History (as in note 5), 386. 7 S. Runciman, A History of the Crusades, 3 vols. (Cambridge, 1952–54), 3:480. 8 G. Barraclough, “Deus le volt?” New York Review of Books, 21 May 1970, 16. 9 J. Ward, “The First Crusade as Disaster: Apocalypticism and the Genesis of the Crusading Movement,” in Medieval Studies in Honour of Avrom Saltman, Bar-Ilan Studies in History 4 (Ramat-Gan, 1995), 255. Cf. on the current unfavorable view of the crusades M. Balard, Les Croisades (Paris, 1988), 9; Riley-Smith, “History” (as in note 6), 1–2; and the review of T. Jones and A. Ereira, Crusades, by M. Evans, D. Green, and J. M. B. Porter in Nottingham Medieval Studies 39 (1995): 201. 10 C. Erdmann, The Origin of the Idea of Crusade, trans. M. Baldwin and W. Goffart (Princeton, 1977), 8, 349; E. Delaruelle, Ide´e de croisade au moyen-aˆge (Turin, 1980), 23; and J. Riley-Smith, What Were the Crusades? (London-Basingstoke, 1977), 22–33, who stressed the recurrence of “the ideas of liberation (another word for recovery) and defence” (23) and said that “a crusade, whenever and against whomsoever it was aimed, was regarded as being essentially defensive” (29). See also J. Flori, “Guerre sainte et re´tributions spirituelles dans la 2e moitie´ du XIe sie`cle,” RHE 85 (1990): 627–28, on the concept of the legitimacy of recovering wrongly taken lands.

[ 4 ] Historiography of the Crusades varying terminology and biblical passages.11 Guibert of Nogent stressed the apocalyptic and millenarian aspects, and Ekkehard of Aura the supernatural and physical phenomena that preceded and accompanied the crusade. Many writers had their own heroes. The roles of Godfrey of Bouillon and Peter the Hermit were central for Albert of Aachen; Bohemund of Taranto in the anonymous Gesta Francorum; his nephew Tancred for Ralph of Caen; Raymond of St. Gilles for Raymond of Aguilers; Baldwin of Boulogne for Fulcher of Chartres; and Godfrey of Bouillon again in the crusader epics, which dominated the popular perception of the crusades down to the nineteenth century. Odo of Deuil in his history of the Second Crusade concentrated on the activities of Louis VII of France, and the accounts of the Third Crusade in the Estoire de la guerre sainte of Ambroise and the Itinerarium regis Ricardi glorified the role of Richard I of England. The greatest of all crusader historians, William of Tyre, wrote his Chronicon from the point of view of a Latin Christian born and living in the east in order, he said, to record “for the everlasting memory of the faithful of Christ” the way in which God “wanted to relieve the long-lasting oppression of His people.”12 Innocent III in his crusading bull Quia maior of 1213 asked how anyone could “know that his brothers, Christian in faith and name, are held in dire imprisonment among the perfidious Saracens and most profoundly subjected by the yoke of servitude, and not take effective action for their liberation . . . And indeed the Christian peoples held almost all the provinces of the Saracens until the times of the blessed Gregory.”13 Even more strikingly, the fourteenth-century Castilian magnate Don Juan Manuel wrote in his Libro de los estados that the Muslims had conquered and held many lands that had belonged to Christians “who had been converted to the faith of Jesus Christ by the apostles. And on this account there is war between the Christians and the Muslims, and will be war until the Christians have recovered the lands that the Muslims seized from them, since there would be no war between them with regard to the law nor the religion (secta) that they hold.”14 While the accuracy and realism of these views can be questioned, they reflect 11 P. Alphande´ry, “Les citations bibliques chez les historiens de la premie`re croisade,” RHR 90 (1929): 139–57. 12 William of Tyre, Chronicon, 1.10, in CC continuatio mediaeualis 63:124. See P. Edbury and J. G. Rowe, William of Tyre: Historian of the Latin East, Cambridge Studies in Medieval Life and Thought 4.8 (Cambridge, 1988), 41. Von Sybel, Geschichte (as in note 4), 148–63, trans. 311–31, classified the sources for the First Crusade (149, trans. 312) as sources (Quellen), legends (Sage), and William of Tyre, whose unique importance as both a primary source and a secondary writer entitle him to a special place. 13 G. Tangl, Studien zum Register Innocenz’ III. (Weimar, 1929), 90, trans. L. and J. Riley-Smith, The Crusades: Idea and Reality, 1095–1274, Documents of Medieval History 4 (London, 1981), 120. See J. Riley-Smith, “Crusading as an Act of Love,” History 65 (1980): 177–92, citing this bull (191). P. J. Cole, The Preaching of the Crusades to the Holy Land, 1095–1270 (Cambridge, Mass., 1991), 105, referred to Innocent’s “juridical concept of the crusade as military service for Christ.” R. Ro¨hricht, Kleine Studien zur Geschichte der Kreuzzu¨ge, Wissenschaftliche Beilage zum Programm des Humboldts-Gymnasiums zu Berlin 58 (Berlin, 1890), 9–11, analyzed the motives for the crusades given in papal bulls, including their necessity. 14 Don Juan Manuel, Libro de los estados, ed. R. B. Tate and I. R. MacPherson (Oxford, 1974), 53; see xl, dating it 1328 with subsequent revisions. According to V. Cantarino, “The Spanish Reconquest: A Cluniac Holy War against Islam?” in Islam and the Medieval West: Aspects of Intercultural Relations, ed. K. I. Semaan (Albany, N.Y., 1980), 98, “The Spanish Reconquest remained in its essence a territorial struggle.” See also R. A. Fletcher, “Reconquest and Crusade in Spain, c. 1050–1150,” Transactions of the Royal Historical Society 5.37 (1987): 31–47, who said that reconquest was really conquest (46–47).

Giles Constable [ 5 ] the attitude of most Christians in the Middle Ages and throughout the first period of crusading historiography. The importance of irredentism in motivating the crusades has been emphasized by many scholars, including Islamists like Norman Daniel, who said that “every Christian reference to lands that had once been Christian, and particularly to the Holy Land, must be understood to have been made on the assumption that these were lost provinces belonging by right to the Latin Church.”15 The process of what has been called the affabulation of the First Crusade, by which it became a “work of collective imagination” rather than historical reality,16 can be seen already in the earliest accounts, which reflected the view of the crusade as it developed, and perhaps as it should have been, rather than as it actually was. They were influenced in particular by the capture of Jerusalem and the establishment of the Latin Kingdom and crusader states, which were marks respectively of the success and the permanence of the undertaking.17 This can be seen in the use made of the Gesta Francorum by Guibert of Nogent, Baldric of Bourgueil, and Robert of Rheims, and of Fulcher of Chartres by William of Malmesbury, and also in Albert of Aachen’s Liber christianae expeditionis pro ereptione, emundatione, restitutione sanctae Hierosolymitanae ecclesiae, which was written about 1130 and was long considered the most reliable account of the crusade, but which depends heavily on legendary material, especially concerning Peter the Hermit. Caffaro di Caschifellone, writing in the mid-1150s, in addition to stressing the Genoese contribution to the First Crusade, saw its origins in a visit to Jerusalem by Godfrey of Bouillon, who on his return went with Raymond of St. Gilles and eleven other knights (to one of whom the archangel Gabriel appeared) in order to plan the rescue of the Holy Sepulcher from the Muslims.18 The history of the crusades thus became part of the ongoing propaganda, both official and popular, for the crusading movement,19 and it is often impossible to distinguish clearly between what would today be called primary sources and secondary accounts, because the historical and contemporary concerns of the writers overlapped. “By the 1140s,” according to Riley-Smith, “the crusading experiences of previous generations, and pride in them, had been locked deeply in the collective memory of some cousinhoods.”20 Pope Eugene III said at the beginning of Quantum predecessores, which opened 15 N. Daniel, Islam and the West (Edinburgh, 1960), 109, who went on to say that “this was more than a general way of thinking. It had juridical and liturgical expression.” See also C. Cahen, “L’Islam et la Croisade,” in Comitato internazionale di scienze storiche: X Congresso internazionale di scienze storiche, Roma 4–11 settembre 1955. Relazioni, vol. 3, Storia del Medioevo (Rome, 1955), 629. 16 See, on affabulation, P. Alphande´ry, La chre´tiente´ et l’ide´e de croisade, ed. A. Dupront, Evolution de l’humanite´ 38 and 38 bis (Paris, 1954–59; repr. 1995), 1:154. 17 E. O. Blake, “The Formation of the ‘Crusade Idea,’” JEH 21 (1970): 11–31, who stressed that the actual course of events contributed to “a developing sense of pattern” (21), and J. Flori, “Mort et martyre des guerriers vers 1100. L’exemple de la premie`re croisade,” CahCM 34 (1991): 121–39, who argued against the view that the idea of martyrdom evolved in the course of the crusade. 18 Caffaro di Caschifellone, De liberatione civitatum orientis, in Annali genovesi di Caffaro e de’ suoi continuatori, ed. L. Belgrano et al., Fonti per la storia d’Italia 11–14 bis (Rome, 1890–1929), 1:97–124. 19 J. M. Powell, “Myth, Legend, Propaganda, History: The First Crusade, 1140–ca. 1300,” in Autour de la Premie`re Croisade: Actes du Colloque de la Society for the Study of the Crusades and the Latin East (Clermont-Ferrand, 22–25 juin 1995), ed. M. Balard (Paris, 1997), 127–41. 20 J. Riley-Smith, The First Crusaders, 1095–1131 (Cambridge, 1997), 102.

[ 6 ] Historiography of the Crusades the Second Crusade, that “we learn from the account of former men and we find written in their deeds how greatly our predecessors the Roman pontiffs labored for the freedom of the eastern church,” and he went on to say that Urban II, thundering “like a sacred trumpet,” had summoned “the sons of the Roman church” from various parts of the world to free Jerusalem and the Holy Sepulcher “from the filth of the pagans.”21 The view of the crusades found in accounts written later in the Middle Ages depends to a great extent on the sources used, but they were always presented as a response to the external attacks of the Muslims and pagans or to the internal threats of heretics and schismatics. The Turkish victories in the fifteenth century stimulated a new interest in the crusades in writers who were ostensibly dealing with the past but were really concerned with the present.22 Philip the Good of Burgundy’s devotion to the Holy Land was inspired by political ambition as well as personal piety, and his image of himself as the successor of Godfrey of Bouillon was fostered by his reading of the vernacular epics of the crusades.23 Eneas Sylvius, the future Pope Pius II, abbreviated the sections on the crusades in Flavio Biondo’s Decades and referred to the crusaders as “our Christians,” and Benedetto Accolti’s Historia Gotefridi seu de bello a Christianis contra barbaros gesto pro Christi sepulchro et Judea recuperandis, which appeared in 1464, was designed to promote a new crusade against the Turks, who had recently taken Constantinople. It was included, presumably for this reason, among the primary sources published in the Recueil des historiens des croisades.24 Accolti and other humanist historians hoped to find in the accounts of previous crusades, particularly the First, both guidance and inspiration for the contemporary campaigns against the Turks. Even the Jewish chronicler Joseph ben Joshua ben Meir, writing in the first half of the fifteenth century, wanted “the children of Israel to know what they [the Christians] have done unto us” and saw the Muslims as the instruments of divine vengeance on the Christians.25 In the sixteenth century the crusades tended to move into the past and to be treated as part of national history, but crusading ideology was kept alive not only by the advances of the Turks but also by the wars of religion. Etienne le Blanc wrote an essay in 1522 to show that Louis IX “had not destroyed the Kingdom [of France] for his holy voyage overseas,” and toward the end of the century Franc¸ois de la Noue and Rene´ de Lucinge

21 Otto of Freising, Gesta Friderici I. imperatoris, 1.35, ed. G. Waitz, MGH, ScriptRerGerm, 3d ed. (HannoverLeipzig, 1912), 55. 22 L. Schmugge, Die Kreuzzu¨ge aus der Sicht humanistischer Geschichtsschreiber, Vortra¨ge der Aeneas-SilviusStiftung an der Universita¨t Basel 21 (Basel-Frankfurt, 1987), and N. Housley, The Later Crusades, 1274–1580: From Lyons to Alcazar (Oxford, 1992), 84, 99–100 (“From 1453 . . . the crusade became a simple matter of selfinterest”), 385, 388, 420 (“The fact that the golden age of crusading was, by about 1450, beginning to be viewed in historical perspective, did not mean that the crusade itself had become history.”). 23 J. Paviot, “La de´votion vis-a`-vis de la Terre Sainte au XVe sie`cle: L’exemple de Philippe le Bon, duc de Bourgogne (1396–1467),” in Autour de la Premie`re Croisade (as in note 19), 401–11. 24 RHC, HOcc, 5:525–620. See Von Sybel, Geschichte (as in note 4), 329–30; R. Black, Benedetto Accolti and the Florentine Renaissance (Cambridge, 1985), 224–85, esp. 230 and 237; and Schmugge, Kreuzzu¨ge (as in note 22), 12–13. 25 The Chronicles of Rabbi Joseph ben Joshua ben Meir, the Sphardi, vol. 1, trans. C. H. F. Bialloblotsky, Oriental Translation Fund (London, 1835), 325, no. 436.

Giles Constable [ 7 ] made use of crusading rhetoric in their polemics against the Turks.26 Catholics and Protestants both saw themselves as soldiers of Christ fighting a holy war in defense of Christianity against the forces of evil.27 Pope Gregory XIII in 1580 offered the same indulgence given to crusaders to the Holy Land to the Irish who joined the expedition against Queen Elizabeth.28 Echoes of crusading ideology continued in the seventeenth century, as in the Civil War in England, and down into modern times, when any ideological enterprise can be called a crusade, like Lloyd George’s The Great Crusade and Dwight Eisenhower’s Crusade in Europe. Reality changed after the battle of Lepanto in 1571, however, and the huge success of the fictional account of the First Crusade in Tasso’s Gerusalemme liberata, which was published in 1581, shows how far history had moved from the realm of fact into that of fantasy, where it remained, in popular mentality, until well into the nineteenth century.29 The second period of crusading historiography was meanwhile ushered in by the appearance in 1611 of the important collection of primary sources on the crusades edited by Jacques Bongars under the title Gesta Dei per Francos sive orientalium expeditionum et regni Francorum Hierosolimitani historia and in 1639 of Thomas Fuller’s Historie of the Holy Warre, which has been called, in spite of its prejudices, the first serious general history of the crusades to treat them as fully in the past and to raise the question of their legitimacy.30 It has a remarkable frontispiece showing various groups of crusaders setting out from Europe and returning from Jerusalem ravaged by the attacks of the angel (owing to their perfidy and falsehood), the Turks, and death (Fig. 1), and an equally remarkable introductory poem, signed only with the initials J. C., explaining the frontispiece and concluding that: Those that escap’d, came home as full of grief As the poore Purse is empty of relief. Fuller was a Protestant minister and wrote from a strongly anti-Catholic point of view. The opposite is true of Louis Maimbourg, whose pro-Catholic Histoire des croisades came out in the 1670s, with a dedication to Louis XIV, and was frequently reprinted and translated into several languages. It is marked, according to Von Sybel, by the author’s 26 E. A. R. Brown, “A Sixteenth-Century Defense of Saint Louis’ Crusades: Etienne le Blanc and the Legacy of Louis IX,” in Cross Cultural Convergences in the Crusader Period: Essays Presented to Aryeh Grabois on His Sixty-fifth Birthday, ed. M. Goodich, S. Menache, and S. Schein (New York, 1995), 21–48, and M. J. Heath, Crusading Commonplaces: La Noue, Lucinge and Rhetoric against the Turks, Travaux d’humanisme et renaissance 209 (Geneva, 1986). 27 A. Wang, Der “Miles Christianus” im 16. und 17. Jahrhundert und seine mittelalterliche Tradition, Mikrokosmos 1 (Frankfurt a. Main, 1975), and P. Rousset, “L’ide´ologie de croisade dans les guerres de religion au XVIe sie`cle,” SZG 31 (1981): 174–84. Many of the same biblical texts cited by the historians of the crusades were used in the 16th century. V. J. Gellhaus, “Franzo¨sische Kreuzzugsideen und Weltfriedensbewegung im Zeitalter der Aufkla¨rung” (diss., Munich, 1934) deals mostly with the 17th and 18th centuries. 28 H. Pissard, La guerre sainte en pays chre´tien (Paris, 1912), 173–74. 29 Schmugge, Kreuzzu¨ge (as in note 22), 46 n. 142, and E. Siberry, “Tasso and the Crusades: History of a Legacy,” JMedHist 19 (1993): 163–69. 30 J. E. Bailey, The Life of Thomas Fuller, D.D. (London-Manchester, 1874), 173–81; Von Sybel, Geschichte (as in note 4), 163, trans. 332; and Boehm, “Gesta” (as in note 5), 63–64.

[ 8 ] Historiography of the Crusades self-confidence, religiosity, and “a trace of modern good sense,” but his hesitation between enthusiasm and skepticism was replaced in the eighteenth century by “a strong, relentless opposition.”31 For the rationalist writers of all denominations in the age of Enlightenment, the crusades were inspired by religious zeal, worldly motives, and ecclesiastical interference in secular affairs. Voltaire in his book on the crusades, which came out in 1751 and was incorporated (with some changes) into his Essai sur les moeurs, called the crusaders adventurers and brigands who were moved by “the thirst for brigandage,”32 and for Edward Gibbon their principle was “a savage fanaticism,” though he expressed some grudging admiration for their spirit and achievements.33 “The historical writing of the Enlightenment,” said Boehm, “cultivated with regard to the crusades a one-sidedness of treatment from which the nineteenth century only slowly freed itself”34 and which persisted longer in the United States than in Europe. Ralph Waldo Emerson recorded in his journal in 1826 that the crusades had taken their place in public opinion “among the monuments of folly and tyranny” and wrote to Charles Emerson in 1828 about the “shrill and evil sound” of “a fanatic voice saying ‘It is the voice of God.’”35 By this time the tide of opinion had turned in Europe, ushering in the third period of crusading historiography. A sympathetic attitude toward the Middle Ages, including the crusades, emerged in the late eighteenth and early nineteenth centuries under the influence of romanticism and nationalism and can be seen in the favorable depictions of the crusaders in literature, art, and music, especially the novels of Sir Walter Scott, several of which dealt with the crusades.36 They were “a holy war, purifying the Holy Land” for Thomas Rowley and a response to “the call of piety and honour” for Kenelm Digby, who converted to Catholicism in 1825 and whose Broad Stone of Honour, published in 1822, was widely read in the nineteenth century.37 The enthusiasm for medieval literature in France at this time has been called “a mythological revolution,” and the theme of the crusaders’ return frequently appeared in both literature and art.38 31 Von Sybel, Geschichte (as in note 4), 163–64, trans. 332–33; Gellhaus, “Kreuzzugsideen” (as in note 27), 86–87; and Boehm, “Gesta” (as in note 5), 64–66. 32 Voltaire, Essai sur les moeurs, 54, in Oeuvres comple`tes de Voltaire, ed. L. Moland (Paris, 1877–85), 11:442. J. H. Brumfitt, Voltaire Historian (Oxford, 1958), 68, said that Voltaire was “delighted to be able to show that the crusades were not the result of lofty religious motives, but of a desire for plunder.” See Von Sybel, Geschichte (as in note 4), 164–65, trans. 334; Gellhaus, “Kreuzzugsideen” (as in note 27), 90–92; and the introduction by H. Berr to Alphande´ry, Chre´tiente´ (as in note 16), 1:viii–ix. 33 E. Gibbon, The History of the Decline and Fall of the Roman Empire, chap. 61, ed. D. Womersley (London, 1994), 3:727; see also chap. 58 (ibid., 563 n. 19), where he said that some critics called Bongars’s collection Gesta diavoli per Francos, and the editor’s introduction in 1:xcix–c. See Gellhaus, “Kreuzzugsideen” (as in note 27), 95–97. 34 Boehm, “Gesta” (as in note 5), 70. Cf. Zaborov, Istoriografija (as in note 4), 76–144, on the Enlightenment treatment of the crusades. 35 The Journals and Miscellaneous Notebooks of Ralph Waldo Emerson, vol. 3, 1826–1832, ed. W. Gilman and A. Ferguson (Cambridge, Mass., 1963), 18, and The Letters of Ralph Waldo Emerson, ed. R. L. Rusk, vol. 1 (New York, 1939), 246. 36 J. Dakyns, The Middle Ages in French Literature, 1851–1900 (Oxford, 1973), 1–28, and E. Siberry, “Images of the Crusades in the Nineteenth and Twentieth Centuries,” in Oxford History (as in note 5), 365–85. 37 K. Morris, The Image of the Middle Ages in Romantic and Victorian Literature (London, 1984), 46, 105. 38 Dakyns, Middle Ages (as in note 36), 4, 17, 254–56.

1 Frontispiece to T. Fuller, Historie of the Holy Warre (Cambridge, 1640)

2 Illustration by Gustave Doré for the 1877 Paris edition of J. Michaud, Histoire des Croisades

Giles Constable [ 9 ] In historical writing this shift was marked by the appearance of two important multivolume histories of the crusades, one in Germany and the other in France. The first, by Friedrich Wilken, was published between 1807 and 1832 and is still of scholarly value; the second, by J. F. Michaud, appeared between 1812 and 1822 and was often reprinted, including an edition published in Paris in 1877 with a series of illustrations by Gustave Dore´ that mark a high point in the religious and nationalistic enthusiasm for the crusades in France (Fig. 2).39 The decision of King Louis Philippe to include the family names of French participants in the crusades in the Salle des Croisades at Versailles produced a flood of forged crusading charters, which still occasionally mislead historians.40 A more serious scholarly note was struck by the three volumes of pie`ces justificatives that accompanied Michaud’s history and even more by his four volumes of translated sources, including one from the Arabic. The Recueil des historiens des croisades, which includes editions of primary sources in Latin, Greek, Arabic, Armenian, and Old French and is still a standard work of reference, was officially launched by the Acade´mie des Inscriptions et Belles lettres in 1824.41 In the late 1830s Leopold von Ranke gave “the first impulse” to a critical examination of the sources for the crusades in his seminar at the University of Berlin, and his student Von Sybel put the study of the First Crusade on a new scholarly basis in his Geschichte des ersten Kreuzzugs, which included, as mentioned above, the first considerable study of the historiography of the crusades.42 In the second half of the nineteenth century, crusading studies continued to flourish in Germany, where the names of Reinhold Ro¨hricht and Heinrich Hagenmeyer in particular come to mind, and in France, where Paul Riant founded the Socie´te´ de l’Orient latin in the 1870s. Important contributions were also made by scholars in England, Italy, and, somewhat later, the United States, where the study of the crusades was promoted by the teaching of Dana C. Munro.43 The first task of these scholars was to prepare critical editions of the sources, assess their value, and to establish the facts of the history of the crusades. This prepared the way for the appearance in the first half of the twentieth century of some new general histories, addressed to the public as well as to scholars. Among the most influential of these were the two works, both in three volumes, of Rene´ Grousset, published in 1934–36, and Steven Runciman, whose History of the Crusades was completed in 1954. These are basically narrative accounts, but Grousset as an Orientalist and Runciman as a Byzantinist both saw the crusades in terms of east-west

39 On these works see Boehm, “Gesta” (as in note 5), 73–74, and on Wilken, Von Sybel, Geschichte (as in note 4), 167–72, trans. 339–43, and on Michaud, ibid., 173–78, trans. 345–53, and Zaborov, Istoriografija (as in note 4), 179–211. 40 G. Constable, “Medieval Charters as a Source for the History of the Crusades,” in Crusade and Settlement: Papers Read at the First Conference of the Society for the Study of the Crusades and the Latin East and Presented to R. C. Smail, ed. P. W. Edbury (Cardiff, 1985), 73 and references in n. 3. 41 According to H. Dehe´rain, “Les origines du recueil des ‘Historiens des croisades,’” JSav, n.s., 17 (1919): 260–66, the first initiatives went back to the end of the 18th century. 42 Von Sybel, History (as in note 4), iii. See E. Fueter, Geschichte der neueren Historiographie, Handbuch der mittelalterlichen und neueren Geschichte, Abt. 1 (Munich-Berlin, 1936), 535–36; G. P. Gooch, History and Historians in the Nineteenth Century, new ed. (Boston, 1959), 120–21; and Zaborov, Istoriografija (as in note 4), 212–27. 43 The Crusades and Other Historical Essays Presented to Dana C. Munro by His Former Students, ed. L. J. Paetow (New York, 1928). See also Mayer, “America” (as in note 3), 38.

[ 10 ] Historiography of the Crusades relations, between either the Christians and Muslims or the Latins and Greeks.44 The new international Society for the Study of the Crusades and the Latin East has held meetings in Cardiff, Jerusalem, Syracuse (N.Y.), and Clermont-Ferrand. Meanwhile, a group of American scholars based at the University of Pennsylvania and later at Wisconsin undertook a major collaborative history of the crusades, which was published in six volumes between 1955 and 1989. It is interesting to follow the development of this work from its conception in the 1930s and 1940s down to its completion and to compare its coverage with that in the shorter, but also collaborative, Oxford Illustrated History of the Crusades. The Wisconsin history, as it is called, deals in considerable detail with the factual history of each crusade, down to the fifteenth century, and has a volume each on art and architecture and on the impact of the crusades respectively in the east and in the west. The original plan was to devote the fifth volume to political and economic institutions, agricultural conditions, crusading propaganda, western missions, religious minorities, and social history. The sixth was intended to include an atlas and gazetteer, but it turned out to be something of a catch-all and includes, in addition to other material, a chapter on crusader coins and a long bibliography. The extensive treatment given in this work to the late medieval crusades, to art and architecture, and to the impact of the crusades in the east reflected the development of crusading studies during the late nineteenth and early twentieth centuries.45 The Oxford history devotes yet more space to crusading in the late Middle Ages, to the military orders (to which the Wisconsin history gave comparatively little attention, aside from a chapter on the Teutonic Knights), and above all to the ideology and spirituality of the crusades, which has been a subject of major interest to crusader historians since the publication in 1935 of Carl Erdmann’s Entstehung des Kreuzzugsgedankens, which appeared in English in 1977 as The Origin of the Idea of Crusade.46

II. Recent Trends in Crusading Historiography More than any other single work written in the twentieth century, Erdmann’s book changed the direction of crusading studies.47 He had been trained as a theologian, studied under Paul Joachimsen at Munich, and worked on the Papsturkunden series and at the 44 On Grousset and his “colonialist” point of view, see especially Boase, “Recent Developments” (as in note 3), 116–22; H. Berr in his introduction to Alphande´ry, Chre´tiente´ (as in note 16), 1:ix; Cardini, “Studi” (as in note 3), 82–83; Mayer, “America” (as in note 3), 41 (esp. n. 18); and on Runciman, Young, “Crusades” (as in note 3), 87–97, and Cardini, “Studi” (as in note 3), 83–86. 45 See Mayer, “America” (as in note 3), 42–44, and D. Queller, “Review Article: On the Completion of A History of the Crusades,” International History Review 13 (1991): 314–30, who discussed a number of other recent works on the crusades. 46 C. Erdmann, Die Entstehung des Kreuzzugsgedankens, Forschungen zur Kirchen- und Geistesgeschichte 6 (Stuttgart, 1935); see note 10 above for the translation, which has additional notes by the translators. 47 Erdmann, as his title says, concentrated on the origin of the idea of crusade, not its later history, which had already been studied, among others, by O. Volk, Die abendla¨ndisch-hierarchische Kreuzzugsidee (diss., HalleWittenberg, 1911), who covered the popes from Leo IX to Gregory IX (omitting, somewhat oddly, Urban II), and he broke, according to J. Richard in his introduction to Delaruelle, Ide´e (as in note 10), vii, with the thesis that “the crusade was explained above all by its end: the Holy Land.” Barraclough, “Deus le Volt?” (as in note

Giles Constable [ 11 ] Monumenta Germaniae Historica. He therefore combined the tradition of German intellectual history or Geistesgeschichte, which emphasized the ideas underlying the observable events of history, with a rigorous training in source-criticism. He was not alone in his interest and approach. Etienne Delaruelle and Paul Alphande´ry in particular wrote along parallel lines at almost the same time as Erdmann, though their works were published later. The series of articles by Delaruelle published between 1941 and 1954 under the title “Essai sur la formation de l’ide´e de croisade,” and republished together in 1980, originated in a thesis presented at the Institut Catholique in Paris in 1935, the same year Erdmann’s book appeared.48 Alphande´ry’s two posthumous volumes on La chre´tiente´ et l’ide´e de croisade, published in 1954 and 1959, were based on his lectures at the Ecole des hautes e´tudes before his death in 1932.49 As a historian of religion, Alphande´ry was interested in the spontaneous and charismatic aspects of the crusades, of which he found the essence in the expedition led by Peter the Hermit and in the so-called popular crusades, to which the Wisconsin and Oxford histories devoted comparatively little attention, aside from a chapter on the Children’s Crusade in the Wisconsin history. Contemporaries for the most part had no clear concept of the ideology of crusading, which lay to a great extent beneath the surface of the events. There was not even a single generally accepted term for a crusade.50 At their origins, and indeed throughout the Middle Ages, crusades were usually referred to by terms, both in Latin and the vernacular, indicating movement or travel, such as peregrinatio, iter, via, expeditio, and later passagium, and the corresponding verbs, often combined with a reference to Jerusalem, the Holy Land, the Holy Sepulcher, or the cross, and in the vernacular with outre mer or u¨ber meer. They often expressed a religious engagement or desire (negotium, bellum, causa, opus, voluntas, or later simply crux), and referred to its sacred character or to God, Christ, or Jerusalem. Although the early crusaders were sometimes referred to as signed with or bearers of the cross,51 the cross did not become the mark of crusading, as distinct from pilgrimage, until the end of the twelfth century. The earliest known use of crozada is in Spain and southwestern France in the early thirteenth century, but it remained rare, as did croiserie and croisade, and crusade was not common in English before the eighteenth

8), 14 n. 3, called Erdmann’s book “by far the most important contribution to crusading history in the last generation.” For some reservations see J. Gilchrist, “The Erdmann Thesis and the Canon Law, 1083–1141,” in Crusade and Settlement (as in note 40), 37–45, and M. Bull, “The Roots of Lay Enthusiasm for the First Crusade,” History 78 (1993): 355–59. 48 E. Delaruelle, “Essai sur la formation de l’ide´e de Croisade,” Bulletin de litte´rature eccle´siastique 42 (1941): 24–45, 86–103; 45 (1944): 13–46, 73–90; 54 (1953): 226–39; 55 (1954): 50–63; reprinted in his Ide´e (as in note 10), 2–127, with an introduction by J. Richard. 49 An interesting account of Alphande´ry and of Dupront’s edition of his work is found in M. Balard’s postscript to the reprint (Paris, 1995), 565–93. Balard praised the work especially for its concern with the interior history of the crusade and with the history of collective mentality and psychology (575), but said that since the 1980s it has been more cited than used, especially by English-speaking scholars. 50 The single best discussion of this topic, which is mentioned by many scholars, is in D. A. Trotter, Medieval French Literature and the Crusades (Geneva, 1988), 31–70. See also Rousset, Ide´ologie (as in note 6), 51–57; RileySmith, What Were the Crusades? (as in note 10), 12; and A. Dupront, Du sacre´: Croisades et pe`lerinages. Images et langages (Paris, 1987), 239–63. 51 M. Markowski, “Crucesignatus: Its Origins and Early Usage,” JMedHist 10 (1984): 158, and C. Tyerman, “Were There Any Crusades in the Twelfth Century?” EHR 110 (1995): 575.

[ 12 ] Historiography of the Crusades century.52 The fourteenth-century French crusading propagandist Philippe de Me´zie`res called the crusade “the hunt of God . . . to capture the rich prize,”53 and for Fuller in the seventeenth century it was simply the holy war. The participants in the crusades were normally referred to in the early sources as pilgrims or Christians or, depending on the writer, as milites Dei or Christi, pauperes, or Hierosolymitani and later as cruciferi and crucesignati, though some of these terms could also apply to pilgrims. In Old French sources the crusaders were called pe`lerins, croise´s, or Franks. Collectively they were the populus, plebs, gens, militia, or exercitus Dei, and their enemies were infideles, barbari, and pagani. Nothing in this terminology clearly distinguished the crusades from pilgrimages, and it offers little or no guidance to scholars seeking to define a crusade. Those who want a strict definition mostly agree on the importance of taking the cross, making a vow, and the granting by the papacy of spiritual and worldly privileges, though whether the promise of forgiveness from sins applied to eternal as well as temporal punishments is uncertain. They disagree, however, on the centrality of the objective of a crusade. The socalled traditionalists hold that a true crusade must be directed toward the east, either to assist the Christians there or to liberate Jerusalem and the Holy Sepulcher,54 whereas for the so-called pluralists the defining feature of a crusade, whatever its objective, is papal authorization. The traditionalists ask where a crusade was going and therefore hold that the crusades basically ended with the fall of the crusader states in the east. The pluralists, on the other hand, ask how a crusade was initiated and organized and thus extend the history of the crusades not only geographically but also chronologically, down to recent times.55 Both approaches present problems. The traditionalists reject, and even regard as a corruption of legitimate crusading, any crusade not directed toward the east, including those in Spain and northern Europe, and those against heretics, schismatics, and other enemies of the church, even when they were called by the papacy and rewarded by spiritual privileges. These present no difficulty for the pluralists, who find it hard to fit into their definition the “popular” crusades, which were neither authorized nor sup-

52 See the charter of October 1212 in S. A. Garcia Larragueta, El gran priorado de Navarra de la orden de san Juan de Jerusalen: Siglos XII–XIII, vol. 1, Estudio preliminar (Pamplona, 1957), 149, no. 145, and P. Ho¨lzle, Die Kreuzzu¨ge in der okzitanischen und deutschen Lyrik des 12. Jahrhunderts, Go¨ppinger Arbeiten zur Germanistik 278 (Go¨ppingen, 1980), 34 (and 31–34 generally). For English, see The Oxford English Dictionary, 2d ed., vol. 4 (Oxford, 1989), 85. 53 Philippe de Me´zie`res, Letter to King Richard II, ed. and trans. G. W. Coopland (Liverpool, 1975), 101. See J. Williamson, “Philippe de Me´zie`res and the Idea of Crusade,” in The Military Orders: Fighting for the Faith and Caring for the Sick, ed. M. Barber (Aldershot-Brookfield, Vt., 1994), 358–64. 54 For Erdmann, Origin (as in note 10), xxi, “Jerusalem was the immediate goal of the campaign (Marschziel ), but liberation of Eastern Christianity from the infidel remained the fundamental aim of the war (Kampf- or Kriegsziel)”; cf. xxviii and 348, saying that pilgrimage was “a late addition.” 55 On the distinction between the traditionalists and the pluralists, see Riley-Smith, in Oxford History (as in note 5), 8–10, and Housley, Later Crusades (as in note 22), 2–3. Among modern scholars the leading traditionalist is H. E. Mayer and the leading pluralist is J. Riley-Smith, whose Atlas of the Crusades (New York–Oxford, 1991) reflects this approach in its breadth of coverage. An attempt to bridge the gap is made by J. Richard, Histoire des croisades (Paris, 1996), on which see Riley-Smith in the Times Literary Supplement, 2 May 1997, 28, who says that Richard “does not quarrel with the pluralist definition” but sees “the earlier enterprises” as “Crusades par excellence.”

Giles Constable [ 13 ] ported by the papacy but which for some scholars embody the essence of crusading.56 Both groups are uncertain what to do with the so-called pre- or proto-crusades, which were neither directed toward the east nor summoned by the pope. I have myself been counted among the pluralists owing to my article showing that contemporaries regarded the expeditions against the Wends and Muslims on the Iberian peninsula as part of the Second Crusade,57 but I am reluctant to exclude the “popular” crusades or to deny that at least a spiritual orientation toward Jerusalem was an essential aspect of crusading. Von Ranke was the first, so far as I know, to distinguish between what he called the hierarchical or official and the popular impulse (Moment) of crusading.58 Erdmann also stressed “the fundamental distinction between the hierarchical and popular ideas of crusade,” where his use of Ideen in place of von Ranke’s Moment reflected his interest in ideology.59 The traditionalists and pluralists both tend to look at the official aspects of crusading, but another group of scholars adhere to what may be called a spiritual or psychological definition that emphasizes the inner spirit and motives of the crusaders and their leaders. Alphande´ry said that “throughout the west the crusade was a project swept along by eschatological forces, the idea of the proximate coming of AntiChrist, the conquest of the last days, the belief in the dwelling place of the saints in Jerusalem.”60 For Delaruelle the crusade was a permanent miracle that “originally appeared as a moment of collective exaltation, like a ‘prophetic’ deed by which a man of God announces to an entire people that an hour has come, like the meeting with the Savior . . . , a privileged moment without a tomorrow.”61 These writers and their followers see the crusades as a religious groundswell of the socio-religious elect, the pauperes, humiles, and others who made themselves children for the sake of God.62 For them the only true crusade was the First, which was marked by widespread religious enthusiasm and popular response.63 Some have posited two First Crusades: one 56 P. Raedts, “The Children’s Crusade of 1212,” JMedHist 3 (1977): 300: “Any definition of the crusade is unsatisfactory if it does not include the hordes who streamed toward Jerusalem in the wake of Peter the Hermit, Emicho of Leiningen and so many others.” See F. Cardini, “Per una ricerca sulle crociate popolari,” Quaderni medievali 30 (1990): 156–67. 57 G. Constable, “The Second Crusade as Seen by Contemporaries,” Traditio 9 (1953): 213–79. 58 L. von Ranke, Weltgeschichte, vol. 8, Kreuzzu¨ge und pa¨pstliche Weltherrschaft (XII. und XIII. Jahrhundert) (Leipzig, 1898), 71, 80. 59 Erdmann, Origin (as in note 10), 269, 355 n. 2, and introduction, xxxv. 60 Alphande´ry, Chre´tiente´ (as in note 16), 1:97, 177, 194. Dupront, Du sacre´ (as in note 50), 290, described the crusade as “a march to the meeting with the Second Coming at the end of time” and gave further references to Alphande´ry. 61 Delaruelle, Ide´e (as in note 10), 246. 62 Alphande´ry, Chre´tiente´ (as in note 16), 2:36–40, 67, 127–47; Delaruelle, Ide´e (as in note 10), 122; P. Rousset, “L’ide´e de croisade chez les chroniqueurs d’Occident,” in X Congresso (as in note 15), 560–61; and M. Mollat, Les pauvres au moyen-aˆge (Paris, 1978), 95, who saw the crusade as “par essence et essentiellement” an affair of the poor. See V. Epp, Fulcher von Chartres: Studien zur Geschichtsschreibung des ersten Kreuzzuges, Studia humaniora 15 (Du¨sseldorf, 1990), 242–50, on the use of the term pauperes in the work of Fulcher of Chartres, who used it in a positive as well as a derogatory sense. G. Miccoli, “Dal pellegrinaggio alla conquista: Poverta` e ricchezza nelle prime crociate,” in Poverta` e ricchezza nella spiritualita` dei secoli XI e XII, Convegni del Centro di studi sulla spiritualita` medievale 8 (Todi, 1969), 45–80, warned against referring to “the poor as such” in the First Crusade and argued that in the course of the 12th century symbolic/spiritual poverty changed to literal poverty. 63 Rousset, “L’ide´e” (as in note 62), 547, and Ide´ologie (as in note 6), 19, 61, and W. Goez, “Wandlungen des Kreuzzugsgedankens in Hoch- und Spa¨tmittelalter,” in Das Heilige Land im Mittelalter: Begegnungsraum zwischen

[ 14 ] Historiography of the Crusades official, led by the princes who responded to the appeal of Urban II, and the other popular, led by Peter the Hermit, whose traditional role as the initiator of the First Crusade (which is based on the chronicle of Albert of Aachen and depicted on the frontispiece to Fuller’s Historie) has recently found some defenders against the attacks of the nineteenth-century scholars who asserted the priority of the official crusade.64 Alphande´ry was particularly interested in the visions, miracles, and apocalyptic signs that accompanied the First Crusade. By the time of the Fourth Crusade, he said, “the animating center of the deed of crusade tends to become a symbol . . . the crusade is enclosed in interior combat.” The whole movement petered out after the Children’s Crusade, which still expressed “the deep life of the very idea of crusade,” and came to an end with Frederick II’s negotiated recovery of Jerusalem and the resumption of the “tolerated” pilgrimages of the early Middle Ages.65 There is, finally, a group of historians who can be called generalists and who broadly identify the crusades with holy war and the justification of fighting in defense of the faith—the astonishing effort, as Michel Villey put it, to baptize war.66 They emphasize in particular the traditional concept of the just war, the ideal of Christian knighthood that emerged in the tenth century, the regional movements known as the Peace and Truce of God and designed to protect particular categories of people and to prevent fighting at certain times, and the efforts of the popes in the eleventh century to mobilize the milites sancti Petri to support and defend the papacy. Ernst-Dieter Hehl, in an article entitled “Was ist eigentlich ein Kreuzzug?” (What essentially is a crusade?), rejected both the traditionalist and pluralist definitions of a crusade as too restrictive and argued that a crusade was a war fought at the order of and with the authority of God—“a Deo auctore war”—and that Urban’s innovation was to fit the crusade into “a historicaltheological schema” or “a theology of war.”67 According to this view, the essential features of a crusade were to carry out the will of God on earth and thus to win forgiveness for sins, with or without papal approval. Jerusalem was thus spiritualized, and in practice a crusade could be directed against any perceived enemies of God, even though the crusade to the east continued, as Christopher Tyerman put it, to provide “the language of crusading.”68 In the middle of the thirteenth century, the canon lawyer Hostiensis, Orient und Okzident, ed. W. Fischer and J. Schneider (Neustadt a. d. Aisch, 1989), 34, who described the First Crusade as “sui generis.” 64 J. Flori, “Une ou plusieurs ‘premie`re croisade’? Le message d’Urbain II et les plus anciens pogroms d’Occident,” RH 285 (1991): 3–27; M. Bull, Knightly Piety and the Lay Response to the First Crusade: The Limousin and Gascony c. 970–c. 1130 (Oxford, 1993), 256; Ward, “First Crusade” (as in note 9), 264–65; and, on Peter the Hermit, E. O. Blake and C. Morris, “A Hermit Goes to War: Peter and the Origins of the First Crusade,” Studies in Church History 22 (1985): 79–107; M. D. Coupe, “Peter the Hermit—A Reassessment,” Nottingham Medieval Studies 31 (1987): 37–45; Ward, “First Crusade” (as in note 9), 285–87; and D. Lohrmann, “Albert von Aachen” (as in note 2), 150–51. 65 Alphande´ry, Chre´tiente´ (as in note 16), 2:110, 147, 257. Goez, “Wandlungen” (as in note 63), 42, said that there was no real passagium generale after 1228/9. 66 M. Villey, “L’ide´e de croisade chez les juristes du moyen-aˆge,” in X Congresso (as in note 15), 593. 67 E.-D. Hehl, “Was ist eigentlich ein Kreuzzug?” HZ 259 (1994): 297–336 (citations on 301 and 307). 68 C. Tyerman, “The Holy Land and the Crusades of the Thirteenth and Fourteenth Centuries,” in Crusade and Settlement (as in note 40), 108. Cf. Raedts, “Children’s Crusade” (as in note 56), 301, who said that “I shall consider as crusades all expeditions, armed or unarmed, whose participants took a vow and intended to liberate

Giles Constable [ 15 ] while reserving “the vow of the cross” to the pope, wrote that “If the crusade across the sea (crux transmarina) is and should be preached in order to acquire or recover the Holy Land, then the crusade against the schismatics on this side of the sea (crux cismarina) should be preached all the more strongly in order to preserve the unity of the church. . . . For the Son of God did not come into the world and suffer the cross to acquire land but to redeem captives and to recall sinners to repentance.”69 This points toward a broad definition of a crusade, as Riley-Smith put it, as “a holy war fought against those perceived to be the external or internal foes of Christendom for the recovery of Christian property or in defense of the Church or Christian people,” and carrying with it, one might add, an expectation, implicit or explicit, of forgiveness of sins for those who participated.70 The view of modern historians who see the crusades as the beginning of European colonialism and expansion would have surprised people at the time. They would not have denied some selfish aspects, including a search for salvation and a desire to escape unwelcome obligations and to find a new life away from home, but the predominant emphasis was on defense and the recovery of lands that had once been Christian and on the self-sacrifice rather than the self-seeking of the participants. Since there was no clear concept of the crusades, however, their character changed over time. On the one hand, they became more institutionalized as their various features were defined by the popes and canon lawyers of the thirteenth century.71 Prudence and efficiency rather than enthusiasm became the prerequisites for a crusade, and there was an increasing stress on organization, regulations, fiscal arrangements, and administrative routines.72 Disagreements continued, however, even over the nature of an official crusade. In the thirteenth century, when the popes used the crusades against any enemies of the church, Hostiensis said that some people held that it was unjust and dishonest to take the cross against Christians, and at the Council of Basel in 1420 Alonso of Cartagena argued that a holy war must be against infidels.73 At the same time, however, the crusades were spiritualized and internalized, as in the crusading sermons where “the idea of soldiering for Christ is tied inextricably to the idea of the crusade as an imitation of Christ” and to “a moral and spiritual renewal” leading the crusader “to Christ’s cross of suffering and physical death in battle.”74 This Jerusalem and other holy places from Moslem rule or to defend them for Christendom,” and Housley, Later Crusades (as in note 22), 45, 47, 49. 69 Henry of Susa (Hostiensis), Summa aurea, 3.19 (Venice, 1574; repr. Turin, 1963), 1141. See Villey, “Ide´e” (as in note 66), 568, 578–81. 70 J. Riley-Smith, The Crusades: A Short History (New Haven-London, 1987), xxviii. 71 The canonists, according to Villey, “Ide´e” (as in note 66), 593, created the concept and theory of crusade because they wanted to assure the institution “of a long and solid future.” See the elaborate definition of a crusade in M. Purcell, Papal Crusading Policy: The Chief Instruments of Papal Crusading Policy and Crusade to the Holy Land from the Final Loss of Jerusalem to the Fall of Acre, 1244–1291, Studies in the History of Christian Thought 11 (Leiden, 1975), 10–11. 72 Delaruelle, Ide´e (as in note 10), 246, discussed whether the crusade should be seen as an event or an institution. 73 Henry of Susa, Summa aurea, 3.19 (as in note 69), 1141, and Cantarino, “Spanish Reconquest” (as in note 14), 100. 74 Cole, Preaching (as in note 13), 124–25; see 172–73 on “the penitential idea of the crusade.”

[ 16 ] Historiography of the Crusades concept of the crusade has been called penitential and imitationist. True crusading never became an institution, according to Alphande´ry, who said that by the thirteenth century “the monk-knight of the preceding age disappears before the vir spiritualis, poor, weak, predestined to the glory of the saints.”75 Not many scholars today would go as far as this, but there is a tendency to move away from the factual history of the crusades and their growing definition—some would say deformation—at the hands of the popes and canonists and to take a more flexible view of the crusade as an event rather than an institution.76 “Crusading is coming to have the appearance of a spectrum of enterprises,” wrote Riley-Smith, “each with its own personality, united by common elements.”77 And in an article provocatively entitled “Were There Any Crusades in the Twelfth Century?” Tyerman argued that “what we call ‘the Crusades’ in fact covered a fragmented series of military and religious activities that lacked coherence” and that “the First Crusade only appeared as the beginning of a coherent movement retrospectively when that movement existed, after 1187.”78 This raises the question, to which no satisfactory answer has so far been given, of the numbering of the crusades, which differs widely both in the sources and in subsequent histories. After the first few crusades, Mayer wrote: “The numbering of the crusades lacks all consistency. Many scholars do not count the Damietta crusade at all and for them, Frederick II’s crusade of 1228–9 is the fifth and Saint Louis’s first crusade (1248– 50) the sixth. Others count the Damietta crusade but not Frederick II’s. Still others count the Damietta crusade as the fifth, Frederick II’s as the sixth, and Saint Louis’s as the seventh.”79 Mayer recently wrote that “Even the numbering of crusades I–IV is a dubious affair. It is accepted by everyone, but it only counts the general expeditions in which more or less all of Europe was involved. It made people blind to the smaller crusades.”80 No one, among either the traditionalists or the pluralists, assigns a number to the crusades in Spain or northeastern Europe, the Children’s or other “popular” crusades, the Albigensian crusade, the Mongol crusade of 1241, and other expeditions that seem to meet the definition of a crusade and are commonly referred to as such. Although there was in the twelfth century an awareness of previous crusades, as the arenga to Quantum predecessores shows, there was no practice of numbering them. When Ordericus Vitalis in the 1130s referred to the crusade in 1107 as “the third expedition of the westerners to Jerusalem,” he was presumably counting those in 1096–97 and

75 Alphande´ry, Chre´tiente´ (as in note 16), 2:112, 160, 163. See also his “Citations bibliques” (as in note 11), 149. These views are well on the way to what has been called the myth or metahistory of the crusade: see Rousset, Ide´ologie (as in note 6), 211–13; G. Fedalto, Perche` le crociate (Bologna, 1986), 70; and Dupront, Du sacre´ (as in note 50), 34–35. 76 See the definitions of a crusade by Raedts, cited in note 56 above, and D. Nicol, The Crusades and the Unity of Christendom, Friends of Dr. Williams’s Library 40th Lecture (London, 1986), 5, who (writing as a Byzantinist) called it “a campaign for the liberation of a faraway place dreamed up in a fog of pious romanticism.” 77 Riley-Smith, “History” (as in note 6), 10. 78 Tyerman, “Were There Any Crusades” (as in note 51), 554, 558, 566 n. 3. He went on to say that “Innocent III transformed one sort of ecclesiastical warfare into juridical crusading.” 79 H. E. Mayer, The Crusades, trans. J. Gillingham, 2d ed. (Oxford, 1988), 314 n. 117. See also Bull, “Roots” (as in note 47), 354. 80 Personal letter, 26 Feb. 1997.

Giles Constable [ 17 ] 1101–2 as the first two.81 It is possible, however, to see the entire period from 1095 until 1107, and even later, as part of the response to the appeal for the First Crusade, of which the message spread slowly and to which participants responded at different times.82 The numbers given to the crusades by later writers during the first period of crusading historiography also deserve to be studied.83 Fuller, who ushered in the second period, called the crusades “the Holy Warre” in the singular, but in his chronological table he distinguished thirteen voyages (or pilgrimages, as he called them in the text) between 1095 and 1269, counting separately, in addition to the presently numbered crusades, the expeditions of 1101, Henry of Saxony in 1197, the king of Hungary in 1216, Theobald of Navarre in 1239, and Richard of Cornwall in 1241. Maimbourg and Gibbon both counted seven crusades, but Wilken used no system of numbering, and some modern scholars identify the crusades simply by date. This approach has led to the reformulation and reexamination of various questions concerning crusading. Among these is motivation, which was traditionally considered to include both secular and religious motives, of which the respective importance was assessed differently by scholars according to their religious beliefs and the standards of their times. To these should be added an emotional or psychological element that was neither specifically religious nor secular.84 “Sentiment, not strategy, has always been the dominant factor in the affairs of Palestine,” said the reviewer of a recent book on the background of Palestinian-Israeli disputes. “An understanding of the problems of Arabs and Jews [or of Muslims and Christians] in the Holy Land should begin by clearing the mind of the confusing clutter of strategy and by focusing on the sentiments, or, to put it more precisely, the passionate feelings, of those concerned.”85 Very little is known about the sentiments of the crusaders, let alone their passionate feelings, but they certainly felt a fierce loyalty to Christ and a sense of outrage that his patrimony and tomb were held by infidels and could be visited by Christians only on sufferance. Many years ago Adolf von Harnack said that “the enthusiasm of the Crusades was the direct fruit of the monastic reform of the eleventh century,”86 and Erdmann associated the concept of holy war with the efforts of the popes, and especially Gregory VII, to free the church from the control of laymen. If lay investiture and lay possession of ecclesiastical property and revenues were abhorrent to the reformers, how much more so the control by the Muslims 81 Ordericus Vitalis, Historia ecclesiastica, 5.19, ed. M. Chibnall, Oxford Medieval Texts (Oxford, 1969–80), 3:182. See Riley-Smith, First Crusaders (as in note 20), 9 and 109, citing references in other sources to the “first” and “second” expeditions. 82 See the works cited in note 64 above, especially Bull, Knightly Piety. 83 See Schmugge, Kreuzzu¨ge (as in note 22), 34, on the numbering of the crusades by Carion and Bullinger, both of whom counted twelve expeditions (reckoning the Third Crusade as three). 84 On the levels of motivation among participants in the First Crusade, see J. Flori, La premie`re croisade (Brussels, 1992), 225–29. 85 M. E. Yapp, reviewing D. Hiro, Sharing the Promised Land, in the Times Literary Supplement, 14 Feb. 1997, 28. 86 A. von Harnack, Monasticism: Its Ideas and History and the Confessions of St Augustine, trans. E. E. Kellett and F. H. Marseille (London, 1901), 87. Barraclough, “Deus le Volt?” (as in note 8), 12–13, said that “the Crusades, far from being a unique or isolated phenomenon, were only a particular manifestation of a great spiritual crisis without parallel until the sixteenth century,” and Miccoli, “Pellegrinaggio” (as in note 62), 46: “La crociata, nei suoi aspetti ideologici, nasce come tipica espressione della riforma gregoriana.”

[ 18 ] Historiography of the Crusades of the most sacred of Christian shrines. This is not to say that all crusaders were religious reformers, but the view that they were motivated largely by greed and self-interest has been to some extent replaced by an acceptance of their sincerity and idealism, combined with a recognition that altruistic and selfish motives were unconsciously mixed in the minds of individual crusaders. According to Riley-Smith, “There can be little doubt that those who took the cross, and the families who helped to finance them, were moved on the whole by idealism. The only explanation for their enthusiasm seems to be that Urban’s message encountered the laity’s growing aspirations and the hand stretched out by the Church to lay people was suddenly grasped.”87 Not all scholars accept this idealistic and somewhat defensive stance, and several recent writers have pointed out the failure of historians to take account of psychological, sociological, and economic theory in studying the motivation of the crusaders. In an article on “The Motives of the First Crusaders: A Social Psychological Analysis,” published in the Journal of Psychohistory in 1990, the crusades were seen as a way of resolving the tension, or endemic cognitive dissonance, between the religious ideals and worldly violence of medieval society. Urban’s crusade offered “the new reconciliation” between the desire for salvation and the need to fight and became a mass movement because it met a widespread psychological need.88 John Ward studied the First Crusade in terms of disaster theory, using the so-called Foster scale, and argued that it was a remedial “disasterreaction” to the “inflammatory millenarian force,” Muslim advances in the east, and consciousness of sin in the late eleventh century. He called this “a ‘postmodern deconstruction and reconstruction’” of the nineteenth-century myth of the official, hierarchial crusade and concluded that “Crusade is thus interesting not so much for what it was, . . . but for what contemporaries thought it would be.”89 A very different view was put forward by four economists, none of them specialists on the crusades, in an article entitled “An Economic Interpretation of the Medieval Crusades” published in the Journal of European Economic History in 1992. They see the crusades in terms of “contemporary economic theory” as “(a) a supply-side response to the attempts of the Moslems and Turks to raise a rival’s (the Latin church’s) cost in maintaining the credibility of its product and (b) a demand-side attempt by the medieval church to maintain and maximize the value of its wealth by expansion of market area and monopoly control.” Although the crusades “claimed to be primarily motivated by ideological or theological fervor,” they were in fact “an essential part of a wealth-maximizing strategy” both by the church, of which the monopoly over salvation—called “a pure credence good”—was threatened by the Muslims, and by individual crusaders who hoped to make their fortunes in the east.90 The views of Voltaire and Gibbon have thus been revived by modern economic theorists. Another old question that has been addressed with new interest in recent years is the 87

J. Riley-Smith, The First Crusade and the Idea of Crusading (London-Philadelphia, 1986), 153. J. R. E. Bliese, “The Motives of the First Crusaders: A Social Psychological Analysis,” Journal of Psychohistory 17 (1990): 393–411 (quotation on 400). 89 Ward, “First Crusade” (as in note 9), 253–92 (quotations on 259 and 288–89). 90 G. M. Anderson et al., “An Economic Interpretation of the Medieval Crusades,” Journal of European Economic History 21 (1992): 339–63 (quotations on 340 and 342). 88

Giles Constable [ 19 ] background and origin of the crusades. Some scholars believe that they sprang almost out of nothing, like Athena out of the head of Zeus, as Alphande´ry put it, calling them “the almost spontaneous outpouring of a prodigious power of collective animation.”91 Others stress the extensive prehistory, reaching back to the early Christian tradition of pilgrimage, the development of the theories of martyrdom and the just war, the Byzantine campaigns against the Muslims, the changing role in Christian spirituality of Jerusalem, both heavenly and earthly, the emergence of chivalric values and the effort to harness them to the interests of local law and order by the Peace and Truce of God, and the policy of the papacy, which Erdmann examined especially in the second half of the eleventh century. The recent defense of the authenticity of the crusading encyclical attributed to Sergius IV, which has long been considered a forgery, would carry back the papal prehistory of the crusades another half century,92 and raises the question of to what extent the campaigns in Spain in the mid-eleventh century, and later in southern Italy, should be considered authentic crusades.93 This research has important implications for the study of the First Crusade, which has been subjected to intensive investigation owing to recent commemoration of its ninehundredth anniversary. Among the many questions that have been raised are not only those, mentioned above, of the content of Urban’s crusading appeal, especially the relative importance he gave to helping Christians in the east and to freeing Jerusalem; the role of Peter the Hermit and his army; the massacres of the Jews (were they an intrinsic part of the crusade or an aberration?); the influence on the chronicles and histories of the development of the crusade;94 but also how Urban’s message was spread and the crusading armies gathered; the participation in the expeditions and how they were financed;95 the nature of the oaths taken by the crusading leaders to the Byzantine emperor;96 and the relation of the Latin histories and chronicles to the Old French crusading

91 Alphande´ry, Chre´tiente´ (as in note 16), 1:9. Cf. Flori, Premie`re Croisade (as in note 84), 231, on whether the crusade was an end or a beginning. 92 H. M. Schaller, “Zur Kreuzzugsenzyklika Papst Sergius’ IV.,” in Papsttum, Kirche und Recht im Mittelalter: Festschrift fu¨r Horst Fuhrmann zum 65. Geburtstag, ed. H. Mordek (Tu¨bingen, 1991), 135–53, with references to previous literature on this disputed topic. 93 See among other works Cantarino, “Spanish Reconquest” (as in note 14), and Fletcher, “Reconquest” (as in note 14), 31–47, who denied that the campaigns in Spain were crusades before the 12th century. 94 According to Erdmann, Origin (as in note 10), 147, “The crusading idea became articulate only after it had developed in real life.” See E. O. Blake, “Formation” (as in note 17), 11–31, who referred to “a developing sense of pattern” (21), and C. Morris, “The Aims and Spirituality of the First Crusade as Seen through the Eyes of Albert of Aachen,” in Saints and Saints’ Lives: Essays in Honour of D. H. Farmer, Reading Medieval Studies 16 (Reading, 1990), 99–117: “The ideas of the crusaders were further, and deeply, shaped by the remarkable experiences of the expedition itself” (99). On the impact of the conquest of Jerusalem on the concept of the crusade, see Goez, “Wandlungen” (as in note 63), 38–39; and on the dispute over the extent to which the idea of martyrdom emerged during the crusade, see Daniel, Islam (as in note 15), 314–15; H. E. J. Cowdrey, “Martyrdom and the First Crusade,” in Crusade and Settlement (as in note 40), 46–56; J. Flori, “Mort et martyre” (as in note 17), 121–39; and J. France, “The Destruction of Jerusalem and the First Crusade,” JEH 47 (1996): 1–17. 95 See F. A. Cazel, “Financing the Crusades,” in the Wisconsin History (as in note 3), 6:116–49, and J. Riley-Smith, “Early Crusaders to the East and the Costs of Crusading, 1095–1130,” in Cross Cultural Convergences (as in note 26), 237–57. 96 See J. Pryor, “The Oaths of the Leaders of the First Crusade to Emperor Alexius I Comnenus: Fealty, homage—pi´sti", doulei´a,” Parergon, n.s., 2 (1984): 111–41; Mo¨hring, “Kreuzzug und Dschihad” (as in note

[ 20 ] Historiography of the Crusades cycle, some parts of which are older and closer to the historical sources than was once thought.97 Among the most important developments in crusading studies in the second half of the twentieth century has been the attention given by scholars such as John La Monte, Joshua Prawer, Jean Richard, Hans Mayer, and Jonathan Riley-Smith to the history of the Latin Kingdom and other crusader states. Prawer, in the introduction to his Crusader Institutions, remarked on the shift of interest from “the Crusades as a movement to the history of Crusader establishments in the East” and to “the European colonies on the eastern shores of the Mediterranean,” especially their constitutional, legal, cultural, ecclesiastical, social, and economic history.98 This can already be seen in the Wisconsin History of the Crusades, which includes chapters not only on the political and institutional but also on the cultural history of the crusader states. Mayer has in particular studied the role of the church and ecclesiastical institutions and the chancery of the Latin kings of Jerusalem.99 The old view of the Latin Kingdom as a classic feudal state, which was based to a great extent on legal sources, has been increasingly replaced by a view that puts significantly greater emphasis on the power of the monarchy.100 Meanwhile, there has been a vigorous debate over the question of whether or not the Latin states should be regarded as colonies in the modern (and characteristically pejorative) sense of the term.101 A similar array of questions surrounds the history of the later crusades, including the waning enthusiasm and growing criticism, which is usually seen as developing from a relatively few isolated voices in the twelfth century, beginning with the reaction to the failure of the Second Crusade, into the chorus of doubts reflected in the memoirs solicited by Pope Gregory IX in preparation for the Council of Lyons in 1274.102 Norman Housley in The Later Crusades, however, maintained that “fundamental questioning of the validity of crusading existed from the start of the movement and was at its strongest 3), 367–68; and R.-J. Lilie, Byzantium and the Crusader States, 1096–1204, trans. J. C. Morris and J. E. Ridings (Oxford, 1993), 8–28. 97 See J. Richard, “L’arrie`re-plan historique des deux cycles de la croisade,” in Les e´pope´es de la croisade: Premier colloque international (Tre`ves, 6–11 aouˆt 1984), ed. K.-H. Bender, Zeitschrift fu¨r franzo¨sische Sprache und Literatur, Beihefte, n.F., Heft 11 (Stuttgart, 1987), 6–16. Cf. Trotter, Medieval French Literature (as in note 50), 27. 98 J. Prawer, Crusader Institutions (Oxford, 1980), xi. Cf. his The Crusaders’ Kingdom: European Colonialism in the Middle Ages (New York–Washington, D.C., 1972). 99 H. E. Mayer, Bistu¨mer, Klo¨ster und Stifte im Ko¨nigreich Jerusalem, Schriften der MGH 26 (Stuttgart, 1977), and Die Kanzlei der lateinischen Ko¨nige von Jerusalem, Schriften der MGH 40 (Stuttgart, 1996). 100 S. Tibble, Monarch and Lordships in the Latin Kingdom of Jerusalem (Oxford, 1989), with references to previous literature, and J. Phillips, Defenders of the Holy Land: Relations between the Latin East and the West, 1119–1187 (Oxford, 1996). 101 Barraclough, “Deus le volt?” (as in note 8), 16. The question of the crusades as a colonial enterprise and the crusader states as colonies was posed by J. Prawer, “The Roots of Medieval Colonialism,” in The Meeting of Two Worlds: Cultural Exchange between East and West during the Period of the Crusades, ed. V. Goss and C. Bornstein, Studies in Medieval Culture 21 (Kalamazoo, Mich., 1986), 23–38, and debated at a symposium held in Jerusalem in 1984 and published under the title “The Crusading Kingdom of Jerusalem—The First European Colonial Society,” in The Horns of Hattin: Proceedings of the Second Conference of the Society for the Study of the Crusades and the Latin East. Jerusalem and Haifa, 2–6 July 1987, ed. B. Z. Kedar ( Jerusalem-London, 1992), 341–66. 102 P. A. Throop, Criticism of the Crusade: A Study of Public Opinion and Crusade Propaganda (Amsterdam, 1940) and, most recently, E. Siberry, Criticism of Crusading, 1095–1274 (Oxford, 1985), with references to other works.

Giles Constable [ 21 ] in the mid-twelfth century. . . . Fewer such basic doubts were expressed in the thirteenth century, when the crusade was safely enclosed in the armour plating of a just war framework by such canonists as Hostiensis and pope Innocent IV.”103 For some scholars, however, the framework was empty before the armor plating was put in place, and true crusading ended with the Fourth Crusade, of which the diversion to Zara and Constantinople has been the subject of scholarly dispute for more than a century.104 “The question is basically an unfruitful one,” said Mayer, “and will probably never be settled, yet even today there is no sign that the flood of literature on the subject will dry up.”105 This is owing not only to the number and complexity of the sources, which lend themselves to diverse interpretations, but also because the Fourth Crusade stands for so many critical issues in the history and nature of the crusades. The turning of the crusades against Christians in particular raised major questions for contemporaries as well as for later writers on the crusades,106 not only the traditionalists who regard the liberation of Jerusalem and the Holy Land as the essential objective of the crusades, but also the pluralists, because Innocent III disapproved of the Fourth Crusade, though later popes declared and promoted crusades against Christians. Some scholars still see the crusades in primarily European terms and as of relatively little importance to Islam.107 Even an Islamist like Claude Cahen described the crusades as “a western phenomenon” and “a western fact,” and Francis Robinson called them “mere pinpricks” from the point of view of Islam.108 In parts of the Islamic world, however, they had a profound influence almost from their inception,109 and the impact of the crusades in the east inspired the critical views of scholars like Runciman and much of the current hostility to the crusades. Scholars also disagree over the extent to which missionary work and conversion played a part in the crusades. The desire to exalt (exaltare) and expand (dilatare) the Christian faith is found in the earliest crusading sources, including the Old French and Middle High German epics,110 but conversion seems to have played a comparatively small role in motivating the crusades before the thirteenth century, when they began to be seen as “the instrument for opening a country to missionizing.”111

103

Housley, Later Crusades (as in note 22), 377. See the bibliographical articles by D. Queller and S. Stratton, “A Century of Controversy on the Fourth Crusade,” Studies in Medieval and Renaissance History 6 (1969): 235–77, and C. Brand, “The Fourth Crusade: Some Recent Interpretations,” MedHum, n.s., 12 (1984): 33–45. 105 Mayer, Crusades (as in note 79), 201. 106 Pissard, Guerre sainte (as in note 28), and N. Housley, “Crusades against Christians: Their Origins and Early Development, c. 1000–1216,” in Crusade and Settlement (as in note 40), 17–36. 107 In addition to the works of Alphande´ry (as in note 16) and Erdmann (as in note 10), see H. E. J. Cowdrey, “The Genesis of the Crusades: The Springs of Western Ideas of Holy War,” in The Holy War, ed. T. P. Murphy (Columbus, Ohio, 1976), 13. 108 C. Cahen, Orient et occident au temps des croisades (Paris, 1983), 7, 179–80, cf. 188–89, and F. Robinson, reviewing B. Lewis, The Middle East, in the Times Literary Supplement, 8 Dec. 1995, 3. 109 E. Sivan, L’Islam et la Croisade: Ide´ologie et propagande dans les re´actions musulmanes aux croisades (Paris, 1968). 110 E. R. Curtius, “Der Kreuzzugsgedanke und das altfranzo¨sische Epos,” Archiv fu¨r das Studium der neueren Sprachen 169 (1936): 53–54. 111 B. Z. Kedar, Crusade and Mission: European Approaches toward the Muslims (Princeton, 1984), 169. See A. Cutler, “The First Crusade and the Idea of ‘Conversion,’” The Muslim World 58 (1968): 57–71, 155–64, and Goez, “Wandlungen” (as in note 63), 42–43. 104

[ 22 ] Historiography of the Crusades Almost every aspect of the crusades has thus been reinterpreted, often from very different points of view, in recent years, and both the learned world and the general public show a voracious appetite for works on the crusades. More perhaps than any other phenomenon of European history, the crusades hold up a mirror to how the west sees itself and is seen by others, and as the angle of the mirror changes, so does the view of the crusades. In the Middle Ages, and down to the end of the nineteenth century, they were part of contemporary history, and the Muslims, pagans, heretics, and schismatics were seen as presenting a threat—real or imagined—to the stability of the west. After the nineteenth century they moved increasingly into the past and were regarded either with aversion or, later, with an admiration and nostalgia that grew into a myth, at the same time heroic and barbaric, that has still not been dissipated by research. Meanwhile, new myths and hostilities have been generated by the effort to relate the crusades to developments in the modern world and to see them from a non-western point of view. There is no reason to believe that this process of revision is near an end or that any agreement concerning the nature and impact of the crusades will ever be reached, given the changing concerns of contemporary society. Today no less than in the past, therefore, writings on the crusades must be interpreted in the light of the differing positions from which they were written. Institute for Advanced Study

The Idea of the Jiha¯d in Islam before the Crusades Roy Parviz Mottahedeh and Ridwan al-Sayyid The jiha¯d, often loosely translated as “holy war on behalf of Islam,” is usually understood to be a fairly stable idea in Islamic law. In that standard reference book, The Encyclopaedia of Islam, the very considerable scholar Emile Tyan tells us that the notion of jiha¯d stems from the principle that Islam, “along with the temporal power which it implies, ought to embrace the whole universe, if necessary by force.” Moreover, according to Tyan the jiha¯d is “a religious duty” and “has a perpetual character.” Tyan mentions only one major premodern figure of the Sunnı¯ tradition, Sufya¯n al-Thawrı¯, who disagreed with this view. Sufya¯n al-Thawrı¯, a thinker of the second/eighth century, regarded the jiha¯d as an obligation only as a defensive war.1 In fact, differences about the status and nature of jiha¯d are a marked feature of early Islamic law, and details about the conduct of jiha¯d continue to reflect historical circumstance throughout the history of Islamic law in the Middle East. It is important to remember that there is no authoritatively codified Islamic law before the nineteenth century .. Therefore, we have an accretional body of law (or more accurately, of legal thinking) in which minority opinions are preserved and in which there may be more than one widely held or “normative” position. This essay attempts to give some idea of the divergence of opinion on the “universal” and “perpetual” nature of the jiha¯d in the . ija¯z, the province of Western Arabia that contains first Islamic century in Syria and the H Mecca and Medina. It then tries to show how, in Iraq in the second half of the eighth century, certain normative theories of jiha¯d were accepted which continued to have widespread acceptance through and beyond the period of the Crusades. However, as there are literally hundreds of legal books and other genres of literature that deal with the jiha¯d in the centuries before the Crusades, well over half of which exist only in manuscript, this essay in no way attempts to give a full survey of the normative theories of jiha¯d in the pre-crusading period or to weigh the comparative importance of these theories at different periods. We start with a review of the primary sources available to reconstruct the earliest discussions of jiha¯d. The first half of the second Islamic century (which ends in 766 ..) saw the emergence of two genres of writing about jiha¯d which either mixed history and legal thought or attempted to set down rulings for war that relied upon historical prece1

E. Tyan, “Djiha¯d,” EI 2, 2:538–40. We wish to thank Michael Bonner, Fred Donner, and Hossein Modarressi for their very useful comments.

[ 24 ] The Idea of the Jiha¯d in Islam before the Crusades dents. The first were the books on the military expeditions organized by the Prophet in the Medinan period, and some of these books also included the military expeditions of the early caliphs. The second were the books on the conduct of state, or siyar. The books on military expeditions—most of which, unfortunately, are no longer extant—were written by such important figures as al-Zuhrı¯ (d. 124/741), ‘Urwah b. al-Zubayr (d. 94/ 714), and Mu¯sa¯ b. ‘Uqbah (d. 141/758). In fact, the late Martin Hinds suggests that before Ibn Hisha¯m (d. 213/828 or 218/833), all transmitters of biographical accounts of the Prophet were primarily concerned with this genre, including Ibn Hisha¯m’s primary source, Ibn Ish. a¯q (d. 150/767), who is usually regarded as the first major biographer of the Prophet.2 While the concern in this genre on military expeditions appears to be historical, these works were undoubtedly also seen as sources for the Islamic rules of war, as we understand from the advice of the jurist Ma¯lik b. Anas (d. 172/795) to his students to refer to the magha¯zi, or accounts of exemplary early campaigns, by Mu¯sa¯ b. ‘Uqbah and not to those of Ibn Ish. a¯q. Among the books on the conduct of state (siyar) from the second Islamic century there is a lost book by the very celebrated Syrian jurist al-Awza¯‘ı¯ (d. 157/773), of which parts are preserved in a refutation written by Abu¯ Yu¯suf (d. 182/798)3 and long extracts in the above-mentioned Ibn Ish. a¯q. There is also a lost book by Muhammad al-Nafs alZakı¯yah (d. 145/762) of which many fragments survive in the writing of later Zaidi authors.4 Another book of this genre is ascribed to the celebrated jurist Abu¯ H . anı¯fah (d. 150/767), which we also know through the above-mentioned treatise of Abu¯ Yu¯suf refuting al-Awza¯‘ı¯. In fact, Abu¯ Yu¯suf represents his own work as completing the groundwork laid by Abu¯ H . anı¯fah. A yet further work of this genre is the Siyar of Abu¯ Ish. a¯q al-Faza¯rı¯ (d. ca. 185/801), of which a section has been found in Morocco.5 While this survey is far from complete, it shows that we know in fragmentary or, occasionally, in full form literature of the second half of the century, much of which is in dialogue with the opinions of authors or “authorities” of the early second century, the end of the Umayyad period. These early surviving books on the conduct of state are united by two concerns: first, their discussion of the virtues of jiha¯d; and, second, their discussion of legal rulings related to jiha¯d, a subject that often occupies by far the largest part of such texts. At the end of the second half of the second Islamic century (which ends in 820 ..), a new genre of writing on the legal aspects of war appears, namely, the books on the khara¯j, or land tax, and on amwa¯l, the public finances of the Islamic community. The earliest example of this genre appears to be a composition written by a vizier of the caliph al-Mahdı¯ (158/774–169/785).6 The earliest extant book is the well-known treatise on khara¯j by a famous pupil of Abu¯ H . anı¯fah who became supreme judge in the period of the caliph Ha¯ru¯n al-Rashı¯d, the Abu¯ Yu¯suf mentioned above. Other such works of the

2

M. Hinds, “Al-Magha¯zı¯,” EI 2, 5:1161. Abu¯ Yu¯suf Ya‘qu¯b bin Ibra¯hı¯m, Al-Radd ‘ala¯ Siyar al-Awza¯‘ı¯, ed. Abu¯ l-Wafa¯’ al-Afgha¯nı¯ (Cairo, n.d.). 4 These fragments have been gathered by Ridwan al-Sayyid and will be published in the near future. 5 Al-Faza¯rı¯, Kita¯b al-Siyar, ed. F. Hama¯dah (Beirut, 1986). 6 See A. Ben-Shemesh, Taxation in Islam (Leiden, 1969), 8–9. 3

Roy Parviz Mottahedeh and Ridwan al-Sayyid [ 25 ] ¯ dam (d. 821/206) and by Abu¯ ‘Ubayd third Islamic century were written by Yah. ya¯ b. A al-Qa¯sim b. Salla¯m (d. 224/838). This genre continued to exist for many centuries; and . anbali author Abu¯ al-Faraj ‘Abd althe Kitab al-Istikhra¯j fi Ah. ka¯m al-Khara¯j by the H Rah. ma¯n Ibn Rajab (d. 795/1392) is, perhaps, the last book of this type to achieve prominence among Sunnı¯ jurists in the Arab world. It should be noted that after the appearance of books on the land tax and public finances, the books on the conduct of state and on the military expeditions of the Prophet appear to dwindle. However, books specifically on the jiha¯d increased. These, by and large, fall into two categories. One category consists of specific sections of the collections of Prophetic tradition, or separate books, on the “virtues of the jiha¯d.” The Kita¯b al-Jiha¯d by ‘Abd Alla¯h b. al-Muba¯rak al-Marwazı¯ (d. 181/797) must surely be one of the earliest examples of such individual books.7 Another category consists of the chapters on jiha¯d in the law books. This early literature raises many questions, only a few of which are addressed here. The view of Sufya¯n al-Thawrı¯ (d. 161/778), who, as we mentioned above, believed that only the defensive jiha¯d 8 was obligatory, is not as idiosyncratic as Tyan has led us to believe. It has already been noted by Jacqueline Chabbi that there is some divergence . ija¯zı¯ and the Syrian schools on this question. She points out that the between the H Muwat..ta’, written by the Medinan Ma¯lik b. Anas (d. 179/795), seems in the version compiled by al-Shayba¯nı¯ (d. 189/804) to lack any endorsement of warfare on the frontier in a context of jiha¯d. She concludes: It is thus possible to suppose that in the mid second/eighth century, the Medinan editor (or, at least, his H . anafı¯ editor, a generation later) may have belonged to a tendency which was skeptical about warfare on the frontier; particularly with regard to the purity of the intentions of the fighters. . . . In the Cordovan recension (but not that of al-Shayba¯nı¯) there is furthermore attributed to Ma¯lik the transmission of a h. adı¯th, according to which the most scrupulous piety (ablutions, attendance at the mosque, continued observance of prayer) would be the true riba¯.t. . . . This does indeed seem to represent a position which would effectively have been professed by Ma¯lik. . . . It may be wondered whether these traditions do not allow the supposition of a conflict of representation between traditionalists at the end of the second/eighth century. These indications could permit the fixing of the time when the ideology of jiha¯d, professed by circles yet to be identified, began to stress the meritorious aspect of military service on the frontier, while in other circles there was manifest opposition to this new point of view (possibly from the people of Arabia, i.e. of ‘Iraq . . .). If such was the case, it could be said that this conflict would, as if symbolically, have divided those who, of quietist tendency, aspired to make muja¯wara [“living close to the Ka‘ba”] . . . from those who aspired to make riba¯.t (. . . “dwelling on the frontier”). This latter would have professed a new type of activism.9 . amma¯d (Beirut, 1977). ‘Abd Alla¯h b. al-Muba¯rak, Kita¯b al-Jiha¯d, ed. N. H Al-Sarakhsı¯, Sharh. al-Siyar al-Kabı¯r, ed. S. al-Din al-Munajjid (Cairo, 1958), 1:187. 9 J. Chabbi, “Riba¯.t,” EI 2, 8:495. 7 8

[ 26 ] The Idea of the Jiha¯d in Islam before the Crusades There is a fair amount of evidence to support Chabbi’s hypothesis. Chabbi has already noticed that Ibn Qutayba (d. 276/889) tells us that al-Fud. ail b. ‘Iya¯d. (d. 187/863), who died as a muja¯wir, told an anecdote unfavorable to the destiny of frontier warriors,10 and al-Dhahabı¯ confirms this account.11 In fact, Chabbi’s supposition that the very early Ma¯likı¯ school, usually understood to express the majority opinion in the H . ija¯z, did not believe in the obligatory nature of certain kinds of aggressive war, seems borne out by statements from Ma¯lik himself quoted in the extremely important foundational text of Ma¯likı¯ law, al-Mudawwanah, compiled from Ma¯lik’s teachings by Sah. nu¯n (d. 240/854). These passages indicate that, in whatever circumstances Ma¯lik did approve of jiha¯d, he was extremely cautious—even doubtful—about the legitimacy of a Muslim offering his services in border warfare led by the Syrian Umayyads, presumably because of questions on the legitimacy of their rule. Ma¯lik is asked several times: Do you see any harm in fighting the jiha¯d against the Byzantines alongside these rulers (wula¯t)? He repeatedly says, “There is no harm in doing so,” twice justifying his stand by mentioning the Byzantine success at Mar‘ash (Germanikeia), presumably referring to its destruction by Constantine V in 129/746. The implication of the passage is that for this major H . ija¯zı¯ jurist, fighting the jiha¯d with the Syrian Umayyads was in no sense a duty of a Muslim, only a permissible act that was to some degree meritorious, especially because of the general danger to Islamic territory.12 A neglected but extremely valuable source for this subject is al-Mus.annaf, a book composed by ‘Abd al-Razza¯q al-S. an‘a¯nı¯ (d. 211/826), which consists of materials he gathered from his teachers in the H . ija¯z and in Syria. An analysis of this material allows us to understand how jiha¯d as obligatory aggressive war came to be the prevalent opinion in the second half of the second/eighth century. ‘Abd al-Razza¯q mentions a group of highly respected H . ija¯zı¯ jurists in the circle of Ibn Jurayj (d. 150/762) who rejected the idea that the jiha¯d was obligatory for all; and they seem, moreover, to have given primacy to other religious acts.13 Yet the Syrian jurists quoted by ‘Abd al-Razza¯q, perhaps reflecting the determination to make progress on the Byzantine frontier in the first half of the second Islamic century, were quite naturally attracted to the idea that aggressive war was obligatory. So in the Mus.annaf we see that in Syrian circles pious stories circulated about the importance of being a frontier warrior in Syria, and of warfare by sea as well as by land.14 A jurist of this period, the above-mentioned ‘Abd Alla¯h b. al-Muba¯rak, was alarmed by what he considered to be the bad conditions of the Muslims on the Syrian frontier and its critical points (thughu¯r). He wrote, as already mentioned, a book on jiha¯d, and he 10

Ibid., 8:496. Al-Dhahabı¯, Siyar A‘la¯m al-Nubala¯’ (Beirut, 1990), 8:421–22. 12 Al-Mudawwanah (Cairo, 1326), 3:5. Michael Bonner suggests that the polemical tone in some verses ascribed to Fud. ail b. ‘Iya¯d. reflects their disagreement over this matter; see Ibn Taghribirdı¯, Al-Nuju¯m al-Z . a¯hirah (Cairo, n.d.), 2:103–4. 13 ‘Abd al-Razza¯q, Al-Mus.annaf (Beirut, 1983), 5:171–72. It might be argued that the “pietist” jurists are merely supporting the later normative view that jiha¯d is an obligation of the community as a whole and not of individuals; but their tone and emphasis on other religious duties seem to make a further point. It should not, however, be imagined that they in any way whatsoever opposed all jiha¯d. 14 Ibid., 5:172–73. 11

Roy Parviz Mottahedeh and Ridwan al-Sayyid [ 27 ] also devoted a section of his still extant book on asceticism to the jiha¯d, which he considered an ascetic practice.15 He was a Khurasanian from Marw who came to Syria to study with the “jurist of Syria” par excellence, as al-Awza¯‘ı¯ was called; and he supported alAwza¯‘ı¯ in his dispute with the jurists of the H . ija¯z about the virtues of jiha¯d. It should be remembered that, because of the troubled internal state of the Islamic empire at the end of the Umayyad period, the central government of the caliphate neglected the frontier. It is in this context that al-Awza¯‘ı¯ and ‘Abd Alla¯h b. al-Muba¯rak wrote. Several of the key jurists of the next generation, especially al-Sha¯fi‘ı¯ (d. 204/820), wrote on this subject and proved to be in most respects followers of al-Awza¯‘ı¯ and ‘Abd Alla¯h b. al-Muba¯rak. Hence the Syrian doctrine of jiha¯d was transformed into the normative doctrine of the majority of Iraqis. The more general acceptance of the Syrian school reached its peak in the thought of al-Sha¯fi‘ı¯, who elevates the destruction of unbelief to be the primary justification for jiha¯d.16 Nevertheless, as we have seen, the opinions of the pre-Sha¯fi‘ı¯ jurists continued; at least up to the seventies of the second Islamic century, it is clear that some of the jurists did not see jiha¯d as an individual or communal duty. This essay considers jiha¯d and its justifications among early jurists. We have tried to . ija¯z, more show that there is evidence that a belief in the defensive jiha¯d existed in the H especially in Medina, where at least an earlier generation had reason to resent the removal of the capital from that city. Furthermore, even if they accepted an offensive jiha¯d, they had objections to participating in it if it were led by the Syrian Umayyads, who were illegitimate leaders of the Muslims in the eyes of many H . ija¯zı¯ jurists. Varieties of this earlier H . ija¯zı¯ attitude continue to reappear in Islamic thought when circumstances favored it. For example, Rashı¯d al-Din Fad. l Alla¯h, the famous early fourteenth-century vizier to the Mongolian Ilkhanid government of the Middle East, argues that the socalled verse of the sword (Qur’a¯n IX:29) is specifically directed toward Arab polytheists,17 a view in accord with much contemporary scholarship; for, as Tyan says, “according to a view held by modern orientalist scholarship, Muhammad’s conception of the jiha¯d as attack applied only in relation to the peoples of Arabia.”18 Yet this legal context must be related to wider contexts that have been suggested by more recent studies of jiha¯d. First, it is now clear that in their earliest examples the siyars, or books on the conduct of state, and magha¯zı¯, or literature on exemplary early campaigns, were mixed. Only in their later examples, when law and history became more distinct disciplines, did siyar and magha¯zı¯ become distinct genres.19 Furthermore, in the process of the development of law, although there was probably a great deal of common 15

Kita¯b al-Zuhd wa-l-Raqa¯’iq (Malegaon, 1966). M. Khadduri, The Islamic Law of Nations: Shaybani’s Siyar (Baltimore, 1966), “Translator’s Introduction,” 58. While Khadduri’s contention that Sha¯fi‘ı¯ was the “first” to formulate this doctrine that “the jiha¯d had for its intent the waging of war on unbelievers for their disbelief” is open to doubt, Sha¯fi‘ı¯’s stamp of approval for this doctrine seems to have made this additional motivation for jiha¯d much more commonly accepted. 17 D. Krawulsky, “Fı¯ al-H . arakiyah al-Ta’rı¯khı¯yah al-I¯dı¯u¯lu¯jı¯yah li al-Jiha¯d fi al-Isla¯m,” Al-Ijtiha¯d 12 (1991): 127–31. 18 Tyan, “Djiha¯d,” 538. 19 M. Bonner, “Some Observations Concerning the Early Development of Jihad on the Arab-Byzantine Frontier,” Studia Islamica 75 (1992): 5–31. 16

[ 28 ] The Idea of the Jiha¯d in Islam before the Crusades ground in the very earliest period because of the close interaction of the original Muslim community, local schools developed, and the differences of at least some members of the . ija¯zı¯ school with Syrian schools become apparent in this study. Nevertheless, in early H this early period, as in later periods, the schools interact; and when a normative view was formed in Iraq, “minority” opinions continued to be transmitted by the tradition. Second, the gradual emergence of normative jiha¯d theory must also be seen as a function of the adjustment of the early Islamic world to an apocalypse that never conclusively happened. Roman and Sassanian traditions of war already had established the idea of victory as divine confirmation; and, given the apocalyptic atmosphere that pervades much of the Qur’a¯n, the early Islamic conquests seemed confirmation that Islam was destined to create a universal state. Yet Constantinople and a significant part of the Byzantine Empire remained unconquerable. Muslim jurists became more interested in fiscal problems—in particular the status of land—as determined by the earlier conquests. In this way the genres on “land tax” and “state finances” came into existence and were more cultivated than was the literature on campaigns, which passed into biographies and histories, while the genre on “the conduct of state” became a relatively (though not uniformly) static section of general law books. We believe, though it cannot be demonstrated here, that the transition to a formal legal theory of war changed jiha¯d from a theory primarily based on historical memories of the battles fought in the time of the Prophet and the early Islamic period to a more precisely defined and normative theory rooted in very specific events in the life of the Prophet and very specific interpretation of Qur’a¯nic verses. In the theory based on Qur’a¯nic verses, an attempt was made to organize the relevant verses in chronological order so that the so-called verse (or verses, according to some) of the sword, which made war perpetual and a permanent obligation of the Islamic community, came last and therefore abrogated verses that could clearly have allowed a different development of the law. Incidentally, T . abarı¯, who opens his work on The Divergences of the Jurists with a careful list of which Qur’a¯nic verses supersede which others, when he comes to the so-called verse of the sword in this Qur’a¯nic commentary gives no indication that it supersedes other verses.20 Third, as a state of relative equilibrium is established on the Byzantine frontier, the concept of the “realm of Islam” and the “realm of war” comes into being, which recognizes the temporary failure of the Islamic conquests to become universal. The division is found as early as the first half of the second/eighth century in Muhammad al-Nafs azZakı¯yah.21 The division appears by its nature to be more an expression of something that jurists had to deal with after it had occurred and not as an expression of what they thought should happen.22 By the long reign of the Umayyad Caliph Hisha¯m (105/724– 125/743), Muslim armies suffered setbacks in Western Europe and Central Asia as well 20 T . abarı¯, Ikhtila¯f al-Fuqaha¯’ (Cairo, 1933), 1–21, and Ja¯mi‘ al-Baya¯n ‘an Ta’wil al-Qur’a¯n (Cairo, 1968), 3:109–10. 21 See Ridwan al-Sayyid’s forthcoming publication of his works. 22 F. M. Donner, in his fundamentally important and innovative essay on the jiha¯d (“Sources of Islamic Conceptions of War,” in Just War and Jihad, ed. J. M. Kelsay and J. T. Johnson [New York, 1991], 50 and n. 88, p. 67), dates the appearance of the two abodes to the late 2nd century ../late 8th century .., a point that he promises to develop in a further publication.

Roy Parviz Mottahedeh and Ridwan al-Sayyid [ 29 ] as on the Byzantine frontier. At the same time, the relative stabilization of the frontier led to a truce and arrangements for furtherance of trade. By the time of al-Sha¯fi‘ı¯ a juridical theory of a third abode, “the realm of treaty relations,” had emerged.23 These juridical developments had become necessary to deal with a new situation and seem to reflect rather than precede the appearance of this situation. Fourth, jiha¯d had become an element of formalized piety, in part to keep alive the momentum lost by the caliphate as an instrument of conquest and in part as an attempt to “spiritualize” a deferred apocalyptic event. The Syrian tradition of al-Awza¯‘ı¯ had seen a continuous line of leaders from the Prophet through the Umayyads whose rule was justified by their active expansion of the realm of Islam, thereby fulfilling injunctions and prophecies of the earliest Islamic period. These leaders represent a continuous tradition of the uninterrupted practice of the Muslims, which in one view should be a major basis for law. But the apocalypse had not come. In contrast to al-Awza¯‘ı¯, his equally belligerent Khurasanian contemporary ‘Abd Alla¯h b. al-Muba¯rak emphasized the spiritual discipline and merits of jiha¯d. Quotations directly from the Prophet take pride of place in his work, and his interest in asceticism (on which he wrote a separate book) is never far away. No wonder that, during the Second Crusade, Ibn ‘Asa¯kir gave public readings of ‘Abd Alla¯h b. al-Muba¯rak’s Kita¯b al-Jiha¯d in Damascus, which inspired Abu¯ al-H . asan b. Munqidh, brother of the famous Usa¯mah, to volunteer to help raise the siege of Ascalon.24 While ‘Abd Alla¯h b. al-Muba¯rak tried to integrate the role of government, society, and the individual in jiha¯d, the subsequent propagation of a popular jiha¯d literature created a strangely balanced problem for later Islamic leaders. The majority of jurists regarded jiha¯d as lawful only if led by a legitimate Muslim ruler,25 in particular a legitimate caliph/ imam, on whose exact identity fewer and fewer of the jurists were willing to pronounce in any decisive way. Yet many military leaders exploited the image of jiha¯d in popular piety by saying that they owed their legitimacy, at least in part, to their successful pursuit of jiha¯d. At some point such a leader, if successful enough, might try to imply that he was sanctioned by some caliph or even, either obliquely or directly, to imply that he was a caliph. In this way he might win the overt support of some of the jurists. Jiha¯d had become an unpredictable variable in the internal politics of Islamic lands just as it had become in their relations with non-Islamic lands. Harvard University and Lebanese National University

23

H. Inalcı¯k, “Dar al-‘Ahd,” EI 2, 2:116. E. Sivan, L’Islam et la Croisade (Paris, 1968), 75, cited in Bonner, “Some Observations,” 20. 25 Donner, “Islamic Conceptions,” 41. Other important recent literature on jiha¯d in European languages includes P. Crone, “The First-Century Concept of Higra,” Arabica 41 (1994): 352–87; K. Yahya Blankinship, The End of the Jihad State: The Reign of Hisham ibn Abd al-Malik and the Collapse of the Umayyads (Albany, 1994); M. Bonner, Aristocratic Violence and Holy War: Studies in the Jihad and the Arab Byzantine Frontier (New Haven, Conn., 1996); A. Morabia, Le Gihad dans l’Islam me´die´val: Le “combat sacre´” des origines au XIIe sie`cle (Paris, 1993); A. Noth, Heiliger Krieg und heiliger Kampf in Islam und Christentum (Bonn, 1966); and a forthcoming book by Reuven Firestone. 24

Defenders of the Christian People: Holy War in Byzantium George T. Dennis For most civilized people the term holy war is a contradiction in terms. What religious motive could possibly transform the widespread destruction and the slaughter of thousands of human beings into a holy and meritorious act? But, as we know, religion has all too often served as a pretext for violence. Before going any further, however, we should agree upon a definition of holy war. Three criteria, I think, are essential. A holy war has to be declared by a competent religious authority, the obvious examples being a Christian pope or a Muslim caliph. The objective must be religious; again, two obvious examples are the protection or recovery of sacred shrines or the forced conversion or subjection of others to your religion. There could, of course, be other goals. Finally, those who participate in the holy war are to be promised a spiritual reward, such as remission of their sins or assurance of a place in paradise.1 In the world around the Mediterranean, two forms of holy war did emerge. First, the Muslim jiha¯d. Much has been written about this, and I wish only to point out its salient features.2 Jiha¯d is a religious duty for the Muslim community to propagate Islam, employing coercion of various sorts as needed, until the whole world professes Islam or is subject to its laws. At times, especially when the caliph, or other religious authority, proclaims it, this obligation takes the form of armed conflict. Those who die in the struggle are acclaimed as martyrs and are believed to go straight to paradise. The doctrine of jiha¯d may be traced to the earliest days of Islam, although maybe not directly to Muhammad himself. The jiha¯d did not become one of the five “pillars” of Islam, but it was kept alive by preaching and the attractiveness of the ideal of martyrdom and paradise and the more tangible rewards of booty and plunder. In essence, it was aggressive and bent on conquest. Of course, not every war waged by Muslim powers, including those against 1 See M. Canard, “La guerre sainte dans le monde islamique et dans le monde chre´tien,” RAfr 79 (1936): 605–23, repr. in Byzance et les musulmans du Proche Orient (London, 1973), no. ; V. Laurent, “L’ide´e de guerre sainte et la tradition byzantine,” RHSEE 23 (1946): 71–98; N. Oikonomides, “The Concept of ‘Holy War’ and Two Tenth-Century Byzantine Ivories,” in Peace and War in Byzantium: Essays in Honor of George T. Dennis, S.J., ed. T. Miller and J. Nesbitt (Washington, D.C., 1995), 62–86; T. P. Murphy, ed., The Holy War (Columbus, Ohio, 1976). 2 See Canard, “Guerre sainte”; E. Tyan, “Djiha¯d,” EI 2 (Leiden, 1961), 2:551b–553a; J. Kelsay and J. T. Johnson, Just War and Jihad: Historical and Theoretical Perspectives on War and Peace in Western and Islamic Traditions (New York, 1991).

[ 32 ] Defenders of the Christian People nonbelievers, was a holy war. Many were simply tribal, ethnic, or even national conflicts whose roots often went back to pre-Islamic times. In Western Europe the idea of a holy war developed later and for different reasons. So much has been written about this that there is no need to enter into detail.3 First, we must remember that what we call a crusade was, especially during the first century or so, a pilgrimage, and those who took part in it were pilgrims; it was a holy journey (iter, passagium), not a holy war. It was regarded primarily as defensive, that is, armed escorts were to protect pilgrims on their way to the sacred shrines of Christendom and were to recover or defend the holy sites in Palestine. This defensive character differentiated it from jiha¯d, as did the fact that it did not advocate the forceful imposition of Christianity upon others. In subsequent centuries, admittedly, and for some participants it did take on a more belligerent character. One need only recall the so-called Albigensian crusades or the one that sacked Constantinople in 1204. Still, the notion of using force to convert the infidel was, with few exceptions, foreign to Christianity, East and West. But the Crusades were proclaimed by the highest religious authority in the West, the pope; they were directed toward a religious end, the protection of fellow Christians in the East and the recovery and defense of the holy places; and those who took part were promised religious rewards, particularly the remission of sin. For the Byzantines, it must be said at the outset, both ideas and forms of holy war— jiha¯d and crusade—were abhorrent.4 They absolutely rejected both. First, the jiha¯d. They did not understand it. What motivated the armies of Islam, as the Byzantines saw it, was the hope of booty and a barbaric love of fighting. According to Leo VI, “The Saracens do not campaign out of a sense of military service and discipline, but rather out of a love of gain and license or, more exactly, in order to plunder on behalf of their faith.” 5 Leo dismisses them as “barbarians and infidels” concerned only with plunder.6 Immense multitudes of them come from Syria and Palestine, “oblivious to the dangers of war, intent only on looting.” 7 Byzantine authors, from the seventh to the fourteenth century, 3 See J. Riley-Smith, The First Crusade and the Idea of Crusading (London, 1993), and, in general, S. Runciman, A History of the Crusades, 3 vols. (Cambridge, 1951–54); K. M. Setton, A History of the Crusades, 2d ed., 6 vols. (Madison, Wisc., 1969–89); A. S. Atiya, The Crusade: Historiography and Bibliography (Bloomington, 1962); H. E. Mayer, Bibliographie zur Geschichte der Kreuzzu¨ge (Hannover, 1960); this comprises 5,362 titles, and the number of works on the Crusades has surely doubled since then. For continuing study of the Crusades, consult the annual Society for the Study of the Crusades and the Latin East: Bulletin (1981–97). 4 Canard, “Guerre sainte”; Laurent, “L’ide´e de guerre sainte”; A. Laiou, “On Just War in Byzantium,” in To Hellenikon: Studies in Honor of Speros Vryonis Jr., ed. S. Reinert et al. (New Rochelle, N.Y., 1993), 1:156–77; G. Dagron, “Byzance et le mode`le islamique au Xe sie`cle a` propos des ‘Constitutions tactiques’ de l’empereur Le´on VI,” CRAI (Paris, 1983): 219–43. Byzantine rhetoric about holy war, though, has led some modern scholars to refer to the luckless campaign of Manuel I against the Turks in 1176 as a sort of crusade: R.-J. Lilie, Byzantium and the Crusader States, 1096– 1204, trans. J. C. Morris and J. E. Ridings (Oxford, 1993), 211–14; P. Magdalino, The Empire of Manuel I Komnenos, 1143–1180 (Cambridge, 1993), 95–98. 5 Leonis VI Tacticae constitutiones 18.24, PG 107:952 (hereafter Taktika). Book 18 is also edited by R. Va´ri, “Bo¨lcs Leo Hadi Taktika´janak XVIII Fejezete,” in G. Pauler and S. Szila´gyi, A Magyar Honfoglala´s Kutfo¨i (Budapest, 1900), 11–89. 6 Taktika 18.128; PG 107:976. 7 Taktika 18.132; PG 107:977.

George T. Dennis [ 33 ] repeat these accusations, as they profess their utter repugnance for the doctrine of jiha¯d. In their polemics against Islam they vehemently criticize the jiha¯d as little more than a license for unjustified murder and a pretext for pillaging.8 And, while the Byzantines, when the opportunity arose, may have indulged in their share of massacre and looting, they did not excuse it in the name of religion. As far as the Crusades are concerned, it suffices to listen to Anna Komnene, who abhorred both the movement and many of its participants.9 Still, some Byzantines welcomed the Westerners at first. They were, after all, fellow Christians, although perhaps somewhat careless in their teachings and practices. Emperor Alexios treated them in a civil, almost cordial manner, although he was always nervous about what they might do, and he provided them with military assistance through Asia Minor. But, in general, the Byzantines never seemed to understand why all those Western knights and their followers were marching through their land. Restoring Jerusalem to Christian rule was perhaps a laudable objective, but was it worth such an immense effort, fraught with so many perils and uncertainties and carried out with such brutality? Constantinople, after all, was the New Jerusalem, the true holy city. The Byzantines, always practical, were far more interested in possessing Antioch because of its important strategic position than in holding Jerusalem with all its sentimental value. Pilgrimage they understood and warfare they understood, but the conjoining of the two they did not understand. They would have been utterly appalled at the preaching of St. Bernard and his call for the extermination of the infidel (delenda penitus), as well as his assertion that killing an enemy of Christ was not homicide, but malecide.10 And what would they have thought of the rule he drew up for the Templars, monks who wielded lethal weapons in battle?11 The Byzantines soon came to believe that the warriors from the West had nothing less in mind than the conquest of the empire, and the events of 1204 proved they were right. Ultimately, they came to hate the Latins as much or even more than the Muslims. If the Latins ever referred to their eastern expeditions as “holy war,” that term, it is clear, would not have been appreciated by the Byzantines. Now, to the main point. I have already indicated that the Byzantines did not have any concept of a true holy war, although this will be qualified below. Byzantine writers did use the term holy war (hieros polemos), but only in reference to one of the three “sacred wars” waged over the possession of the oracle of Apollo at Delphi; these occurred in 590, 449, 355–347, all .. Most Byzantine references, such as the Souda (I.191), allude 8 A. T. Khoury, Pole´mique byzantine contre l’Islam, VIII–XIII s. (Leiden, 1972), 243–59; W. Eichner, “Die Nachrichten u¨ber den Islam bei den Byzantinern,” Der Islam 2 (1936): 131–62, 197–244. 9 Anne Comne`ne, Alexiade, ed. and trans. B. Leib, 3 vols. (Paris, 1937–43), book 10, 5–11: vol. 2:205–36. 10 De laude novae militiae, in S. Bernardi opera, vol. 3, ed. J. Leclercq and H. M. Rochais (Rome, 1963), 204– 39, esp. chap. 3, p. 217; epistola 457, opera, vol. 8 (Rome, 1977), p. 433; et al. 11 See De laude, passim; Alexiade 10.8.8; vol. 2:218. Constantine Stilbes strongly criticized the Latin clergy for engaging in combat and killing the enemy, including other Christians, and for teaching that those who died in war went directly to heaven: J. Darrouze`s, “Le me´moire de Constantin Stilbe`s contre les Latins,” REB 21 (1963): 50–100, esp. 69–77. In 1250 Emperor John Vatatzes told Frederick II that it was scandalous for priests to carry weapons and fight in battle: F. Miklosich and J. Mu¨ller, Acta et Diplomata graeca medii aevi sacra et profana, 6 vols. (Vienna, 1860–90), 3:72–73, no. 18.

[ 34 ] Defenders of the Christian People to the second one, apparently following Thucydides (1.112) and Aristophanes (Aves 556). The term holy war is used, as far as I can determine, by ancient and Byzantine writers only in connection with those wars. In one sense, however, all Byzantine wars were holy because the emperor was holy, and it was by his authority and sometimes under his leadership that wars were waged. They were declared by the emperor and fought on behalf of the empire. They were imperial wars, fully in the Roman tradition. Their essential character did not change because the legions now entered battle under the sign of the cross. Their prayers for God’s blessing and other religious practices did not make their wars specifically holy or religious, as has sometimes been maintained.12 From time immemorial, religion has played a role in warfare. One people offers sacrifice to its gods before going into battle and, upon emerging victorious, will topple the statues of the other people’s gods and set up its own. Are these religious wars, or are they simply tribal conflicts motivated by revenge, plunder, or the acquisition of land or slaves? The invocation of deities is basically an additional means of assuring victory, of enlisting the aid of powerful allies and shifting the balance in your favor. Consider the Trojan War. Not only were gods and goddesses called upon with prayer and sacrifice, but they participated directly in the fighting. Yet nobody calls the Trojan War a holy war. Consider, too, those conflicts that have often been cited as precedents and inspirational models for Christian holy wars, I mean those waged by the people of Israel, as related in the books of Joshua, Judges, Kings, and elsewhere. Do they really qualify as religious wars? Were they not primarily armed conflicts between seminomadic tribes struggling to acquire land? Their god may grant them victory or deny it, but, in the final analysis, the fundamental motivation and objective of most of those wars were not primarily religious, those of the Maccabees perhaps being an exception. How many wars, then, waged later by Christians and Muslims were truly religious wars, not to mention holy wars? Were they not, to a large extent, tribal or feudal conflicts with a lot of religious trappings? In trying to categorize a conflict as religious or holy, we might ask: Are they fighting this war primarily for religious reasons? If little or no religious motivation were present, would they still be fighting? The Crusaders provide a good example. Nobody in his right mind, even in the Middle Ages, would leave the comforts of home, pack up all his belongings, and march off for two thousand kilometers, endure incredible hardships, and face the very real threat of death unless he were religiously motivated. While there were some, like Bohemond, who may have had less lofty motives, the majority of the Crusaders gained no strategic, economic, or political advantage, especially during the first hundred years. They marched off to the East for what they regarded as a religious act, if not a duty. For them, this was surely a holy war. On the other hand, the long campaigns of Herakleios against the Persians, sometimes depicted as a prototypical crusade, abounded in religious elements.13 The Persians had 12 See the detailed study by A. Kolia-Dermitzakes, Ho Byzantinos “hieros polemos” (Athens, 1991); also the review by W. Kaegi, Speculum 69 (1994): 518–20. 13 William of Tyre begins his account of the Crusades with the reign of Herakleios: Willelmi Tyrensis Chronicon, CC continuatio medievalis 63–63, ed. R. Huygens (Turnhout, 1986), 1.1:105; trans. E. A. Babcock and A. C. Krey, A History of Deeds Done beyond the Sea, by William Archbishop of Tyre, 2 vols. (New York, 1943), 1:60.

George T. Dennis [ 35 ] destroyed churches, massacred Christians, and taken away the holy cross from Jerusalem; they must be punished and the cross restored. The patriarch prayed for victory and blessed the troops as they marched out under the standard of the cross. Religion played a major role throughout the conflict. But, even if these religious motivations had not been present or had not been so prominent, Herakleios would almost certainly have still gone to war. His wars were waged as much for strategic advantage and territory as for religion. The wars of Herakleios were but one phase of the geopolitical conflict between the Romans and the Persians that had been going on for six hundred years. These were imperial wars, not holy wars. Although religious rhetoric and ritual were prominent and pervasive, subsequent Byzantine wars, those of Nikephoros Phokas in the tenth century, for example, or those of the Komnenian emperors in the twelfth, were first and foremost imperial wars. That their objectives sometimes coincided with religious ones did not alter that basic characteristic. Finally, it should be noted that the same religious practices were observed by the Byzantine armed forces whether they were facing a non-Christian or a Christian enemy. War cries, such as “God help the Romans,” “The Cross is victorious,” do not transform the nature of a particular war. Religious shouts and symbols are used to instill confidence in the individual soldier and to raise the morale of the army. Religious services, especially the eucharistic liturgy, are meant to comfort the soldier and to prepare him to risk his life.14 Chaplains still conduct religious services for modern armies, but that does not sanctify their conflicts. Athletes often join in prayer before a game, but we do not talk of a holy football game or a holy soccer match. The church certainly prayed for victory, but it rejected the request of Nikephoros Phokas to have fallen soldiers honored as martyrs.15 The cross was displayed on the standards, or used in place of a standard, to remind the troops of God’s protection and that they were fighting for a Christian nation.16 Through the centuries, the cross, it may be noted, has been depicted on many banners in wars that have been far from holy. The cross displayed on the flags of several modern nations does not tell us anything about the religious sensibilities of its citizens; Great Britain has three crosses on its flag. The Byzantine attitude toward war can best be understood in the context of the way in which they viewed the world and life in general. This world and the life it bore were fragile and transitory. The only permanent reality was to be found in another world, the kingdom of heaven. The empire on earth was a mere reflection of that in heaven, and 14 See G. Dennis, “Religious Services in the Byzantine Army,” Eulogema: Studies in Honor of Robert Taft S.J., Studia Anselmiana 110 (Rome, 1993): 107–17. 15 Ioannis Scylitzae Synopsis historiarum, ed. I. Thurn (Berlin, 1973), 274.62–67; see P. Viscuso, “Christian Participation in Warfare: A Byzantine View,” in Peace and War in Byzantium (as in note 1), 33–40. Some soldiers were honored as martyrs, such as the Forty-two of Amorion, but that was because they chose to die rather than deny their faith. Three liturgical offices (akolouthiai) that have come down to us do not provide evidence for a Byzantine holy war; rather, they are prayers that God may look kindly on the faithful soldiers who have died in war, that he may forgive their sins and receive them into Paradise: L. Petit, “Office ine´dit en l’honneur de Nice´phore Phocas,” BZ 13 (1904): 398–419; A. Pertusi, “Una acolouthia militare inedita del X secolo,” Aevum 22 (1948): 145–68; T. De´torakes and J. Mossay, “Un office byzantin ine´dit pour ceux qui sont morts a` la guerre, dans le Cod. Sin. Gr. 734–735,” Le Muse´on 101 (1988): 183–211. 16 See G. Dennis, “Byzantine Battle Flags,” ByzF 8 (1982): 51–63.

[ 36 ] Defenders of the Christian People the emperor was called to imitate the Lord of heaven. Under God, he was to assure the well-being of his subjects and protect them from all dangers, within and without. The church had a different role. Jesus had told his followers that he could call upon legions of angels to save himself from death,17 but he did not do so, and neither would his church. Unlike its Latin sister, the Byzantine church left the call to arms and the waging of war, even against the most pernicious and destructive heretics and infidels, to the imperial government. But it took the lead in another kind of struggle, one for the souls of the faithful, a struggle not against human enemies but against cosmic powers and superhuman forces of evil.18 For Byzantine Christians this was a form of warfare that could be called holy, although I have not found explicit use of that term. The concept of the Christian being involved in a war against the forces of evil goes back, of course, to St. Paul, if not before.19 While every Christian had to withstand the onslaughts of the devil, the monks were the frontline troops in the war against the legions of Satan. Night and day, according to Gregory of Nazianzos, the monk must fight the spiritual war (pneumatikos polemos).20 Chrysostom tells his audience that the war against demons is difficult and never ending.21 Spiritual combat is a regular theme in the vitae of the saints.22 Demons in a variety of shapes, from hyenas to dragons, viciously attacked saints Theodore of Edessa, Gregory of Dekapolis, Joseph the Hymnographer, John Psychaites, Isidore, abbess Sarah, and many others.23 Story after story is told of their incessant struggles against the forces of sin and darkness. The demons, for their part, took warfare seriously. They appear in full battle array, in phalanxes of cavalry and infantry that wheeled about in formation. They wore iron breastplates and carried bows and arrows and other missiles.24 They began their advance against St. Ioannikios in proper order, although making a tremendous racket; they drew up in formation, shouted their war cry, and shot a steady stream of arrows at him. All of this he repelled by the sign of the cross. Under their commander (strategos) Satan, the demons arrayed themselves in their phalanxes in a proper battle line ( parataxis), just as the armed forces of the emperor do, and charged against Constantine the Jew.25 As the military manuals prescribe, they feigned retreat, shouted insults from afar, regrouped, 17

Matt. 26:53. Eph. 6:12. 19 E.g., Rom. 7:23; Eph. 6:16–20; 1 Thess. 5:6–8; 1 Tim. 6:12; 2 Tim. 2:4. 20 Oratio 2, 91; PG 35:495. 21 In s. Eustathium, PG 50:599. 22 See P. Bourguignon and P. Wenner, “Combat spirituel,” DSp; T. Sˇpidli´k, Spirituality of the Christian East (Kalamazoo, Mich., 1986), 233–66. 23 F. Dvornik, Vie de s. Gre´goire le De´capolite et les Slaves mace´doniens au IXe sie`cle (Paris, 1926), 47, 31; cf. Vita of Joseph the Hymnographer by Theophanes, ed. A. Papadopoulos-Kerameus, Sbornik grecˇeskikh i latinskikh pamjatnikov ˇ itie ize vo sv. otca nasˇego Feodora arkhiepiskopa kasajusˇˇcikhsja Fotija patriarkha, vol. 2 (St. Petersburg, 1901), 41; Z Edesskogo, ed. I. Pomjalovskij (St. Petersburg, 1892), 67, 1–31; P. Van den Ven, “Vie de s. Jean le Psichaı¨te,” Le Muse´on 21, n.s., 3 (1909): 103; (Isidore) Apophthegmata Patrum, PG 65:97; (Sarah) ibid., 229. Research in this area was greatly facilitated by the Dumbarton Oaks Hagiographical Database; for her assistance in its use the author is especially grateful to Dr. Stamatina McGrath. 24 AASS, Nov. 2.1:395c–396a. 25 AASS, Nov. 4:640. 18

George T. Dennis [ 37 ] and attacked again. The saint beat them off with a wooden cross made on the spot, but the effort left him exhausted. A monk in Skete heard a battle trumpet sound as the demons prepared to attack him and force him to quit his prayers.26 To confront such adversaries, the monk had to be a soldier. Symeon reminds his monks that they have been called to fight against invisible foes. They have enlisted and taken their place in the ranks of Christ’s soldiers.27 The monks did not wait to be attacked; they did not simply hold the fort, but took the war into the devil’s territory and fought him on his own turf, in the desert and in other wild, abandoned locations. Many made a point of settling in the desert where the demons lived.28 Daniel the Stylite learned that demons were hiding in an old church. He immediately went in to fight them “as a brave soldier strips himself for battle against a host of barbarians,” holding the invincible weapon of the cross.29 What, then, about the visible, tangible wars waged by the Byzantines with armor and weapons made of solid iron and steel, and against other human foes? No Byzantine treatise on the ideology of war, whether a holy or a just war, has come down to us, and it is unlikely that any was ever written. One must glean what one can from the military manuals and the histories. Although there were occasional rhetorical flourishes in admiration of valor and bravery on the field of battle, and although they were dependent on military means for their survival, the Byzantines, in the words of a retired combat engineer in the sixth century, regarded war “as a great evil and the worst of all evils.”30 “We must always prefer peace above all else,” wrote Leo VI, “and refrain from war.”31 For them war was not the “politics by other means” of Clausewitz, but was the last resort. The threat of overwhelming force was preferable to the actual use of such force, and in this, it may be noted, they displayed a striking continuity with the ancient Romans. They sought to obtain their objectives by diplomacy, bribery, covert action, paying tribute, or hiring other tribes to do the fighting. Only when all else had failed were they to take up arms. And even then they tried to avoid a frontal assault and concentrated on wearing out the foe by light skirmishing, clever strategy, and adroit maneuvering. They were reluctant to wage war on both moral and practical grounds. Killing, even when deemed justifiable, was evil—one need only recall the famous, if rarely observed, canon of St. Basil which declared that soldiers who had killed in battle were to be refused communion for three years.32 On the practical side, war was both hazardous and expensive. All this is consistent with the remarkable centrality of defense in Byzantine strategic theory and practice. One American military scholar wrote of a sixth-century tactician: Pratum spirituale, PG 87:3017; M. J. Rouet de Journel, Le Pre´ spirituel, SC 12 (Paris, 1946), 152, p. 204. Syme´on le nouveau the´ologien, Cate´cheses, ed. B. Krivocheine, SC 96.1 (Paris, 1963), 3, 129–34, p. 290. Stratiotes Christou and the Latin miles Christi are very commonly used to designate a monk, but they can also be used for professional soldiers: see, e.g., Kolia-Dermitzakes, Hieros polemos, 257. 28 See, e.g., Evagrius, Praktikos, Traite´ pratique ou le moine, ed. A. and C. Guillaumont, SC 171 (Paris, 1971), 505; A.-J. Festugie`re, Les moines d’Orient, 4 vols. (Paris, 1961–65), 2:101 ff. 29 H. Delehaye, Les saints stylites (Brussels, 1923), chap. 15, p. 15. 30 G. Dennis, Three Byzantine Military Treatises, CFHB 25 (Washington, D.C., 1985), 20–21. 31 Taktika 2, 45; ed. R. Va´ri, Leonis imperatoris Taktika, 2 vols. (Budapest, 1917–22), libri –, 43 (hereafter Va´ri); entire work in PG 107:669–1120; Va´ri, 1:40; PG 107:696. 32 Saint Basile, Lettres, ed. Y. Courtonne, vol. 2 (Paris, 1961), ep. 188, 13, p. 130. 26 27

[ 38 ] Defenders of the Christian People “He has a distinctly defensive mind, and sees so clearly what the enemy may do to him that he has no time to think of what he may do to the enemy.”33 The Byzantines were not a warlike people and, in fact, this led the Crusaders to accuse them of cowardice. Their entire attitude toward war was colored by their emphasis on defense and, in this respect, certainly differed from the crusade and the jiha¯d, both of which were aggressive by nature. Even the offensive campaigns into enemy territory of Herakleios, Nikephoros Phokas, John Tzimiskes, and Basil II were aimed at recovering and protecting regions that rightfully belonged to the Roman Empire. In the Byzantine world, war was not, as sometimes in the West, a lethal playing field on which so-called noblemen displayed their prowess and sought glory. In itself, war was not a good or meritorious act, and it was certainly not “holy.” How, then, did they justify war? “The purpose of all wars is peace.” So wrote Aristotle long ago, and in the eleventh century Anna Komnene quoted him in explaining why her father Alexios had to devote so much time and energy to warfare.34 She also makes it clear that, as with an individual, so a nation was entitled to use force in defending itself. Alexios was also, in her mind, justified in taking military action to recover lost territory, to force compliance with a sworn treaty, or to avert a greater evil.35 Other writers, when they do advert to the causes of war, seek to justify it much as Anna. Perhaps the clearest and most deliberate explanation of the Byzantine view of war is that put forth by Leo VI in the beginning of his Tactical Constitutions, very early in the tenth century. While the emperor’s highest priority was to see to the peace and prosperity of his subjects, he realizes that, to assure this, he must maintain the armed forces in good order and promote the study of tactics and strategy. Why must war take up so much of the emperor’s energies? “Out of reverence for the image and the word of God, all men ought to have embraced peace and fostered love for one another instead of taking up murderous weapons in their hands to be used against their own people. But since the devil, the original killer of men, the enemy of our race, has made use of sin to bring men around to waging war, contrary to their basic nature, it is absolutely necessary for men to wage war in return against those whom the devil maneuvers and to take their stand with unflinching resolve against nations who want war.” Eventually, he hopes, “peace will be observed by all and become a way of life.”36 The Byzantines were not to wage war against other peoples, Leo wrote, unless those others should initiate hostilities and invade our territory. “Then,” he addressed the commander, “you do indeed have a just cause, inasmuch as the enemy has started an unjust war. With confidence and enthusiasm take up arms against them. It is they who have provided the cause and who have unjustly raised their hands against those subject to us. Take courage then. You will have the God of justice on your side. Taking up the struggle on behalf of your brothers, you and your whole force will be victorious. . . . Always make sure that the causes of war are just.”37 33

Dennis, Three Treatises, 83 n. 1. Alexiade 12.5.4; vol. 3:68. The reference is to Aristotle, Politics 7.13.8. 35 Cf. Laiou, “Just War,” 156–65. 36 Taktika, prooemium, 3; Va´ri, 1:4; PG 107:673. 37 Taktika 2.46; Va´ri, 1:40; PG 107:696. 34

George T. Dennis [ 39 ] The Byzantine wars were not “holy” wars, but just wars, imperial wars. They were waged to defend the empire or to recover land that rightfully belonged to it. The soldiers put their lives on the line for the emperor and the people subject to him, the Christian people. They were to “struggle on behalf of relatives, friends, fatherland, and the entire Christian people.”38 Toward the end of the tenth century another military author spoke up on behalf of the men on the eastern frontier who “choose to brave dangers on behalf of our holy emperors and all the Christian people. They are the defenders and, after God, the saviors of the Christians.”39 In conclusion, then, Muslims believed force might be used to bring all people under the sway of Islam; Western knights believed that they were called not only to defend but to “exalt” Christianity and that attacks on its enemies could be holy and meritorious. The Byzantines believed that war was neither good nor holy, but was evil and could be justified only in certain conditions that centered on the defense of the empire and its faith. They were convinced that they were defending Christianity itself and the Christian people, as indeed they were. The Catholic University of America

38

Taktika 18.19; Va´ri, 1:21; PG 107:949. Late in the 12th century, archbishop Euthymios Malakes of Patras, in a court oration, has the soldiers of Manuel I echo these same sentiments: “We labor on behalf of religion and campaign on behalf of God; we do no injustice to foreigners but do battle for what belongs to us.” He has the emperor take the lead in the struggle and profess his readiness to die on behalf of the Christian people. Euthymiou Malake ta sozomena, ed. K. G. Mpones, 2 vols. (Athens, 1937–49), 2:31.5–8; 52.10–13. 39 Dennis, Three Treatises, chap. 19, pp. 216–17.

The Land of War: Europe in the Arab Hero Cycles M. C. Lyons Historians may perhaps object to the Aristotelian distinction that separates history, as the study of particulars, from poetry, whose aim is to produce universals.1 This does, however, provide a convenient setting for a definition of hero legend, in which the particulars of historical fact are mixed with the universals of folklore and fantasy. Whether or not such a literary form can serve any serious academic purpose beyond its own boundaries must, in part, depend on the reasons why it was created in the first place. Here it can be noted that in pre-Islamic and early Islamic poetry, in whose tradition there is very little narrative, there are frequent references to tribal gatherings and evening conversations in which factual narratives were presented and embroidered and stories exchanged. At times such stories revolved around universally identifiable motifs of folklore. In a tribal context, however, of more immediate relevance is the self-identification of the clan and of its members, positively in relation to its own merits and negatively in contrast to the demerits of its opponents. Where these opponents represent a nation rather than a rival clan, the positive identification is more widely shared, and so it is that the pre-Islamic battle fought by Bakr against a Persian army at Dhu¯ Qa¯r serves poets like an Arab Bannockburn as a source of general pride for at least the next two and a half centuries.2 Here although the negative, represented by the enemy, provides a necessary contrast to the heroic positive, its emotional coloring is not a uniform shade of black. Persians were on the wrong side at Dhu¯ Qa¯r, but in both early poetry and later hero legend individual Persian rulers are presented as figures of dignity and splendor. Similarly, while the expression “the land of war” serves to distinguish non-Muslims from Muslims, what is perhaps its earliest intrusion into Arabic poetry is emotionally neutral. Miskı¯n alDa¯rimı¯ writes: “I stay in Da¯r al-H . arb as long as I am not humiliated there, but if I fear disgrace in my dwelling, I leave it.” 3 This, from the pre-Islamic poet al-Shanfara onward, is the familiar boast of the proud man who will only remain even among his own clan as long as his dignity is respected. Tribal identification is an obvious starting point for many of the extant Arabic hero 1

De arte poetica 1451b6. For Islamic poetic references to Dhu¯ Qa¯r, see Abu¯ Tamma¯m, Dı¯wa¯n, ed. M. A. ‘Azza¯m (Cairo, 1951), 1:195.25, 215.28; al-Akht.al, Dı¯wa¯n (Beirut, 1969), 226.4, 317.4; al-Farazdaq, Dı¯wa¯n (Beirut, 1960), 2:111.3. 3 Dı¯wa¯n (Baghdad, 1970), 25.2. 2

[ 42 ] Europe in the Arab Hero Cycles cycles, as where the ‘Antara legend combines the glories of its eponymous hero with praises of the clan of ‘Abs, while the heroes of the cycle of Dha¯t al-Himma are from Kila¯b. Over the centuries, however, the development of the cycles has accommodated a wider audience, and their printed form represents a continuous accretive narrative process. As a result of this, facts of history and geography are inevitably distorted, jumbled, or misinterpreted, but the corruptions represent not a conscious rewriting of history but what has been assimilated into the collective memory of the storytellers and their unsophisticated listeners. If Europe in general, and the Crusades in particular, were of real significance to these audiences, traces of this, positive or negative, should still be detectable.4 Because of the nature of the tradition, it is no real surprise to find that even ‘Antara, the quintessential desert warrior of pre-Islamic days, is given a European role, ostensibly dated to the second half of the fifth century. He first encounters Franks when he is fighting on the side of the Persians against Byzantines reinforced by the Frank alKhalı¯ja¯n, who has brought an army of two hundred thousand men “to aid the servants of the Cross and to destroy the fire-temples of the Persians,” with the declared intention of “staying in Syria and visiting Jerusalem.” Al-Khalı¯ja¯n accuses the Byzantine emperor of hypocrisy and tells his Franks that, when he has finished with the Arabs, he proposes to take over “the kingship of all the Messiah’s lands.” The episode ends when he himself is killed by ‘Antara.5 . a¯rith, the Byzantine client king of Damascus, ‘AnLater, following the murder of al-H tara espouses the cause of al-H . a¯rith’s nine-year-old son, defeats a Byzantine rebel, and is invited to Constantinople by the grateful emperor, Caesar. On being told of this, Lı¯lima¯n, Lord of the Isles, sends a message to Caesar, who is addressed as “king of the Christians,” demanding that the guest be seized. It is explained that Lı¯lima¯n is not a Byzantine subject, that his lands are a forty-day journey distant, and that they extend for another four months’ journey in length and breadth. There follows an extended sequence of adventures in which ‘Antara, accompanied by Caesar’s son Herakleios, sails out to deal with a sequence of European enemies. Of these the most formidable is the giant king of Andalus, twelve cubits tall, who speaks “in Frankish and the languages of Andalus” and who proposes to seize both Rome and Constantinople and then to attack the H . ija¯z, ‘Ira¯q, Khurasa¯n, and Transoxiana. Like al-Khalı¯ja¯n, he too is killed by ‘Antara, who now moves back through North Africa before returning home via Constantinople, which he reaches after a twenty-day voyage with fair winds from the Oases Island.6 As can be be seen, this account is hardly a storehouse of historical fact. Its proper 4 The evidence presented here is derived from anonymously edited and undated publications designed for a semiliterate market in the early decades of the twentieth century. Manuscripts exist, but an effective study of their traditions has still to be completed. In general, the position is that of the Arabian Nights, whose manuscript tradition can be extended to the 9th century and whose early popularity is recorded in the 12th. As both the Arabian Nights and the hero cycles represent accretive oral tradition, their value as historical sources does not lie in the exactitude of their dating but in their function as filters for popular attitudes and traditions. 5 Because of the difficulties associated with the printed texts of the cycles, all references to them refer to the sections of the relevant sı¯ras into which they are divided in The Arabian Epic, vols. 2 and 3, ed. M. C. Lyons (Cambridge, 1995) (hereafter Epic). The ‘Antara reference here is to sections 45–46. 6 Epic, 80–82.

M. C. Lyons [ 43 ] names can at times be recast into European forms, but except in the case of Herakleios, who was well known to the Arabs, they can scarcely be related to real people. As for geography, it would be naive to base investigation on a fluid text whose details may be purely imaginary. All that is confirmed here is that the Mediterranean was known to lead to island, or peninsular, kingdoms, these being “the islands of the Franks” to which, according to Ibn Shadda¯d, Saladin hoped to extend the boundaries of Islam.7 Less common is the distinction drawn between Andalusians and Franks, while the reference to the languages spoken by the Andalusian king underlines a significant difficulty that the Arabs experienced with regard to Europe. At a time when Richard Coeur de Lion was easily able to have the captured draft of al-Harawı¯’s geographical work identified for him,8 the Muslims were complaining that they needed relays of interpreters for the interrogation of prisoners, each translating into a language that the next in the chain could understand. In what may be thought of as the political content of the cycle, Byzantines and Persians are presented as natural rivals, each with their own client kings, who normally find it convenient to live at peace. Equilibrium in the region is upset by Franks, more formidable fighters than the Byzantines, who wish not merely to conquer but to colonize, and in this context Arabs are natural supporters of whoever opposes them. Here there are, of course, historical precedents, as when the Goths attacking Constantinople after the battle of Hadrianople in 378 .. were turned back by a so-called Saracen force,9 and it is true that the term “Franks” could apply in retrospect to Westerners in general, ¯ mid during its siege by Sapor. It must be suggested, however, that such as the Gauls at A ‘Antara’s adventures do not merely reflect old memories, transmitted in part by the Qur’a¯n, of Byzantine/Persian wars. Rather, the formidable Western seafarers who acknowledge the emperor as “the king of Christendom” but are prepared to attack him and take his lands, are of a later vintage and their introduction reflects, if not the Crusades themselves, at least the disturbing influence that the Crusaders exerted on the balance maintained between Islam and Byzantium. The Sı¯rat H . amzat al-Pahlawa¯n, whose hero, a pre-Islamic Meccan paladin, is unhistorically represented as overthrowing the Persian Empire, is of less importance here. At an early stage of his career, H . amza is sent to collect dues from Persian tributary states, which include Aleppo, Tripoli, Beirut, Sidon, Tyre, and Acre.10 He is welcomed at Constantinople by the wise king Stephanos, and he then visits Greece and the country of Caesar, “king of the Ru¯ma¯n.” He fights his way down the Levant coast and moves through Egypt to North Africa. Here he finds “The City of Andalus,” whose army helps him in a battle fought, surprisingly, against the Persians at Tangier.11 Where geography is concerned, while H . amza’s Constantinople is correctly placed, Greece appears to lie to the east of the Aegean. Andalus is in the south of the Mediterra-

Al-nawa¯dir al-sult.a¯nı¯ya, ed. J. al-Shayya¯l (Cairo, 1962), 22. Al-Harawı¯, Kita¯b al-isha¯ra¯t ila¯ ma‘rifat al-ziya¯ra¯t, ed. J. Sourdel-Thomine (Damascus, 1953), 3. 9 Cf. Cambridge Medieval History, ed. H. M. Gwatkin and J. P. Whitney (Cambridge, 1924– ), 1:252. 10 Epic, 8 f. 11 Epic, 15. 7 8

[ 44 ] Europe in the Arab Hero Cycles nean, and the capital of the Ru¯ma¯n king is more likely to be Anatolian Caesarea than Rome. In this context the chief point of interest is that the Ru¯ma¯n, unlike the Byzantines, are hostile to the Arabs, as are the cities of the Levant coast. Although these were among the territories recovered from the Persians by Herakleios, their resistance to H . amza would fit more appropriately into a distorted memory of the Crusades. The published texts of the Banı¯ Hila¯l cycle can be dismissed briefly. The eponymous Hila¯l is represented as one of the companions of Muhammad, and the fortunes of his clan are due in part to the defeat by one of his descendants of an invasion aimed at Mecca and led by a Byzantine emperor.12 Later, Franks occur in an Arabian Nights–type story of a bride searching for her missing bridegroom who has been carried off from the coast by a party of raiders. Familiar folklore motifs dictate the development of the story, but reality intrudes where a Frankish leader says: “our prisoners herd pigs, grind salt, carry stones, and row ships,” a remark that would apply to both sides of the Mediterranean through centuries of piracy and conflict.13 The Sı¯rat Dha¯t al-Himma is an extended paean of praise for the Banu¯ Kila¯b, who are represented as the only effective guardians of the northern frontier of the Islamic empire between the reigns of the Umayyad ‘Abd al-Malik and the ‘Abba¯sid al-Wa¯thiq. Within the text the framework of Islamic geography, from Khurasa¯n to Andalus, is clear and reasonably accurate, and real sites mix with the imaginary on the eastern shore of the Dardanelles. Outside these borders, the story’s heroes travel south to Africa and east to the Indian Ocean, while to the west and north are lands where fact has been dislodged by fantasy. Sicily provides an obvious landfall on a passage to the west, but it is found to be within a fortnight’s voyage of a Christian kingdom that is itself a year-and-a-half’s journey from Constantinople. Greece is, not unreasonably, described as a land of narrow passes, but it also contains a forest frequented by lions and snakes.14 The forty Castles of the Sun, reached by ship from Anatolia and set in the seventh clime, may be on the Sea of Azov, if it can be thought that this was visited by Indian cotton merchants.15 The Island of Ru¯bı¯s is forty days’ sail from the Island of Mines, after which it takes two months to get to the Atlantic and the Farthest Cape, beyond which lies an island that exports both gold and pearls.16 In the north are found the Qara¯zima, whose country is as far from the Land of the Slavs as Malat.ya is from Constantinople. Its roads are blocked by snow in the winter, but its ships are said to trade with both India and the West.17 This geography can at times be linked to religion. Thus islands in the far north are ruled by Magian kings, recalling al-‘Udhri’s description of Ireland as “the only (remaining) Magian base in the whole world.” The editor of Abu¯ H . a¯mid of Granada, referring to a similar entry in Idrı¯sı¯, suggests a connection with what he calls “the German fire-cult,” and it may have acquired a racial rather than a religious significance by the 12

Epic, 3. Epic, 6–8. 14 Epic, 44 f. 15 Epic, 73. 16 Epic, 139. 17 Epic, 92. 13

M. C. Lyons [ 45 ] time that it is applied to Normans in Sicily.18 In the Sı¯rat Dha¯t al-Himma, however, the Magians are fire-worshipers pure and simple. For the rest, all other European kings are Christians, and when Hı¯lu¯n, a king of the Franks, refuses to answer the summons of Emperor Manuel to join in an expedition against the Muslims, his mind is changed by an order from “the Patriarch.”19 In a similar incident, Kundafru¯n—presumably a corruption of “the Count of . . .”—joins Manuel after having been threatened with excommunication. More remarkably, in the account of the country of the Qara¯zima there is an apparent reference to the homoousia/homoiousia debate. The historical account of relations between Islam and Europe starts with a romanticized version of Maslama’s attack on Constantinople in 717 ..,20 while later Mu‘tas.im’s capture of Amorium is correctly dated.21 For the rest, there is a seesaw pattern of unspecific raids, interspersed by periods of truce, but in all this it is clear that Byzantium enjoys favored enemy status. When Leon, whose name at least is borrowed from Leo III, is attacked by the king of Portugal, it is Fa¯.tima, the amazonian heroine of the cycle, who comes to his rescue,22 while the caliph Ha¯ru¯n al-Rashı¯d allies himself with one of Leo’s ¯ mid.23 The Franks, huge, clean-shaven men, are at successors to fight Muslim rebels at A times a formidable component of Byzantine armies. They are distinguished from the Venetians, who also give unhistorical support to the Byzantines,24 while on their own Western ground Franks campaign without success in Spain.25 The presence of a betterdefined crusading stratum is indicated by the repetitive adventures of a series of Western, or specifically Frankish, kings. Of these the first is Bah. ru¯n,26 who collects ten thousand ships for an attack on Constantinople, where he marries the emperor’s daughter before proceeding to march against the Muslims. Bah. ru¯n is followed by Arma¯nu¯s from the Outer Isles,27 who is joined by Syrian bedouin in his eastern campaign, and he is then deposed by the formidable Bohemond.28 After Bohemond has fallen, yet another Frank, Mı¯la¯s, is instructed in a dream to rescue Jerusalem and to sit on the throne of Constantinople.29 His venture is short-lived, but to end the series another, and more serious attack, is launched by Bohemond’s son, Michael.30 The Sı¯rat Baibars is a conglomeration of narrative motifs each attached to one of the various narrative strands of which the cycle is composed. Of these strands, that of wonder tales has its own geography, the Land of Wonders itself being set beyond the territories 18

Al-Qazwı¯nı¯, Atha¯r al-bila¯d wa-akhba¯r al-‘iba¯d (Beirut, 1960), 577.11. Ed. C. E. Dubler (Madrid, 1953), 243; this work contains an excellent summary of what was known by Arab geographers and travelers of the geography of the West. 20 Epic, 62. 21 Epic, 8 f. 22 Epic, 137 f. 23 Epic, 53 f. 24 Epic, 104, 170. 25 Epic, 69. 26 Epic, 111. 27 Epic, 117. 28 Epic, 142. 29 Epic, 148. 30 Epic, 163. 19

[ 46 ] Europe in the Arab Hero Cycles of Byzantium and Persia in a region where there are two suns and two moons.31 Elsewhere this geography is allowed to borrow familiar names. A galleon with a golden prow and a stern of silver carries Baibars in six months to England, where time is suspended so that, although he has been entertained for twelve years, he is returned to Alexandria on the day after he left.32 A second English episode introduces another Kundafru¯n, Lord of all the English islands, who attacks Aleppo.33 More generally, islands and distant lands are hostile to the Muslims. The queen of the archipelago of al-Manı¯‘a prepares a fleet with which to attack them,34 and similarly, the king of the Islands of Gold demands the hand of the Byzantine emperor’s daughter as a preliminary to an eastern campaign.35 Baibars is taken as a prisoner to the City of Snow near the land where the sun sets, and it is in his absence that Alexandria is attacked by a five-decked ship with seven masts, sent by the king of Seville, which is now found to be in Christian hands.36 Ships, in fact, are of particular importance in this cycle, and the Muslims are fortunate to have the help of Abu¯ Bakr of Tangier, a sailor “unique in his age,”37 as well as a seafaring saint whose skiff miraculously tows a whole Muslim fleet.38 This allows them to deal with a number of European cities which make their only appearance in the cycles. Thus Genoa is attacked three times;39 and the city of Catalan, with its two harbors, is burned.40 The kings of the Franks and of the Ru¯m are called to a crusading muster at Barcelona, which the Muslims proceed to capture.41 One of Baibars’ companions is sent to Rome to collect ransom money, and for his return journey narrative geography allows him, albeit with the loss of all but one of his force, to travel by land.42 Portugal supplies the story with a halfhearted villain who is eventually converted to Islam; its king pays tribute to Baibars and is only unwillingly and unsuccessfully led to oppose him.43 Rhodes is ruled by a hostile cannibal.44 Cyprus is one of Baibars’ tributaries,45 and the Muslims retaliate against the holding of their captives in Macedonia by capturing its capital city.46 The historical Baibars enjoys only an attenuated narrative strand in his own cycle. In this his questionable rise to power leaves him suspected of having murdered Tu¯ra¯nsha¯h during St. Louis’ crusade, a charge the dead man himself is made to refute. Details of 31

Epic, 144. Epic, 115 f. 33 Epic, 168. 34 Epic, 162. 35 Epic, 199. 36 Epic, 193. 37 Epic, 62. 38 Epic, 115. 39 Epic, 41, 68, 104. 40 Epic, 89. 41 Epic, 98. 42 Epic, 61 f. 43 Epic, 99. 44 Epic, 136. 45 Epic, 208. 46 Epic, 73. 32

M. C. Lyons [ 47 ] the fighting at al-Mans.u¯ra and Faras Ku¯r are given, but St. Louis himself has left no personal mark, the invaders being led by four unnamed Frankish kings, all of whom are killed.47 Earlier in his career Baibars had fought against the Franks of Syria where, after a quarrel over the division of grain revenue, a Frankish leader Sarjawı¯l was captured in an abortive attack on Damascus and then allowed to ransom himself.48 This same pattern of private feuding accompanied by a readiness to accept ransom or tribute is followed throughout the confused accounts of Syrian campaigns, where towns are taken from the Christians but are then promptly found back in their hands. These campaigns cover the whole of the Levant coast from Tripoli down to al-‘Arı¯sh, whose lord, Franjı¯l, buys immunity from al-Malik al-S. a¯lih. , who is on his way to confront the Mongols.49 At the center of the pattern, both historical and geographical, is Aleppo. In the Sı¯rat Dha¯t al-Himma, enemy attacks follow the Byzantine invasion route down the Euphrates toward Baghdad. Here, however, the cycle’s heartland is Egypt. When Aleppo is attacked by the Mongols in the reign of al-Malik al-S. a¯lih. , who is represented as a saintly simpleton, it has to be explained to him that its loss will involve that of all Syria, from where the enemy will attack Egypt.50 Antioch is used as a Christian base from which Aleppo can be threatened, on one occasion in conjunction with the Mongols.51 A more elaborate plan involves a synchronized strike from the sea against Alexandria and Damietta and a land attack on Aleppo,52 while immediately after Baibars’ death the king of Rome fails in an attempt to recover Antioch.53 It would, of course, be easy to multiply such details, but there must be a limit to the acceptance of any such accumulation of faulty geography and distorted myth-history. It remains to be asked whether in evidence of this type there is anything that can be used to throw light on the relationship of the Crusades to the Muslim world, a relationship that has been dismissed by some scholars as insignificant, while elsewhere in recent approaches to the pope it has been magnified as a continuing source of friction between Islam and the West. It is obvious that, for the storytellers, relations with the the European world begin with Byzantium, either as represented by its client-kings in Syria or by the splendor of Constantinople itself, and one stratum of the cycles clearly contains fossilized remains of the pre-Islamic Byzantine/Persian wars. Another holds maritime memories. Among these are landsmen’s tales that admit no difference between champions sitting on horseback in front of their armies or sailing out alone from their fleets to challenge their enemies. European sea power, however, based not only in vaguely described islands but in Venice, Genoa, and the Spanish littoral as far as Portugal, made it reasonable for the audiences throughout the period of the cycles’ development to welcome the general reassurance that here too the Muslims have the upper hand. 47

Epic, 42. Epic, 10. 49 Epic, 34. 50 Epic, 33. 51 Epic, 123. 52 Epic, 203. 53 Epic, 215. 48

[ 48 ] Europe in the Arab Hero Cycles Beyond this there are certainly remains of a crusading stratum, eroded though this may be. In the Sı¯rat Baibars it is made clear that the closure of the church of the Holy Sepulcher will provoke a Christian reaction.54 After their abortive muster at Barcelona, the kings of Christendom are summoned to a holy war by the pope, who thus confirms his position as “the Messiah’s deputy.” St. Louis’ attack is neither more nor less serious than many others, but a more obvious threat is posed by the misplaced Franks in the Sı¯rat ‘Antara, specifically because they want to stay in Syria. Traces remain of the Latin Kingdom and the County of Tripoli in the towns held by Franks in the Sı¯rat Baibars, but for the storytellers Jerusalem itself is comparatively unimportant, being described baldly as “a noble place in which other prophets prayed as well as Muhammad.”55 In fact, it is Constantinople and Egypt that are the prime targets for Frankish attacks. Throughout all this, Europe and the Franks add piquancy to the conventional formulae of identification. The fictional Byzantines were scarcely ever allowed to be a serious threat to the Muslims, and so to maintain the interest of the audiences more impressive enemies were needed. In the Sı¯rat Dha¯t al-Himma, the bravest of the Franks, the grand duke, is “a very wicked and guileful man,”56 but he comes low on the scale of strength and savagery in comparison with Bohemond, who imprisons his own father and can drive his fist clean through a man’s body from breast to backbone,57 while another angry king crushes the skull of his son with his bare hands.58 Queens are no less formidable, among them being Karna, who is strong enough to lift a camel,59 and Miriam, who when displeased cuts through one of her own retainers with a single blow.60 That even enemies like these can be defeated adds to the status of the cycles’ heroes and enhances the glories of Islam. But violent as they are, they represent elemental forces of aggression, which are matched by similar forces on the side of the Muslims—this being the point of the equation of Muslim Spain to the Latin Kingdom made by the contemporary polymath al-Qazwı¯nı¯.61 By contrast, the real villains are those who release these forces for their own ends, in the case of the Sı¯rat Dha¯t al-Himma this being done by the recreant Muslim ‘Uqba, and in the Sı¯rat Baibars by the false Christian Juwa¯n. In neither case are these men motivated by religion, but rather by a restless wickedness that impartially hates both sides and wants to destroy stable society. For the external observer this underlines what appears to be an important point. The Muslims, as such, are distinguished from their enemies in the same general way that the Land of War is distinguished from the lands of Islam, and this distinction can be seen as an extension of Ibn Khaldu¯n’s loose principle of “group feeling.” It must be noted, however, that it is not the Muslims as a whole who repel Franks and Byzantines in the Sı¯rat Dha¯t al-Himma, but the clan of Kila¯b, while similarly in the Sı¯rat Baibars the effective 54

Epic, 36. Sı¯rat Dha¯t al-Himma 268. 56 Epic, 112. 57 Sı¯rat Dha¯t al-Himma 142. 58 Sı¯rat Dha¯t al-Himma 128. 59 Sı¯rat Dha¯t al-Himma 92. 60 Sı¯rat Baibars 162. 61 Atha¯r al-bila¯d wa-akhba¯r al-‘iba¯d (Beirut, 1960), 503.22. 55

M. C. Lyons [ 49 ] groupings are those of the Kurds, followed later in the work by the Isma¯‘ı¯lı¯s. Historically, Nu¯r al-Dı¯n, and more particularly Saladin, tried to universalize “group feeling” by enlisting it in the service of a holy war. In propaganda terms the comparative failure of this attempt can be seen in the ungrateful reaction of the ‘Abba¯sid caliph to the recovery of Jerusalem, a triumph that, in turn, has left no impression on the extant hero cycles. More immediately, when transferred to the battlefield, “group feeling,” as embodied in Saladin’s volunteers and his lightly armed bedouin, proved now to be of only limited use, the battles being won or lost by the contingents of highly trained cavalry and archers.62 In spite of the fictional Kila¯bı¯s, in practice we are no longer dealing with clans that are their own armies and still less with citizen forces such as those of the early Greek poleis. Warfare has become increasingly professionalized, and, as Saladin discovered early in his career, in such a context success depended on money. Toward the end of his fictional career, Baibars is shown as suffering a series of disasters in an attempt to defend Aleppo from fire-worshiping Mongols. In spite of the danger, the Aleppans themselves are lukewarm in their support for him, saying that they will open their gates to whichever side wins. Baibars cannot reequip his troops, as his treasury is empty, and when he tries to raise money at Damascus for what is, after all, a war against the infidel, he is told: “you are a king. . . . Kings meet one another and fight for their positions, and we serve whoever is on the throne.” He is refused a loan by merchants who claim that this would upset their trade, and when as a last resort he decides to raise money by a tax, he is warned by the religious leader, the shaikh al-Nu¯rı¯, that this is unlawful. When he presses the point, al-Nu¯rı¯ goes so far as to curse him, as a result of which he is temporarily blinded.63 Although all this is fiction, the obvious point is that the audience is supposed to accept its verisimilitude. It might be argued pragmatically that, although there must be some emotional validity in the self-identification of the cycles’ audiences with the heroes of Islam, money comes nearer to their own experience of reality. Thus the universal divide between rich and poor is underlined by the Cairene crowds in the Sı¯rat Baibars, who believe that “only the poor die; these kings have the water of life.”64 On this line it could be suggested that the difference between the importance of the Crusades in the French chansons de geste and their limited role in the Arabic hero cycles depends directly on the difference of audience, comprising in the one case the consumers and in the other the producers of wealth. This, however, would be a dubious generalization, and a more fruitful line of investigation may be to add to identification the concept of identity. The original building blocks of the hero cycles were the bedouin clans where the clan’s identity is a product of the individual clansmen who identify with it. In urban settings, the cycles show us that the place of the clan as a unit is taken by groupings organized in accordance with some common factor, such as district, guild, or sect, whose

62 The weight of the Crusaders’ charge, together with the length of their lances and their defensive armor, could not be countered, as ‘Ima¯d al-Dı¯n noted, by those who were without experience of Frankish fighting; Kita¯b al-faih. al-qussı¯ fı¯’l-fath. al-Qudsı¯, ed. Comte C. de Landberg (Leiden, 1888), 97. 63 Epic, 188. 64 Epic, 87.

[ 50 ] Europe in the Arab Hero Cycles sense of identity is strong, but for whom the importance of identification has been diluted.65 In such a context, Europe and, by extension, the actions of the Europeans can be identified in terms of the general concept of the Land of War, within which there is little, if any, individual identity. The challenge to Crusader historians is to avoid being restricted to a similarly generalized identification in respect of Islam. The preaching of the Crusades sought to rally the single, optimistically identified, entity of Christendom against Islam. With similar optimism Saladin’s secretary, ‘Ima¯d alDı¯n, commenting on the threat posed by Frederick Barbarossa, wrote: “the Muslims will rise up for us.”66 Later, however, Saladin was forced to ask: “where are the Muslims? God forbid that they should be abandoning Islam.”67 In the same way that “Christendom,” as a shared identification, failed to restrict the internal rivalries of the Crusaders, so contemporary Arab accounts confirm what is suggested by popular narrative, that the threat posed by the Crusades was not met by a united response, either among the competing leaders or at a lower social level, at which the Qa¯d. ¯ı al-Fa¯d. il complained of the prevalence of “acts of disobedience to God” and “the scent of sedition.”68 The society that confronted the Crusaders was a complex cellular structure. Translated into medical terms, the body of Islam was faced at this period with a serious threat from a virus that it could instantly identify as belonging to the category of “nonself.” To counter this it employed defense mechanisms, which were, of course, already in place, but which now can be argued to have developed a form of gigantism that, in turn, removed them from the category of self to what approximated nonself. In other words, the effective defense against the Crusaders was provided, in the main, by non-Arabs, whether Turks, Kurds, negroes, or Mamluks. The hero cycles suggest that, although these groups were accepted, and appreciated, as fighting for Islam, they were not integrated into the main body of civilian society. The Mamluk Baibars, for instance, has to be provided with a noble Islamic pedigree to win acceptance, and in his cycle it is not the Mamluks but the Syrian Isma¯‘ı¯lı¯s who show themselves to be the lions of Islam. The most popular, or specifically the most widespread hero of the cycles, is not an irresistible fighter but the poor man’s friend, the Man of Wiles. In the same way that in the chansons de geste Charlemagne has an uneasy and at times hostile relationship with his paladins, so in the cycles orthodox authority in the form of the caliph or the sultan may fall out with Islamic heroes, and in these cases it is the Man of Wiles who is invariably, and invincibly, in the right.69 This suggests that one effect of the Crusades was to provide a clear and immediate test of the emotional validity of a catchall identification of “Islamic” society, a test that proved negative. On a practical level they helped to develop, and perhaps to

65 The account of Baibars’ rise to power in his sı¯ra gives a number of details of urban organization which tend to confirm this point. The inhabitants of the “ideal” urban quarter founded by Baibars have no tribal, racial, or specifically religious identification. Its identity has been superimposed on it by Baibars and shows itself through cooperation. 66 Kita¯b al-faih. al-qussı¯ fı¯’l-fath. al-Qudsı¯, 216. 67 Abu¯ Sha¯ma, Kita¯b al-raud. atain fi akhba¯r al-daulatain, ed. M. H. M. Ah. mad and M. M. Ziya¯da, 2 vols. (Cairo, 1956–62), 2:157. 68 Ibid., 166; Ms. Munich 402, fol. 104. 69 See above for Ha¯ru¯n al-Rashı¯d’s alliance with the Byzantines against al-Bat..ta¯l, the Arab Maugis.

M. C. Lyons [ 51 ] unbalance, its defense mechanisms. In general, the cycles suggest that for many social groupings the main fear was of an upset in the delicate social equilibrium that allowed them to maintain their own identity derived from or expressed in patterns of life that were set at a level below that of power, wealth, and culture. Fictionally the Crusades did not produce such an upset, but whether the military organization that they helped to produce, leading most obviously to the Mamluk state, intensified a “we and they” social dichotomy must be the subject of another inquiry. In turn this must cover on a wider basis of evidence the questions of identity and identification, insofar as these are both the products of any given society and among its most potent shaping forces. Pembroke College, Cambridge, England

Byzantium through the Islamic Prism from the Twelfth to the Thirteenth Century Nadia Maria El-Cheikh

The historiography of Muslim-Byzantine relations suffers from major gaps. The two major figures who have dealt with the subject are A. A. Vasiliev and Marius Canard. Vasiliev’s monumental work, Byzance et les Arabes, concentrates on the military and political relations between the Byzantines and the Arabs. Marius Canard’s publications are more varied; his bibliography includes pioneering works on various facets of the socioeconomic, diplomatic, and cultural relations between the two empires.1 Important gaps remain, however, concerning the life and conditions on the Arab-Byzantine frontier,2 certain aspects of sociocultural and economic interactions,3 and the respective images of the two warring societies. This study addresses the latter topic, the Islamic view of Byzantium during the twelfth and thirteenth centuries. The literature on this subject is still limited in spite of the growing number of published works on various aspects of East-West interactions and mutual perceptions during medieval times.4 Works instigated by both the debate on Orientalism and the activation of the Christian-Muslim dialogue have also tackled the issue of perception and representation, but only a few have handled it from the perspective of Arab1 A. A. Vasiliev, Byzance et les Arabes, French edition by H. Gre´goire and M. Canard (Brussels, 1950), 2 vols. For the bibliography of Marius Canard, see M. Lefort, “Notes et documents: Index de l’oeuvre historique de Marius Canard,” Arabica 22.2 (1975): 180–211, and F. Daftary, “Marius Canard (1888–1982): A BioBibliographical Notice,” Arabica 33 (1986): 251–62. 2 Advances have been made in recent years on this subject; see, for instance, J. F. Haldon and H. Kennedy, “The Arab-Byzantine Frontier in the Eighth and Ninth Centuries: Military Organization and Society in the Borderlands,” Recueil des travaux de l’Institut d’Etudes Byzantines, Acade´mie Serbe des Sciences et des Arts 19 (1980): 79–116; and C. E. Bosworth, “Byzantium and the Syrian Frontier in the Early Abbasid Period,” in Fifth International Conference on Bila¯d al-Sha¯m: Bila¯d al-Sha¯m during the Abbasid Period, ed. A. Bakhit and R. Schick (Amman, 1991), 2:54–62. 3 The work of Marius Canard, in particular, “Les relations politiques et sociales entre Byzance et les Arabes,” DOP 18 (1964): 35–56; “Quelques a` coˆte´ dans l’histoire des relations entre Byzance et les Arabes,” in Studi orientalistici in onore di Giorgio Levi Della Vida, vol. 1 (Rome, 1956), 98–119; and “Le ce´re´monial fatimide et le ce´re´monial byzantin,” Byzantion 21 (1951): 355–420, has increased our knowledge in basic ways. 4 N. Daniel, Islam and the West (Edinburgh, 1960), with the new updated French version, Islam et occident (Paris, 1993); R. W. Southern, Western Views of Islam in the Middle Ages (Cambridge, Mass., 1962); B. Lewis, The Muslim Discovery of Europe (New York, 1982).

[ 54 ] Byzantium through the Islamic Prism Byzantine relations.5 Works on Byzantine perceptions of the Arabic-Islamic culture, people, and history remain few,6 while studies on Arab perceptions of Byzantium have witnessed a slight cumulative increase in recent years.7 By the eleventh century, a certain tradition depicting Byzantium and the Byzantines had been elaborated in the Arabic-Islamic sources.8 Later sources juxtaposed contemporary perceptions side by side with earlier views. A major characteristic of the ArabicIslamic sources of the twelfth and thirteenth centuries is an approach that favored compilation and systematization. As a result, in a wide variety of sources, there is often duplication and reproduction from earlier works. In order to delineate the Muslim representation of Byzantium in the twelfth and thirteenth centuries, I review a collection of important and well-defined prose sources from this period, including universal chronicles and local histories, geographical works and biographical dictionaries, as well as monographs on individuals and dynasties.9 I have avoided the genre of the hero cycles, such as Sı¯rat ‘Antar and Sı¯rat Dha¯t al-Himma, because of the uncertainties concerning the identity of the authors, the date of composition, and the manuscript tradition.10 Not only are these texts filled with additions and interpolations, they also do not exist in critical editions. There is no doubt that once these obstacles are surmounted, the epics will provide a great amount of material relevant to the popular Arab medieval mentality. 5 On Orientalism, in addition to E. Said, Orientalism (New York, 1978), see H. Djait, Europe and Islam, trans. from the French by P. Heinegg (Berkeley, 1985); and M. Rodinson, La fascination de l’Islam: Les e´tapes du regard occidental sur le monde musulman (Paris, 1980). On the Christian-Muslim dialogue, see W. M. Watt, MuslimChristian Encounters: Perceptions and Misperceptions (London, 1991); and W. Bijlefeld, “Christian-Muslim Studies: Islamic Studies and the Future of Christian-Muslim Encounter,” in Christian-Muslim Encounters, ed. Y. Haddad and W. Haddad (Gainesville, 1995), 13–40. 6 See, for instance, J. Meyendorff, “Byzantine Views of Islam,” DOP 18 (1964): 114–32; V. Christides, “The Image of the Pre-Islamic Arab in the Byzantine Sources” (Ph.D. diss., Princeton University, 1970); A. Ducellier, “Mentalite´ historique et realite´s politiques: L’Islam et les Musulmans vus par les Byzantins du XIIIe`me sie`cle,” ByzF 4 (1972): 31–63; S. Vryonis, “Byzantine Attitudes towards Islam during the Late Middle Ages,” ByzMetabyz 2 (1981): 263–86. 7 Andre´ Miquel’s seminal work, La ge´ographie humaine du monde musulman jusqu’au milieu du XIe`me sie`cle (Paris–The Hague, 1976–80), 3 vols.; A. Shboul, Al-Mas‘u¯dı¯ and His World (London, 1979), and “Arab Attitudes toward Byzantium: Official, Learned, Popular,” in Kathegetria: Essays Presented to Joan Hussey for Her 80th Birthday (London, 1988), 111–29; So. . el-Attar, “Contemplaciones initiales sobre el tema bizantino en la cultura arabe,” Byzantion Nea Hellas 7–8 (Santiago, 1985): 209–26; as well as various articles published in GraecoArabica. 8 See N. M. El-Cheikh, “Byzantium Viewed by the Arabs” (Ph.D. diss., Harvard University, 1992). 9 For a brief presentation of the 12th- and 13th-century Arabic sources, see M. H. M. Ahmad, “Some Notes on Arabic Historiography during the Zengid and Ayyubid Periods: 521/1127–648/1258,” in Historians of the Middle East, ed. B. Lewis and P. M. Holt (London, 1962), 79–97. 10 See H. T. Norris, The Adventures of Antar (Warminster, 1980), 4, who states that the sı¯ra of ‘Antar was drafted between 1080 and 1400 but includes both earlier and later material. Bridget Connelly states, in Arab Folk Epic and Identity (Berkeley-Los Angeles, 1986), 8, that “the authorship and the date of these works are generally obscure.” Marius Canard, in a series of articles on Dha¯t al-Himma, concluded that the epic is formed of two different cycles of different periods and origin and that “it is impossible to give an exact date for the composition of the romance.” See “Dhu¯l Himma,” EI 2, and “Delhemma: Epope´e arabe des guerres arabobyzantines,” Byzantion 10 (1935): 283–300. M. C. Lyons, in “The Crusading Stratum in the Arabic Hero Cycle,” in Crusaders and Muslims in Twelfth-Century Syria, ed. M. Shatzmiller (Leiden, 1993), 147–61, states that the literary construction of the cycles “represents oral accretive tradition, based on the manipulation of narrative standard ingredients.” See also the paper by M. C. Lyons, in this volume.

Nadia Maria El-Cheikh [ 55 ] It is necessary to take into account the limitations of such an approach. One such factor is that Arab culture, during these centuries, was mainly a palace culture.11 We see the age through the eyes of the senior bureaucrats and of the ‘Ulama¯’ who had drawn close to the centers of power.12 The information and views are thus limited socially. Another important consideration to keep in mind is that although the individual contemporary authors could emphasize a particular aspect of Byzantium, or bring out a new facet, they still conveyed entrenched images extant in a wide variety of sources. I try to distinguish between those authors who related firsthand accounts, having visited Byzantine territories during this period, and those who relied solely on secondhand information, whether oral or written. The present task is thus to look at the texts of the twelfth and thirteenth centuries in order to detect continuities and changes in these depictions, in light of the new historical context created by the onslaught of the Crusades. The period of the twelfth and thirteenth centuries was a crucial epoch for the Muslim world, which was in a state of political fragmentation. Syria was divided among rival Turkish amirs who were squandering their resources in internal wars, while Fatimid Egypt was trying to maintain its hold on Palestine. This political disintegration prevented any viable resistance to the Crusades, especially given the ideological division that split loyalties between the Abbasid Caliphate and the Fatimid Imamate.13 The Arab authors were there to witness the conquest of Jerusalem by the Crusaders and the establishment of various Frankish principalities. However, many authors were also there to watch the Zankid reconquest, the triumph of the armies of Saladin, and the consolidation of the Sunni restoration. The twelfth and thirteenth centuries were also the period during which the Byzantine Empire was politically and militarily weak. Following a brilliant period in the late tenth and the early eleventh century, disintegration set in. Although Alexios I Komnenos (1081–1118) strengthened the empire, his achievements and those of his successors were not enduring and were followed by a collapse of the Byzantine state.14 What distinguishes this period from earlier ones, therefore, is that the twelfth and thirteenth centuries witnessed the gradual decline of the Byzantine state contrasted with the establishment of the Crusaders in the East. During this period, the Byzantines alongside the Muslims became the target of the Crusaders’ offensive. The appearance of these Christians with an agenda and a mission so completely different from the traditional Byzantine policies and outlook was bound to affect the Islamic image of Byzantium. This study aims at delineating this multifaceted image by tracing both the recurrent traditional themes as well as new representations that arose in the course of the twelfth and thirteenth centuries. The main aspects relate to the Muslim appreciation of Byzantine ori11 See N. Faris, “Arab Culture in the Twelfth Century,” in A History of the Crusades: The Impact of the Crusades on the Near East, ed. K. M. Setton et al. (Madison, Wisc., 1985), 5:3–32. 12 T. Khalidi, Arabic Historical Thought in the Classical Period (Cambridge, 1994), 191–92. 13 For an overview of the situation in Syria during the 12th century, see N. Elisse´eff, “The Reaction of the Syrian Muslims after the Foundation of the First Latin Kingdom of Jerusalem,” in Shatzmiller, Crusaders and Muslims in Twelfth-Century Syria (as in note 10), 162–72. 14 G. Ostrogorsky, History of the Byzantine State (New Brunswick, N.J., 1969), 320 ff.

[ 56 ] Byzantium through the Islamic Prism gins, beauty, artistic skills, knowledge, character and morals, as well as the Muslim authors’ observations concerning the Byzantine territory, particularly Constantinople. The Byzantines are still referred to in our texts as al-Ru¯m. Sometimes, especially during the early period of the Crusades, the term al-Ru¯m was also used to refer to the Franks. Upon the arrival of the Crusaders, the Arabs tended to confuse them with the Byzantines. Mostly, however, and especially with time, the new term al-Ifranj was coined for the Franks, reflecting an emerging distinction between Byzantines and Crusaders. It did not take long for a new image of the Ifranj, dissociated from that of the Byzantines, to emerge. However, both the terms Ifranj and Ru¯m were at times used to mean Christian, in general. Banu¯ al-As.far is another name that continues to be used by Arab authors in the twelfth and thirteenth centuries to refer to the Byzantines.15 Arab Muslim authors continue to trace the origins of the Ru¯m back to Abraham and attempt to explain the reference Banu¯ al-As. far. Ibn Manz. u¯r (d. 711/1311), in his famous dictionary Lisa¯n al-‘arab, defines the Ru¯m as a known people who are traced back to Esau, son of Isaac.16 Ya¯qu¯t (d. 626/1229) repeats information found in earlier sources which trace the Ru¯m mainly through Esau, son of Isaac, son of Abraham. This same Esau, in one story, married Basma, daughter of Ishmael, and since Esau was blond, his wife brought into the world al-Ru¯m.17 Hence the Ru¯m were called Banu¯ al-As. far because they were blond.18 Similarly, the twelfth-century geographer Muh. ammad al-Zuhrı¯ traces the origin of the Ru¯m back to Abraham and his son Isaac, making a clear distinction between the Ru¯m and the Yuna¯niyyu¯n (ancient Greeks).19 Another continuity with the earlier image of the Byzantines is connected with their physical beauty. The cosmographer al-Qazwı¯nı¯ (d. 682/1285) states that the Byzantines are mostly white, with blond hair and sturdy bodies.20 Similarly, the geographer Ibn Sa‘ı¯d (d. 678 or 685/1274 or 1286) stresses the whiteness and blondness of the Byzantines, stating that the inhabitants of the sixth climate are characterized by extreme whiteness, blue eyes, and blond hair and they often have freckles on their faces.21 These physical attributes were highly valued, as can be deduced from various adab works and special manuals, which delineate the prevalent Arab typology of beauty.22 The Arab authors of the twelfth and thirteenth centuries reiterated the now entrenched belief of the Byzantines’ unequaled skill in building, craftsmanship, and paint15 Ignaz Goldziher has a brief discussion of the various colors the Arabs used to refer to other people in Muslim Studies, trans. by C. R. Barber and S. M. Stern (London, 1967), 2:243–45. 16 Ibn Manz. u¯r, Lisa¯n al-‘arab (Beirut, 1992), 12:258. 17 Ya¯qu¯t, Mu‘jam al-bulda¯n (Beirut, 1957), 3:9. See also Shams al-Dı¯n al-Dimashqı¯, Nukhbat al-dahr fı¯ ‘aja¯’ib al-birr wa akhba¯r al bah. r, ed. A. Mehren (Leipzig, 1923), 258. 18 Ya¯qu¯t, Mu‘jam al-bulda¯n, 3:98. 19 Muhammad al-Zuhrı¯, Kita¯b al-jughra¯fiya¯, ed. M. Hajj Sadok, BEODam 2 (1968): 201. 20 Zakariyya¯ al-Qazwı¯nı¯, Atha¯r al-bila¯d wa akhba¯r al-‘iba¯d (Beirut, n.d.), 530. 21 Ibn Sa‘ı¯d, Kita¯b al-jughra¯fiya¯, ed. I. al-‘Arabı¯ (Beirut, 1970), 177. 22 S. ala¯h. al-Dı¯n al-Munajjid, Jama¯l al-mar’a ‘inda al-‘arab (Beirut, 1969), 67–68. The beauty of the Byzantines is related in earlier Arabic sources, such as S. a¯‘id. al-Andalusı¯ (d. 426/1070), T. abaqa¯t al-umam (al-Najaf, 1967), 13.

Nadia Maria El-Cheikh [ 57 ] ing. Ibn Jubayr (d. 614/1217), the Andalusian traveler whose rih. la (voyage) took place between 1183 and 1185, confirms, through his personal observations, the Byzantines as supreme builders. He includes his own original descriptions but also reproduces older traditions. Having visited the Citadel in Cairo, Ibn Jubayr states that the stones of the surrounding trench are carved with a talent that makes it a lasting wonder. This was the handwork of the Byzantine captives “who are irreplaceable for such building skills.”23 Ibn Jubayr stresses these Byzantine skills again in the city of H . arra¯n, where he marvels at the main mosque and particularly one of its great domes (qubba), built by the Ru¯m.24 Upon seeing the Great Mosque of Damascus, Ibn Jubayr repeats the story found in the geography of al-Maqdisı¯ (d. 390/1000) to the effect that the Umayyad caliph ordered the Byzantine emperor in Constantinople to send him twelve thousand artisans to help with the mosaic work.25 Al-Idrı¯sı¯, who completed his work at the court of King Roger II of Sicily in 548/1154, also describes the inimitable dome of the Cordoba mosque decorated with colored and golden mosaics sent to Caliph al-Na¯.sir by the ruler of the “great Constantinople.”26 The Arabic sources are tireless, indeed, in their admiration of Byzantine craftsmanship and artwork: Al-Harawı¯ (d. 611/1215), whose descriptions are based on his own travels, states that there are in Constantinople bronze and marble statues, columns, and marvelous talismans “the like of which are not to be found in the lands of the Muslims.”27 Not only are the Byzantines great builders, their painting skills are unequaled. Ibn Jubayr, mentioning the pre-Islamic Byzantine church of Mary in Damascus, describes it as being magnificently constructed, housing “marvelous paintings that bewilder the mind and transfix the gaze. It is a marvelous spectacle.”28 Echoing a long tradition dating back to al-Ja¯h. iz. (d. 254/868) and Ibn al-Faqı¯h (d. 291/903), which saw the Byzantines as “the most skilled nation in painting,”29 al-Qazwı¯nı¯ praised the Byzantines using very similar words, stating that “they have in painting great skills; they paint the human being laughing or crying, happy or sad.”30 This view is reiterated by the physician bibliographer Ibn Abı¯ Us.aybi‘a (d. 668/1270) in ‘Uyu¯n al-anba¯’ fı¯ .tabaqa¯.t al-at.ibba¯’, which includes a reference found in earlier texts to a correspondence between the caliph of Cordoba, alNa¯.sir ‘Abd al-Rah. ma¯n (300–350/912–961), and the Byzantine emperor in the year 337/948. The correspondence was accompanied by gifts that included the book of Di. ussayn Nas..sa¯r (Cairo, 1955), 20. Ibn Jubayr, Rih. la, ed. H Ibn Jubayr, Rih. la, 233. 25 Ibn Jubayr, Rih. la, 249. Al-Maqdisı¯, Ah. san al-taqa¯sı¯m fı¯ ma‘rifat al-aqa¯lı¯m, ed. M. J. de Goeje (Leiden, 1967), 157. 26 Muhammad al-Idrı¯sı¯, Nuzhat al-mushta¯q fı¯ ikhtira¯q al-afa¯q (Leiden, 1970–84), 5:576. 27 Abu¯ Bakr al-Harawı¯, Kita¯b al-isha¯ra¯t lı¯ ma‘rifat al-ziya¯ra¯t, ed. J. Sourdel-Thomine (Damascus, 1953), 56. 28 Ibn Jubayr, Rih. la, 272. 29 The famous passage in al-Ja¯h. iz. , Kita¯b al-akhba¯r wa kayfa tas.ih. h. , text and translation by C. Pellat in JA 255 (1967): 86: “Their painters paint human beings without leaving any detail out, for the Byzantine painter is not satisfied with the painting until he turns the figure into a young man, a middle-aged man, or an old man; he then makes the figure handsome and charming and then makes it laughing or crying; the painter even manages to distinguish in his painting between on the one hand a sarcastic smile and a shy one and on the other hand between gaiety and the laughter of a delirious person.” The passage is reproduced almost identically in Ibn alFaqı¯h, Kita¯b al-bulda¯n, ed. M. J. de Goeje (Leiden, 1885), 136. 30 Al-Qazwı¯nı¯, Atha¯r, 531. 23 24

[ 58 ] Byzantium through the Islamic Prism oscorides, painted in the “amazing Byzantine style” (al-t.as.wı¯r al-ru¯mı¯ al-‘ajı¯b).31 The twelfth- and thirteenth-century sources, therefore, by a combination of personal observation and hearsay or copying, persevere in the earlier idealization of the Byzantines as master artists and craftsmen. The later sources thus reproduce positive comments concerning the Byzantines’ origins, beauty, and artistic skills found in the earlier Arabic-Islamic sources. Absent, however, from our texts is the discussion of Byzantine versus Greek learning that used to permeate earlier texts. The texts of the ninth, tenth, and eleventh centuries include a discussion concerning the role of the Byzantines in the scientific and philosophical knowledge passed on to the Muslims. At issue was the extent to which the Byzantines should be credited for the learning of the ancient Greeks. The twelfth- and thirteenthcentury sources do not emphasize these points of contention, limiting themselves to a few references that confirm the Byzantines in their role as repositories of ancient Greek knowledge. The Egyptian Ibn al-Qift.¯ı (d. 646/1248) copies the story found in the tenthcentury al-Fihrist of Ibn al-Nadı¯m, who includes an anecdote concerning Caliph alMa’mu¯n’s request of ancient books found in bila¯d al-Ru¯m.32 The Damascene historian al-Jazarı¯ (d. 793/1338) mentions an original story that has the merchant ‘Abdalla¯h describing the libraries of Hagia Sophia, where one can find “all the sciences” and books embodying the names of cities, rivers, and sources.33 Otherwise, the debate on Byzantine knowledge, a salient theme of the earlier centuries, is dropped from the later sources. Of course, much of the previous discussion was included in a general reappraisal of the “merits of various nations,” within the context of the Shu‘u¯biyya controversy opposing Persians and Arabs within the Muslim empire.34 With this literary controversy now long gone, a number of related themes no longer appear in our later texts. The real rupture from the earlier image appears, however, in conjunction with the character and morals of the Byzantines. While the earlier sources are replete with accusations of sexual immorality, lack of generosity, treachery, and women’s improper behavior, the twelfth- and thirteenth-century sources surprise us by their silence. The absence of such criticisms constitutes a significant breach, since the earlier negative image concentrated primarily on the character, morals, behavior, and customs of the Byzantines. AlQazwı¯nı¯ alone has a word on their character, and it is positive: “they are playful and joyful.”35 What a complete turnabout from the earlier Arabic sources, which, if they praised the Byzantines for their beauty and artistic skills, reserved all their animosity and vehemence for their character and morals. By eliminating such comments, the texts of 31 Ibn Abı¯ Us.aybi‘a, ‘Uyu¯n al-anba¯’ fı¯ .tabaqa¯t al-at.ibba¯’, ed. N. Rid. a¯ (Beirut, 1965), 493–94. The reference is to the De materia medica of Dioscorides. An analysis of the paintings of an Arabic Dioscorides manuscript dating from 1229 is found in R. Ettinghausen, Arab Painting (Geneva, 1977), 67–74. 32 ‘Alı¯ b. al-Qift.¯ı, Ta¯rı¯kh al-h. ukama¯’, ed. J. Lippert (Leipzig, 1903), 29–30. 33 Al-Jazarı¯, Jawa¯hir al-sulu¯k fı¯ al khula‘a¯’ wa al-mulu¯k, based on the Paris manuscript no. 6730, fols. 91–94. The translation of the pertinent sections on Constantinople is found in M. Izeddin, “Un texte arabe ine´dit sur Constantinople byzantine,” JA 246 (1958): 453–57. Jean Sauvaget has analyzed the content in La Chronique de Damas d’al-Jazari (Paris, 1949). 34 H. T. Norris, “Shu‘u¯biyyah in Arabic Literature,” in The Cambridge History of Arabic Literature: Abbasid Belles-Lettres (Cambridge, 1990), 31–47. 35 Al-Qazwı¯nı¯, Atha¯r, 530.

Nadia Maria El-Cheikh [ 59 ] the twelfth and thirteenth centuries end by providing a less variegated, more consistent, and clearly more positive image of the Byzantines. In addition to our texts’ elaboration on various facets of the Byzantine persona, twelfth- and thirteenth-century Arabic Islamic sources on Byzantium include physical descriptions of bila¯d al-Ru¯m and especially of Constantinople. Here again the surviving earlier view is juxtaposed side by side with the evolving image that reflects the new contemporary situation. Bila¯d al-Ru¯m, that is, the Byzantine Empire, continues to be highly praised. Exaggerating its geographical extent, al-Zuhrı¯ states that bila¯d al-Ru¯m extends from Constantinople in the East to Barcelona in the West.36 Ya¯qu¯t gives its frontiers: They have the Turks, Khazars, and Rus on the north and east, in the south their limits are al-Sha¯m and Alexandria, and in the west they border the sea of al-Andalus.37 Al-Qazwı¯nı¯ states that bila¯d al-Ru¯m is a great country and kingdom. The reason for its survival, he claims, lies both in its great distance from bila¯d al-Isla¯m and in the strength of its rulership. Its survival, as opposed to the disappearance of the great Persian Empire, was predicted by the Prophet Muhammad, who said: “For Persia no thrusts and no Persia after that; whereas al-Ru¯m with the many horns, as a generation passes, another one succeeds it.”38 One of the distinguishing characteristics of Bila¯d al-Ru¯m in our sources is its high fertility. Al-Qazwı¯nı¯ states that “it is a great land, most fertile,” and al-Zuhrı¯ mentions that bila¯d al-Ru¯m is a fertile land, as are all the other cities of al-Ru¯m, which abound in crops, cattle, fruit, and vineyards.39 The Aleppine historian Kama¯l al-Dı¯n Ibn al-‘Adı¯m (d. 660/1262) corroborates these authors in his statement about Aleppo: “Aleppo is a country scarce in fruits, vegetables, and wine, except for what it imports from bila¯d alRu¯m.”40 Among the most important cities are Amorium and especially Antioch, which al-Zuhrı¯ describes as one of the greatest cities of bila¯d al-Ru¯m.41 Topping the list, by far, however, was Constantinople. The Arab authors of the twelfth and thirteenth centuries called the Byzantine capital al-Qust.ant.iniyya; they are nevertheless also aware of the site’s old name, Byzantium. Ya¯qu¯t relates that Constantinople was built by one of their kings who is referred to as Buzant.¯ı.42 Both Ibn al-Athı¯r and Ya¯qu¯t mention a third name, that of Istanbul.43 Gathering their information from earlier Arabic sources, our authors underline the historical importance of the transfer of the Roman capital from Rome to Constantinople in the fourth century .. They also stressed the watershed represented by both Christianiza36

Al-Zuhrı¯, Kita¯b al-jughra¯rfiya¯, 228. Ya¯qu¯t, Mu‘jam al-bulda¯n, 3:98. 38 Al-Qazwı¯nı¯, Atha¯r, 530. 39 Al-Qazwı¯nı¯, Atha¯r, 530, and al-Zuhrı¯, Kita¯b al-jughra¯fiya¯, 228. 40 Ibn al-‘Adı¯m, Bughiyat al-h. alab fı¯ ta¯rı¯kh h. alab, ed. S. Zakka¯r (Damascus, 1408 . . ), 1:61. 41 Al-Zuhrı¯, Kita¯b al-jughra¯fiya¯, 239. 42 Ya¯qu¯t, Mu‘jam al-bulda¯n, 3:98. 43 Ibn al-Athı¯r, al-Ka¯mil fı¯ al-ta¯rı¯kh (Beirut, 1979), 1:330, and Ya¯qu¯t, Mu‘jam al-bulda¯n, 4:247. Istanbul was mentioned early on by al-Mas‘u¯dı¯ in al-Tanbı¯h wa al-ishra¯f. 37

[ 60 ] Byzantium through the Islamic Prism tion and the establishment of the empire’s capital in Byzantium and connected the two events with Constantine the Great. Ya¯qu¯t, for instance, states that Constantine the Great moved to Byzantium and built a wall around it and called it Qust.ant.iniyya, and it is their da¯r al-mulk (ruling capital) until today.44 Similarly, the geographers Ibn Sa‘ı¯d, al-Zuhrı¯, and al-Dimashqı¯ confirm that Constantinople was built by Constantine, who made it the city of the caesars.45 Constantinople occupied a unique place in the Byzantine Empire, well reflected traditionally in the Muslim sources, which continued to confirm its exceptional political, economic, and cultural importance despite the historical developments and mutations it underwent during this period. The Muslim authors of the thirteenth century were aware of the two momentous events in the recent history of the city, namely, the massacre of the Latins in 1182 and the conquest of the city by the Latins in 1204. Concerning the events of 1182, Ibn Jubayr relates the following confused and inaccurate story: The report had it that the Sovereign of Constantinople had died, leaving his kingdom to his wife and young son. But his cousin usurped the throne, killed the widow, and seized the boy. The usurper fell in love with the boy’s sister, who was famed for her beauty. Yet he could not marry her since it was forbidden for the Ru¯m to take their kinswomen in marriage. Impetuous love, blind and deafening desire . . . impelled him to take her and go to Prince Mas‘u¯d, Sovereign of Konia. . . . The two of them embraced Islam . . . and got married . . . then with the backing of Muslim armies he entered Constantinople, slaying some fifty thousand of its inhabitants. The Muslims seized Constantinople, and all its money was transported to Amir Masu¯d. . . . This conquest is one of the signs of the Hour, ashra¯.t al-sa¯‘a.46 Ibn Jubayr, here, has the Muslims conquering Constantinople in 1182. Of course, no Muslim army captured Constantinople until the Ottoman conquest of 1453, and it was the Latin inhabitants who were slaughtered by the Byzantines. As for the last phrase on ashra¯.t al-sa¯‘a, it belongs to the Muslim apocalyptic literature that developed very early on in connection with the military expeditions against the Byzantine capital in the seventh and eighth centuries. Numerous traditions going back to the Prophet Muhammad made its conquest one of the six portents of the Hour signaling the approaching end of the world.47 These traditions, found in the earliest compilations, are naturally included in the twelfth- and thirteenth-century texts, albeit in an altered form at times. The text of Jama¯l al-Dı¯n b. Wa¯.sil (d. 697/1298) provides the following significant 44

Ya¯qu¯t, Mu‘jam al-bulda¯n, 4:247. Ibn Sa‘ı¯d, Kita¯b al-jughra¯fiya¯, 184; Muhammad al-Zuhrı¯, Kita¯b al-jughra¯fiya¯, 234; and al-Dimashqı¯, Nukhbat al-dahr, 259. 46 Ibn Jubayr, Rih. la, 327–29. Andronikos was the cousin of Emperor Manuel, and he did kill his widow as well as the young emperor Alexios II. However, Andronikos did not fall in love with his cousin, although he was famous for his adventurous love affairs. The girl he married, after he became emperor, was the thirteenyear-old widow of Alexios II, daughter of Louis VII. 47 See M. Canard, “Les expe´ditions arabes contre Constantinople dans l’histoire et la le´gende,” JA 208–9 (1926): 61–121; and S. Bashear, “Apocalyptic and Other Materials on Early Muslim-Byzantine Wars: A Review of Arabic Sources,” JRAS, 3d ser., 1 (1991): 173–207. 45

Nadia Maria El-Cheikh [ 61 ] alteration: “A letter from Nu¯r al-Dı¯n to al-Mustad. ¯ı’ states: Constantinople and Jerusalem are both in the gloom of deep darkness, waiting for the crow of familiarity. God the allhigh, by his generosity will bring close the harvest of both conquests for the Muslims.”48 The addition of Jerusalem to the well-known tradition on the conquest of Constantinople, coinciding as it did with the period of the Crusades, was not accidental. This was a new development that had been absent in the earlier sources. The resurgence of the idea of the sanctity of Jerusalem in the twelfth- and thirteenth-century Muslim texts is directly linked to its occupation by the Crusaders. Jerusalem’s new situation made it the target of the military campaigns and the main focus of the political and ideological propaganda of the Muslims, rendering it also a main subject of the literature.49 The novel emphasis on Jerusalem is, indeed, also reflected in the Qur’a¯nic exegetical . abarsı¯ (d. 548/1153) explains literature of this period. The Shiite exegete Abu¯ ‘Alı¯ al-T the opening verses of su¯rat al-Ru¯m in a totally new way. These verses read: “The Greeks (al-Ru¯m) have been vanquished in the nearer part of the land; and after their vanquishing, they shall be victors in a few years . . . and on that day the believers shall rejoice in God’s help.”50 Al-T . abarsı¯ states that “the believers will rejoice for the expulsion of the Persians from Jerusalem and not for the Byzantine victory over the Persians.”51 The conquest of Jerusalem, not of Constantinople, now became the crowning achievement of the Muslim conquests. Constantinople was relegated, for a while and in certain texts, notably religious ones, to a secondary position, having lost its place d’honneur in the new discourse that reflected the new prevailing conditions in the Near East. As for the conquest of Constantinople in 1204, Ibn al-Athı¯r relates the political circumstances that led to the Latin takeover of the city, describing how the Franks took the money and gold of the churches, even the adornments on the crosses, icons, and Bibles.52 The Franks in the city, who were numerous, around thirty thousand . . . , rose, with the help of the Franks who were besieging the city, throwing fire, time and again, thus burning one-fourth of the city. They entered the city and ravaged it for three days, killing and plundering. The Byzantines were all either killed or became destitute. A group of Byzantine aristocrats sought refuge in Haghia Sophia but were followed by the Franks, and although a number of priests, monks, and abbots came out, begging them with the crosses and Bibles they were carrying, the Franks disregarded them, killing them all and plundering the church.53 The Arabic sources are thus aware of the systematic pillaging of the city. In addition to the killings and plundering of treasures, Ibn al-Athı¯r points to the most striking act, the plundering of Hagia Sophia. He also explains how Baldwin of Flanders was crowned Jama¯l al-Dı¯n b. Wa¯.sil, Mufarrij al-kuru¯b fı¯ akhba¯r banı¯ ayyu¯b, ed. J. al-Dı¯n al-Shayya¯l (Cairo, 1953), 1:253. E. Sivan, “The Sanctity of Jerusalem in Islam,” in Interpretations of Islam: Past and Present (Princeton, 1985), 75–106. Ka¯mil al-‘Asalı¯ lists 45 manuscripts of the 12th and 13th centuries dealing with the fad. a¯’il al-quds literature in his Makht.u¯.ta¯.t fad. a¯’il bayt al-maqdis: Dira¯sa¯t wa bibliu¯ghra¯phya¯ (Amman, 1981). 50 Qur’a¯n, 30:1–5; trans. in A. J. Arberry, The Koran Interpreted (Oxford, 1964). 51 . abarsı¯, Majma‘ al-baya¯n fı¯ tafsı¯r al-Qur’a¯n (Beirut, 1961), 19–25:7. Abu¯ ‘Alı¯ al-T 52 Ibn al-Athı¯r, al-Ka¯mil, 12:190. 53 Ibn al-Athı¯r, al-Ka¯mil, 12:191. 48 49

[ 62 ] Byzantium through the Islamic Prism in Constantinople, although the Ru¯m never recognized his authority, making Nicaea the provisional capital of the Byzantine Empire, headed by al-Ashkarı¯, the Lascarid.54 Ibn al-Athı¯r is not alone in grasping the consequential political developments that had occurred in the Byzantine Empire. Ya¯qu¯t likewise states that “today, [Constantinople] is in the hands of the Franks.”55 Ibn Wa¯.sil also mentions the conquest of Constantinople by the Latins: In this year (600) the al-Ifranj left their lands in great crowds and conquered it. Constantinople remained with the al-Ifranj until 660, when the Ru¯m took it back.56 Thus the Arab authors were not only aware of the conquest of Constantinople by the Latins in 1204, but appreciated the magnitude of the event. What was the Arabs’ image of Constantinople during the twelfth and thirteenth centuries, when so much had come to pass? Statements in our sources confirm earlier descriptions of the greatness and uniqueness of Constantinople. Al-Harawı¯ states that “Constantinople is a city greater than its reputation,”57 and al-Qazwı¯nı¯ proclaims that “Nothing was ever built like it, neither before nor after,” and even “if it is no longer that way . . . it remains a great city.”58 Ya¯qu¯t states that stories concerning Constantinople’s greatness and beauty abound.59 Such remarks were not unique to the Arab authors. Western authors were similarly bedazzled by the great city. Geoffrey de Villehardouin, writing on the conquest of Constantinople by the Latins in 1204, stated: “Many of our men, I may say, went to visit Constantinople, to gaze at its many splendid palaces and tall churches and view all the marvelous wealth of a city richer than any other since the beginning of time.”60 However, the Western image of Constantinople, during the period of the Crusades, contains inherent negative components well reflected in Odo of Deuil who, after praising Constantinople’s richness and glory, draws a pejorative conclusion: “In every respect she [Constantinople] exceeds moderation; for just as she surpasses other cities in wealth, so too does she surpass them in vice.”61 Unlike the Western view of Constantinople, which moves from praise to denigration, the Arabic texts of the twelfth and thirteenth centuries rarely, if ever, include negative comments. If earlier Arabic texts, dating from the ninth and tenth centuries, had accused Constantinople of arrogance and pride,62 twelfth- and thirteenth-century texts describe the city’s opulence without judgmental undertones. The only outright negative comments found in our sources are repetitions of earlier traditions which point to the city’s 54

Ibn al-Athı¯r, al-Ka¯mil, 12:192. Ya¯qu¯t, Mu‘jam al-bulda¯n, 4:247. 56 Ibn Wa¯.sil, Mufarrij al-kuru¯b, 3:160. 57 Al-Harawı¯, Kita¯b al-isha¯ra¯t, 57. 58 Al-Qazwı¯nı¯, Atha¯r, 603. 59 Ya¯qu¯t, Mu‘jam al-bulda¯n, 4:247. 60 Joinville and Villehardouin, Chronicles of the Crusades, trans. M. R. B. Shaw (New York, 1984), 76. 61 Trans. in J. P. A. Van der Vin, Travelers to Greece and Constantinople (Istanbul, 1980), 2:519. This negative view reaches its highest level after the capture of Constantinople by the Latins in 1204. See A. Ducellier, “Une mythologie urbaine: Constantinople vue d’Occident au Moyen Age,” Me´langes de l’Ecole franc¸aise de Rome: Moyen aˆge et temps modernes 96 (1984): 405–24. 62 See, for instance, the early 10th-century geographer Ibn al-Faqı¯h, Kita¯b al-bulda¯n, 146: “And so God promised her [Constantinople] suffering before the Last Judgment and said: By my power and majesty I will remove your jewels, your silk, your wine, and your bread, and will leave you with no song to sing.” 55

Nadia Maria El-Cheikh [ 63 ] wretchedness: Ibn al-‘Adı¯m cites a h. adı¯th from Abu¯ Hurayra: “Four cities in this world are from paradise: Mecca, Madina, Jerusalem, and Damascus. And four cities from fire: Rome, Constantinople, Antioch, and S. an‘a¯’.”63 This h. adı¯th is also found in the bibliographical work of the Damascene Ibn ‘Asa¯kir (d. 571/1176), although slightly modified: the four cities of paradise are Mecca, Madina, Jerusalem, and Damascus; and the four . abariyya¯, Antioch, and S. an‘a¯’.64 cities of hell are Constantinople, T Aside from this tradition, the writings on Constantinople that we find in the twelfthand thirteenth-century Arabic-Islamic texts describe the Byzantine capital either in neutral or positive terms. Politically, economically, and culturally, the Arab authors still hold Constantinople in the highest regard. Al-Idrı¯sı¯, writing in the mid-twelfth century, before the Latin conquest of the city, states: “Constantinople is prosperous, having markets and merchants, and its people are affluent.”65 This impression of the city as prosperous and economically active continues to permeate our sources in the thirteenth century, that is, after the disaster of 1204. Al-Jazarı¯ mentions the arrival of the merchant ‘Abdalla¯h b. Muhammad in Damascus in 692/1293. ‘Abdalla¯h had lived in Constantinople for twelve years during the reign of Andronikos II Palaiologos (1282–1328), having left Syria during the Mongol invasions. Asked by the father of the narrator whether it was permissible for a h. ajj, pilgrim, to establish himself among the Ifranj (here in the sense of Christian), the merchant answered: “My brother, if I were to tell you about this city, you will understand better, and you will realize that those who inhabit it have nothing to fear. They can do whatever they please and at the same time make considerable profit.”66 Asked to describe it, he said: It is a great city, comparable to Alexandria, on the seashore, and it takes one morning to cross it from end to end. There is a place as large as two-thirds of Damascus, surrounded by walls with a gate, which is reserved especially for the Muslims to live in. There is equally a similar place for the Jews. . . . There are one hundred thousand minus one churches. . . . When I asked I was told that the ruler of Constantinople had an equal number of kings at his service, each with his own church. He completed the number by building the Great Church. . . . It is one of the most considerable and marvelous buildings that we can see.67 ‘Abdalla¯h lived in Constantinople after the Byzantine restoration in 1261, when the city had regained some of its earlier opulence. Having made an extended stay in the city, he had the time and the opportunity to experience it as a reality. ‘Abdalla¯h describes a wealthy, healthy, and powerful Constantinople, cosmopolitan, with a plethora of monuments and churches, and having recovered its important economic role. Nothing in his description accuses or condemns: quite the contrary, he divulges the pride of one who 63

Ibn al-‘Adı¯m, Bughiat, 1:97. Ibn ‘Asa¯kir, Ta¯rı¯kh madinat dimashq, ed. S. al-Dı¯n al-Munajjid (Damascus, n.d.), 2.1:209–10. A third h. adı¯th names the following five cities of paradise: Jerusalem, H . ims, Damascus, Jibrı¯n, Dhofar; five from hell: Constantinople, al-Tunna, Antioch, Tadmur, and San‘a¯’ (ibid., 211). 65 Al-Idrı¯sı¯, Nuzhat al-mushta¯q (as in note 26), 5:801. 66 Izeddin, “Un texte arabe ine´dit.” 67 Ibid. 64

[ 64 ] Byzantium through the Islamic Prism has lived in a great city. ‘Abdalla¯h, like all our authors, has only praise for Constantinople. Indeed, in spite of the catastrophes that befell the Byzantine capital in the late twelfth and thirteenth centuries, it remained, in our texts, a model of affluence and material magnificence. The twelfth- and thirteenth-century Arabic-Islamic sources include a few major sites that repeatedly appear in Muslim sources. Although the authors copied one another, it is these sites, whether or not actually visited by our authors, that seem to have served as prototype symbols of Constantinople.68 The list includes the Golden Gate, the Column of Justinian, the Column of Theodosius, the Horologium, the Bronze Horses, and Hagia Sophia. According to al-Harawı¯, “the great church is Haghia Sophia.”69 Al-Qazwı¯nı¯ gives the following description: The king’s church has a golden dome and ten gates: six of gold and four of silver. The place where the king stands is 4 by 4 arms’ lengths and is ornamented with pearls and rubies. . . . All the walls of the church are covered with gold and silver. There are twelve columns, each four arms’ length, and on top of each is a statue of a human, a king, a horse, a lion, a peacock, an elephant, or a camel. Next to it is a container which, when filled, brings the water up to the statues. On Palm Sunday . . . they fill the divisions of the container with oil, wine, honey, rose water, and vinegar, which are all scented. . . . As the container is covered, the liquids flow out from the mouths of the statues.70 Similarly, al-Dimashqı¯ describes “the Great Church, where it is said that an angel resides and where lies a colossal high altar with huge doors and columns.”71 Al-Jazarı¯ also includes a description of Hagia Sophia provided by the merchant ‘Abdalla¯h: “The church is one of the most considerable and marvelous buildings that we can see. The place in which they stand for praying is surrounded by grills. . . . On the walls of this church are represented all the cities of the world and all the crafts. When one of them wants to choose a craft for his son, he takes him to the walls and shows him the crafts.”72 Another main monument is the column of Justinian. Ibn Jubayr, al-Qazwı¯nı¯, Ya¯qu¯t, and al-Jazarı¯ all provide lengthy descriptions of the column of Justinian which are essentially the same version. Ibn Jubayr states that on top of a bronze column is placed the equestrian statue of Constantine. Constantine has his right hand upward and his palm opened as if pointing toward the realm of Islam. In his left hand he holds a globe. “Opinions vary concerning this monument. Some think that the globe is a talisman with power 68 Ian Richard Netton has called a number of sites that appear in Muslim descriptions of Cairo “the medieval tourist paradigm.” I. R. Netton, “Tourist Adab and Cairene Architecture: The Medieval Paradigm of Ibn Jubayr and Ibn Bat..tu¯.ta,” in Literary Heritage of Classical Islam: Arabic and Islamic Studies in Honor of James Bellamy, ed. M. Mir (Princeton, 1993), 275–84. 69 Al-Harawı¯, Kita¯b al-isha¯ra¯t, 56. 70 Al-Qazwı¯nı¯, Atha¯r, 603–4. 71 Al-Dimashqı¯, Nukhbat al-dahr, 227. 72 Al-Jazarı¯ in Izeddin, “Un texte arabe ine´dit.” ‘Abdalla¯h was mistaken about the location of the libraries. See R. Janin, Constantinople byzantine (Paris, 1950), 161–62.

Nadia Maria El-Cheikh [ 65 ] to inhibit the enemy from invading their country; others say that the globe holds an inscription that says: I have possessed the world and I held it in my hand like this globe; I left it without carrying off anything.”73 Al-Harawı¯ similarly talks about the talismanic nature of the statue: “in his hand is a talisman that prevents the enemy from invading the country.”74 One other monument mentioned relatively often is the column of Theodosius. AlHarawı¯ includes its description in his work: “There is a white marbled column in the market . . . entirely covered with three-dimensional sculptures of admirable skill.” It is surrounded by a grill that includes a talisman. If one climbs to the top of the column, he can have a panoramic view of the city in its entirety. “In my kita¯b al-‘aja¯’ib I will . . . talk about the veneration the inhabitants of this land profess for it and for the figures that cover it.”75 Although al-Harawı¯ mentions the talismanic element of this column, he does not explain what kind of power it holds. The Byzantine texts seem to imply that the column announced the future of the city.76 Al-Qazwı¯nı¯ gives some details of the Horologium, the clock that so impressed the Arab authors: “In the lighthouse of Constantinople there is a Horologium which is made up of twelve doors, each representing an hour. At every hour, one of the doors opens and a statue comes out. . . . The Byzantines say that it is the work of the wise Bı¯na¯s” (Apollonius).77 Al-Qazwı¯nı¯ also attributes to Apollonius the creation of the three bronze horses, a talisman located at the gate of the imperial palace to prevent the horses of the city from making noise or neighing.78 The majority of the monuments and statues mentioned in our sources are endowed with talismanic or magical power. The talismanic objects found in Constantinople are not unique to it. The talismanic protection of antique cities is a theme of medieval Arabic literature. Al-Qazwı¯nı¯ states in the introduction to his geography that strange talismans were created by the wise philosophers, h. ukama¯’, for the defense of cities,79 and various earlier authors such as Ibn al-Faqı¯h and al-Mas‘u¯dı¯ (d. 345/956) mention several talismans in various ancient Near Eastern cities.80 Indeed, Rasa¯’il ikhwa¯n al-safa¯’ include the science of talismans,81 and the tenth-century Ibn al-Nadı¯m wrote at the beginning of the section dealing with books of magic that “one group of philosophers and servants of the stars assert that they have talismans based on [astronomical] observations.” Ibn alNadı¯m further states that “this art is divulged openly among the philosophers” and goes on to give a biographical entry of Apollonius the Wise, “one of the people of Tyana, in 73 Ibn Jubayr, Rih. la, 49. For similar descriptions see Al-Qazwı¯nı¯, Atha¯r, 605, and Ya¯qu¯t, Mu‘jam al-bulda¯n, 4:348–49. On the column, see Janin, Constantinople byzantine, 78–80. 74 Al-Harawı¯, Kita¯b al-isha¯ra¯t, 49. 75 Al-Harawı¯, Kita¯b al-isha¯ra¯t, 49. See Janin, Constantinople byzantine, 84–85. 76 Janin, Constantinople byzantine, 85. 77 Al-Qazwı¯nı¯, Atha¯r, 605–6. 78 Al-Qazwı¯nı¯, Atha¯r, 606. 79 Al-Qazwı¯nı¯, Atha¯r, 8. 80 See G. Calasso, “Les remparts et la loi, les talismans et les saints: La protection de la ville dans les sources musulmanes me´die´vales,” BEODam 44 (1993): 83–104. 81 J. C. Burgel, “On Some Religious, Cultural and Social Implications of the Occult Sciences in Medieval Islam,” in The Feather of Simurgh (New York-London, 1988), 27–52.

[ 66 ] Byzantium through the Islamic Prism the Byzantine territory” and “the first to initiate speech about talismans.”82 Apollonius was referred to, in various early Arabic-Muslim sources, as .sa¯h. ib al-t.ilas.ma¯t, or “father of talismans.” In fact, the reputation of Apollonius was such that several Arabic works dealing with astrology, alchemy, and magic, among others, have been falsely attributed to him.83 It was the connection between the “Byzantine” Apollonius and talismans that probably led to the prevalent belief in the Islamic sources of the excessive presence of talismans in the Byzantine capital.84 This assumption continued unabated and even increased in the texts of the twelfth and thirteenth centuries, which managed to assign talismanic power to a relatively larger number of monuments. This has partly to do with the proliferation of books on cosmography such as those of al-Qazwı¯nı¯ and al-Dimashqı¯, which, with their wondrous elements, became popular starting in the late twelfth century. In addition to all these typically Byzantine monuments, the geographers of the twelfth and thirteenth centuries mention the Islamic monuments of the city. They reproduce the legend of Abu¯ Ayyu¯b al-Ans.a¯rı¯, a Companion of the Prophet, who participated in the expedition against Constantinople in the year 48–49/668–9. Al-Qazwı¯nı¯ mentions his tomb, beneath the walls of Constantinople, and says that its soil is venerated by the Byzantines, who go there in their prayer for rain during drought. Ibn al-Athı¯r and alDimashqı¯ also mention the tomb of Abu¯ Ayyu¯b who fought in Badr and was with the fourth caliph, ‘Alı¯ b. Abı¯ T . a¯lib, in S. iffı¯n.85 The second important Islamic monument mentioned in our sources is the mosque connected with the name of Maslama, the leader of the most important expedition against Constantinople, in 97–99/715–717. This mosque is also mentioned in conjunction with the restoration of Constantinople to the Byzantines in 660/1261. Ibn Jubayr states that the Byzantine emperor rebuilt Maslama’s mosque in 455/1263. Ibn ‘Abd alZ . a¯hir (d. 692/1291) states that while the ambassador from Egypt was touring Constantinople with al-Ashkarı¯, the Laskarid, they came to the mosque built by Maslama. Saladin had wanted at one time to reconstruct this mosque, but the Byzantines had refused. According to Ibn ‘Abd al-Z . a¯hir, God postponed this deed, so that it would be God’s . a¯hir, and a glory for his state.86 reward for al-Z Thus, as late as the late thirteenth century, and despite the recent destruction of the city by the Latins, Constantinople’s symbolic importance had not diminished. Rebuild82 Ibn al-Nadı¯m, al-Fihrist (Cairo, n.d.), 443–44, 448; trans. in B. Dodge, The Fihrist of al-Nadı¯m (New York, 1970), 2:726. 83 See Pseudo-Dionysius, Kita¯b sirru al-khalı¯qa wa .sin‘at al-t.abı¯‘a: Kita¯b al-‘ilal, ed. U. Weisser (Aleppo, 1979), 10. 84 This, and the fact that Byzantine authors down to the 12th century continued to describe the talismans that Apollonius erected in various cities. See W. L. Dulie`re, “Protection permanente contre des animaux nuisibles assure´e par Apollonius de Tyane dans Byzance et Antioche: Evolution de son mythe,” BZ 64 (1970): 247–77. 85 Al-Qazwı¯nı¯, Atha¯r, 606; Ibn al-Athı¯r, al-Ka¯mil, 3:93; and al-Dimashqı¯, Nukhbat al-dahr, 227. 86 Ibn ‘Abd Al-Z . a¯hir, al-Rawd. al-za¯hir fı¯ sı¯rat al-malik al-Z . a¯hir, ed. ‘Abd al-‘Aziz al-Khuayt.ir (Riyadh, 1976), 129, 131. For a very similar passage, see M. M. Taher, “La mosque´e de Constantinople a` l’e´poque byzantine d’apre`s un manuscrit arabe (BN de Paris),” Byzantiaka 11 (1991): 117–27.

Nadia Maria El-Cheikh [ 67 ] ing this mosque in Constantinople brought glory and prestige to the Muslim ruler and symbolized the extent of his power and influence. Equally important is the development of amicable relations between the restored Byzantine Empire and the Egyptian state. Faced with the same enemies, the Byzantine-Egyptian alliance served as a counterweight to Western, Mongol, and Turkish threats. The good relations extended into the reign of Sultan Qala¯wu¯n, who exchanged sworn undertakings in 1281 with Emperor Michael VIII (1259–82) in which they agreed to maintain love and friendship without limit of time.87 A few sources refer to two additional Muslim tombs in Constantinople. Both alDimashqı¯ and al-Harawı¯ mention a tomb for a descendant of al-H . ussayn, son of ‘Alı¯ b. 88 Abı¯ T . a¯lib and grandson of the Prophet Muhammad, while al-Zuhrı¯ mentions the tomb of Abu¯ ‘Ubayda b. al-Jarra¯h. , a Companion of the Prophet, who played a pivotal role in the history of early Islam and who, furthermore, belongs to al-‘ashara al-mubashshara, the ten believers to whom paradise was promised. Al-Zuhrı¯ states that Abu¯ ‘Ubayda b. alJarra¯h. died during Maslama’s expedition and was buried in front of the walls of Constantinople and that to the present day the Ru¯m light candles at his tomb.89 Although AlZuhrı¯ may be confusing the tombs of Abu¯ Ayyu¯b and Abu¯ ‘Ubayda,90 the proliferation of the names of such prominent and revered personalities around Constantinople testifies to the high esteem and quasi-veneration in which Constantinople was held in our sources. Manuela Marin has mentioned the importance of the presence of Islamic monuments in Constantinople in providing a symbolic possession of the city.91 The existence of Islamic monuments linked to major figures of Islamic mythology Islamizes and sacralizes the Byzantine capital, providing Constantinople with the semblance of a Muslim genealogy.92 It is in fact the presence of these burials within the walls of the city that permitted the inclusion of Constantinople in a repertoire of places of pilgrimage like that of al-Harawı¯. Constantinople is clearly at the center of the picture. The great majority of references to the Byzantine Empire are really references to Constantinople. The Arab authors concentrated on the marvels of the city and on its Islamic monuments: talking about Constantinople was talking about a number of monuments that subsumed the entire city. Thus the idea of Constantinople and the fascination felt about it remained at the center of our literature. The Byzantine Empire continued to be defined by its capital. The Arab authors are moved by Constantinople’s diversity, opulence, the many marvels to which they impute talismanic powers, and, of course, the monuments endowed with Muslim religious connotations. In their descriptions, the city is harmonious, orderly, and 87 For the text of the treaty in English, see P. M. Holt, Early Mameluk Diplomacy (1260–1290): Treaties of Baybars and Qalawu¯n with Christian Rulers (Leiden, 1995), 122–28. 88 Al-Dimashqı¯, Nukhbat al-dahr, 227, and al-Harawı¯, Kita¯b al-isha¯ra¯t, 56. 89 Al-Zuhrı¯, Kita¯b al-jughra¯fiya¯, para. 117. 90 M. Marin, “Ru¯m in the Works of Three Spanish Muslim Geographers,” Graeco-Arabica 3 (1984): 109–17. 91 M. Marin, “Constantinopla en los geografos Arabes,” Erytheia 9.1 (1988): 49–60. 92 G. von Grunebaum, “The Sacred Character of Islamic Cities,” in Me´langes Taha Hussein (Cairo, 1962), 25–37; reprinted in Islam and Medieval Hellenism: Social and Cultural Perspectives (London, 1976).

[ 68 ] Byzantium through the Islamic Prism organized. The authors do not talk about dilapidated, shabby Constantinople. It is as if the city survived all its catastrophes intact.93 Constantinople succeeded in maintaining its reputation and prestige despite the fact that this period witnessed both its conquest by the Latins and the development of fad. a¯’il al-quds (merits of Jerusalem) literature.

Conclusion This investigation has revealed both continuities and changes in the twelfth- and thirteenth-century Muslim representation of the Byzantines. The texts persist in extolling Byzantine beauty and artistic skills, and the fascination that the Arab writers had for Constantinople continues unabated, although the city underwent decisive mutations announcing its decline.94 At the same time, the old philosophical and moral slurs accompanied by empty rhetorical posturing have no longer any place in our texts. The Byzantines are not the target of accusations, for the Crusaders, with their “vulgarity” and their terrifying ways, were deemed unworthy inferiors in comparison with the sophisticated and refined Byzantines. The prince of Shayzar, Usa¯mah b. Munqiz. , wrote in his memoirs: “When one comes to recount cases regarding the Franks, he cannot but glorify Allah and sanctify him, for he sees them as animals possessing the virtues of courage and fighting but nothing else.”95 If, earlier on, the Muslims had been shocked by Byzantine “immorality,” they now seem to be more deeply shocked by Frankish “immorality” and behavior. Thus, whereas in the earlier image, the character, morality, and practices of the Byzantines were a main subject of Arabic-Islamic texts, the moral character of the Byzantines, with personal details and blatant judgments, are now absent. The Homo byzantinus seems to have been replaced by the Crusader, and the rhetoric on Byzantium, designed for home consumption, now shifted to the Franks. The omission of negative descriptions, with respect to the Byzantines, particularly on the more abstract moral plane, is itself a sign of the dawn of a new age. The new wave of crusading conquests forced Muslim authors to view their traditional Byzantine foes, now greatly weakened, in a new perspective. The new political and military realities transformed the relations between Byzantium and Islam, hence the new appreciation. Once the Muslims understood the religious and psychological motivations of the Franks, they were quickly able to distinguish between al-Ru¯m and al-Ifranj. In addition, the new political alliance between the Byzantines and Saladin,96 and later on between the Byzantine and Mamluk states, redirected Muslim animosity toward the Franks, while the Byzantines came to be viewed in much friendlier terms. 93

In fact, the appearance of the city had changed, even before the sack of 1204. See P. Magdalino, The Empire of Manuel I Komnenos, 1143–1180 (Cambridge, 1993), 112–13. 94 A. Ducellier, “Apoge´e et de´clin d’une capitale,” in A. Ducellier and M. Balard, eds., Constantinople: 1054– 1261 (Paris, 1996), 18–38. 95 P. Hitti, An Arab-Syrian Gentleman and Warrior in the Period of the Crusades: Memoirs of Usa¯mah Ibn-Munqidh (London, 1987), 161. 96 C. Brand, “The Byzantines and Saladin, 1184–1192: Opponents of the Third Crusade,” Speculum 37 (1962): 167–81.

Nadia Maria El-Cheikh [ 69 ] In addition to being influenced by events that conditioned their time and shaped the resulting perceptions, the texts continue as well a tradition inherited from previous ages. The presence of the Crusaders in the Near East was no doubt the major influence determining the selection of the information on Byzantium and the resulting diversion from the previous pattern. The sources reiterate earlier positive themes, discontinue a few negative motifs, and incorporate their new perceptions colored and accentuated by different filters and prisms affecting the traditional image. Our sources not only perceive Byzantium with more indulgence but also show positive appreciation. Our sources, influenced by contemporary developments, reveal a prismatic view that conveys the complexities of the new age. American University of Beirut

The Crusaders through Armenian Eyes Robert W. Thomson The impact of the Franks on various aspects of Armenian life and culture in the principality, then kingdom, of Cilician Armenia has long attracted attention. Significant changes in Armenian social institutions and religious attitudes were brought about by direct contacts with Western Europe in the twelfth and thirteenth centuries, and these have been studied with varying degrees of thoroughness over many years. One less tangible aspect of Crusader influence, however, has not attracted the same attention as developments in art, religion, language, and other facets of life in Cilicia.1 That is the way in which Armenians, both those in Cilicia who came into personal contact with the Crusaders and those in Greater Armenia who experienced the Crusades less directly, interpreted the arrival of the Frankish armies. How were the Crusades fitted into an Armenian worldview? By the time of the Crusades there was a long-standing tradition of historical writing in Armenian, going back to the fifth century. Over a period of more than six hundred years, Armenian writers had had to come to terms with numerous vicissitudes.2 Even if the arrival of the Crusaders was generally interpreted in positive terms—rather than as yet another calamity, like the arrival of the Turks—nonetheless, it caused a fairly radical upset in the eastern Mediterranean which merited more than a casual reference. Previous upheavals, like the Muslim advance of the seventh century, or the arrival of the Turks in the eleventh, had been placed by Armenian writers in a broad framework.3 1 There is an excellent and up-to-date general bibliography in P. Halfter, Das Papsttum und die Armenier im fru¨hen und hohen Mittelalter, Forschungen zur Kaiser- und Papstgeschichte des Mittelalters, Beihefte zu J. F. Bo¨hmer, Regesta Imperii, 15 (Cologne-Weimar-Vienna, 1996). For linguistic change add J. J. S. Weitenberg, “Literary Contacts in Cilician Armenia,” in East and West in the Crusader States, Orientalia Lovaniensia Analecta 75 (Louvain, 1996), 63–72; and for developments in miniature painting, S. Der Nersessian, Miniature Painting in the Armenian Kingdom of Cilicia, 2 vols., DOS 31 (Washington, D.C., 1993), and the dissertation of Helen Evans, “Manuscript Illumination at the Armenian Patriarchate in Hromkla and the West” (New York University, Institute of Fine Arts, 1989). 2 For early Armenian historical writing see J.-P. Mahe´, “Entre Moı¨se et Mahomet: Re´flexions sur l’historiographie arme´nienne,” REArm 23 (1992): 121–53. For details of editions and translations of Armenian historians, see R. W. Thomson, A Bibliography of Classical Armenian Literature to 1500 .., CC (Turnhout, 1995). 3 Sebe¯os, writing near the end of the 7th century, is the first Armenian historian to discuss the Islamic conquests; see commentary and discussion in R. W. Thomson, J. Howard-Johnston, and T. Greenwood, The Armenian History Attributed to Sebeos, Translated Texts for Historians 31 (Liverpool, 1999). For the Armenian response to the Turkish invasions, see M. Canard and H. Berbe´rian, Aristake`s de Lastivert: Re´cit des Malheurs de la Nation arme´nienne, Bibliothe`que de Byzantion 5 (Brussels, 1973).

[ 72 ] The Crusaders through Armenian Eyes Historians had looked backwards to seek out links with past events in Armenian history (real or imagined), and forwards with visions of deliverance from present foes and the inauguration of an era of peace and well-being. When in turn the Mongols came upon the scene, the Armenians already possessed a framework of interpretation that could be adjusted and adapted to these new circumstances. How, then, were the Crusades integrated into an Armenian scheme of things? Interpretations took some time to emerge. Until the late twelfth century there is a surprising paucity of references to the Crusaders and a general lack of interest in the West as such. Some Armenians played a role in the new developments, which in Cilicia completely changed the character of traditional Armenian life. But the majority, those in Greater Armenia, had little direct experience of the Crusaders. So the “Armenian eyes” to which the title of this paper refers are not necessarily those of actual witnesses to events in Cilicia and northern Syria. Here I shall pay more attention to those Armenians who reflected on the Crusades and tried to integrate them into their own approach to the past, than to Armenian descriptions of events in Syria and Palestine. But first a few words on the immediate reaction of Armenians before the historians had had time to consider these broader issues. Since Armenian scribes had the welcome habit of adding detailed colophons to the manuscripts they were copying, we have some interesting comments on the arrival of the Crusaders which are contemporary with the events described.4 The first Armenian to describe the Crusaders was writing in 1098 on the Black Mountain near Antioch. In his colophon to a collection of canon law,5 a certain John describes the coming of “the western nation of heroes” who had crossed the sea in order to expel the foreign tyrants who held in dishonor the holy [places] and priests of God. The Westerners are not specifically called “Crusaders,” but the scribe does note that “they had set before themselves the sign of Christ’s Cross.” The following year the scribe Aaron, copying a Gospel in Alexandria,6 dates his book to the patriarchate of Lord Gregory—that is, Grigor II Vkayase¯r, who was then residing in Egypt7 —and to the empire of the valiant race of Romans, “who are the Franks, who

4 For a general overview of Armenian colophons and their characteristics, see A. K. Sanjian, Colophons of Armenian Manuscripts, 1301–1480: A Source for Middle Eastern History, Harvard Armenian Texts and Studies 2 (Cambridge, Mass., 1969). The colophons quoted below are taken from the collections published by A. S. ˇ B darer (Erevan, 1988), for the 5th to 12th centuries; and Mat‘evosyan, Hayeren Jer˙agrac’ Hisˇatakaranner, E–Z G. Yovsep‘ean, Yisˇatakarank‘ Jer˙agrac‘ (Beirut, 1951), down to 1250. 5 Mat‘evosyan, 117–19, Yovsep‘ean, 259–64. The MS is now in New Julfa, no. 542. This famous colophon was first brought to general attention by P. Peeters, “Un te´moignage autographe sur le sie`ge d’Antioche par les croise´s en 1098,” Miscellanea historica in honorem Alberti de Meyer, vol. 1, Recueil de travaux d’histoire de l’Universite´ de Louvain, 3d ser., 22 (Louvain, 1946), reprinted in his Recherches d’histoire et de philologie orientales, SubsHag 27 (Brussels, 1951), 2:164–80. 6 Mat‘evosyan, 119–21. The MS, a Gospel, is now in Erevan, Matenadaran no. 288. 7 This Gregory, son of Gregory Magistros Pahlavuni, had been consecrated catholicos in 1065, but he left his administrative duties to a coadjutor and devoted his life to the study of hagiography. “Vkayase¯r” means “lover of martyrs.” He spent much of his life traveling in search of new texts to translate into Armenian. See A. Kapoian-Kouymjian, “Le Catholicos Gre´goire II le Martyrophile (Vkayase¯r) et ses pe´re´grinations,” Bazmavep 132 (1974): 306–25; and S. B. Dadoyan, The Fatimid Armenians, Islamic History and Civilization, Studies and Texts 18 (Leiden, 1997), 85–90.

Robert W. Thomson [ 73 ] had come at God’s instigation.” Aaron describes the battle for Antioch and notes more briefly the capture of Jerusalem, when through the power of Christ, a host of the impious had been put to the sword. Less flattering to the Franks is a later colophon of 1130,8 whose author dates his work both to the Armenian era and to the “tyranny” of the Franks under Count Joscelin. Here “tyranny” [br˙nakalut‘iwn] is the same word as is used of Saladin’s rule in Egypt, Damascus, and Edessa in a colophon of 1183.9 The Crusader kings are occasionally named by scribes. Thus Frederick Barbarossa is the Alaman, while Henry VI becomes “Heri, emperor of old Rome.”10 Their country of origin is rarely specified, though in 1198 a scribe in the monastery of Hałbat in northern Armenia mentions kings who had come from the regions of Italy, and even across the Atlantic Ocean, that is, from England.11 However, in Nerse¯s Sˇnorhali’s Lament on the Fall of Edessa, we read that the Latins came from Dalmatia.12 “Dalmatia” for Italy was quite common in Armenian, and Latin was commonly called “Dalmatian.” According to the Armenian version of the Gospels, the inscription over the cross of Christ was written in Hebrew, Greek, and “Dalmatian.”13 Armenian authors generally refer to the Westerners as Franks, Romans, or Latins. The equivalent of the term “Crusader” is hardly ever applied to them. In his Lament for the Fall of Edessa, Nerse¯s Sˇnorhali does refer in a general way to the “worshipers of the holy Cross,” who failed to come to the support of Count Joscelin II.14 But only once does the Armenian word “Crusader” appear. A Bible written in 1270 includes a colophon that is taken from the end of Nerse¯s of Lambron’s Commentary on the Twelve Prophets. It is there noted: “Jerusalem fell at the completion of ninety years of the rule of the Romans. . . . Then the king of Rome who is German, and other kings, came to recover Jerusalem as xacˇ‘e˘nkalk‘”—which means “those who had taken up the cross.”15 A similar expression, xacˇ‘akir, “one who bears the cross,” is used in Armenian from the fifth century.16 It is applied to monks or ascetics and is a calque on the Greek stauropho´ros. That is the modern 8 Mat‘evosyan, 158. The MS, a copy of Ełisˇe¯’s History, is now in Erevan, Matenadaran no. 4515. It was written in the region of Tluk‘ at the monastery of St. Paul; the site is not known and does not appear in M. Thierry, Re´pertoire des monaste`res arme´niens, CC (Turnhout, 1993). 9 Mat‘evosyan, 235–36. The MS, a Gospel, is now in Erevan, Matenadaran no. 3133. 10 Mat‘evosyan, 301–2. The MS, a copy of the Revelation of John, is now in Erevan, Matenadaran no. 10480; although it is dated to 1286, it reproduces the comments of Nerse¯s of Lambron, who had translated the Revelation from Greek in 1198. 11 Mat‘evosyan, 296. The MS, a copy of the Commentary on Jeremiah by Mxit‘ar Gosˇ, is now in Erevan, Matenadaran no. 2606. 12 I. Ke´chichian, Nerse`s Sˇnorhali: La complainte d’Edesse, Bibliotheca Armeniaca 3 (Venice, 1984), 41, being line 518 of the poem. There are many editions of the Armenian text; see Thomson, Bibliography, 179. 13 Luke 23:38; John 19:20. For “Dalmatia” cf. the History of Agat‘angełos, para. 874, where St. Gregory and King Trdat arrive in Rome to greet Emperor Constantine: “By land and sea they hastened on their way until they arrived in the empire of the Italians, in the land of the Dalmatians, in the royal capital of Rome.” Armenian text with facing translation in R. W. Thomson, Agathangelos: History of the Armenians (Albany, N.Y., 1976), 409. This legendary encounter will play a prominent role in later Armenian tradition; see further below. 14 Ke´chichian, La complainte, 33, line 393. 15 Mat‘evosyan, 249. The MS, a Bible, is now in Erevan, Matenadaran no. 345. 16 The earliest example is in Koriwn’s Life of Masˇtoc‘, chap. 4. Armenian text in M. Abełyan, Vark‘ Masˇtoc‘i (Erevan, 1941; repr. Delmar, N.Y., 1985), 38; trans. in G. Winkler, Koriwns Biographie des Mesrop Masˇtoc‘, OCA 245 (Rome, 1994).

[ 74 ] The Crusaders through Armenian Eyes Armenian term to render “Crusader,” but except for this colophon I have not found it in medieval authors. Colophons do not provide a very solid basis for generalizations, given the haphazard survival of Armenian manuscripts and the fact that only those in major collections have been published. Nonetheless, it is noticeable that scribes pay relatively little attention to the Crusaders. The historians, however, began to see broader significance in the arrival of the Franks than the instigation of God and the punishment of the infidels. Matthew of Edessa, whose own work goes down to 1136, is the first to offer an interpretation. He bases it on the Bible and on earlier Armenian traditions. The arrival of these Franks was in fulfillment of the prophecy of the Armenian patriarch Nerse¯s the Great at the time of his death and of Daniel’s vision of the beast [Daniel 7:7]. The Continuator to Matthew even claims that the subsequent recapture of Jerusalem by the infidels was also predicted by Nerse¯s.17 This Nerse¯s lived in the fourth century, but his first biographer gives no indication that he was possessed of such extraordinary prophetic powers. According to the author of the Epic Histories or Buzandaran (widely, if incorrectly, known as P‘awstos), Nerse¯s had rebuked king Arsˇak for his evil ways and bade him redeem himself from the wrath of God, so that “this unfortunate realm of Armenia may not perish on account of you! For I have seen a vision that perdition and destruction are advancing on this doomed realm of Armenia.”18 Our fifth-century author had in mind the division of Armenia into Roman and Iranian sectors, which occurred in 387, some fourteen years after Nerse¯s’ death. But naturally this hint was enough to encourage later writers to update the prediction. In the tenth century an elaborate Life of Nerse¯s was composed; it is attributed to a priest Mesrop from Vayoc‘ Jor in Siunik‘.19 In turn this Life was subjected to later revisions. Unfortunately we possess no study of the progressive recensions of the text, but it was from a version of the later Life of Nerse¯s, not from the fifth-century Epic Histories, that Matthew of Edessa took his prediction, which had already been removed from its place in the original narrative and put to the time of Nerse¯s’ death. According to this later form of the prediction, after fifty years the priestly line of St. Gregory and the royal line of the Arsacids would come to an end. This refers to the abolition of the monarchy in 428. One hundred and fifty years after that the Persians will capture Jerusalem and take the Cross captive, which is not very exact reckoning in order to bring us to 614. When the Cross is returned the Greeks will no longer rule over Jerusalem; the Ishmaelites will replace them. The latter will subject the Greeks to tribute until the time of the coming of the valiant race of the Romans, who are called Franks,

17 Matthew of Edessa, 2.109 and 124. Armenian text, Patmut‘iwn Matt‘e¯osi Ur˙hayec‘woy ( Jerusalem, 1869), 306, 325; trans. in A. E. Dostourian, Armenia and the Crusades: The Chronicle of Matthew of Edessa (Lanham, Md., 1993). Continuator, 41; Armenian text, 539. 18 Buzandaran, 4.13. Armenian text, P‘awstosi Buzandac‘woy Patmut‘iwn Hayoc‘ (Venice, 1933); trans. in N. G. Garsoı¨an, The Epic Histories (Buzandaran Patmut‘iwnk‘), Harvard Armenian Texts and Studies 8 (Cambridge, Mass., 1989). 19 Patmut‘iwn srboyn Nersisi Part‘ewi, Sop‘erk‘ 1 (Venice, 1853); trans. in J.-R. Emine, “Ge´ne´alogie de la famille de saint Gre´goire et vie de saint Nerse`s,” in Collection des historiens anciens et modernes de l’Arme´nie, ed. V. Langlois, vol. 2 (Paris, 1969), 21–41.

Robert W. Thomson [ 75 ] who will capture Jerusalem and remove the subjection of the Greeks. As for Armenia, numerous afflictions will befall the country, including the oppression of the “Archers.” This is an expression usually associated with the Mongols, but it is also used of the Seljuks.20 Eventually salvation will come from the Romans. Matthew of Edessa’s language is so vague that it is not clear whether in his mind Daniel’s fourth beast represents the Crusaders, who will devour and trample the enemies of Christ, or whether it stands for Muslim rule, now to be destroyed. The first Armenian to identify the Muslims with Daniel’s fourth beast was Sebe¯os who, writing some time soon after 660, tried to come to terms with this new power in the East.21 Sebe¯os, however, has no optimistic forecast of deliverance from these new disasters. He was unacquainted with the prophecies contained in the Apocalypse of Pseudo-Methodius, which was composed in Syriac in the last decade of that century. Although an Armenian version of that Apocalypse has been attributed to the eighthcentury Armenian scholar Step‘annos of Siunik‘,22 only much later was the key motif of the last things incorporated into Armenian thought. One of the later adaptors of the Life of Nerse¯s not only interpreted the kingdom of the new Romans, the Crusaders, as the fourth kingdom of Daniel’s vision; he also explained that it would be destroyed by AntiChrist, preceding the final coming of Christ in glory. Matthew of Edessa, however, was not influenced by this later version of the Life of Nerse¯s. He does not interpret the Crusaders as harbingers of the last things. An even older Armenian tradition involving the West, long predating the Crusades, was now associated with these new circumstances. This is the supposed meeting between the newly converted kings Constantine and Trdat of Armenia, who were accompanied by their patriarchs, Silvester and Gregory the Illuminator. Whether the original story is an adapted reminiscence of the visit to Nero’s court of Trdat I, or whether it reflects some subsequent treaty between Rome and Armenia, is unclear.23 In any event, with certain variations, it is enshrined in Armenian literature from the earliest texts of the fifth century. It may be cited in a military context to recall earlier Roman assistance to Armenia, or in an ecclesiastical context to demonstrate the orthodoxy of Armenian faith and practice, mutually confirmed by these Christian monarchs even before the Council of Nicaea. The arrival of the Crusaders, normally called “Romans” by Matthew and most 20 For the name see R. P. Blake and R. N. Frye, “History of the Nation of the Archers (the Mongols) by Grigor of Akanc‘,” Harvard Journal of Asiatic Studies 12 (1949): 269–399, esp. 384 n. 1. 21 Sebe¯os, 177. In this edition of the Armenian text, Abgaryan has changed the order of the last section as found in the MSS, putting the prediction of disasters to the very end of the History. According to the MSS, it seems that Sebe¯os ended his first draft with a prophecy of devastation, but later returned to his text in order to add a brief description of the Muslim civil war, ending with the triumph of Muawiya in 661. The original pessimistic tone was thus modified: Sebe¯os regarded the imposition of the new re´gime as bringing a de facto settlement. See the English translation, as in note 3 above. 22 But see further below at notes 35 and 54. 23 The story first appears in Agat‘angełos; see note 13 above. But in the earliest version of the visit, the bishop of Rome is called Eusebius. Silvester does not enter the Armenian tradition until Movse¯s Xorenac‘i (of disputed date, but probably 8th century) and the Arabic version of the V recension of Agat‘angełos. For the latter see G. Garitte, Documents pour l’e´tude du livre d’Agathange, ST 127 (Vatican City, 1946). The evidence of these and later texts has been collected in R. W. Thomson, “Constantine and Trdat in Armenian Tradition,” Acta Orientalia 50 (1997): 201–13.

[ 76 ] The Crusaders through Armenian Eyes other Armenian historians,24 immediately made this old tradition topical. The story of the meeting of Trdat with Constantine underwent elaborate developments in Cilicia; and at the same time a major theme of the Pseudo-Methodian Apocalypse was woven into the tale, namely, the introduction of the Romans into predictions of the coming of Anti-Christ and the last things. In Greater Armenia during the twelfth century the Crusaders were still not viewed in prophetic terms. The historian Samuel of Ani—here relying on the lost Chronicle of John the Deacon, also of Ani—briefly describes the perfidious attitude of Alexios II, “the son of Satan,” toward the Crusaders.25 These Romans had come to seek vengeance for the troubles suffered by the Christians at the hands of the Scythians (Turks), Persians, and Arabs. In Samuel’s time, Ani was under Georgian control, and although he mentions the capture of Jerusalem and refers to later events in Cilicia and Palestine, in the broader Armenian perspective the Crusades were not the main focus of his attention. There is certainly no suggestion that the Crusades fitted into a grand scheme of things that had been dimly apprehended by Patriarch Nerse¯s many centuries earlier. In Cilicia, on the other hand, an unknown author of the late twelfth century did develop an elaborate scenario combining earlier Armenian traditions with apocalyptic ideas from Pseudo-Methodius and other writers, in which the arrival of the Crusaders plays a major role.26 Disguised as a homily, this document was ascribed to Epiphanius of Salamis (to whom were also attributed the apocryphal Lives of the Prophets, for example) rather than to an Armenian authority such as Nerse¯s or his grandson Sahak.27 With many reminiscences of the fall of Ani to the Byzantines and the arrival of the Turks, the homily refers back to the visit of Trdat and Gregory to Rome at the summons of Constantine.28 However, the author adds the interesting information that the two kings and their patriarchs proceeded to visit Jerusalem, where they divided the holy places between the two parties. The holy sites are not here identified, though later versions of the story will be very precise on this point. Coming to events of the future, the author predicts that in due time a new Constantine and a new Trdat and Gregory will be born. They will trample on the infidels and come to Jerusalem, inaugurating a period of peace. Eventually Anti-Christ will arise, only to be destroyed before the Second Coming. But the author does not know the time of these last things. We have here the first intimation in Armenian of the themes found 24

E.g., Samuel of Ani, Kirakos of Ganjak, Smbat the Constable; but not Vardan. Samuel of Ani, Hawakmunk‘ i groc‘ patmagrac‘ (Eˇmiacin, 1893), 120; trans. in M. Brosset, “Samouel d’Ani: Tables chronologiques,” in Collection des historiens arme´niens, vol. 2 (St. Petersburg, 1876). The later Kirakos quotes this passage, changing “son of Satan” to “son of Belial.” Kirakos Ganjakec‘i, Patmut‘iwn Hayoc‘ (Erevan, 1961), 107; trans. in M. Brosset, “Histoire de l’Arme´nie par le vartabed Kiracos de Gantzac,” in Deux historiens arme´niens (St. Petersburg, 1870), 1–194. 26 Pseudo Epiphanii Sermo de Antichristo, ed. G. Frasson, Bibliotheca Armeniaca 2 (Venice, 1976). 27 The prophecy of Nerse¯s I has been discussed above. The prophecy of Sahak, predicting the restoration of a king in the Arsacid line and a patriarch descended from St. Gregory, appears in the History of Łazar P‘arpec‘i, Patmut‘iwn Hayoc‘ (Tiflis, 1904; repr. Delmar, N.Y., 1985), 29–37; trans. in R. W. Thomson, The History of Łazar P‘arpec‘i (Atlanta, 1991). A Greek version also exists: G. Garitte, “La vision de S. Sahak en grec,” Le Muse´on 71 (1958): 255–78. Sahak’s vision is also discussed in A. Hovannisyan, Drvagner hay azatagrakan mtk‘i patmut‘yan, vol. 1 (Erevan, 1957), which was not available to me. 28 Pseudo-Epiphanius, 5.226. Note the summons, which is not the version in the Armenian Agat‘angełos (known as Aa; see Garitte, Agathange, 1–4). 25

Robert W. Thomson [ 77 ] in the so-called Letter of Concord, which purports to be the actual text of the pact signed by Constantine and Trdat in Rome. This Letter exists in various recensions and has not yet received a proper edition. It is a strongly pro-Western document, probably a translation of a French original, representative of the extreme pro-Roman party in Cilician Armenia. The printed versions of the Letter of Concord, which is some thirty pages long, have considerable variations of detail. The Venice edition of 1695 bears the confident title: “Letter of Friendship and Unity of the great emperor Constantine and saint Silvester the supreme Pope, and Trdat king of Armenia and saint Gregory the Illuminator, composed in the year of the Lord 316.”29 Three points are significant for our immediate purpose: Silvester ordains Gregory and makes him patriarch of all Armenians, wherever they may be.30 Constantine grants Bethlehem to “my dear brother” Trdat.31 And Constantine refers to future misfortunes befalling Armenia. However, “salvation will come from the Lord and help from our own progeny,” says the emperor, foretelling the rise of a new Constantine.32 These ideas are developed by some of the thirteenth-century Armenian historians. The diversion of the Fourth Crusade to Constantinople at the beginning of that century did not have many immediate echoes in Armenia. The scribe Gregory at the patriarchal see of Hr˙omkla, copying in 1204 the Commentary on Luke by Ignatios vardapet, noted that King Leo had heard of the fall of Constantinople to the Crusaders by special messenger from Baldwin. Gregory states: “At first we were saddened; then we remembered the prophecy of saint Nerse¯s, and we await the liberation of Jerusalem and all this land.” He wisely added: “But God will take care of the future.”33 Although it is another half century before Armenian historians echo such hopes for the liberation of Jerusalem, prophetic expectations were clearly in circulation, as we hear from a non-Armenian source. Friar William of Rubruck spent the years 1253–55 in a lengthy voyage to the Mongol khan. In his report to King Louis IX, he notes that on his return journey he stopped at Nakhchawan on the Araxes, where an Armenian bishop told him that: “[The Armenians] have two prophets. The first is the martyr Methodius, who . . . made a full-blown prophecy about the Ishmaelites, which has come to fruition in the Saracens. The name of the other prophet is Acacron. When dying he uttered a prophecy about a race of archers who would come from the north, saying that they would conquer the whole of the East. . . . They will occupy the countries from the north down to the south and advance as far as Constantinople.” To summarize: the Franks in Jerusalem will fall upon the Tartars with the aid of the Armenians, and occupy Tabriz. All the peoples of the East 29 Translated from the Armenian title page of Lettera dell’Amicitia e dell’Unione (Venice, 1695). Another version in K. V. Sˇahnazareanc‘, Dasˇanc‘ T‘łt‘oc‘ k‘nnut‘iwn u herk‘ume˘ (Paris, 1862), 11–30; partial version in C. Galanus, Conciliatio ecclesiae Armenae cum Romana, Pars  (Rome, 1690), 31–35. 30 There was an earlier tradition in the Syriac and Karshuni versions of Agathangelos that Gregory was ordained in Rome, but by Leontius (the bishop of Caesarea in Aa). See M. van Esbroeck, “Le re´sume´ syriaque de l’Agathange,” AB 95 (1977): 291–358, esp. 344. For the Karshuni version derived from the Syriac, see M. van Esbroeck, “Un nouveau te´moin du livre d’Agathange,” REArm 8 (1971): 13–167. 31 Letter, 15. 32 Letter, 16. 33 Yovsep‘ean, col. 697; the MS is now Jerusalem no. 334.

[ 78 ] The Crusaders through Armenian Eyes and the unbelievers will be converted, and such peace will reign in the world that the living will say to the dead: “Alas for you, who have not lived to see these times.” William of Rubruck says that he had first heard this prophecy in Constantinople from Armenians who lived there, and that throughout Armenia it is credited with the same reliability as the Gospel. The Armenian bishop also told him: “Just as the souls in Limbo were waiting for the coming of Christ so as to be set free, so are we waiting for your coming in order to be delivered from this slavery we have been in for so long.”34 Methodius is not named by Armenian writers until the very end of the thirteenth century, when Step‘annos of Siunik‘ quotes a lengthy passage from the Apocalypse attributed to him.35 Acacron, however, is known in Armenian as Agadron or Agat‘on. His prophecy has its origin in the well-known prediction of St. Nerse¯s, with wild elaborations of the Mongol period, and echoes themes found in the Letter of Concord. It too refers to the coming of Constantine and Trdat to Jerusalem: we learn that Constantine built the church of the Holy Sepulcher, and Trdat that of the Nativity in Bethlehem. At a later time a new Constantine, with an army that includes troops descended from the Armenians who escorted Trdat to Rome, will liberate the Christians from the bondage of the Archers, that is, the Mongols. Later events leading up to the reign of Anti-Christ and the Second Coming follow the pattern of the treatise of Pseudo-Epiphanius.36 Such ideas, however, were slow to be integrated into the mainstream of Armenian historiography, as a brief look at historians writing after William of Rubruck will show. The History of Vanakan vardapet is unfortunately lost. This went down to 1265 and dealt in particular with the Mongol invasions. Much of its information was probably incorporated into the works of his pupils Kirakos of Gandzak and Vardan, but they do not specify what is his and what is their own. A brief anti-Greek tract attributed to Vanakan refers to the story of Gregory’s visit to Rome. The only new detail not found in the Letter of Concord is that when the two kings divided the holy places, Trdat received St. James, the Armenian cathedral, and Golgotha, but that they threw lots for the Anastasis. Constantine lost, but St. Gregory allowed the Romans to have it in order to expand the building.37 Kirakos in his own History picks up the theme of Nerse¯s’ prophecy. He repeats the claim that Nerse¯s had foreseen the ruin of Armenia at the hands of the Archers, “which we have now seen with our own eyes,” says Kirakos, referring to the Mongols.38 These disasters are harbingers of Anti-Christ. But Kirakos makes no link with the Crusaders, whose activities he mentions only sporadically. According to him, these Romans had come via Thrace to Asia Minor in order to 34 Quoted from the translation by P. Jackson and D. Morgan, The Mission of Friar William of Rubruck, Hakluyt Society (London, 1990), 265–67. 35 See below. 36 See the summary in A. K. Sanjian, “Two Contemporary Armenian Elegies on the Fall of Constantinople, 1453,” Viator 1 (1970): 223–61, esp. 232–34. The Armenian text published by B. Sargisean in Bazmave¯p (Venice, 1913) was unavailable to me; he gives extracts and a lengthy discussion of the text in his Usumnasirut‘iwnk‘ hin ktakarani anvawer groc‘ vray (Venice, 1898), 177–207. 37 Vanakan, Ban Hawatali, text in M. Mut‘afean, “Ban Hawatali erki hełinaki xndire˘,” Sołakat‘ (Istanbul, 1995): 156–64. 38 Kirakos Ganjakec‘i, 231.

Robert W. Thomson [ 79 ] avenge the afflictions brought on the Christians by the Scythians (Turks) and Persians. On their way they had been much vexed by the son of Belial called Alexios. But those who survived his deceit captured Antioch and Jerusalem; this he dates to 1097.39 Reaching the year 1139, he notes that the Catholicos Gregory III went to Jerusalem in order to worship at the holy places. He says nothing about the Council of Jerusalem held in 1141,40 but reports that, while there, the Catholicos Gregory and the Franks renewed the old compact of Trdat and Gregory with Constantine and Silvester. As noted earlier, this famous encounter in Rome is a perennial feature of Armenian tradition and was cited frequently from the fifth century onwards in various contexts. Unlike his teacher Vanakan, Kirakos does not say that the two kings actually went to Jerusalem, but rather that Constantine and Silvester gave the Armenians relics of the apostles and also rights in Jerusalem: not only the church of St. James and Golgotha, but also a place for the liturgy in the Anastasis. Gregory was allowed to hang a lamp over the tomb, and he prayed that on the feast of Easter this lamp would be lit without visible light.41 This is not the first mention of the miraculous Fire on Holy Saturday in Armenian sources, for disputes about the correct date for that feast date from the divergence of the calendars in the seventh century.42 The importance of Kirakos’ claim is that after Jerusalem had been recaptured by the Muslims, Armenians were basing their claims to the holy places on rights bestowed by the builder of those churches. More surprising is his acknowledgment of the title of patriarch coming from Rome. This runs contrary to earlier Armenian theories of patriarchal autonomy and sits uneasily with the general attitude of writers in Greater Armenia toward the pope. However, Kirakos does not make any connection between the new Romans—that is, the Crusaders—and the old Romans of the time of Constantine. Nor does he tie the Mongols, harbingers of AntiChrist, to any hope of deliverance from the West. Explaining the origin of the Crusades, his fellow pupil Vardan also notes that the prophecy of St. Nerse¯s had been fulfilled, though he does not indicate what the prophecy had actually predicted. However, the immediate cause of the first Crusade, asserts Vardan, was the affair of the French count’s eye. “While the Scythians were ruling over Jerusalem and levying fees on those who visited the holy tomb, a certain count, Frank by race, happened to be jostled in the throng and became angered. For that reason he was beaten, until one of his eyes was knocked from its socket and fell out. Picking it up, he placed it in his wallet and took it to Rome, where he showed it. He made many burn with rancor. So their most noble men, leaders of numerous troops, came forth, whose names were as follows.” Vardan then gives the names of several leading Crusaders, which he has taken from the History of Matthew.43 39

Kirakos, 106. Kirakos, 117. For the council see the description in Halfter, Das Papsttum, 127–29. 41 Kirakos, 11. 42 See A. K. Sanjian, “Cr˙azatik ‘Erroneous Easter’—A Source of Greco-Armenian Religious Controversy,” Studia Caucasica 2 (1966): 26–47. 43 Vardan Arewelc‘i, Hawak‘umn Patmut‘ean (Venice, 1862; repr. Delmar, N.Y., 1991), 110; trans. in R. W. Thomson, “The Historical Compilation of Vardan Arewelc‘i,” DOP 43 (1989): 125–226. Cf. Matthew, 308. 40

[ 80 ] The Crusaders through Armenian Eyes The story of the count’s eye is also found in the Armenian version of the Syrian patriarch Michael’s Chronicle, where the count is identified as St. Gille. Michael’s famous work in twenty-one books, composed soon before his death in 1199, was rendered into Armenian by a Syrian named Isˇox with the help of Vardan himself. As with the Armenian version of the Georgian Chronicles made anonymously about the same time, Vardan and Isˇox not only greatly abbreviated the original, but also made various additions of their own. Michael himself, in the original Syriac, does not mention Raymond of St. Gille’s eye; nor does this story appear in the works of Bar Hebraeus. It seems to be of Armenian origin. On the other hand, in their Armenian version of Michael, Isˇox and Vardan did not add any reference to the well-known prophecy of Nerse¯s.44 One might also note in passing a further Armenian addition to the original Syriac of Michael’s Chronicle regarding the Crusades: Edessa was captured by the Crusaders before Jerusalem because King Abgar had believed in Christ before Jerusalem did. This reflects the old Armenian tradition that Abgar was himself an Armenian, his entourage were members of Armenian noble families, and thus Armenian faith in Jesus Christ predates that of other churches.45 As one might expect, the Chronicle of Smbat, brother of king Het‘um, is well informed about the Crusades. He notes that fuller details about these princes of the Romans may be found in the Frankish historians.46 Smbat’s own close involvement with the West is reflected in his adaptation of the earlier Armenian law code of Mxit‘ar Gosˇ to incorporate westernizing practices in Cilicia.47 But Smbat offers no reflections on the significance of the Crusades, either in terms of the Armenian past or with expectations of the future. His chronicle was completed soon after 1272,48 by which time the Crusaders had been reduced to a fraction of their earlier significance. He does include one passing reference to the prophecy of Nerse¯s, but that concerns the fall of Ani to the Byzantines back in 1044.49 44 ˇ amanakagrut’iwn ( Jerusalem, 1871), 402; trans. (from a different Armenian Armenian text of Michael, Z recension) in V. Langlois, Chronique de Michel le grand (Venice, 1868). Although Vardan had named St. Gilles among the crusading princes, it is only in the Armenian version of Michael that he is associated with the count who lost his eye. For the complicated textual history of the Armenian text, see A. B. Schmitt, “Die zweifache armenische Rezension der syrischen Chronik Michaels des Grossen,” Le Muse´on 109 (1996): 299–319; and for adaptations introduced by the translators Vardan and Isˇox, F. Haase, “Die armenische Rezension der syrischen Chronik Michaels des Grossen,” OC, n.s., 5 (1915): 60–82, 271–84. For similar alterations in the Armenian version of the Georgian Chronicles, see R. W. Thomson, Rewriting Caucasian History: The Medieval Armenian Adaptation of the Georgian Chronicles, Oxford Oriental Monographs (Oxford, 1996), xlv–l. 45 ˇ amanakagrut’iwn, 404. The first historian to think of Abgar as Armenian was Movse¯s Xorenac‘i. Michael, Z See his History, book 2.26–33, for the adaptation of the earlier legend concerning the conversion of Edessa. 46 Smbat Sparapet, Taregirk‘ (Venice, 1956), 100, under the year 545 (⫽ 1096 ..). For the different recensions of this Chronicle, see S. Der Nersessian, “The Armenian Chronicle of the Constable Smpad or of the ‘Royal Historian,’” DOP 13 (1959): 143–68, reprinted in her Etudes byzantines et arme´niennes, vol. 1 (Louvain, 1973), 352–77. The translation by G. De´de´yan, La chronique attribue´e au conne´table Smbat, Documents relatifs a` l’histoire des croisades publie´s par l’Acade´mie des Inscriptions et Belles-Lettres 13 (Paris, 1980), begins at p. 179 of the Armenian text under the year 608 (⫽ 1159 ..). 47 Armenian text, translation, and commentary in J. Karst, Sempadscher Kodex aus dem 13. Jahrhundert oder Mittelarmenisches Rechtsbuch (Strassburg, 1905), 2 vols. 48 That is the date of the last entry; Smbat died in 1276. 49 Smbat, Taregirk‘, 42.

Robert W. Thomson [ 81 ] Nor does Mxit‘ar Ayrivanec‘i, whose Chronicle reaches the year 1289, pay much attention to the Crusaders. They came to “help the Christians”; but Mxit‘ar has no reference to the prophecy of Nerse¯s or to any connection between these new Romans and the old pact with Constantine.50 The last major historical work to be considered is that composed in 1299 by Step‘annos, metropolitan of the large province of Siunik‘ in northeastern Armenia. His History is a history of his own province; he does not concern himself with the fate of Armenians elsewhere. When, for example, he mentions in passing the fall of Jerusalem to Saladin in 1187, the important feature of the ensuing massacres was the martyrdom of a bishop from Siunik‘ who happened to be visiting the Holy Sepulcher.51 Step‘annos was obviously not very well informed about events that far from his episcopal see, for he calls Saladin a “Khorazmian”—though perhaps he just means “Turk.” And when Akka fell in 1291, Step‘annos confuses the name with Achaia in Greece, since he refers to it as “the populous city of the Corinthians.”52 Step‘annos condemns the catholicos, then Grigor VII, for following the Greek calendar when celebrating Easter in 1293. He devotes many more lines to this dispute over the Holy Fire on Holy Saturday than to the fall of Hr˙omkla to the Mamluks the previous year.53 In any event, Step‘annos has no interest in mentioning the arrival of the Crusaders or in describing their final collapse; nor does he link their presence to any Armenian traditions. And when he engages in theological disputation, either on matters of liturgical practice or the decisions of councils, it is the Greek church that is condemned. The Franks and the West are ignored. Step‘annos does not bring up the famous prediction of Nerse¯s, except to note that it was fulfilled back in the fourth century when Armenia was divided. But for the first time we find in Armenian a long quotation from the Pseudo-Methodian Apocalypse.54 This comes immediately after a description of the life and work of his eighth-century predecessor, Step‘annos, bishop of Siunik‘, who is famous for his translations of the PseudoDionysian corpus and other Greek texts. It is not clear whether the earlier Step‘annos was supposedly responsible only for the long extract quoted or for the whole of the Apocalypse. In any event, the extract cited by the second Step‘annos comes from the Greek version of Methodius, not the original Syriac. Which Step‘annos is responsible for additions that bring in Armenia is also unclear. Be that as it may, this long quotation is not integrated into Step‘annos Orbelean’s own History, being a kind of appendix to the chapter on the earlier Step‘annos. It refers to the Romans overcoming the Ishmaelites and bringing peace to east and west. But then Gog

50 Mxit‘ar Ayrivanec‘i, Patmut‘iwn Hayoc‘ (Moscow, 1860), 59, under the year 1094; trans. in M. Brosset, “Histoire chronologique par Mkhithar d’Airivank, XIII S,” Me´moires de l’Acade´mie impe´riale des sciences de St. Petersbourg, ser. 7, vol. 13.5 (1869). Mxit‘ar is familiar with the pact, as known to Kirakos, and he adds the figure of 70,000 escort (as in Aa 873) under the year 287. 51 Step‘annos Orbelean, Patmut‘iwn nahangin Sisakan (Moscow, 1861), 252; trans. in M. Brosset, Histoire de la Siounie par Ste´phannos Orbe´lian (St. Petersburg, 1864). 52 Step‘annos, 245. 53 Step‘annos, 248–49. 54 Step‘annos, 89–94.

[ 82 ] The Crusaders through Armenian Eyes and Magog, releasing all the wild peoples who had been confined to the north by Alexander the Great, will devastate the world and the son of perdition will appear. This will be followed by the Second Coming of Christ, the extermination of the impious, and the ascension of the just to heaven like luminous stars. However, Step‘annos does not in any way make the prophecies relevant to later events, either to the Crusades or to the Mongol invasions. Such apocalyptic ideas and hopes for help from the West—always disappointed but tenacious nonetheless—persisted long after the time of the Crusades. The fall of Constantinople in 1453, for example, was commemorated in various elegies, a genre with a long history in Armenian literature. Ar˙ak‘el of Bitlis drew on the traditions just described: he looks forward to the recovery of “Stamboul” by the Franks, who will then go on to Jerusalem and drive out the infidels. The victorious Franks will be guided to Armenia by descendants of those Armenian soldiers who stayed behind in Rome on the occasion of Trdat’s visit. Thus will Armenia be liberated.55 And those hoping for Western intervention in later centuries, even down to modern times, echo similar aspirations.56 To return to the Crusaders. Their immediate impact in Cilicia was of course allembracing; Western traditions deeply influenced Armenian social life, religious practice, legal codes, linguistic idiom, and artistic achievement. Little of this is reflected in Armenian historiography of the time. Those Armenians who were writing outside Cilicia do not concern themselves much with events there. Even when they do discuss westernizing habits, it is generally to condemn them. As in the past, Armenian authors did not offer a perspective that embraced all Armenian viewpoints. There was never any one “Armenian attitude” to the Crusades, any more than there had been one attitude toward Byzantium or toward the other major powers of the Near East. What is noticeable, however, is the attempt by some Armenian historians to integrate the Crusades into the broader progression of Armenian history. Traditions from the earliest authors of the fifth century—when Armenians first began to reflect on their past— were picked up, elaborated, and brought into connection with contemporary events. The Franks were not only seen in terms of past Armenian prophecy but were also integrated into predictions for the future derived from foreign sources. The Crusaders had not liberated Armenia from the Turks or from the Mongols. But because of their Western origins they could be associated with that most ancient of Armenian aspirations: the return of the golden age, namely, the reign of a new Trdat from the Arsacid family and a new Gregory from the line of the holy Illuminator. The Latinophiles might emphasize the role of Pope Silvester. Most Armenians, however, hoped for renewed military success in a descendant of the first Christian emperor of Rome, Constantine, who so long before had entered into an eternal alliance with the first Christian king of their own country. Oxford University 55

See Sanjian, “Two Contemporary Armenian Elegies.” See the discussion for the 16th–19th centuries in N. G. Garsoı¨an, “Reality and Myth in Armenian History,” in The East and the Meaning of History, Universita` di Roma “La Sapienza,” Studi Orientali 13 (1994), 117–45. 56

Latins and Franks in Byzantium: Perception and Reality from the Eleventh to the Twelfth Century Alexander Kazhdan† When scholars write about relations between the West and Byzantium in the Middle Ages, they naturally emphasize the contrasts between the two societies:1 Catholicism versus Orthodoxy, feudalism versus “totalitarian” regime, predominantly oral culture versus consistent textuality, barter economy versus uninterrupted circulation of coins, poetization of warfare versus the ideology of peace, a list of oppositions that could be continued almost to infinity. It is unclear whether these contrasts are in fact part of the reality of the medieval world or were spawned by the confessional intolerance of nineteenth-century historiography. Fortunately, this problem lies far beyond the limited tasks of this study. At first sight, the political and ecclesiastical events of the eleventh and twelfth centuries confirm the traditional opinion: this period begins with a theological clash and the so-called division (schism) of the church in 1054, and it ends with the sack of Constantinople by the Crusader army in 1204. Who could deny that the two worlds stood in opposition to each other? Political conflicts are, however, a questionable litmus test for the evaluation of cultural relations, and the picture of an unbridgeable gap between the Greek empire and western kingdoms often trips over humble reality. We should not forget that the Latin West, in its turn, was split into various units of power, and the “internecine” struggles between the papacy and the German emperors, or between France and England, determined medieval structures much more than the Byzantine siege of Latin Antioch. As modern study of the medieval past advances, it becomes more and more doubtful that the dispute between Cardinal Humbert and Patriarch Michael Keroularios in 1054 (let alone that of Photios and Pope Nicholas I almost two hundred years earlier) created an actual schism, a separation of the two churches and two cultures.2 The Byzantines forged alliances with 1 This contrast that existed despite certain “points of convergence” is masterfully delineated by A. Laiou, “Byzantium and the West,” in Byzantium: A World Civilization, ed. A. Laiou and H. Maguire (Washington, D.C., 1992), 67–78, in the form of a narrative of the journey of two imaginary Westerners to Byzantium. Cf. M. Whittow, “How the East Was Lost: The Background to the Komnenian Reconquista,” in Alexios I Komnenos, ed. M. Mullett and D. Smythe (Belfast, 1996), 56–63. 2 See esp. P. Lemerle, “L’Orthodoxie byzantine et l’œcume´nisme me´die´val: Les origines du ‘schisme’ des Eglises,” BullBude´ (1965): 228–46.

[ 84 ] Latins and Franks in Byzantium some western and northern neighbors against other such neighbors, married western notables,3 accepted western warriors not only as defenders of their borders but also as saints of the Greek Orthodox church,4 and developed theological and secular intellectual exchanges with the allegedly schismatic West. The relations were too complex to reach a plausible conclusion based on common sense alone.5 I shall focus here on several aspects of Western-Byzantine relations in the eleventh and twelfth centuries, those that seem to me either ignored or misinterpreted. While not denying that there was a substantial element of hostility in Constantinople toward the West, and vice versa, I hope to demonstrate that the traditional concepts of contrast and hostility are insufficient to describe the relationship. I begin with terminology.6 We usually describe the western world, in contrast to Byzantium, with the single, unifying term Latins. The Latin term Latini is an old one, which in Roman usage originally designated the population of Latium. It was extended by Roman law to include certain groups within the empire who were granted restricted legal rights: the so-called Junian Latini were manumitted slaves who were considered free during their lifetime but reverted to slave status at death, so that their properties were inherited by their patrons as peculium. Justinian I abolished the status of Latini in 531,7 and the term disappeared together with the institution. Thus the ethnonym Latinoi is not used ca. 800 by the Greek historians Theophanes the Confessor and Patriarch Nikephoros. For Nikephoros, the western world lies outside his geographical horizons: after mentioning such northern neighbors as the Avars, Huns, Bulgarians (and Kotragoi), Khazars, Scythians, Slavinians (and twice Slavs), the patriarch neglected the inhabitants of Central and Western Europe; even his land of the Goths is located in the Crimea.8 Theophanes knows all the northern neighbors (Avars, Bulgarians, Huns, Scythians, etc.) mentioned by Nikephoros, as well as numerous tribes active in the late Roman period to the west and north of the empire (Celts and Galli, Goths, Germans, Gepids, Vandals and Alemans, western Galatians and Iberians, Alans), and several peoples (Franks, Sicilians) and territories (Carthage, Spain, Gallia, Italy, Calabria, Campania) of later centuries, but no Latins. Nor does pseudo-Genesios, in the tenth century, refer to Latins, although he occasionally names some western territories (Spain, Italy, Sicily, Longibardia or Longibarbia [Lombardy], Frankia), and includes in the army 3 F. Tinnefeld, “Byzantinische auswa¨rtige Heiratspolitik vom 9. zum 12. Jahrhundert,” BSl 54 (1993): 21– 28. From the Western point of view, the problem has been studied by G. Wolf, “Die byzantinischabendla¨ndische Heirats- und Verlobungspla¨ne zwischen 750 und 1250,” ADipl 37 (1991): 15–32. 4 C. P. Kyrris, “The ‘Three Hundred Alaman Saints’ of Cyprus,” in The Sweet Land of Cyprus, ed. A. Bryer (Nicosia, 1993), 203–35. 5 This complexity is stressed by H.-G. Beck, “Byzanz und der Westen im 12. Jahrhundert,” Vortra¨ge und Forschungen 12 (Konstanz, 1969): 227–41, reprinted in his Ideen und Realita¨ten in Byzanz (London, 1972), no. . 6 The problem of the Greek terminology of the Crusades was recently studied by A. Kolia-Dermitzaki, ¨ B 41 (1991): 163–88, but from a “Die Kreuzfahrer und die Kreuzzu¨ge im Sprachgebrauch der Byzantiner,” JO different viewpoint. The author argues that the Byzantines perceived the Crusades as an ordinary military expedition, no different from others. 7 A. Steinwenter, in RE 12 (1925): 922. 8 C. Mango, ed., Nikephoros, Patriarch of Constantinople: Short History, DOT 10 (Washington, D.C., 1990), para. 42.7.

Alexander Kazhdan [ 85 ] of Thomas the Slav certain northern (Slavs, Huns, Alans) and western (Vandals and Getae) tribes.9 Richer is the ethnogeographic information of Genesios’ contemporaries, the continuators of Theophanes (I consider all six books of Theophanes Continuatus as a group, notwithstanding the complexity of their authorship). These historians were aware of many western territories and peoples, such as Italy, Sicily, Calabria, Langobardia, Francia, Gallia, Spain, Andalusia; without hesitation, they call the Spanish Arabs Spanoi,10 while the general term Latinoi makes no appearance in the Continuatio. On the other hand, the Continuators know various northern tribes; in a passage of the fifth book (Vita Basilii), we find Croatians, Serbs (Serbloi), Zachlumians, Terbuniotai, Kanalitai, Diokletianoi (from Dukla), and [A]rentanoi who are given the general label of “Scythians.”11 The same term is applied in the Continuatio to other ethnic groups, Bulgarians12 and Rus’, the latter being flatly defined as a Scythian tribe.13 We may surmise that Theophanes Continuatus had some idea (possibly influenced by an ancient tradition) of the unity of Byzantium’s northern neighbors, but did not envisage the unity of the western peoples. In the treatise “On the Administration of the Empire” attributed to Constantine VII and produced at approximately the same time as the Continuatio, we find numerous names of northern peoples (Avars, Bulgarians, Serbs, Croatians, Dalmatians, Pechenegs, Rus’ and their tributaries, Turks [⫽ Hungarians], etc.); the compiler gives the same list of Southern Slavic tribes14 as is rendered by the Continuator, omitting only the generalization “Scythian.” The author is very cautious in applying the term “Scythian”: only once does he use it in a generic sense, referring, as does the Continuatio, to the Rus’ as one of the “northern and Scythian tribes.”15 The term appears two more times,16 but in both cases is applied to a limited ethnogeographical area. The western lands and peoples appear frequently: Spain and its parts (Lusitania, etc.), Italy and its parts, especially often Langobardia and the Venetians, Gallia, Germanoi, Franks, Gepids, Goths (and Visigoths), Vandals. The general term Latinoi, however, is missing from the treatise. Nor is it found in the Chronicle of John Skylitzes written ca. 1100, which does mention such western peoples as Spaniards, Italians, Calabrians, and Franks. The chronicler also refers to various northern neighbors, including Varangians, Rus’, Pechenegs, Hungarians, Serbs, Croatians, and Bulgarians, applying the term “Scythian” primarily (but not exceptionally) to the Rus’. Byzantine historians from Theophanes to Skylitzes were well aware of the disunity of the western world. The data collected by the Dumbarton Oaks Hagiography Project, which includes the vitae of saints who died between 700 and 1000, records the use of the ethnonym Latinoi

9

Genesii Regum libri quattuor, ed. A. Lesmu¨ller-Werner and J. Thurn (Berlin-New York, 1978), 24.18–19. Theophanes Continuatus, ed. I. Bekker (Bonn, 1838), 474.22 (hereafter TheophCont). 11 TheophCont 288.19–20. 12 TheophCont 11.17, 420.4. 13 TheophCont 196.6–7. 14 De administrando imperio, vol. 1, ed. G. Moravcsik and R. Jenkins (Washington, D.C., 1967), chap. 29.56–58 (hereafter DAI ). 15 DAI, chap. 13.25. 16 DAI, chaps. 43.2, 53.129. 10

[ 86 ] Latins and Franks in Byzantium only rarely: all examples are gleaned from the stories of holy men of South Italy (the biographies of Christopher and Makarios, Elias Speleotes, Sabas the Younger), and all refer to the Latin language as contrasted with Greek. Another Italian hagiographer, who wrote about Neilos (Nilus) of Rossano (d. 1004) in the first half of the eleventh century, speaks of Latinoi, contrasting them with the Greek population of South Italy.17 To the best of my knowledge, the Greek term Latinos first appears as a generic appellation of Westerners in a patriarchal decision of 1054;18 in most passages the compiler of the document speaks of the “Italian language” or “Italian characters,” and only in a section translated from Latin is the term Latinos used. Michael Attaleiates, at the end of the eleventh century, used the term Latinos several times, but the meaning of the ethnonym is confused: his Latins are evidently good warriors,19 but their ethnic identity is obscure. The Albanoi and Latins in Italy20 echo the ancient nomenclature that is in general typical of Attaleiates, who, for instance, explains the contemporary name Nemitzoi as designating ancient Sarmatians.21 We shall see later that Attaleiates employed the ethnic term Latinis where his contemporaries spoke about Franks or Normans. After the First Crusade the situation changed quickly: Anna Komnene, according to the index to the Alexiad prepared by P. Gautier,22 used the noun Latinos 97 times and 4 times the adjective Latinikos. Even more frequent, in the Alexiad, is the ethnonym Keltos (176 times plus 20 cases of the adjective Keltikos), applied interchangeably with the word Latinos. Anna also employs specific designations such as Frank (20 times plus 14 times the adjective Frankikos), Italos (7 times), and rarely Germanos/Germanikos. Niketas Choniates, in his Chronike diegesis, employs the term Latinos (as noun and adjective) 127 times, the feminine form Latinis 4 times, and the adjective Latinikos 26 times. Other (specific?) ethnic names used by Anna are rare in Choniates, except for Italos (17 times): Keltos appears once, and once its adjective, in the expression Keltiko` n ge´ no",23 while Frank is found three times. On the other hand, Alemans are common (52 times), and new designations ( Ij taliw´ th", Fraggi´sko", Ij ggli'no") are employed here and there. John Kinnamos prefers specific, sometimes confusing ethnonyms (such as Alemans, Germans, Italians), but even he several times speaks of Latin armies, Latin customs, Latin peoples, and Latin language.24 In other words, before the twelfth century, the Byzantines saw the West as composed of separate territories and distinct peoples (Italians, Spaniards, Germanic tribes, and so on), while the concept of Latin language entered Greek literature of the late tenth century in a specific area, South Italy. By the twelfth century, the notion of Latin peoples (and of Latin habits) was firmly established: wrongly or rightly, Byzantine intellectuals began to consider the West as a unified entity. 17 G. Giovanelli, Bi´o" kai` politei´a tou' oJsi´ou patro` " hJmw'n Nei´lou tou' Ne´ ou (Grottaferrata, 1972), 112.8. Nilus is described as using the “Hellenic” (p. 113.12–13) and “Roman” (p. 114.1) languages. 18 Les regestes des actes du patriarcat de Constantinople, fasc. 3, ed. V. Grumel ([Constantinople], 1947), no. 869. 19 Michaelis Attaliotae Historia, ed. I. Bekker (Bonn, 1853), 35.12–13, 46.22 (hereafter Attal.). 20 Attal. 9.12. 21 Attal. 147.1. 22 Anne Comne`ne, Alexiade, ed. P. Gautier, vol. 4 (Paris, 1976), 73 f (hereafter Alexiad). 23 Nicetae Choniatae Historia, ed. J. L. van Dieten (Berlin-New York, 1975), 264.64 (hereafter NikChon). 24 Ioannis Cinnami Epitome, ed. A. Meineke (Bonn, 1836), 19.6, 47.9, 146.19, 199.17, 208.20, 267.20, 282.5 (hereafter Kinnam.).

Alexander Kazhdan [ 87 ] Along with this progression from the notion of the “Latin language” to the “Latin people[s]” and “Latin habits,” the Byzantines began to distinguish particular features (beyond linguistic differences) typical of the Westerners or “Latins.” Like Kinnamos, Anna speaks of Latin habits (e“qh).25 Among the characteristic traits of their behavior is first of all their martial prowess: Anna constantly mentions Latin troops, strateu´ mata and fa´ lagge", their “irresistible attacks,” their horses and weapons, especially long spears and the so-called tzangra, and defines them as “noble.”26 Several times Anna describes their habit of swearing an oath and characterizes them as arrogant, avaricious, and greedy for power;27 she knows that the Latins differ from the Byzantines in their interpretation of theological problems and believe that the Roman pope is the head of the whole universe.28 In passing, Anna notes that an educated Latin is as strange a figure as a Hellenized Scythian.29 Choniates presents a similar, albeit more differentiated picture of the Latins. He speaks, of course, about the Latin language30 and—what is more substantial for our purpose—Latin peoples either in the singular or plural, ge´ nh31 and fu´ lai.32 Thus Choniates considered the Latins as a unity that, in its turn, consisted of a plurality of peoples or tribes. The Crusaders who conquered Syria are characterized as a unity, “the Latins”;33 at the same time, he notes that Constantinople was populated by the Latins of various peoples (diaforogenw'n).34 Since Bertha of Sulzbach is described as a Latin by her ge´ no",35 the Germans were included in this category. Choniates applies the ethnonym “Latin” also to the Normans of Sicily.36 Like Anna, Choniates emphasizes the military role of the Latins: he speaks of the Latin strateu´ mata, stratiwtiko´ n, strato´ ", stratia´ , duna´ mei", sti'fo", su´ ntagma, strato´ pedon, moi'ra.37 He is aware of the Latin retainers and Latin mercenaries.38 Some Latins appear in his narrative as infantry soldiers, others as mounted warriors, or fighting in heavily armed phalanges.39 Like Anna, Choniates emphasizes that the Latins bore spears which they used skillfully.40 Choniates respects the military prowess of the Latins: he 25

Alexiad 3:79.28–29, 117.22. Alexiad 2:97.10–11. 27 On the oath, see Alexiad 2:105.21, 215.4–5, 232.25–26; on their arrogance, 2:212.20–21, 229.27–28, 3:18.12; on their avariciousness, 2:211.18–19, 3:148.21; on their greed for power, 2:212.25–28. 28 On theology, see Alexiad 2:218.18–20; on the pope, 1:48.12–14. 29 Alexiad 2:218.4–5. 30 E.g., NikChon 200.88, 308.2. 31 In singular, NikChon 178.7, 350.47; in plural, 392.64, 612.87. 32 NikChon 537.57. 33 NikChon 417.56–57. 34 NikChon 247.31. 35 NikChon 147.86. 36 E.g., NikChon 360.48. 37 For strateu´ mata, see NikChon 67.52, 550.34–35, 600.46–47, 619.26; stratiwtiko´ n, 108.53; strato´ ", 556.71; stratia´ , 365.56; duna´ mei", 556.78, 563.67, 568.71; sti'fo", 391.41, 532.28, 616.58, 633.55; su´ ntagma, 618.12, 630.64–65, 639.83–84; strato´ pedon, 615.25, 616.40, 624.8; moi'ra, 637.30. 38 For Latin retainers, see NikChon 238.94; for mercenaries, 386.86. 39 For infantry soldiers, see NikChon 379.91–92 (and cf. oJplitiko´ n, 628.18); mounted warriors, 542.68, 568.71, 605.63–65, 616.47; fa´ lagge", 572.76–77 (cf. 640.8). 40 NikChon 610.88, 35.2. 26

[ 88 ] Latins and Franks in Byzantium speaks of their invincible force and precision in battle;41 the Byzantines fled from the Latins like a herd of deer from a roaring lion and foolishly mistook the Latin security for cowardice.42 And once more he stresses that the Latins acted in a certain way not because of cowardice but following the principles of military technique.43 Only in passing does Choniates note that the Latins espoused a different creed, but he does not spare negative adjectives to characterize the “accursed” Latins.44 They were cruel, treacherous, stupid, unrestrained in speech, arrogant, unreliable, ambitious for glory and profit, and hostile toward the empire.45 Choniates notes that they wore a garment different from that of the Byzantines, and mentions a cape that was a product of Latin wool-spinning;46 disdainfully he calls the Latins cobblers and beef-eaters;47 he ridicules their ravenous appetite and metaphorically laments the victims of Latin jaws.48 Only once does Choniates deviate from this biased image of the Latins and mention that some of them were graceful.49 Choniates emphatically contrasts the Latins and Byzantines. “Between us and them,” he says, “lies an open gap, and we are separated in our views and diametrically opposed.”50 The historian presents the Byzantines as gentle and modest, whereas the Latins are supercilious, boastful, arrogant, and stupid. His stereotype of the Latin completely coincides with another stereotype favored by Choniates, that of the barbarians, which encompasses the Latins. The barbarian in the Chronike diegesis is also cruel, arrogant, greedy, unstable (unreliable), and foulmouthed. Choniates supplements these negative qualities, common with the Latins, with two more features that he attaches specifically to the “barbarians,” not Latins, even though some of his “barbarians” were Westerners: noisiness and lack of culture.51 Thus the stereotype of the Latin was molded by Niketas Choniates on the stereotype of the barbarian. But was Choniates as exceedingly anti-Latin as he seemed in forming this stereotype? Let us compare the description of the Second Crusade by Kinnamos and Choniates.52 Kinnamos’ conclusion is simple: the Crusaders were responsible for all the misunderstandings and conflicts, whereas the position of the Byzantine administration toward them was justified at all points. Choniates is much more cautious in his evaluation of the event: he not only emphasizes the manliness of the Crusaders but condemns Man41

NikChon 203.66, 618.19. On the Byzantine flight, see NikChon 561.21–22; Latin security, 614.2. 43 NikChon 644.33–35. 44 NikChon 576.92, 301.21. 45 On Latin cruelty, see NikChon 301.6; on their treachery, 623.57; stupidity, 39.39, 164.60; unbridled speech, 392.61; arrogance, 612.56, 625.20; unreliability, 540.21–22, 622.25 (cf. 602.4–5); ambition, 550.40; hostility to Byzantines, 551.44. 46 NikChon 537.42, 557.20–21. 47 NikChon 317.15–16, 594.75. 48 NikChon 560.86, 571.48. 49 NikChon 557.22–23. 50 NikChon 301.27–28. 51 On Latin noisiness, see NikChon 26.69, 398.17; on their lack of culture, 155.73, 322.44, 653.94, etc. 52 See also C. Asdracha, “L’image de l’homme occidental a` Byzance: Le te´moignage de Kinnamos et de Choniates,” BSl 44 (1983): 31–40, and S. Rakova, “Eastern and Western Man in Nicetas Choniates,” EtBalk 29.4 (1993): 55–63. 42

Alexander Kazhdan [ 89 ] uel I’s unfair actions toward the participants in the Second Crusade, and acknowledges that the Byzantines were deceiving the Western knights.53 Even more astonishing is Choniates’ panegyrical portrait of Frederick Barbarossa.54 Among the virtues of the Roman emperor he includes his noble origins, reason, invincibility in battles, and extraordinary love of Christ, and even compares Frederick with the apostle Paul. Choniates praises the handsomeness, manliness, reason, and strength of Conrad of Montferrat; he lauds the Antiochene notable Baldwin who died bravely at Myriokephalon, Raymond of Poitiers, prince of Antioch, and Peter of Bracieux.55 Near the end of his book, he recollects a Venetian merchant who became his friend, after Choniates had helped him and his wife;56 one of the manuscripts preserves the name of the man, Dominicus, who “dealing with barbarians as a barbarian” saved Choniates and his family. In another passage the historian blatantly defines as foolishness the undiscriminating Byzantine animosity toward Westerners, the lack of desire to distinguish between the bad and good among them.57 While “Latins” was a generic term, the ethnonym “Frank” seems to have had a predominantly specific meaning. Theophanes follows the principles of late antique ethnology when he lists Alans and Germanic tribes as allies of the Vandals; the Germanoi, he adds, are now called Franks.58 This definition goes back to Prokopios59 and continued to be used in later centuries: both the author of the book De administrando imperio and Attaleiates identified the Germanoi as Franks.60 Theophanes perceived Pipinus (Pepin), who routed the Arabs, as a king of the Franks, and Karoulos (Charlemagne) as another king of the Franks.61 The ethnic affiliation of his Frankoi is beyond doubt; their land is defined as Frankike.62 Like Theophanes (and Prokopios), Constantine Porphyrogenitos construed the Franks as a specific people among the “alien tribes”: while the regulation imposed by Constantine the Great forbade marriages between foreigners and the emperor of the Romans, the Franks were excluded from this regulation due to their constant interaction with the Romans.63 Constantine’s vision of the Franks expanded, however, beyond the borders of former Gallia: the Franks who subdued “the Croatians in Delmatia” are the subjects of “Otto, the great king of Frankia and Saxia” (Saxonia), while Venetians are called “Franks from Aquileia.”64 In any event, his Franks are a specific people located primarily north of the Alps or in northern Italy. 53

NikChon 66.18–33; cf. 568.74–76. NikChon, 416.29–50; cf. F. Boehm, Das Bild Friedrich Barbarossas und seines Kaisertums in den auslaendischen Quellen seiner Zeit (Berlin, 1936), 49–59. 55 On Conrad, see NikChon 201.93–95; Baldwin, 181.7–13; Raymond of Poitiers, 116.58–59; Peter of Bracieux, 601.85–86. 56 NikChon 588.13–16, 27–29. 57 NikChon 552.79–81. 58 Theophanis Chronographia, ed. C. de Boor, vol. 1 (Leipzig, 1883), 94.24–25 (hereafter Theoph.). 59 Wars 3:3.1; cf. 5:11.29, 12.8, in Procopii Caesariensis opera omnia, ed. J. Haury, rev. G. Wirth, vol. 1 (Leipzig, 1962), 317.16–17; vol. 2 (1963), 63.3, 64.9–11. 60 DAI, chap. 25.29–30; Attal. 148.1. 61 Theoph. 403.14–15, 455.20. 62 Theoph. 402.22, 403.9. 63 DAI, chap. 13.111–19. 64 On Otto, see DAI, chap. 30.71–79, and cf. chap. 31.86–87; on the Venetians, see chap. 28.4–5. 54

[ 90 ] Latins and Franks in Byzantium The meaning of the ethnonym “Franks” changed by the end of the eleventh century.65 According to Skylitzes,66 five hundred “Franks” under the command of [H]ardouin [of Milan] joined the army of the Byzantine general George Maniakes, who was campaigning in Sicily; Skylitzes described them as originating from Transalpine Gallia, but from Italian sources we learn that they were Normans.67 While William of Apulia calls them Galli,68 other Italian chroniclers speak plainly of Maniakes’ Norman allies.69 Skylitzes goes on to state that the Franks behaved in Italy as in a conquered land and that Maniakes was ordered to expel them;70 he also writes that Constantine IX sent his allies, the Franks and Varangians, against the Turks.71 Thus, in the chronicle of Skylitzes, the Frankoi are separated from Frankia or Gallia; the ethnic term is applied, as we may gather from the context, primarily to the Normans of South Italy. The term “Franks” is common in the Alexiad of Anna Komnene. Sometimes she used the terms “Franks” and “Latins” interchangeably: thus she writes that Alexios I wanted to join the Latin expedition against the Turks but was apprehensive when he saw the innumerable troops of the Franks.72 In another passage, Anna describes a band of Crusaders whom Alexios ordered to join “the other army of the Latins,” but they did not want to “be united with the Franks.”73 In most cases, however, the Franks are construed as a particular part of the generic mass of Latins. Anna relates a conversation between the emperor and an arrogant Latin when Alexios, with the help of a Latin interpreter, asked the man about his specific origin. “I am,” answered the knight, “a pure Frank of a noble lineage.”74 The emperor, she says, explained to Saint Gilles what the Latins should expect along their route; he also expressed some doubts about the intentions of the Franks.75 Anna speaks of the troops of the Franks and Celts, and lists Lombards, all the Latins, Germans, and “our” (tou` " kaq∆ hJma'") Franks.76 In Anna’s terminology, the Franks have no connection with the land called Frankia. The king of Frankia, the count of Frankia Stephen, and probably “all the counts in Frankia” are French,77 while the Franks, where they are identifiable, are predominantly Normans. The “large army of the Franks” under the command of Bohemond is evi65 On the application of the term Franci to Normans in South Italy, see A. Gallo, Aversa Normanna (Naples, 1938), 7 f. 66 Ioannis Scylitzae Synopsis historiarum, ed. J. Thurn (Berlin-New York, 1973), 425.10–12 (hereafter Skyl.). 67 On Ardouin and his Normans, see F. Chalandon, Histoire de la domination normande en Italie et en Sicile (Paris, 1907), 91 f; J. Gay, L’Italie me´ridionale et l’empire byzantin (Rome, 1904), 453 f; M. Mathieu in Guillaume de Pouille, La geste de Robert Guiscard (Palermo, 1961), 268. 68 Guillaume de Pouille, ed. Mathieu, book 1.195, 204. 69 Gaufredus Malaterra, De rebus gestis Rogerii . . . , in Raccolta degli storici Italiani 5.1 (Bologna, 1928), 12.25 and 31; Amato di Montecassino, Storia de’ Normanni (Rome, 1935), 67 f; Leo Mariscanus, Chronica monasterii Casinensis, MGH, SS 7 (Hannover, 1846), 675. 70 Skyl. 427.47, 55. 71 Skyl. 471.17–18, 474.6–7. 72 Alexiad 3:11.9–12. 73 Alexiad 3:36.23–28. 74 Alexiad 2:229.26–230.1. 75 Alexiad 2:235.9–12. 76 Alexiad 3:82.26–27, 52.23–24. 77 For the king of Frankia, see Alexiad 2:213.1, 3:53.15; for the count of Frankia, 3:27.21; for “all the counts” in Frankia, 2:207.22.

Alexander Kazhdan [ 91 ] dently a Norman contingent; the Frankish navy a Norman fleet; and the famous Frankish warrior Roger a Norman commander.78 Choniates realized the disjunction between the country called Frankia and the people called Franks, and introduced the ethnonym Fraggi´skoi to designate the French,79 although he had no clear perception of their nationality and described them as formerly called Flemish, Flami´one".80 Kinnamos was more archaizing and termed the French Germanoi, while the Germans in his history were Alemans. According to Anna, Bohemond and his “Franks” caused serious troubles to the empire of Alexios but, unlike the Latins, they are rarely characterized in a negative fashion by the princess-historian. Two exceptions are her reference to their irascibility and their habit of breaking oaths.81 Unlike other Latins, she perceived the Franks as “ours,” at least in theory, and mentions a contingent of Franks that followed her father during a military expedition.82 The Latins in general were perceived by Anna as dangerous aliens, while the Normans (“Franks”), in her day, were more than simple mercenaries; some of them entered the ranks of the Byzantine ruling class. The imperial chrysobulls of the second half of the eleventh century often contain a formula exempting monasteries from billeting foreign military contingents.83 This formula avoids the generic term “Latins” as well as archaic designations such as Celts, Alemans, or Germanoi, but itemizes various specific ethnonyms, including Franks. The edict of 1060 comprises a limited list of foreigners: Varangians, Rus’, Saracens, and Franks; in 1073 the enigmatic Koulpingoi (Kolbjagi of the Lex Rossica) are added, in 1074 Bulgarians. In 1082 the formula includes Rus’, Varangians, Koulpingoi, English, and Germans (Nemitzoi ); the absence of the Franks (the Normans of South Italy) in the heat of the war against Robert Guiscard is understandable. They reappear in the chrysobull of 1086, and the edict of 1088 supplements the list with two new ethnic groups—Alans and Abasgians (most probably Georgians in general). The role of individual “Latins” or Westerners in Byzantine society of the eleventh and twelfth centuries has been well studied,84 and we may enumerate the outstanding “Latins” active in the empire during this period. According to Kekaumenos, Byzantine emperors of the first half of the eleventh century (he explicitly names Romanos III) did not grant any Frank or Varangian the title of patrikios.85 Even Kekaumenos, however, men78 For the army under Bohemond, see Alexiad 3:125.12; for the Frankish navy, 1:147.20; for Roger, 3:101.14, 117.20–21, 120.25. 79 On Frankia, see NikChon 417.67; for Frankiskoi, see 588.21, 32; 596.43; 597.64; 647.4. 80 NikChon 553.93. 81 On Norman irascibility, see Alexiad 3:162.18–19; on breaking oaths, 3:146.8. 82 Alexiad 1:24.18–19. 83 For a survey of them, see A. Kazhdan and B. Fonkic´, “Novoe izdanie aktov Lavry i ego znachenie dlja vizantinovedenija,” VizVrem 34 (1973): 49. 84 R. Janin, “Les Francs au service des Byzantins,” EO 29 (1930): 61–72; Marquis de la Force, “Les conseillers latins du basileus Alexis Comne`ne,” Byzantion 11 (1936): 153–65; A. Kazhdan, Social’nyj sostav gospodstvujushchego klassa Vizantii XI–XII vv. (Moscow, 1974), 200–205, 212–18; E. Jeffreys, “Western Infiltration of the Byzantine Aristocracy: Some Suggestions,” in The Byzantine Aristocracy, ed. M. Angold (Oxford, 1984), 202–10; W. McQueen, “Relations between the Normans and Byzantium: 1071–1112,” Byzantion 56 (1986): 436 f, 444–47, 465–67. 85 G. G. Litavrin, Sovety i rasskazy Kekavmena (Moscow, 1972), 280.5–6 (hereafter Kekaum.).

[ 92 ] Latins and Franks in Byzantium tions a certain Peter, nephew of “the basileus of Frankia” (or of “the king of the Germans”), who settled in Byzantium and whom Basil II appointed spatharios and domestikos of the exkoubitai of Hellas.86 Peter is otherwise unknown, but his title and office are relatively minor ones, as Kekaumenos himself emphasizes. Attitudes toward Westerners changed soon after the reigns of Basil II and Romanos III. In the mid-eleventh century, Herve´ Frankopoulos (lit. “a young Frank” or “the son of a Frank”) was one of the leading commanders of the Byzantine army:87 Skylitzes relates how he requested the title of magistros (i.e., one level higher than patrikios), but Emperor Michael VI turned him down with contempt.88 Later Herve´ reached his goal: on a seal he is named magistros, vestes and stratelates of the Orient.89 He headed a contingent of Franks and owned an estate called Dagarabe located in Armeniakon.90 It is probable that “the illustrious magnate called Francopoulos,” known from the chronicle of Matthew of Edessa, who defeated the Turks and was eventually drowned in the Mediterranean by the order of [Constantine X] Doukas,91 was the same Herve´. During the reign of Romanos IV Diogenes, “the Frank Krispinos,” that is, Robert Cre´pin, of the lineage of Grimaldi, had commanded the right wing of the Byzantine army in the battle against the rebel Khachatur;92 later he acted independently and clashed with the imperial troops. He was arrested and exiled to Abydos, and in vengeance his soldiers revolted and pillaged Mesopotamia. Attaleiates and, following him, the Continuator of Skylitzes call Cre´pin a Latinos from Italy.93 The third outstanding Norman commander in Byzantine service at the same time was Roussel de Bailleul, Ourselios/Rouselios of Greek texts. Bryennios calls him a Frank, a lieutenant of Cre´pin, while in Attaleiates and in the Continuation of Skylitzes Roussel appears as a Latin in charge of a Frankish contingent of four hundred men.94 In another passage, however, Attaleiates seems to have applied to Roussel a different ethnic term, Frankos.95 Anna Komnene defines Roussel as a Celt or barbarian, but does not specify him as a Frank.96 On a seal he is titled vestes and defined as Frankopoulos,97 but later he became his own master and refused the high title of kouropalates offered to him by Emperor Michael VII.98 An influential commander, he was the force behind the proclamation of the Caesar John Doukas as emperor in 1074.99 86

Kekaum. 280.15–22. G. Schlumberger, Re´cits de Byzance et des croisades (Paris, 1922), 71–77. 88 Skyl. 484.41–47. 89 G. Schlumberger, Sigillographie de l’empire byzantin (Paris, 1884), 656–60. 90 Skyl. 485.52. 91 Trans. A. E. Dostourian, Armenia and the Crusades: Tenth to Twelfth Centuries (Lanham, Md., 1993), 99–101. 92 Nice´phore Bryennios, Histoire, ed. P. Gautier (Brussels, 1975), 135.5 (hereafter Bryen.). 93 Attal. 122.22, 170.21; E. Th. Tsolakes, JH sune´ ceia th'" cronografi´a" tou' jIwa´ nnou Skuli´tsh (Thessaloniki, 1968), 134.2–3 (hereafter Skyl. Cont.). 94 Bryen. 147.23–24; Attal. 183.11–12, 269.4; Skyl. Cont. 157.6–7. 95 Attal. 186.18. 96 Alexiad 1:10.4, 12.1. 97 Schlumberger, Sigillographie, 660–63. 98 R. Guilland, “Etudes sur l’histoire administrative de l’empire byzantin,” Byzantina 2 (1970): 220. 99 See on him, C. Diehl, “Les aventures d’un chef normand en Orient,” Revue des cours et confe´rences de la Faculte´ des lettres de Paris 20 (7 Dec. 1911): 172–88; D. Polemis, “Notes on the Eleventh-Century Chronology,” BZ 58 (1965): 66–68. 87

Alexander Kazhdan [ 93 ] Less famous than these generals was the “Frank Otton” (Otos), who was granted land by Nikephoros III; some time later, when Otton joined the conspiracy of a certain Pounteses, his land was confiscated.100 As for Pounteses, he is mentioned by Anna Komnene;101 he is usually identified as Raoul de Pontoise, a Norman commander who deserted to the Byzantines soon after 1082, but this identification does not seem to be valid.102 The stream of Norman mercenaries continued under Alexios I. According to Anna, the emperor attracted to his camp Guido (Guy Hauteville), a son of Robert Guiscard (Bohemond’s stepbrother), having promised the Norman money, honors, and a brilliant marriage.103 Guido became Alexios’ councillor: in 1098, when the emperor learned that the Crusaders in Antioch were in a predicament, he summoned first of all “Guido, Bohemond’s brother.”104 In the legendary Tale of Antioch, Guido is presented as a close friend and seneschal of the emperor.105 The brothers Raoul106 and Roger deserted to Byzantium ca. 1080. Anna defined the latter as a noble Frank,107 while Nicholas Kallikles devoted to him a poem “On the grave of the sebastos Rogerios.” According to this poem, Alexios I conferred upon Roger, who originated “from the Frankish land,”108 “seas of gold,” glory, and a high title, as well as imperial connections (he was married to a woman from the lineage of the Dalassenoi, related to the Komnenoi). Possibly he is the same person as Roger, son of Takoupertos (Dagobert), one of the dignitaries who signed the treaty of Devol in 1108.109 He was, most probably, the founder of the NormanoByzantine lineage of the Rogerioi.110 More confusing is the evidence concerning Raoul. Albert of Aachen relates that Rudolf (Raoul?), “the Wolf’s Skin,” and Rotger (Roger), son of Dagobert, Franks by origin, served as Alexios I’s envoys in 1096.111 If Dagobert’s son is, most probably, the abovementioned founder of the lineage of Rogerioi, some scholars consider Rudolf his brother;112 indeed, Roger had a brother Raoul. Albert, however, does not imply that the envoys of 1096 were brothers. It is also difficult to prove that Humbert, son of Graoul, also a signer of the treaty of Devol,113 was a descendant of Roger’s brother. The Raoul

100

Actes de Lavra, vol. 1, ed. P. Lemerle, A. Guillou, et al. (Paris, 1970), no. 45.9–13 (hereafter Lavra). Alexiad 2:22.10–29. 102 See objections by Marquis de la Force, “Les conseillers,” 161 f, supported by McQueen, “Relations,” 444 f. 103 Alexiad 2:51.10–12. 104 Histoire anonyme de la premie`re croisade, ed. L. Bre´hier (Paris, 1924), 142. 105 La Chanson d’Antioche (Paris, 1848), 1:79, 89, 2:156. 106 On him, see F. Chalandon, Essai sur le re`gne d’Alexis Ier Comne`ne (Paris, 1900), 64 and esp. n. 2. 107 Alexiad 3:101.14. 108 Nicola Callicle, Carmi, ed. R. Romano (Naples, 1980), no. 19.10. 109 Alexiad 3:138.28. On this treaty, Ja. N. Ljubarskij and M. M. Frejdenberg, “Devol’skij dogovor 1108 g. mezhdu Alekseem Komninom i Boemundom,” VizVrem 21 (1962): 260–74. The authors abstain from identifying the signers, “natives of the western lands.” 110 On this lineage, see L. Stiernon, “Notes de la titulature et de prosopographie byzantines: A propos de trois membres de la famille Rogerios (XIIe sie`cle),” REB 22 (1964): 184–98. 111 PL 166:415; cf. F. Do¨lger, Regesten der Kaiserurkunden des ostro¨mischen Reiches, vol. 2 (Munich, 1925), no. 1188. 112 A. Ch. Chatzes, OiJ Raou´ l, Ra´ l, Ra´ lai (Kirchhain N.-L., 1909), 10. 113 Alexiad 3:139.1. 101

[ 94 ] Latins and Franks in Byzantium family became a mixed Normano-Byzantine lineage, but they did not play a part commensurate with that of the Rogerioi: in 1138 a certain Leo Raoul compiled a manuscript of Theophylaktos’ commentary on the Gospels;114 much later, in 1195, the sebastos Constantine Raoul-Doukas, a high-ranking relative of the ruling dynasty, was a supporter of Alexios III’s putsch.115 During the reign of Alexios I, many other Westerners held high positions in the Byzantine army and at the imperial court. Constantine [H]oumbertopoulos, Robert Guiscard’s brother, commanded the tagma of the Franks.116 The sebastos Marinos of Naples, in Anna’s words a member of the family of Maistromilioi, was the first among the Byzantine signers of the treaty of 1108;117 he was followed by Roger, son of Dagobert (see above), and Peter Aliphas, described by Anna as a warrior glorious in battles and faithful to the emperor.118 Modern historians of the Crusades usually identify him as the Provenc¸al knight Peter of Aulps.119 The Anonymous History of the First Crusade calls him Petrus de Alpibus,120 thus making the identification plausible; in Orderic Vitalis, however, he appears as Peter de Alfia (that has nothing to do with Aulps) and possibly (in another passage) Petrus Francigena121 (⫽ Greek “Frankopoulos”). Accordingly Marquis de la Force considered him a close relative of Robert, Norman lord of Chiazza and Alifa. The Anonymous History states that omnes seniores granted Peter “a certain beautiful and rich city,” usually identified as Placentia, formerly Comana.122 Peter was the founder of the Byzantine lineage of Petraliphai: Niketas Choniates narrates that four Petraliphas brothers, men of Frankish descent, who settled in Didymoteichon in Thrace, distinguished themselves during the siege of Kerkyra in 1149.123 The later Song of Belisarios preserved a tradition that the Petraliphai were an insignificant family from Didymoteichon,124 but in the days of Manuel I the Petraliphai were far from insignificant: the sebastos Alexios, vestiarites in 1166, was a general,125 as was Nikephoros.126 Later (ca. 1200?), Nikephoros Komnenos-Petraliphas conferred on the monastery of Xeropotamos a sigillion127 in which he is twice titled sebastokrator; there is no ground to question his title, as did B. Ferjancˇic´.128 The sigillion confirms the donation by Nikephoros’ grand114

H. von Soden, Die Schriften des Neuen Testaments, vol. 1.1 (Berlin, 1902), 262. On him, see D. Polemis, The Doukai (London, 1968), 172 f. 116 Alexiad 1:152.2–4. 117 Alexiad 3:101.9–10, 138.28. His name is a distortion of the Latin term magister militum. 118 Alexiad 3:101.15–17. 119 For instance, R. Grousset, Histoire des croisades, vol. 1 (Paris, 1934), 31 n. 2; S. Runciman in A History of the Crusades, ed. K. M. Setton, vol. 1 (Philadelphia, 1955), 297; M. Zaborov, Vvedenie v istoriografiju krestovykh pokhodov (Moscow, 1966), 225. 120 Histoire anonyme de la premie`re Croisade, ed. L. Bre´hier (Paris, 1924), 60. 121 The Ecclesiastical History of Orderic Vitalis, ed. M. Chibnall, vol. 5 (Oxford, 1975), 66; cf. vol. 4 (1973), 34. 122 S. Runciman, A History of the Crusades, vol. 1 (Cambridge, 1953), 191. 123 NikChon 83.81–83. 124 E. Follieri, “Il poema bizantino di Belisario,” in La poesia epica e la sua formazione (Rome, 1970), 632.226–27. 125 PG 140:236; cf. Kinnam. 292.15. 126 Kinnam. 260.22–23. 127 F. Do¨lger, Aus den Schatzkammern des Heiligen Berges (Munich, 1948), no. 33 ⫽ Actes de Xe`ropotamou, ed. J. Bompaire (Paris, 1964), no. 8. 128 B. Ferjancˇic´, “Sevastokratori u Bizantiji,” ZRVI 11 (1968): 147 n. 26. 115

Alexander Kazhdan [ 95 ] mother Maria Tzousmene Komnene; according to J. Bompaire, Maria (died in 1144) was a daughter of the emperor John II; she married Nikephoros’ father. Another Petraliphas, baptized John, was also a sebastokrator, if we believe the late and legendary vita of Theodora Petraliphaina: under the dynasty of the Angeloi he served as a governor of Macedonia and Thessaly.129 Anna lists some other mercenaries who signed the treaty of 1108: Richard of Principat (Printzitas), a brother of Robert Guiscard, and several obscure persons: Guillaume Gantzes, Joffroy Males, and Paul Rhomaios. Marquis de la Force thinks that the latter was not a Byzantine by birth but originated from Rome, a thesis that cannot be substantiated. It is a common perception that Alexios’ grandson Manuel I was a Latinophile.130 Both Niketas Choniates and Eustathios of Thessalonike stressed that Manuel attracted to his court foreigners from various corners of the world; the difference between the two writers consists only in the evaluation of the emperor’s policy. Niketas is unhappy at the privileges bestowed upon the “half-barbarians” (mixoba´ rbaro"), whereas the native Rhomaioi, experienced warriors, were neglected; on the contrary, Eustathios praises Manuel for summoning men from different countries.131 Some Westerners confirm this evidence of the Greek authors. William of Tyre, a writer well informed about the situation in Constantinople,132 eulogizes Manuel I for his extreme love of the Latinus populus; the late emperor, says the archbishop of Tyre, scorned the effeminate Greculi and trusted only the Latins, whose merits and strength he highly appreciated.133 Robert of Auxerre is even more straightforward: according to him, Manuel led only Latins on his expeditions and crowned them with the highest honors of his palace.134 The Continuator of Sigebert of Gembloux writes that Manuel’s son [Alexios II] imitated his father and like him spurned Greeks and loved Latins; he used their advice and support, their theology and legal knowledge.135 In the same vein, Walter Map asserts that Manuel assembled in his capital the Franks of all nations.136 But strangely enough, prosopographical data to some extent contradict or place in question the unequivocal evidence of Greek and Western sources.

129 For the vita of Theodora see PG 127:904. On John Petraliphas, see D. Nicol, The Despotate of Epirus (Oxford, 1957), 215 f. For some additional notes on the Petraliphai, see E. Branousse, review of Guillaume de Pouille, in Athena 65 (1961): 327, and E. Patlagean, “Une sainte souveraine grecque: Theodora impe´ratrice d’Epire (XIIIe sie`cle),” BSl 56 (1995): 453–60. 130 G. Ostrogorskij, Geschichte des byzantinischen Staates (Munich, 1963), 314 f, speaks of Manuel’s “westliche Orientierung” and calls him “Westler.” On the scholarly evaluation of the Komnenoi, see A. Kazhdan, “Zagadka Komninov,” VizVrem 25 (1964): 53–98. Important corrections to the common concept of Manuel’s policy with regard to the Westerners were suggested by P. Magdalino, The Empire of Manuel I Komnenos, 1143– 1180 (Cambridge, 1993), 107 f, who emphasizes that the mixture of races was typical not only of the empire of Constantinople but of several other states of the 12th century. 131 NikChon 209.44–49; Fontes rerum byzantinarum, ed. V. Regel and N. Novosadskij (St. Petersburg, 1892– 1917; repr. Leipzig, 1982), 81.5–11. 132 See on him, A. Kazhdan and M. Zaborov, “Gijom Tirskij o sostave gospodstvujushchego klassa v Vizantii (konec XI–XII v.),” VizVrem 32 (1971): 48–54. 133 Guillaume de Tyr, Chronique, chap. 22:11.14–19, ed. R. Huygens (Turnhout, 1986), 1020. 134 MGH, SS 26:246.49–51. 135 MGH, SS 6:421.40–42. 136 Walter Map, De nugis curialium (Oxford, 1914), 87.

[ 96 ] Latins and Franks in Byzantium Among the leading military commanders under Manuel I as well as his father John II, Western mercenaries were infrequent. Only two significant personages can be ranked among them, but their role was ephemeral and they were active only at the very end of Manuel’s reign: first, in 1176 Baldwin, a brother of Manuel’s second wife, commanded the right wing of the Byzantine army at Myriokephalon, where he fell in battle;137 second, Renier/Rainerius of Montferrat moved to Constantinople in 1179, married Maria, Manuel’s daughter, the next year, and was invested with the title of caesar and given the new name of John Komnenos.138 Some Western chroniclers affirm that Renier was granted as a “fief” the city of Thessalonike and donned a regal crown.139 Most probably, Western informants confused the caesar’s coronet with the imperial headgear. The legend of the regal crown may have originated in Maria’s title, “porphyrogenitas empress” that she bore until the coronation of her younger brother Alexios II.140 In a manuscript in St. Petersburg (Public Library, Greek no. 512), there is a note that characterizes the “purple-born” Maria as the ruler side by side with her brother-emperor.141 The legend also could have perversely reflected a later event, the establishment of the Kingdom of Thessalonike conferred upon Boniface of Montferrat. We may surmise that Renier obtained some rights over Thessalonike,142 or that the couple stayed in Thessalonike for some time after the wedding,143 but all this is no more than a guess. William of Tyre, the most reliable Latin informant concerning affairs in Constantinople, knows nothing about the fief of Thessalonike, and relates only that Manuel married Maria to Renier and rebaptized him John in honor of John II Komnenos.144 P. Magdalino, referring to a document issued by Boniface in 1204, assumes that Renier actually received the “fief” of Thessalonike;145 Boniface, however, speaks in the act of Manuel’s grant of the totum feudum given to his father ( patri meo), that is, William VI of Montferrat, and not to Renier, who was Boniface’s brother. Furthermore, the document does not imply that William’s fief incorporated Thessalonike; Thessalonike appears only in the next clause dealing with the possessiones of the inhabitants of this city. After Manuel’s death Maria and her spouse acquired extraordinary political influence, and even plotted to demote Alexios II (or at least his favorite, the protosebastos Alexios

137

NikChon 180.85–86, 181.7–13. R. Guilland, Recherches sur les institutions byzantines, vol. 2 (Amsterdam, 1967), 32. Cf. G. Usseglio, I marchesi di Monferrato in Italia ed in Oriente durante i secoli XII e XIII, vol. 1 (Monferrato, 1926), 156 f. 139 Roberti de Monte (⫽ Robert of Torigny) cronica, MGH, SS 6:528; cf. as well the chronicles of Sicard of Cremona (MGH, SS 31:173), Salimbene (MGH, SS 32:3), and Albert Milioli (MGH, SS 31:643). 140 See, for instance, the act of 1167: N. Wilson and J. Darrouze`s, “Restes du cartulaire de Hie´ra,” REB 26 (1968): 24.7–12. 141 E. E. Granstrem, “Katalog grecheskikh rukopisej leningradskikh khranilishch,” VizVrem 23 (1963): 173, no. 321. 142 H. Gre´goire, “Empereurs belges ou franc¸ais de Constantinople,” Bulletin de la Faculte´ des lettres de Strasbourg 25.7 (1947): 24 f; C. Brand, Byzantium Confronts the West, 1180–1204 (Cambridge, Mass., 1968), 19. A. Laiou, “A Byzantine Prince Latinized: Theodore Palaeologus, Marquis of Montferrat,” Byzantion 38 (1968– 69): 387, describes Renier’s right as pronoia. 143 S. Runciman, “Thessalonica and the Montferrat Inheritance,” Gregorios ho Palamas 42 (1959): 28. 144 Guillaume de Tyr, Chronique, ed. Huygens, chap. 22:4.38–39, p. 1010. 145 G. L. F. Tafel and G. M. Thomas, Urkunden zur a¨lteren Handels- und Staatsgeschichte der Republik Venedig, vol. 1 (Vienna, 1856), 513; see Magdalino, The Empire, 100 n. 311. 138

Alexander Kazhdan [ 97 ] Komnenos). After bloodshed in the streets of Constantinople caused by their attempt in 1181 or 1182, the parties reached an agreement. The triumph was of short duration: both Renier and Maria perished as soon as Andronikos I took power. We know less about other Western contemporaries of Manuel I, unless we count such Hellenized families as the Rogerioi, Petraliphai, and Raoul. To them we may possibly add the lineage of Gifard (Giphardos), which bore a non-Greek name and could be of Norman descent.146 They served in the army: a certain Gifard was a hypostrategos147 and participated in the war against the Serbs in 1149/50. Alexios Gifard (the same man or another?) was a correspondent of George Tornikes; he held the post of governor of Thrakesion in the 1150s. Ca. 1158 Alexios Gifard went on a diplomatic mission to the sultan Kilic¸ Arslan II.148 Leo Gifard is known only by a seal of the twelfth or thirteenth century; neither his title nor office is indicated.149 The lineage of Gidoi comes to the fore after Manuel’s death: Alexios Gidos held the high post of domestikos of the Orient under Andronikos I and domestikos of the West under Isaac II; in 1194 he led Byzantine troops against the Bulgarians.150 Another Gidos, Andronikos, unsuccessfully commanded a contingent at the beginning of the thirteenth century.151 No text treats the Gidoi as foreigners; their name, however, resembles that of Guido, but it is not possible to demonstrate that they were descendants of the son of Robert Guiscard. W. Hecht even questions the Western origin of the Gidoi.152 Our sources do not always bother to name a Westerner whom they describe. Niketas Choniates narrates how Emperor John II ordered his nephew John to give his horse to an anonymous Western warrior who is characterized as a notable from Italy and then as a Latin.153 Another anonymous Westerner, identified as an “axe-bearer of Celtic origin,” is said to have arrested the fugitive sons of John Batatzes who revolted against Andronikos I.154 In other cases the name is preserved but biographical information is minimal: thus a document of 1162 mentions a threshing-floor in the area of Thessalonike that belonged to the archon John Moliscott;155 the name sounds “Western,” but we know nothing about this modest official and landowner. Comparison of the information about Western generals under the “Latinophile” Manuel I with data on commanders of the second half of the eleventh century leads to an unexpected conclusion. Leaving aside those “Latins” who became assimilated into the Byzantine milieu and only preserved their “barbaric” family names, the Westerners under Manuel lost rather than gained social status: not only did the number of highranking military commanders diminish while the highest military positions fell into the J. Darrouze`s, in Georges et De`me`trios Tornike`s, Lettres et discours (Paris, 1970), 150 n. 1. Kinnam. 94.7; cf. 108.21. 148 Kinnam. 176.10. 149 V. Laurent, Les bulles me´triques dans la sigillographie byzantine (Athens, 1932), no. 59. 150 Guilland, Institutions, 1:408 f. 151 NikChon 641.45–46. 152 W. Hecht, Die byzantinische Aussenpolitik zur Zeit der letzten Komnenenkaiser (Neustadt a.d. Aisch, 1967), 85. 153 NikChon 35.40–51. 154 NikChon 264.64. 155 Lavra, 1: no. 64.36–37. 146 147

[ 98 ] Latins and Franks in Byzantium hands of the Komnenian clan, but the Westerners lacked the independence typical of such pugnacious mercenaries as Herve´ and Roussel; neither are we told about any attempts to create feudal principalities such as had been the aim of several Norman generals before the Komnenian upsurge. The Komnenoi tamed the relentless spirit of earlier “Franks” and transformed them into obedient servants of the empire. Let me emphasize that this conclusion does not mean that the number of Westerners decreased under Manuel. Just the opposite occurred, for the figures we have from the twelfth century—with all their exaggeration—show how enormous was the influx of Westerners into Byzantium. Eustathios of Thessalonike thought that the number of Latins in Constantinople reached sixty thousand.156 Gregory Abu al-Faraj (Bar Hebraeus) counted thirty thousand “Frankish” merchants in the capital, adding, however, that “on account of the great size of the city they were not conspicuous therein.”157 According to Western annals, twenty thousand Venetians fled Constantinople after the hostilities of the 1160s; supposedly more than ten thousand Venetians were arrested in Byzantium in 1172.158 The Genoese chronicle of Caffaro contains more modest figures, calculating the Pisans and Genoese in the “royal city” as one thousand and three hundred respectively.159 Thus the number of Westerners in the empire of the twelfth century was high, but their social status seems to have changed: few were high-ranking military commanders, merchants formed the majority, and prosopographical data demonstrate that the Westerners played an important part in diplomatic service as emissaries and interpreters and as ideological advisers of Manuel. According to my calculations,160 in the mideleventh century at least fifteen “Latin” generals are mentioned (and only one translator); during the reign of Alexios I, twelve generals and five councillors, some of whom could exercise military functions as well. The social pattern of known Westerners in the administration of Manuel I is different: six military commanders, two or three civil officials, five diplomats, and an interpreter. For the short span at the end of the twelfth century, besides the ephemeral activity of Conrad of Montferrat161 (Renier’s brother) and that of the lineage of Gidoi, I traced three Westerner-generals (the evidence is pretty obscure), six diplomats, one translator, and one patriarch of Venetian descent. The list of “Latins in imperial service” during the reign of Manuel I as independently established by P. Magdalino162 follows the same pattern. Besides the Rogerios, Petraliphas, and Raoul families, Magdalino counts two Westerners in military service. The first is the Genoese knight William sent in 1156 to recruit Genoese pirates, that is, not a direct military command; the second is Roger “Sclavone,” dux of Dalmatia and Croatia, lidzios of Manuel I in 1180. Ja. Ferluga identified him as Roger “Sclavus,” the natural son of the count

156

Eustazio di Tessalonica, La espugnazione di Tessalonica, ed. S. Kyriakidis (Palermo, 1961), 34.2. The Chronography of Gregory Abu’l Faraj, trans. E. A. W. Budge, vol. 1 (Oxford, 1932), 358. 158 Historia ducorum Veneticorum, MHG, SS 14:78; see H. F. Brown, “The Venetians and the Venetian Quarter in Constantinople to the Close of the XIIth Century,” JHS 40 (1920): 82. 159 MGH, SS 18:33. 160 Kazhdan, Social’nyj sostav, 217 f. 161 On him, see Th. Ilgen, Markgraf Conrad von Montferrat (Marburg, 1880); Usseglio, I marchesi di Monferrato, 1:150–53; Brand, Byzantium, 18–20, 80–84. 162 Magdalino, The Empire, 222 f. 157

Alexander Kazhdan [ 99 ] Simon, who in 1161 revolted against the Norman king William I and soon went into exile.163 The Norman origin of the man dubbed “Sclavone” is not ascertained, albeit not impossible; it is noteworthy that Roger, like Renier and Baldwin, was active at the very end of Manuel’s reign. Magdalino lists three interpreters, including Leo Tuscus who, with his brother Hugo Eteriano, was not simply one “of Manuel’s agents”: the brothers, professional theologians, served as the emperor’s spiritual advisers.164 Eight men in Magdalino’s inventory are emissaries or diplomatic envoys, the best known of them being the Genoese Baldovino Guercio.165 There is some possibility that the young Baldovino remained in Byzantium as a soldier in the service of John II, although it is hard to imagine that the man who negotiated with Alexios III in 1193 was the same Baldovino who accompanied the Genoese embassy to John II in 1142. His activity is primarily connected, however, with diplomatic missions sent both by Manuel I to the West and by the Ligurian republic to Constantinople. Manuel I was a Latinophile but only to a certain extent: he entrusted the command of the army primarily to his relatives, although in his late years some Western generals may appear in top positions. It is even possible that Turkish commanders had a greater role in Manuel’s army than their Western counterparts.166 Latins of his reign were first of all ordinary soldiers and merchants, while in the administration of the empire they held posts primarily in the diplomatic service. Summing up, we may formulate the following observations. (1) Up to the eleventh century, the Byzantines treated the western world as consisting of separate countries and peoples or tribes; the idea of the unity of the West developed only in the eleventh century and was expressed in the introduction of the generic term Latinoi that encompassed all western peoples. Another generic ethnonym, Keltoi, popular with Anna Komnene, did not withstand the test of time. (2) The term Frankoi was used as a specific designation of individual peoples; to begin with, it was applied to the Germanic tribes settled on the territory of ancient Gallia, then to the population of Gallia in general, and eventually to the Normans of Italy, reflecting the historical fact of their invasion from the land north of the Alps. (3) The Franks/Normans penetrated Byzantine military service en masse in the eleventh century, when Norman mercenaries assumed a leading role in the Byzantine army. Their generals, using ethnic contingents as their base, were more or less independent; they attempted to obtain not only high titles but also territories, and they dared to revolt against the central administration. (4) The Komnenoi managed to suppress this 163 J. Ferluga, Byzantium on the Balkans (Amsterdam, 1976), 209 f and 420; cf. D. Abulafia, “The Crown and the Economy under Roger II and His Successors,” DOP 37 (1983): 13. Previously, Ferluga, Vizantijska uprava u Dalmaciji (Belgrade, 1957), 141 f, was inclined to consider Roger a Slav and a feudal lord of Dalmatia; see, however, objections in a review by M. Klaic, Historijski zbornik 13 (1960): 254. 164 See A. Dondaine, “Hughes Ethe´rien et Le´o Toscan,” Archives d’histoire doctrinale et litte´raire du moyen-aˆge 19 (1952–53): 67–134; cf. his “Hugues Ethe´rien et le Concile de Constantinople de 1166,” HJ 77 (1958): 473–83. 165 On him, see G. W. Day, Genoa’s Response to Byzantium, 1155–1204 (Urbana-Chicago, 1988), 110–13. Day’s statement that Baldovino received a pronoia from Manuel cannot be substantiated, but he was granted some estates. 166 On the Turks under John II and Manuel I, see C. Brand, “The Turkish Element in Byzantium, Eleventh–Twelfth Centuries,” DOP 43 (1989): 5–11.

[ 100 ] Latins and Franks in Byzantium independent spirit of the Norman mercenaries. On the one hand, the Franks entered the elite of Byzantine society, entered into marital contracts with members of the Komnenian clan, created mixed families, and became Hellenized. On the other hand, after the reign of Alexios I, foreigners were deprived of their commanding role in the army and exchanged their military career for diplomatic service; possibly, the last years of Manuel I witnessed the reverse process, and this return to the eleventh-century situation continued during the reign of Andronikos I and the dynasty of the Angeloi. The mercenaries (like the eunuchs),167 who were pushed away from the highest echelons of Komnenian society, then began to recover the positions they had lost under the chivalric Komnenoi. Dumbarton Oaks

167 On the social destiny of eunuchs that chronologically parallels that of Westerners, see A. Kazhdan, “Sostav gospodstvujushchego klassa v Vizantii XI–XII vv. Anketa I chastnye vyvody, VI: Evnukhi,” Anticˇnaja drevnost’ i srednie veka 10 (Sverdlovsk, 1973): 184–94.

The “Wild Beast from the West”: Immediate Literary Reactions in Byzantium to the Second Crusade Elizabeth Jeffreys and Michael Jeffreys The purposes of this study are two: the first and more important is to draw attention to two scrappily published and largely ignored Byzantine poems on the Second Crusade.1 They were written for eyewitnesses of events, with a little hard information caught up in a web of allusive references that require interpretation, typical products of the propaganda machine of Manuel I Komnenos. But the rhetoric is worth analysis, for there is evidence that both poems were written and very likely given some kind of public performance within days, certainly within weeks, of the passage of the Crusade. They may thus be taken as examples of immediate Byzantine reactions to the Crusaders. The less significant purpose is to examine the operation of some of the rhetorical structures of the poems, which, it seems to us, give an unusual degree of insight into the way a twelfth-century rhetorician might use his skills to show his audience how to react to contemporary events. The poems concerned are nos. 20 and 24 in the huge series of poems attributed to “Manganeios Prodromos,” a conventional name invented by modern scholarship to avoid the difficulties of the word anonymous.2 This study assumes that Manganeios Pro1 E. Miller, Recueil des historiens des croisades: Historiens grecs, vol. 2 (Paris, 1881) (hereafter Miller, Recueil), edited only extracts useful for his purposes, some very short, marking poem 20 as from fols. 30r–33r of the manuscript, poem 24 from fols. 35r–37r. Miller wrongly includes in poem 20 the 122 lines of poem 19, which is in a different meter and on another subject. Thus the line numbers he provides for substantial extracts from poem 20 should be reduced by 122 (his numbers are marked here with the prefix Mi). This error is not found in his transcription of Manganeios in Paris, BN suppl. gr. 1219. The extracts he publishes from the two poems are: Recueil, pp. 220–25 (poem 20.17–20 [no Mi line numbers], 56–266 [Mi 178–388]); Recueil, pp. 228–29 (poem 20.287–307 [Mi 409–29]); Recueil, p. 757 (poem 20.329–39 [Mi 451–61]); Recueil, p. 188 (poem 20.355–58 [no Mi line numbers]); Recueil, pp. 757–58 (poem 20.414–22 [Mi 536–44], 479–87 [Mi 601–9], 582–90 [Mi 702–10]); Recueil, pp. 758–59 (poem 24.1–2, 38–39, 88–93, 142–63, 189–90, 212–13, 240–66). 2 On Manganeios Prodromos, see S. Papadimitriou, “ JO Pro´ dromo" tou' Markianou' kw´ diko" XI 22,” Viz¨ B 24 (1975): 95– Vrem 10 (1903): 123–32; W. Ho¨randner, “Marginalien zum ‘Manganeios’ Prodromos,” JO 106; A. Kazhdan and S. Franklin, Studies on Byzantine Literature of the Eleventh and Twelfth Centuries (Berkeley, 1985), 87–93; R. Beaton, “The Rhetoric of Poverty: The Lives and Opinions of Theodore Prodromos,” BMGS 11 (1987): 1–28; P. Magdalino, The Empire of Manuel I Komnenos, 1143–1180 (Cambridge, 1993), esp. 494–500. Quotations from Manganeios Prodromos not taken from Miller, Recueil, are from our forthcoming edition of the corpus. To help elucidate the allusions, without which a modern reader cannot follow the poet’s meaning and intentions, we plan to provide in the edition an additional apparatus of “Keywords” on the page

[ 102 ] The “Wild Beast from the West” dromos was a different person from the most prominent court litte´rateur of the period, Theodore Prodromos.3 The name Manganeios acknowledges this poet’s ultimately successful campaign for admission to the hospice of the Mangana complex. Some of his poems were explicitly written for particular patrons, and many others suggest a framework of patronage. He wrote a great deal for the Sevastokratorissa Eirene, Manuel Komnenos’ sister-in-law, but his talents were also available to other patrons, particularly in the imperial administration.4 Indeed, poem 20 and especially poem 24, as we shall see, were probably imperial commissions. Both poems are textually straightforward. Like the bulk of the corpus they are preserved in one manuscript only, Marcianus graecus XI 22, a thirteenth-century compendium of twelfth-century court rhetoric, perhaps part of a Nicaean attempt to preserve or revive Komnenian court practices. Poem 20 deals with the progress of Conrad III’s German crusading army through Thrace in the autumn of 1147. It opens at the top of fol. 30r, in the middle of a speech: the beginning of the poem has been lost as the result of physical damage to the manuscript. Since the order of folios has also been disturbed, it is impossible even to guess how many lines might have disappeared. However, with 642 lines, poem 20 is already one of the longest in the corpus; it is unlikely that it was originally much longer. The meter is the Byzantine twelve-syllable, less common in Manganeios’ work than its other major meter, the fifteen-syllable. We have a subjective impression that Manganeios is less comfortable with the twelve-syllable than with the fifteen-syllable, since the syntax in the former is often strained and the meaning rather muddy. It is also worth mentioning that Manganeios in his surviving corpus repeats far more twelve-syllable lines than fifteen-syllables, as if he is determined to make maximum use of them once they have been forged.5 The lines run continuously, with no formal divisions into stanzas. Analysis by content, however, suggests that it falls into five sections of unequal length: Lines 1–266: Conrad has reached Philippopolis. The poem narrates clashes around the city (88 ff ) and at Adrianople (109 ff ), culminating in a long description of a below the text, giving references to a “Keyword glossary” at the back of the projected volume. Some of the glossary entries are used experimentally in this study. The poem numbers given here (and in the new edition) are those of E. Mioni, Bibliothecae Divi Marci Venetiae codicum graecorum manuscriptorum catalogus, vol. 3 (Rome, 1970), 116–25. 3 Most evidence for the distinction to be made between the two poets was set out by Papadimitriou, “Pro´ dromo",” 123–32, which our edition will repeat with a few additions. Doubts over Papadimitriou’s conclusions are expressed by Kazhdan and Franklin, Studies, 87–93. The centerpiece of their argument is a new interpretation of Manganeios, poem 37, which Papadimitriou and others (including ourselves) take as a reference to Theodore as a dead colleague—preventing at a stroke any attribution of the corpus to him. However, Kazhdan and Franklin’s reinterpretation of the poem was successfully attacked by R. Beaton, “Rhetoric,” 17–23. Beaton also goes on to equate the two Prodromos poets, but his own reinterpretation of poem 37 is no more convincing than that which he had demolished. The issue will not be examined further here. Internal evidence shows that the two poems under discussion were written in Constantinople in 1147; beyond that, for present purposes their authorship is immaterial. However, it is worth pointing out that the deconstruction undertaken here of the rhetorical structures that underpin the poems shows poor authorial control, which is characteristic of the Mangana corpus but looser than may be observed in most of the works securely attributed to Theodore. 4 Magdalino, Manuel, 510. 5 Many lines of poem 20, for example, recur in poem 108, and there is overlap between poems 67 and 69.

Elizabeth Jeffreys and Michael Jeffreys [ 103 ] flash flood in the camp at Choirobacchoi (131 ff ), where many Germans were swept away in a torrent of mud. Lines 267–286: A transition passage on the forging of the speech by the metaphor of hammering, providing an opportunity to turn attention to tongs as a metaphor for the Virgin, praising her for her miraculous intervention and demonstration of her concern for the imperial city. Lines 287–346: Conrad’s arrival in Constantinople and camping in the suburb of Pikridion, followed by further skirmishes with the Byzantines and his crossing of the Bosphoros to Damalis. Lines 347–473: An encomium of Emperor Manuel, in the persona of the poet. Lines 474–642: A further encomium of Manuel, this time put into the mouth of the City of Constantinople. Poem 24 is much shorter (284 lines), though the difference in length is partly compensated by the fact that it is in fifteen-syllable verse. It is divided into stanzas varying from four to fifteen lines each—a feature that elsewhere in Manganeios’ work suggests a fairly formal level of performance before a public audience. Each stanza is marked in the manuscript by a capital letter at the beginning of its first line. The title suggests that the poem is addressed to the emperor as from the City of Constantinople,6 as a reaction to the crusading armies. It may be summarized as follows (note that the stanzaic division imposes greater fragmentation of sense): Lines 1–22: The City addresses Manuel, congratulating him because the wild beasts, the Crusaders, have fled, terrified by her new teeth and revitalized appearance. Lines 23–50: Manuel has dressed the City like a bride with burgeoning flowers. Lines 51–70: Manuel’s good sense has seen off the enemy. Lines 71–85: The City has been well paid for the cost of bringing him up. Lines 86–105: He has shown the kings from Old Rome the vigor of New Rome. Lines 106–141: Manuel has proved unshakable in the face of attack. Lines 142–149: The crusading armies are depicted begging helplessly for food at the monastery of St. Mamas and on the opposite Asiatic shore of the lower Bosphoros. Lines 160–194: Manuel is the City’s savior, greater than Constantine or Alexander. Lines 195–228: An eclipse marked the eclipse of the kings; New Rome, invigorated by Manuel, is superior to Old Rome; Manuel has been aided by the Virgin. Lines 229–284: Manuel shines brilliantly, surpassing his father and grandfather. Similar events are narrated by Kinnamos and Choniates.7 The three accounts show considerable, though not complete, unanimity. Manganeios adds a few details and adopts a far more hostile point of view toward the Crusaders than the other two. Both Manganeios’ poems show a publicity agenda likely to appeal to employers (or potential employ6 The content of the poem supports the title (not always the case in this corpus): Apo j ` th'" po´ lew" eij" to` n ´a" kai` oJ Fraggi´a". aujtokra´ tora o”te kate´ labon oiJ rJh'ge" oJ Alamani j 7 Ioannis Cinnami Epitome rerum ab Ioanne et Alexio Comnenis gestarum, ed. A. Meineke (Bonn, 1836), 73–80; Nicetae Choniatae Historia, ed. J. L. van Dieten (Berlin–New York, 1975), 63–65. One suspects that the imperial equivalent of press releases for the period were available to all three writers.

[ 104 ] The “Wild Beast from the West” ers) in the imperial administration. But whatever spin he was putting on events, we should remember that he was writing as an eyewitness for an audience of eyewitnesses. When he states, for example, that Manuel replaced the City’s teeth under the threat of attack from the Crusaders and that her walls blossomed unseasonably in September,8 in spite of the vagueness of expression, we should assume some repair of the walls in 1147 and a spectacular display of fabrics to impress the Westerners. Such details command a high degree of credibility. It is the contemporaneity of his accounts that distinguishes Manganeios. In this connection, it is striking that the focus is on Conrad and the German army while the French forces and Louis VII are barely alluded to. The reason may perhaps be because the latter were more amenable, or maybe the events of their passage gave less scope for dramatic representation. When were these two poems written? A reply to this question may use only internal evidence, prefaced by an indication of Manganeios’ practice elsewhere. The preparation of a preliminary edition of the text of all 148 poems of the corpus has led us to the belief that he usually wrote very fast under the pressure of events, and that poems containing historical narrative may usually be dated shortly after the last event mentioned. The last ¨ sku¨activity noted in poem 20 is the Germans’ crossing to Damalis (near the modern U dar),9 and indeed the whole poem is a thanksgiving for the safe removal of this menace from the City’s territory. A poet who shows obvious glee at every sign of the discomfiture of the Germans in Europe could not have ignored the far greater disasters that met them in Asia, had the poem been written late enough to include them. In poem 24 there is a reference to the city’s reaction to the Germans’ arrival in September and to a recent eclipse, perhaps dated to the end of October.10 It would thus seem likely that both poems were written (and probably performed) around this date, as the Crusaders set out into

Peso´ nta" tou` " ojdo´ nta" mou maqo´ nta ta` qhri´a h«lqon wJ" a‘n qhreu´ swsi kai` katafa´ gwsi´ me⭈ ajll∆ wJ" Cristo` " jEmmanouh` l oJ Manouh` l oJ ne´ o" plhro´ donton ajne´ deixe kai` ne´ an me th` n grai'an, kai` tre´ mei tw'n ojdo´ ntwn mou ta` " neara` " ejkfu´ sei", kai` feu´ gei qh` r to` qh´ rama, ka‘n uJlaktei' kai` fu´ sei. 24.38–44 jExh´ nqhsan ta` tei´ch mou kaqa´ per aiJ koila´ de". ` kai` Fra´ gkoi⭈ jExe´ sthsan oiJ ble´ ponte" Alamanoi j h“nqoun wJ" kri´non to` leuko´ n, to` ko´ kkinon wJ" rJo´ don, wJ" kro´ kon to` cruso´ croion, to` pra´ sinon wJ" clo´ hn, th` n oujrano´ croun de` bafh` n wJ" a“nqo" uJaki´nqou, wJ" i“on to` lego´ menon to` katoxe´ o" crw'ma⭈ kai` pa´ nte" katepla´ ghsan th' parado´ xv qe´ a. Cf. Miller, Recueil, 758. 9 20.333–34: Miller, Recueil, 757 [Mi 455–56]. 10 24.47–48 jEgw` d∆ hjrxa´ mhn ajnazh'n ajpo` tou' Septembri´ou, kai` kata` to` fqino´ pwron cloa´ zw kai` nea´ zw. 24.195–99 Me´ ga" ajsth` r ejx oujranou' deinopaqh´ sa" prw´ hn to` fw'" aujtou' sune´ steilen ejpi` pollai'" tai'" w”rai", kai` th` n hJme´ ran e“deixen eJspe´ ran parado´ xw", kai` tw'n fwsth´ rwn tw'n rJhgw'n ejkei´nwn th'" eJspe´ ra" th` n e“kleiyin ejmh´ nusen ejklei´ya" ejn hJme´ ra. V. Grumel, La chronologie (Paris, 1958), 466, lists a solar eclipse for 26 October 1147. 8

24.8–13

Elizabeth Jeffreys and Michael Jeffreys [ 105 ] Asia Minor, and certainly before Conrad was rescued at Ephesos and brought back to Constantinople, where he celebrated Christmas.11 The question of performance is often an insoluble problem in connection with twelfth-century court poetry. Presumably these two texts were to be performed at some gathering or gatherings in Constantinople to celebrate the departure of the Crusaders, since poem 20 uses ejnqa´ de, “here,” in a way that can only mean the City.12 Manganeios could normally not guarantee his work getting a hearing, though in the case of poem 20 he is optimistic, for the speech refers to its own performance.13 It is hard, however, to deduce from either poem any details of the occasion. It is time to look at the two texts more closely, following both the agendas announced at the beginning of this study. We shall examine attitudes shown toward their main characters, Manuel, the City of Constantinople, and the Virgin on one hand, and Conrad and the crusading army on the other, trying to gauge Manganeios’ reactions to the pressure of events and the feelings he expected in his audience, insofar as we are able to discern them. At the same time we shall seek to evaluate the poet’s rhetorical techniques and the frameworks in which he presents the events and personalities of 1147. Since the poems have been published only in scraps, there will inevitably be many quotations in the footnotes to support our analysis. To reduce their number, we shall refer to E. Miller’s publication where we can, despite its faults, and we shall not quote words that seem to us purely conventional (for example, praise applicable to any emperor). The chief conclusion of the second head of the inquiry will point to Manganeios’ use of the smallscale rhetorical reaction at the expense of any overall ideological framework for the poem, leading to inconsistencies at various levels. Evidence for this will be either quoted in full or fully explained with a reference to Miller. Both poems are dominated by the figure of the emperor. The beginning of Manuel’s reign four years earlier had been a moment of great political tension. His father and two of his three older brothers had died in a brief space of time. Isaac—the remaining brother—and others had dreamed of forestalling him on the throne. It is plain that Manuel’s claims needed all possible public exposure and support, and much contemporary work of court poets and rhetoricians may be read as a concerted attempt to achieve this. The huge output of such material at the beginning of the reign has been well analyzed by P. Magdalino, who shows that attempts were made to convert the emperor’s youth into a glamorous asset.14 But by the autumn of 1147, Manuel, now aged twenty-nine, had begun to gain experience; he had waged a not unsuccessful campaign against the Seljuk Masud in Ikonion and had been married for more than a year. The rhetoric bol-

11 Wibald of Stavelot, Epistula 78, ed. P. Jaffe´, Monumenta Corbiensia: Bibliotheca rerum Germanicarum, vol. 1 (Berlin, 1864), 153; cf. F. Chalandon, Les Comne`ne, vol. 2 (Paris, 1912), 308 n. 2. 12 Poem 20.121, 130 (Mi 243, 252): Miller, Recueil, 222. 13 At, e.g., 8.214–17 and 15.137–39 he complains that the emperor did not notice his work and at 49.178–98 that he is fourth in the queue of speakers. Cf. 20.54–55: o”pw" de` tou'ton oi»o" h«n ejfane´ rou oJ nu'n krothqei`" paradhlw´ sei lo´ go". 14 “Eros the King and the King of Amours: Some Observations on Hysmine and Hysminias,” DOP 46 (1992): 197–204; Manuel, 434–54.

[ 106 ] The “Wild Beast from the West” stering his claims to authority no longer needed to be so defensive. Yet one has a sense that the administration had set up a propaganda machine that now continued under its own momentum. Manganeios molds our perception of Manuel both by isolated remarks in the course of the narrative and by formal encomia. The most common reference is to Manuel’s name, which, being one syllable different from Christ Emmanuel (as befits a created being), endows Manuel with a Christlike disposition.15 This leads to references to his virtue, generosity, and tireless work on behalf of the City; he is also compared to David as a youngest son selected ahead of older brothers.16 He is a second Solomon, a second and greater Alexander, the lion’s cub,17 a glorious bridegroom. The overt encomia include similar comments but also focus on his appearance and physique, his good sense, and his only campaign, that against Ikonion.18 Manuel is an unshakable rock amidst a sea of troubles. He cares for the City of Constantinople, tends it and renews it, he is its second founder and greater than the first for he has restored it to its former glory.19 He 15

20.97–100 (Mi 219–22): Miller, Recueil, 221. 20.312–14 su` de` protupw'n cristomimh´ toi" tro´ poi" tou' prwtotu´ pou to` pro´ swpon despo´ tou, oJ cristoklhtw´ numo", oJ pra'o" fu´ sei . . . 24.171 ejn soi` tv' rJu´ sth basilei', tv' mo´ nv cristwnu´ mv. 16 20.353–6 su` ga` r oJ gra´ fwn tou' Daui`d th` n eijko´ na crw´ masi prao´ thto" eujanqesta´ th" kai` ga` r ajlhqw'" wJ" Daui`d e“fu" ne´ o" u”staton a“nqo" uJstero´ rphx tv' cro´ nv . . . Cf. Miller, Recueil, 188. 24.87 kai` ne´ o" ge´ gona" Daui´d, makro´ qumo" kai` pra'o". David appears frequently in the corpus as author of the Psalms and a major character in the narrative books of the Old Testament. His first role is to mark some of Manganeios’ many references to the Psalms in direct speech, which adds variety to the poems’ narrative. The second is as a prototype ruler, making Manuel a new David. The closest comparison is that both gained power despite several older brothers. One may add similarity in strength and military skill and in moral character: patience, modesty, meekness, goodness. Both fought a Goliath (in Manuel’s case the Hungarian Bakchinos), and both are celebrated for routing tens of thousands (cf. 1 Sam. 18:7) and for great achievements in marshy areas. 17 20.414 Sku´ mnon se tou' le´ onto" oujkou'n kale´ sw… “Lion cub” is a warrior image with imperial links (indistinguishable categories for the Komnenoi). The source is biblical (e.g., Gen. 49:9, Deut. 33:22). It usually refers to emperors, but includes sons, a grandson, and a sonin-law. Reference is mostly to Manuel I, sometimes without mention of a lion father. More often it involves Manuel’s father, occasionally glorifying the son from the father, but usually referring to the son’s precocious development to equality with his father, even the son’s superiority. John II is a lion, Manuel a frisky cub who becomes a lion among cubs, then a lion, then even more of a lion than his dead father. 18 20.485–91 o”t∆ ej" po´ lin pe´ fqaka" jIkonie´ wn h’n ejk palaiw'n oujd∆ ejgnwsme´ nwn cro´ nwn oujk ei«den a“llo" oujd∆ ejto´ lmhse fqa´ sai. “Efritton oiJ pa´ nte" ga` r aujth` n wJ" ”Adhn⭈ su` de` fqa´ sa" e“kleisa" aujth'" to` sto´ ma to` n Ke´ rbero´ n te to` n fu´ laka to` n ku´ na e“treya" uJla´ ttonta, mh` da´ knonta´ se. The novelty of the attack on Ikonion is emphasized to an outrageous extent. 19 20.461–67 Su` ga` r polisth` " deutereu´ wn deiknu´ h a“xio" o“ntw" ajndria´ nto" cruse´ ou, eij nu'n parh'n ti" Zeu'xi" h‘ kai` Feidi´a". JO me` n ga` r ajnh´ geiren aujth` n ejk ba´ qrwn

Elizabeth Jeffreys and Michael Jeffreys [ 107 ] is in fact the celestial light that illuminates the New Rome, supported by the Virgin and defending orthodoxy.20 The image that is projected here is of a youthful, vigorous, responsible—and handsome—leader, fully justifying the trust placed in him, for he is already surpassing his father and his grandfather, particularly in his competent dealings with the crusading armies.21 It is also in this context that Manuel is praised for achieving a bloodless victory and for using gold judiciously to achieve it. Next, the City of Constantinople. In the poems we are considering and elsewhere in the corpus, the City is personified, in the ancient manner, as a female figure. In addition to several isolated references, the City in this persona addresses two formal encomia to Manuel, in the last 170 lines of poem 20 and the whole of poem 24. The figure of the City embodies contradictions that enable the poet to stress both the venerable prestige of Constantinople and the particularly vigorous (and apparently youthful) phase she is passing through in Manuel’s reign. The pattern is announced in the first two lines of poem 24: “The City, Lord, shedding the wrinkles of old age and dressed up today as a bride.”22 She is a woman wrinkled with age and lame, an elderly empress,23 but she becomes a young and beautiful bride, brightly dressed.24 She is also Manuel’s nurse who reared him and looks forward to rearing his children,25 and whose efforts in nurturing him are now well recompensed. The City’s transformation is due to Manuel, who has reversed the roles of mother and child by dressing her in brilliant clothing and persuading

su` d∆ ejk ba´ qrwn me´ llousan eij" ptw'sin rJe´ pein e“swsa", ejsth´ rixa" au«qi" hJdra´ sw, hjlpisme´ nou ptw´ mato" aujth` n ejrru´ sw. 24.67–69 kai` kth´ twr prwtodeu´ tero" wJ" krei´ttwn ajnefa´ nh". JO me` n ga` r mo´ non e“ktise, su` de` fqarei'san sw´ zei", katakliqei'san ajnorqoi'", sthri´zei" klonhqei'san . . . 20 20.432–33 tei´na" de` ble´ mma pro` " mo´ nhn th` n prosta´ tin ejf∆ h» to` kleino` n a“stu kauca'tai to´ de . . . 20.548–49 ajspa´ zetai´ sou ta` " krataia` " wjle´ na" do´ ru strefou´ sa" uJpe` r ojrqodoxi´a". 21 20.405–6 Ta` pappika´ soi kata` Lati´nwn kra´ th ajntexeta´ sw toi'" neourgh´ masi´ sou. Cf. 24.246–53: Miller, Recueil, 759. 22 Miller, Recueil, 758. 23 20.516–19 JIdou` ga` r hJ grau'" tv' cro´ nv katekli´qhn oJra'" o”pw" e“kamya kai` ta` " ijgnu´ a", o”pw" de` sugku´ ptousa kei'mai pro` " go´ nu, pw'" kateka´ mfqhn uJpo` talaipwri´a". 20.533–35 Nosei' ga` r aujth` ka´ mptetai´ te tv' cro´ nv kai` th` n palaio´ thta nu'n dei´knusi´ soi wJ" a‘n pare´ xh" cei'ra mhtri` mh` pe´ sh. 20.592 po´ li", trofo´ " soi, basili`" ghrale´ a . . . 24 24.32–33 JO basileu´ " mou sh´ meron wJ" nu´ mfhn me stoli´sa" pasta´ da me` n to` tei'co" mou, qala´ mou" de` tou` " pu´ rgou" . . . 24.36–37 kai` toi'" kuklw´ sasin ejcqroi'" hJ grau'" oJrw'mai ko´ rh kai` nu´ mfh kai` basi´lissa katapepoikilme´ nh. 25 20.525–27 Genh´ somai ga` r crhsi´mh soi kai` pa´ lin⭈ i“sw" ajna´ xw kai` glukei'" sou` " ejggo´ nou"⭈ wJ" ejkgo´ nou" qa´ lyw de` kai` ko´ lpwn e“sw.

[ 108 ] The “Wild Beast from the West” her to put on her cosmetics and conceal her wrinkles.26 He has renewed her and made her young again. Twentieth-century taste may find some of the images grotesque. The rouge with which the City enhances her cheeks is the blood of the Germans,27 an image that does not sit well with praise for Manuel’s bloodless victories. The City had lost her teeth, but Manuel had renewed them—probably, as noted above, a reference to repairs to the city walls. The audience is imbued with a sense of civic pride in ancient tradition, a self-confident perception of renewal, and an appreciation that the city’s fate is intimately connected with the emperor and the ruling dynasty. As the Virgin supports Manuel and comes to the City’s defense, so the City is God-built and protected by God. When we turn to Manganeios’ presentation of Conrad and the crusading army and the motives attributed for their arrival, we find that he assumes that the forces had come with hostile intent. At several points in poem 20, Conrad is said to have planned to capture Constantinople and its territories, to intend a secret attack, and to have designs on the City’s wealth.28 In its present form, poem 20 opens with Conrad’s schemes to impose his faith on Constantinople and to set up a Latin patriarch in the City.29 Poem 24 is predicated on the Crusaders’ hostile intentions: at the outset we were told that they had come hunting the aged City, under the impression that she was now toothless and defenseless. When we look in more detail, we find that Conrad and his army are boorish, bestial, greedy, deceitful, and aggressive.30 Conrad himself is wily, arrogant, and destrucjEn soi` saleu´ w⭈ sh` n trofo` n mh` pari´dh". Xe´ son to` gh'ra", ajfelou' ta` " rJuti´da", ta` gei'ssa moi sth´ rixon, e“nqe" krhpi´da", to` pi´pton ajno´ rqwson, ajnte´ reide´ moi, sto´ lize kai` kai´nize th` n grau'n wJ" ne´ an. 24.26–27 Su` de` stoli´zei" kai` kosmei'", uiJe´ mou, th` n mhte´ ra, kai` kru´ ptei" ta` " rJuti´da" mou kai` periba´ llei" fu'ko". 27 24.210–14 Su` ga` r ejcqroi'" me dunath` n eijrga´ sw kai` sfrigw'san, su´ mou to` gh'ra" e“xesa", e“kruya" ta` " rJuti´da", kai` pa´ lin ajnekai´nisa" kosmh´ sa" kai` foini´xa", kaqa´ per ejruqh´ masin Alamanw' j n toi'" fo´ noi" kai` toi'" ejkei´nwn ai”masin wJ" fu´ kei me lampru´ na". 28 20.61–63 (Mi 183–85): Miller, Recueil, 221. 20.308–10 JO me` n ga` r ejfle´ gmainen oi»" proesko´ pei kai` pro` " to` n o“lbon ajforw'n tw'n ejnqa´ de e“swqen wJ" qh` r ejspara´ tteto brucw'n . . . 20.631–33 Th` n sh` n ga` r aujtoi` katakrath'sai po´ lin kai` pa´ nta" hJma'" ejxalei'yai tou' bi´ou a“rdhn ejbouleu´ santo kaq∆ hJliki´an. 29 20.1–3 Qro´ nv de` tauto´ glwsson ejgkaqidru´ sw⭈ wJ" iJerarcw'n kata` JRwmai´wn e“qo" a“zuma qu´ sei th` n nomikh` n qusi´an. 30 20.14–17 tou` " barbarikou` " ajpoteici´zwn lo´ cou" w”" tisi lo´ cmai" toi'" kataski´oi" do´ loi". Tau'ta progra´ ya" kai` kalu´ ya" ejn sko´ tei th' lanqanou´ sh th'" di´kh" uJpokri´sei . . . 20.6–11 ajlw´ peko" me` n uJpo´ krisin deiknu´ wn ejnto` " de` kru´ ptwn th` n o”lhn skaiwri´an, 26

20.520–24

Elizabeth Jeffreys and Michael Jeffreys [ 109 ] tive.31 The Crusaders are wild boars, Gadarene swine,32 fittingly rolled in mud, wild beasts and insatiable serpents from the West; they are from the land of the evening and fit to be eclipsed. They are innumerable, more than the stars or the sand, they are like ants;33 they are a turbulent sea surging around the rock that is the emperor. Conrad is a fox in disguise, a chameleon, a secret wolf, a savage beast.34 In other words, Manganeios draws on a pejorative vocabulary and, in particular, on pejorative animal imagery. This can become bizarre: toward the end of poem 24, we are told that the dra´ kwn, dragon, the Latin ruler of Antioch, has become a puppy scrabbling at Manuel’s imperial feet.35 We should remember that the only animal image regularly used in connection with Manuel is of the lion cub. The Crusaders are also shown in humiliating situations. Manganeios gloats over the disastrous flood at Choirobacchoi. He takes advantage of the opportunities the name offers for pejorative comments on Bacchic frenzies, on the porcine banquets that the corpses fittingly offer.36 He relishes the ironies of the army’s bitterness in the suburb of Pikridion and starving outside the monastery of St. Mamas.37 Conrad’s pride had provoked divine retribution, and he is reduced to trembling fear.38 The few mitigating features in Manganeios’ account are used only to paint the rest of the picture darker by contrast. For example, though they are not orthodox (as we can see in the opening passage in poem 20 on the use of unleavened bread), the Crusaders are co-religionists nonetheless and not to be killed.39 Though now driven wild like the Gadarene swine, they had previously been rational.40 Conrad is said to be a dedi-

kai` tou' proba´ tou th` n dora` n e“xw fe´ rwn, ajnaire´ thn lu´ kon de` kalu´ ptwn e“sw, kai` to` n camaile´ onta toi'" tro´ poi" gra´ fwn toi'" poiki´loi" crw´ masi tw'n bouleuma´ twn. 20.238 (Mi 360): Miller, Recueil, 224. 20.604–9 ta` " ajkore´ stou" diya´ da" th'" eJspe´ ra" kata` potamo` n w”sper ejxantloume´ nwn pro` " ta` " caru´ bdei" ta` " ajei` kenoume´ na" . . . 31 20.142–43 (Mi 264–65): Miller, Recueil, 222. 32 20.215–19 (Mi 337–41): Miller, Recueil, 224. 33 24.51–54 Ei“dete pw'" ejmau´ risan oiJ ka´ mpoi ajpo` tou' plh´ qou", kai` pw'" bounoi` kai` fa´ ragge" kai` na´ pai kai` koila´ de" uJpe` r ajste´ rwn ajriqmo´ n, uJpe` r qala´ ssh" a“mmon, kata` swro` n ajne´ bruon murmh´ kwn ajmetrh´ twn. 34 20.323–27 jEpei` d∆ oJ pri`n u”poulo", oJ krupto` " lu´ ko", oujk ei«ce kru´ ptein th` n fusikh` n mani´an ajll∆ ajnafando` n kata` th'" ma´ ndra" e“qei qu´ sai te ta` pro´ bata kai` tou` " poime´ na" kai` pa'n spara´ xai toi'" ojdou'sin ajrni´on. 35 24.254–56: Miller, Recueil, 759. 36 20.132–34, 181–89 (Mi 254–56, 303–11): Miller, Recueil, 222–23. 37 24.142–59: Miller, Recueil, 758; cf. J. Pargoire, “Les Saints Mamas de Constantinople,” IRAIK 9 (1904): 261–316, where extensive use is made of Manganeios’ account. 38 20.105–8 (Mi 227–30): Miller, Recueil, 222. 39 20.96 (Mi 218): Miller, Recueil, 221. 40 20.220–23 (Mi 342–45): Miller, Recueil, 224.

[ 110 ] The “Wild Beast from the West” cated Crusader taking no account of God, his armies hymn the cross but destroy the faithful.41 Manganeios’ audience is thus presented with a heroic emperor, a multifaceted city, and bestially aggressive foreigners with few positive features. His audience’s reception of these figures is ruthlessly controlled by the use of loaded adjectives and images, especially animal imagery. Nothing is left to chance in the manipulation of the audience’s sympathies. The text is enmeshed in a web of biblical allusions. These often show little regard for consistency of reference, as may be shown in the opening passage of poem 20.42 The initial justification for the sequence of motifs is that Constantinople is the New Jerusalem (20.19)—a regular part of the Byzantine thought-world.43 If this is so, then Conrad is a Sennacherib, the Assyrian king whose host swept down on Jerusalem (20.20), and may 41

20.58 (Mi 180): Miller, Recueil, 221. 20.620–21 nai´, dusmenei'" e“laune, nai´, tropou' ge´ no" stauro` n me` n uJmnou'n, tou` " de` pistou` " ojllu´ on. 42 20.16–48 Tau'ta progra´ ya" kai` kalu´ ya" ejn sko´ tei th' lanqanou´ sh th'" di´kh" uJpokri´sei ejpestra´ teuse su` n stratia' muri´a ejn th' kaq∆ hJma'" JIerousalh` m ne´ a 20 oJ kai` Senachrei´m te kai` Dwh` k ne´ o"⭈ eij ga` r qusi´a" ajnelei'n ta` " ajrti´a" e“mellen ou»to" ouj sfageu` " h«n ti" a“ra… JO ga` r to` n a“rton mh` Qev' qe´ lwn qu´ ein to` n a“rton aujtv' prosfe´ resqai kwlu´ ei⭈ 25 oJ kwlu´ wn d∆ au« iJerei'" a“rton qu´ ein tou´ tou" ajnairei' kai` latreu´ ei tv' no´ mv⭈ oJ deu´ tero" nu'n JRaya´ kh" kai` kaki´wn oJ kai` lalh´ sa" kata` nou'n kai` krufi´w" kaqairetika` th'" Qeou' koinwni´a" 30 —a“rto" ga` r ejsti`n hJ Qeou' metousi´a— oJ mh` koresqei`" aiJma´ twn kata` Ku'ron, zhtw'n de` tou´ twn th` n me´ qhn katako´ rw", o’" kai` meta´ scoi th'" katadi´kh" Ku'ron h’n hJ To´ muri" hJ Skuqi`" ejkekri´kei 35 eij" ajsko` n aujtou' th` n kefalh` n ejndi´kw" ai”mato" ejmblh´ sasa peplhrwme´ non kai` ããla´ be, Ku're, to` n ko´ ron tw'n aiJma´ twnÃÃ pro` " ajko´ reston aiJmocu´ thn eujsto´ cw" to` rJh'ma toxa´ sasa kaqa´ per be´ lo". 40 JO proskunhth` " tou' Naboucodono´ sor, oJ pro´ sfato" nu'n tou' Sata` n JOlofe´ rnh", oJ kai` strateu´ sa" wJ" oJ pri`n met∆ ejlpi´dwn oi”wn ejkei'no" kai` frenw'n kai` rJhma´ twn kat∆ jEzeki´ou tou' kaq∆ hJma'" kai` ne´ ou, 45 oJ qei`" pro` " u”yo" oujranou' kai` to` sto´ ma o”n ti" jIoudei`q h‘ Qeou' cei`r ejkte´ moi ` r to` kra´ to" tou' kai` pri`n ejktri´yanto" Assou j kai` tou' krataiou' to` qra´ so" leluko´ to". 43 See, e.g., C. Mango, Byzantium: The Empire of New Rome (London, 1980), 207–8, and M. Angold, Church and Society in Byzantium under the Comneni, 1081–1261 (Cambridge, 1995), 508.

Elizabeth Jeffreys and Michael Jeffreys [ 111 ] also become, in the same line, Doeg, Saul’s herdsman.44 For David, read Manuel. Doeg we will return to in a moment. Sennacherib’s opponent in Judaea was Hezekiah, so a little later (20.44) a new Hezekiah (read Manuel) is the hapless victim of invasions. But the invader this time is Holofernes (read Conrad); it may be pedantic to point out that Holofernes was an emissary of Nebuchadnezzar, not Sennacherib, and it was against Sennacherib and not Nebuchadnezzar that Hezekiah organized resistance. We must assume that the overall precision of allusion is of little importance to Manganeios in comparison with the resonance set off in his audience by each opposition. This is one of the more elaborately worked out sets of biblical allusions. Other cases may be brief: for example, the flood at Choirobacchoi parallels Samuel’s experience at Masifat (Mizpah)45 (for Samuel, read Manuel). Or they can be more complex, as demonstrated by the most prominent comments found in connection with the flood at Choirobacchoi. Here Manganeios uses two parallels. The first is with the passage of the Israelites through the Red Sea, an excellent watery precedent.46 But the poet has to work hard to 44

The chief of Saul’s herdsmen (1 Sam. 21–22), responsible (in the Septuagint) for the murder of eighty-five priests suspected of taking David’s side against Saul. On the symbolism of Doeg, see most recently P. Devos, “Doe`k dans l’hagiographie byzantine, chez S. Augustin et dans une lettre de S. Basile,” AB 111 (1993): 69–80. 45 20.256–61 (Mi 378–83): Miller, Recueil, 225. At Mizpah, after Samuel’s sacrifice, God intervened with a thunderous voice (apparently without flood) on Israel’s behalf against the Philistines, leading to the rout of the latter (1 Sam. 7:5–11). However, one line of interpretation (e.g., Athanasios, Expositiones in Psalmos, PG 27:201) makes God’s intervention a violent hailstorm. 46 20.144–80 All∆ j oJ strathgw'n kai` dieuqu´ nwn pa´ lai 145 diaperw'san ta` " jEruqrai´a" di´na" stratia` n ajh´ tthton jIsrahli'tin ejkei'no" au«qi" qaumatourgei' kreitto´ nw". jEkei'se ga` r qa´ lassa pri`n dihre´ qh kai` trista´ tin fa´ lagga Farawni'tin 150 a“rdhn sune´ sce toi'" u”dasin wJ" ta´ foi"⭈ hJ nu'n troph` de` kai` paradoxourgi´a th` n xhra` n uJgra` n kainopoiei' meizo´ nw"⭈ kai` ga` r qalattoi' kai` na´ pa" kai` koila´ da" kai` rJeuma´ twn pe´ lago" ejn pedia´ di 155 aujqwro` n aJploi' kai` to` rJei'qron ajna´ gei uJpe` r kefala` " dusqanw'n ajlasto´ rwn kata` trocia` " kumba´ cwn strwfoume´ nwn. Kai` Faraw` me` n ouj buqi´zei tv' te´ w", pollou` " de` toi'" u”dasi katastrwnnu´ ei 160 wJ" a‘n ejkei'no" sumforai'" ajllotri´ai" maqw` n pro` " oi”an ajntita´ ttetai ko´ rhn porqei'n eJa´ sh th` n meri´da th'" ko´ rh". ”Ora ga` r oi»a qaumatourgei'tai xe´ na⭈ oJ me` n potamo` " ejbruca'to makro´ qen, 165 to` de` stra´ teuma kataplage` n tv' yo´ fv plhqu` n barei´a" sundromh'" strateuma´ twn kai` dou'pon uJpw´ pteusen wJ" ejx iJppe´ wn kai` paracrh'ma qwra´ kwn kai` knhmi´dwn kai` panto` " ajntei´conto sunh´ qou" o”plou 170 kai` to` n si´dhron ejnduqe´ nte", wJ" e“qo", i”stanto pa´ nte" ta` " spa´ qa" ejspasme´ noi gi´gante" aujto´ crhma sidhrendu´ tai

[ 112 ] The “Wild Beast from the West” make it apply to Choirobacchoi. Many of the enemy drowned: for Pharaoh’s men, read the German army; for the Israelites, read the Byzantines. But in Exodus the Red Sea divided and water became dry land before the sea rushed back: at Choirobacchoi, the land was dry and became sea, or at least a flooding river. According to Manganeios, the flood was unprecedented, though Choniates calls it an annual occurrence.47 In Old Testament terms, Pharaoh (read Conrad) was drowned, but now Conrad survived: this, says Manganeios (20.160), was so that he could learn the strength of the Virgin’s support of his enemies. The Lord had worked a greater miracle, a modern improvement on Exodus (20.147, 152). Manganeios is convinced that the present can be better than the past, and gets round the differences between the two situations by the brazen claim that God had now achieved a more spectacular success.48 The rhetorical point that Manganeios was impressing on his audience was an equation of the Byzantines with the Israelites, the Lord’s chosen people, and of their enemies with the Israelites’ enemies. There is no need to explain the moral superiority in the conflict conferred by this equation. But the inconsistency involved is much wider than the list of incongruous parallels given above, and in fact it undercuts the whole of the first part of the poem. The same passage in which Manganeios enthusiastically accepts the Jewish part of the Byzantine heritage, as we have seen, also accuses Conrad, directly and by implication, of Judaizing. The Old Testament appears on Manganeios’ list of positive rhetorical exempla for the claiming of righteousness, while accusations of Jewish religious practices are also prominent in Byzantium as negative exempla, means of condemning nonorthodox Christians.49 Both sets of references seem to be invoked automatically, and there is no overall control of the ideological framework sensitive enough to pick up the clash. There could hardly be a more convincing demonstration of the smallscale, mechanical operation of Manganeios’ rhetorical methods. A second biblical parallel used in connection with the flood at Choirobacchoi involves the Gadarene swine, a reference for which the name Choirobacchoi, with its scope for porcine puns, admirably suited Manganeios’ purposes. This leads to another major tool used by Manganeios: wordplay. Punning is a significant feature of Manganeios’ style, as for many Byzantine writers. Poems 20 and 24 provide ample instances. Some are trivial, like the play on lo´ co" and lo´ cmh in 20.14–1550 or on pepoiqe´ nai and pe´ ponqen in pro` " “Areo" ki´nhsin hujtrepisme´ noi pro` " eijsbolh` n e”toimon iJstw'nte" do´ ru. 175 JRoh'" de` mikra'" ejkdramou´ sh" ojxe´ w" oJ Faraw` prw´ tisto" eujqu` " iJppo´ th" ajntista´ th" e“feugen ajll∆ ouj trista´ th" e”w" to` ku'ma th'" kefalh'" tou' rJo´ ou surreu'san ejstro´ bhse pollou` " ajqli´ou" 180 ejn Coiroba´ kcoi" to` n bi´on leloipo´ ta". 47 Choniates, ed. van Dieten, 64.62–67. 48 Compare other analogies with Alexander and Constantine mentioned above, where Manuel is unhesitatingly called greater than his ancient parallels. Such self-confidence seems particularly common in the 12th century. 49 See, e.g., the discussion by G. Dagron, “Judaı¨ser,” TM 11 (1991): 359–80, at 364. 50 20.14–15 tou` " barbarikou` " ajpoteici´zwn lo´ cou" w”" tisi lo´ cmai" toi'" kataski´oi" do´ loi".

Elizabeth Jeffreys and Michael Jeffreys [ 113 ] 20.122.51 Some have become an almost automatic response, for example, the use of cristomi´mhto" for Manuel, referred to above. In other cases Manganeios is taking serendipitous advantage of opportunities presented by geographical names: for example, the play on de´ rkei and De´ rko" at 20.332 or ajda´ ma" and Da´ mali" at 20.334; or the name of the river that flooded: a blacknamed (melanw´ numo") flood suddenly poured down and hid the army in black death (me´ lani qana´ tv).52 As we know from Choniates, the river that runs over the plain of Choirobacchoi is the Melas, the Black. At times Manganeios develops these chance opportunities extensively. He seizes on the fact that the German armies camp in the district of Pikridion to play with concepts of bitterness:53 Conrad’s bitter mood, caused by his setbacks (20.300), and bitter herbs associated with the Jewish use of unleavened bread (20.302)—for Manganeios a defining feature of nonorthodox Christians, as is clear from the opening of poem 20. Choniates and Kinnamos confirm that the German army was encamped at Pikridion,54 so we need have no suspicion that the demands of rhetoric are taking over the narrative. In other cases we should perhaps be more wary. Let us look at two instances. In poem 20, at lines 31 to 37 (cited in note 42 above), we find wordplay on ko´ ro" and Ku´ ro", “excess” and “Cyrus.” In spite of the loss of the poem’s opening lines, it is clear that what is being expressed is the questioning of Conrad’s motives in coming to the East and, in particular, suspicion that he intends to impose the Latin rite on the Eastern church. Conrad is then denigrated with a series of animal images. But “unleavened bread” has flicked Manganeios’ rhetorical switch marked “Judaism” and has set him off on the series of biblical allusions already discussed, with references to the new Jerusalem (20.19, meaning Constantinople) and Sennacherib (20.20) and the “new Doeg.” Doeg’s relevance here is as a murderer: he stands for Conrad’s violence, although Conrad has not yet in fact inflicted any bloodshed on Byzantine troops. This is no hindrance to Manganeios. Conrad, he says, is proposing to impose the use of unleavened bread, a practice that will prevent the use of a“rto", the true bread (20.24), which is tantamount to murder. One senses that Manganeios may have realized after writing the line that his reference to Doeg went beyond the facts. However, instead of crossing it out, he decides to justify it by the rather strained claim of theological murder. At this point, Manganeios, having exhausted his biblical references, switches his symbolic register to the classical: Conrad is like Cyrus, insatiable. With the Scythian queen Tomyris and a tag from Herodotus comes the punch line, “La´ be, Ku´ re, to` n ko´ ron tw'n aiJma´ twn” (20.37).55 This whole passage shows an escalation of rhetorical paradigms further and further away from the historical situation. A passage in poem 24, lines 145 to 159, is another interesting example where it is not impossible that rhetoric has got the better of geography (see note 37 above). Manganeios depicts the plight of the starving German army, reduced to begging for food. He de-

51

20.122 (Mi 244): Miller, Recueil, 222. 20.332–34 (Mi 454–56): Miller, Recueil, 757; 20.205–7 (Mi 327–29): Miller, Recueil, 224. 53 20.298–307 (Mi 420–29): Miller, Recueil, 229. 54 Choniates, ed. van Dieten, 65.3; Kinnamos, Bonn ed., 75.14. 55 On Tomyris and Cyrus, see Herodotos, 1.205–14. 52

[ 114 ] The “Wild Beast from the West” scribes them crossing the Bosphoros toward its southern end, from the suburb, or harbor, of St. Mamas. Choniates and Kinnamos give no confirmation for this place name: Manganeios seems to be the sole authority for this detail.56 However, the name provides a splendid opportunity for a play on words that puts the threatening foreign hordes into a humiliating perspective, using the Greek “baby talk” word for food, which remains in use to this day: kai` yhlafou'si Ma´ manta kai` Ma´ manta fwnou'si (20.148). Granted the pun, there would be a temptation for Manganeios to shift the point of embarcation to St. Mamas, if indeed it had taken place elsewhere. It must, of course, have been close enough to pass the test of the eyewitness audience to which we have referred. Within a year the wholly negative picture of the Crusaders given in these two poems had to be adjusted in public, for political reasons. Conrad himself, his army decimated, was brought back to Constantinople and healed by Manuel’s medical skill; Conrad’s brother Heinrich Jasomirgott was married to the most eligible unmarried female of the Komnenoi, Theodora, third daughter of the Sevastokratorissa Eirene and the emperor’s deceased brother Andronikos. This German alliance was to be a central plank of Manuel’s Western policy, aimed particularly against Roger II and the dangerous Normans of Sicily. Manganeios, perhaps because of his close connection with the bride’s mother, wrote an official Epithalamion for the wedding. He used stanzas of ten fifteen-syllable verses each, which we believe to be the mark of his most prestigious poems written for public performance: “Dance, Alamania, and leap and shine brilliantly! For the Sevastokrator’s most beautiful daughter is being united to the glorious duke, to his great good fortune, and he is becoming more brilliant from her greater brilliance and much more glorious from her greater glory.”57 The tone is one of almost unalloyed jubilation and triumphant and brilliant ceremonial. The magnificent emperor, Manuel, is bestowing a glorious bride, Theodora, upon a worthy bridegroom, Heinrich. The emperor is a radiant sun, shining brilliantly all around him, while the bridegroom is also a luminary, though a lesser one, “a star who has come from the West to the dawn of daylight”; Manuel is asked not to let the full beams of his radiance blot him out.58 But even so, “the most fortunate duke . . . is advancing from glory to greater glory, he is being promoted from honor to higher honor.”59 There are few shadows in this picture. The bridegroom and his German background See Pargoire, “Les Saints Mamas,” 303; cf. F. Chalandon, Les Comne`ne, vol. 1 (Paris, 1900), 278 n. 2. ´a, co´ reue kai` ski´rta kai` lampru´ nou⭈ 22.1–5 Alamani j tou' ga` r sebastokra´ toro" th' pagkalli´sth ko´ rh oJ dou` x oJ megalo´ doxo" paneutucw'" eJnou'tai, kai` gi´netai lampro´ tero" ajpo` th'" lamprote´ ra", kai` megalodoxo´ tero" ejk th'" ejndoxote´ ra". Ed. C. Neumann, Griechische Geschichtsschreiber und Geschichtsquellen im XII Jahrhundert (Leipzig, 1888), 65–68. 58 22.16 to` n ejx eJspe´ ra" eij" aujgh` n hJmerinh` n ejlqo´ nta. 22.17–20 Mh` pa´ sa" ta` " ajkti´na" sou, mh` pa'san sou th` n ai“glhn, mhde` to` fw'" tou' di´skou sou, mhde` to` pa'n sou se´ la" eij" to` n ajste´ ra sh´ meron ejkpe´ myh" th'" eJspe´ ra", i”na mh` kru´ yh" tv' pollv' fwti´ sou to` n ajste´ ra. 59 22.35–36 ajpo` ga` r do´ xh" sh´ meron pro` " mei´zw bai´nei" do´ xan, ajpo` timh'" eij" krei´ttona timh` n ajnabiba´ zh. 56 57

Elizabeth Jeffreys and Michael Jeffreys [ 115 ] are firmly subordinated to Manuel and the Byzantines: but the praise, though faint, is real. The only negative hints come in the list of Manuel’s martial achievements, for it is said that he has “shown the Kelts too to be boastful skeletons and to have humbled the lofty pretensions of the Kings”—a characteristic pun within a reference to the Second Crusade which was to become a cliche´.60 In the last line of this poem, Manuel becomes the “benefactor of him who did you harm,”61 that is, the bridegroom had been involved in events (surely the crusade) that called for the exercise of benevolence by Manuel. But through Manganeios we have also privileged access to the reactions of the bride’s family at less public moments. In a series of poems written within three years of the marriage, which we believe were performed within the family, the loss of Theodora is described as a major disaster in the life of Eirene the Sevastokratorissa. Eirene’s life was tempestuous, and her reactions to misfortune were always intense, as may be seen from the words of Manganeios and other writers close to her. But her response to the fate of Theodora is shown as unusually and consistently vehement, over several poems and several years: the complaint is that Theodora had been torn from her mother and handed over to a savage monster, to live a gloomy life of exile. Elsewhere, Eirene is consoled for the loss of a golden chick, a nightingale imagined as singing mournfully amid rocks and precipices, surrounded by a flock of birds of prey.62 This is presumably, in part at least, an expression of natural maternal distress that remained keen even years after the marriage. Theodora was the first Komnene bride to be married outside the empire, so that the maternal sensitivities of the ladies of the imperial family were not yet blunted.63 Furthermore, it is possible that Theodora was being sent to an area to which Eirene was personally antipathetic. If Eirene was of Norman background (as we have suggested elsewhere), she may have objected to the use of her daughter to cement a Byzantine alliance with the German dynasty, opponents of the Normans.64 We have no evidence that the maroJ kai` tou` " Pe´ rsa" kaqelw` n kai` to` n aujtw'n satra´ phn, kai` skeletou` " kai` tou` " Keltou` " tou` " ajlazo´ na" dei´xa", kai` tw'n rJhgw'n katabalw` n ta` " uJyhla` " ejpa´ rsei" . . . Western kings (rJh'ge") for Manganeios, including the unrecognised Western Emperor, are naturally of inferior status to the Byzantine Emperor. Reference is positive when the king is an ally, usually linked by marriage, as with the ancestors of Manuel’s wife, Bertha-Eirene, Conrad of Germany and the King of Jerusalem. But kings on crusade are shown negatively, even the same Conrad. The plural, “kings,” becomes shorthand for the Second Crusade, when the Kings of France and Germany are astounded, humbled, beaten off, and routed. 61 22.96 o”ti kai` tou' luph´ santo" ejfa´ nh" eujerge´ th". 62 41.46–55 Ble´ pw se ga` r stugna´ zousan ejn me´ sh carmosu´ nh, kai` to` n paro´ nta glukasmo` n mignu'san ajyinqi´v, o”ti mh` ble´ pei" to` crusou'n ejkei'no sou strouqi´on, to` pro` makrou' sumpetasqe` n kai` mh` sunanazeu'xan. ` ga` r le´ ge moi loipo´ n, cruso´ qrix, crusoci´twn, 50 Alla j ejn poi´v to´ pv to` terpno` n aujli´zetai strouqi´on… Pou' keladei' to` liguro` n th'" ajqumi´a" me´ lo"… Pou' mousourgei' to` penqiko´ n, pou' to` qrhnw'de" a“ dei, ejn fa´ ragxin, ejn o“resin, ejn pe´ trai", ejn sphlai´oi", ejn ajkrwrei´ai", ejn krhmnoi'", ejn to´ poi" kataski´oi"… 63 P. Magdalino, Manuel, 201–17; see also R. Macrides, “Dynastic Marriages and Political Kinship,” in Byzantine Diplomacy: Papers from the Twenty-fourth Spring Symposium of Byzantine Studies, Cambridge, March 1990, ed. J. Shepard and S. Franklin (Aldershot, 1992), 263–80. 64 E. M. and M. J. Jeffreys, “Who Was the Sevastokratorissa Eirene?” Byzantion 64 (1994): 40–68, at 65–66. 60

22.53–55

[ 116 ] The “Wild Beast from the West” riage was unhappy in personal terms. A poem written around 1151, describing a visit by Theodora to her mother in Constantinople, accompanied by one of her sisters, makes purely conventional comments.65 And the marriage was a dynastic success: Theodora’s symbolic role was significant in the establishment of the Babenbergs in Austria and of Austria as an independent political entity.66 This study has provided documentation for an immediately negative Byzantine reaction to the Second Crusade—a reaction of suspicion and intolerance which one might anticipate, but which is otherwise difficult to demonstrate. In at least one part of the ruling dynasty of the Komnenoi, this attitude may be confirmed by its continuation after the end of the Crusade, with Eirene the Sevastokratorissa being encouraged by Manganeios to put aside the diplomatic niceties of the marriage of her daughter and to relive the painful moment of her loss via the demonization of the bridegroom, using the same pejorative animal imagery we have seen in operation in the poems written during 1147: “When did such a union of opposites take place? When did a maiden cohabit with a flesh-eating beast? When did a delicate girl unite with a dragon? When has a tender calf been joined to a wild boar? All this I endured when I saw my tender daughter defiled, when the wild beast from the west was joined with her, and I wept over my living daughter as though she were dead.”67 Exeter College, Oxford, and the University of Sydney

65 Poem 55, dated (probably accurately) by Magdalino to 1150. In the title, Theodora and her sister are returning ajpo` tou' taxeidi´ou (from campaign), i.e., from the imperial army camps which made a kind of alternative court during Manuel’s reign. There is no indication that her sister had been to Austria with her. 66 This subject is dealt with exhaustively by K. J. Heilig, “Die Verwandtschaft der Theodora im byzantinischen Kaiserhaus,” in Kaisertum und Herzogsgewalt im Zeitalter Friedrichs I., ed. T. Mayer, K. Heilig, and C. Erdmann (Leipzig, 1944), 229–71. 67 47.116–23 Po´ te toiau´ th ge´ gone tw'n ejnanti´wn mi´xi"… Po´ te qhri` sunv´ khse parqe´ no" wjmobo´ rv… Po´ te sunh'lqe dra´ konti malqakwta´ th ko´ rh… Po´ te sunh´ fqh moniv' truferwta´ th mo´ sco"… ` kai` tou'to gegono` " uJpe´ sthn oujc eJkou'sa, 120 Alla j kai` to` terpno` n quga´ trion lelumasme´ non ei«don, oJpo´ tan qh` r eJspe´ rio" ejkei´nv sunhnw´ qh, kai` zw'san ajpeqrh´ nhsa kata` nekra` n th` n pai'da. Cf. Miller, Recueil, 768–69.

Byzantine Perceptions of Latin Religious “Errors”: Themes and Changes from 850 to 1350 Tia M. Kolbaba In 1339 Emperor Andronikos III Palaiologos (1328–41) sent the bilingual Calabrian monk Barlaam (ca. 1290–1348) to Avignon. There Barlaam delivered two speeches to Pope Benedict XII (1332–42) about the necessity of a united Christian front against the Turks and the ways in which a reunion of the churches might be achieved. To the pope’s demand that reunion of the churches precede military aid from the West, Barlaam gave the following reply: “It is not so much difference in dogma that alienates the hearts of the Greeks from you, as the hatred that has entered their souls against the Latins,1 because of the many great evils that at different times the Greeks have suffered at the hands of Latins and are still suffering every day. Until this hatred has been removed from them, there cannot be union. In truth, until you have done them some very great benefit, neither will that hatred be dispelled nor will anyone dare to breathe a word to them about union. . . . Know this too, that it was not the people of Greece that sent me to seek your help and union, but the Emperor alone and secretly. Until help is sent to these parts, he cannot let his people see that he wants union with you.” 2 Barlaam thus highlighted the most obvious impact of the Crusades—especially the Fourth Crusade and the Latin occupation of Constantinople from 1204 to 1261—on religious life and religious literature in Byzantium. Everyone agreed that the union of the churches was, in principle, desirable, because everyone knew that Christ’s body, the church, should not be dismembered. But the violent conflict of the Crusades and attempts to force papal primacy on Greeks after 1204 meant that few Byzantine churchmen could negotiate for such a union with any measure of trust and goodwill. So, too, it comes as no surprise that the most scurrilous, least sophisticated kinds of anti-Latin literature increased over time. When Constantine Stilbes (fl. 1182–1204) connects his seventy-five-item list of Latin errors to a list of the atrocities committed in the sack of 1204, the connection seems natural.3 Such a reaction makes sense. That the Crusades 1 As do most Greek writers of his time, Barlaam uses the term Latin as a general term for Westerners. I use the term Latins throughout this paper to refer to Western Europeans who were members of the church that used Latin as its liturgical language. This does not mean that Byzantines themselves always called Westerners “Latins”; see Alexander Kazhdan’s contribution to this volume. 2 Acta Benedicti XII, 1334–1342, ed. A. L. Tau˘tu, Fontes 3, vol. 8 (Vatican City, 1958), doc. 43. Cited and translated in J. Gill, Byzantium and the Papacy (New Brunswick, N.J., 1979), 197–98. 3 J. Darrouze`s, “Le me´moire de Constantin Stilbe`s contre les Latins,” REB 21 (1963): 50–100.

[ 118 ] Byzantine Perceptions of Latin Religious “Errors” led to an increase in the number of virulently anti-Latin texts and in the number of people who agreed with them has been recognized at least since Barlaam’s time. This study, then, goes beyond that obvious effect to investigate whether the Crusades had an impact on more moderate religious texts written by churchmen who negotiated or debated with the Latins. Because there are many Byzantine responses to the Crusades in secular texts, from Anna Komnene’s Alexiad to Doukas’ chronicle of the fall of the City, one might expect to find direct responses to the Crusades in theological literature as well. But a survey of religious discussions with and polemic against the Latins from the middle of the eleventh century through the end of the empire unearthed no reasoned refutation of the idea of holy war and no theological discourses against such Western innovations as the crusade indulgence or monastic knights.4 In short, if the Crusades altered religious literature, they did so indirectly. This study attempts to identify such indirect influence by analyzing some characteristics of Byzantine theological material contemporary with the Crusades. The conclusion will return to the question of whether and how these traits are related to the Crusades. My primary thesis is that Byzantine religious texts that discuss Western Europeans emphasize different issues at different times. To many historians, such a claim may seem obvious, even trite. After all, the cultural gap between Byzantine East and Latin West; the kinds and degree of contact Byzantines had with Latins; the relative wealth, poverty, military power, and sophistication of the two cultures—all of these things changed immeasurably in a millennium or so. Yet an assumption of eternal verities pervades the history of Byzantine religious disagreements with the Western church. For example, many studies assume that the Filioque5 is always the central issue for moderate, reasonable men. But it was not. Concerns changed as the times changed. Furthermore, when placed in their historical context, the issues raised are often related less to the explicit targets of the polemic, the Latins, than to the polemicists themselves and their world. An issue becomes one of the crucial issues in the Greek theological literature only when it becomes a matter for debate within the Orthodox world. This connection removes the Latins from the center of the picture and reveals the extent to which debates explicitly about Latins were implicitly about Byzantines. In other words, a difference between Greeks and Latins became a source of anxiety and the subject of numerous treatises and debates only when Byzantine opinion was divided. Debates about Latin practices and beliefs grew fierce and polarized less because of the intrinsic importance of the issue being debated than because of fundamental doubts about what 4 Such issues do arise rarely in the unreasonable polemic. See, e.g., Darrouze`s, “Me´moire,” para. 27, 38, 60, 61. 5 Starting in Spain in the 6th century, various Western churches added a phrase to the Nicene Creed. Where the creed originally stated, “We believe in the Holy Spirit . . . who proceeds from the Father,” these churches added “and the Son” (Latin: Filioque). This addition was accepted in Frankish areas by the 8th century and in Rome in the early 11th century. Eastern theologians objected both to the unilateral addition to the creed (which could not, they maintained, be amended without an ecumenical council) and to the theological implications of that addition. Discussions of the theology, including theological polemic from both East and West, can be found easily. Good introductions: J. Pelikan, The Spirit of Eastern Christendom (600–1700): The Christian Tradition 2 (Chicago, 1974), 183–98, and J. Meyendorff, Byzantine Theology: Historical Trends and Doctrinal Themes (New York, 1974), 91–94, 180–90.

Tia M. Kolbaba [ 119 ] it meant to be an orthodox, imperial Christian—what it meant to be, as they would have put it, a pious Roman.6 Negotiations and debates within the Empire of the Romans about how to distinguish “us” from “them” were not new in the tenth, or even the eighth, century. From the beginning, Christians were defining themselves against other groups, distinguishing “followers of Christ” from “Jews”; “Orthodox” from Arians, Nestorians, and Monophysites; righteous and orthodox venerators of icons from heretical iconoclasts. None of these distinctions between “us” and “them” was established in a day or even in a decade. All of them took some time and caused some casualties. Some people who considered themselves orthodox Christians had to be thrown out of the church; the tares could not, after all, be allowed to grow with the wheat. In the period of the Crusades, it became important to distinguish “us” Christians of the empire from “them” Latins from the West. But that distinction did not come easily, either. People argued about it for centuries, and their arguments can be partially reconstructed from the materials studied here. The second part of this study discusses the tone of anti-Latin texts. This, too, changes over time, but the change is not a simple descent from moderate, intelligent discussion to hateful, radical polemic. Moderate works exist and exert some influence down to the end, revealing a growing ambivalence about Latin culture and the western world. What this study presents as a matter-of-fact outline still has gaps, and other scholars who study these texts will correct and refine it on points of detail and interpretation. Still, it is time to attempt a survey of the theological literature from these centuries precisely because a great body of work makes it possible to do so with some assurance. We need to draw together what we already know before we can make further progress. The current level of knowledge owes much to the works of Jean Darrouze`s, Joseph Gill, and a long list of other scholars. The sources cited below should indicate my debt to their erudition and painstaking labor. Darrouze`s noted thirty-two years ago that “the history of dogma can only profit from a more exact knowledge of historical context.” 7 He spent most of his life establishing that context, and his work especially has taught us a great deal about which issues dividing East and West were important in which period. Without it, this study would be impossible.

Changing Issues The Ninth Century Photios (patriarch of Constantinople, 858–867 and 877–886) introduces this study, but not because anyone accepts that the “Photian Schism” was permanent and irrevocable; Francis Dvornik refuted that idea fifty years ago. Rather, Photios’ era can reveal the 6 Vocabulary is a problem here—these were debates about what it meant to be Byzantine and Orthodox. Still, we need to keep in mind that these are modern terms; people at the time called themselves “Romans,” “Orthodox,” “pious” (eujsebh´ ")—never “Byzantine,” unless they were distinguishing residents of the capital from other “Romans.” 7 J. Darrouze`s, “Les documents byzantins du XIIe sie`cle sur la primaute´ romaine,” REB 23 (1965): 43.

[ 120 ] Byzantine Perceptions of Latin Religious “Errors” possibilities for a relationship between Rome and Constantinople which was, if not exactly peaceful, certainly different in kind from the relationship of the later Middle Ages. Most importantly, differences with the Western church were not the crucial canonical or theological issues during Photios’ patriarchate. The burning issue was still iconoclasm. From our perspective, a kind of foreshortening makes it obvious that iconoclasm was dead and not to be resurrected. But Photios and his contemporaries knew how the first “restoration of Orthodoxy” had been followed by a revival of iconoclasm. Most had personal memories of that revival. All were still being dragged into arguments about how to punish iconoclasts—some advocating severe sanctions, while others called for oikonomia and forgiveness.8 The quarrels over this issue affect every other quarrel of the period, including the “Photian Schism” with Rome. Beyond iconoclasm, Photios himself joined many other men in writing about other “heretics”: Paulicians, Armenians, Muslims, Bogomils, Monophysites, and others.9 So the quarrel with Rome is only one issue among many in ninth-century Byzantium. Moreover, the Photian Schism did not arise from differences over dogma. Nobody claimed that the pope was not qualified to render a judgment because he was a heretic. Instead, the issue was the canonical authority of the pope within the Eastern church— a question that neither began nor ended with Photios. His predecessor, Ignatios, had had similar problems during his first patriarchate (847–858).10 In the controversy over the legitimacy of Ignatios’ deposition (or resignation) and Photios’ elevation to the patriarchate, both sides appealed to the pope. Photios’ refusal to accept the pope’s judgment was based not on some challenge to the pope’s legal authority, but rather on the pope’s failure to hear any representative of Photios’ side of the case before he made his decision.11 This recognition of Rome’s jurisdiction, with its assumption of Rome’s orthodoxy, is more like the church of the iconoclast period or even of John Chrysostom’s time, than like the church of Michael VIII Palaiologos. In the later period, Rome’s jurisdiction will be challenged on the grounds that the popes, who used to have the authority of a first among equals, lost that authority when they fell into heresy.12 Nevertheless, Photios and some of his contemporaries did object to the Filioque (and other Latin “errors”). Those who maintain that the Filioque has always been the most important issue for thoughtful, moderate men begin with Photios, for he did explicitly state that the Filioque was a heresy and the weightiest issue outstanding between Constantinople and some Westerners: “Moreover, they have not only been discovered transgressing the law in all the above, but they have progressed to the crown of all evils, if there is such a thing. . . . They have also tried, with spurious reasoning, interpolated argument, and an excess of impudence, to adulterate the divine and holy creed which has its impregnable strength from all the synodical and ecumenical decrees (Oh, the subtle deceptions of the Evil One!), for they have added new words, that the Holy Spirit F. Dvornik, The Photian Schism: History and Legend (Cambridge, 1948; repr. 1970), 6 ff. H.-G. Beck, Kirche und theologische Literatur im byzantinischen Reich (Munich, 1959), 520–30. 10 Dvornik, Photian Schism, 19–32. 11 Ibid., part 1, chaps. 2–8. 12 E.g., see the polemicists cited by F. Dvornik, Byzantium and the Roman Primacy, trans. E. A. Quain (New York, 1966), 159–62. 8 9

Tia M. Kolbaba [ 121 ] proceeds not from the Father alone, but also from the Son.”13 But before we portray this statement as the earliest example of Byzantine awareness of Roman heresy, we need to look carefully at its context. Photios discussed the Filioque in an encyclical letter to the Eastern patriarchs (quoted above) and in his Mystagogy of the Holy Spirit. These texts were not attacks on the whole Western church, but refutations of the teaching of Frankish missionaries in Bulgaria. The latter had taught the Bulgars the addition to the creed and quarreled with Byzantine missionaries about it. In fact, the Filioque was not yet being chanted in Rome. When Photios wrote his treatises against the double procession of the Holy Spirit, he had good reason to think that the popes did not accept the doctrine.14 Nor did he challenge the pope’s authority on the grounds that he was a heretic. Both of these things differentiate his position from later opinions. Finally, anti-Latin arguments do not develop sequentially from Photios to 1453. An examination of the transmission of texts shows that Photios’ writings against the Frankish missionaries had little impact. Nobody adopts his arguments on these issues, and few people even refer to his opinions, until late in the thirteenth century. At that time, the Filioque is central to Byzantine polemic for other reasons, to be discussed below.15 The Eleventh Century Anti-Latin arguments do, however, have a continuous life from 1054 on. In the middle of the eleventh century, Byzantine polemicists raised many issues, some of which already had a history. Photios had complained, for example, about Latin Lenten observances and the Latin rite of confirmation, and Michael Keroularios (patriarch of Constantinople, 1043–58) raised these same issues.16 But the most prominent complaint of the middle of the eleventh century had not surfaced in Photios’ period. Among the “Roman” errors Keroularios mentioned is the use of unleavened bread (azymes) in the eucharist. Other texts of the period echoed the theme. In terms of number of words written, or number of treatises written, azymes far outstrip the procession of the Holy Spirit.17 Some who mentioned the Filioque—Peter III of Antioch (1052–56), for example—maintained that the addition was more important than unleavened bread, but their actions belied these words. Peter wrote far more about azymes than about the Filioque. To explain this emphasis, one needs to look behind Byzantine relations with the Western church to stresses within the empire. In general, the eleventh century saw a number of challenges to the definitions of “orthodox” and “Roman.” These were not purely external challenges—enemy attacks on the outer boundaries of Byzantium—but civil wars, causing disagreements among the powerful even at the heart of the empire. In 13

Photius, Epistulae et Amphilochia, ed. B. Laourdas and L. G. Westerink (Leipzig, 1988), 1:43. Dvornik, Photian Schism, 122. 15 This is one of the themes of Dvornik, Photian Schism; see esp. part 2, chaps. 5–6. 16 Details of the complaints about Lenten observance, confirmation, and other issues can be found in T. M. Kolbaba, “Meletios Homologetes ‘On the Customs of the Italians,’” REB 55 (1997): 137–68. 17 J. H. Erickson, “Leavened and Unleavened: Some Theological Implications of the Schism of 1054,” SVThQ 14.3 (1970): 156–58. The best introductions to the azyme controversy are ibid., 155–76, and M. H. Smith III, And Taking Bread . . . Cerularius and the Azyme Controversy of 1054 (Paris, 1978). 14

[ 122 ] Byzantine Perceptions of Latin Religious “Errors” other words, this was one of those periods in Byzantine history, like the sixth century and the iconoclast period, in which people fought over who had the right to define “orthodox” and “Roman.” Ironically, success had caused these fierce fights—the success of Byzantine armies, which had reconquered parts of southern Italy and huge areas of Asia Minor and Mesopotamia in the ninth and tenth centuries. These victories brought peoples into the empire who had been beyond its borders for a century or more. Their reintegration “posed a demographic problem, which the eleventh century transcribed and prolonged into a religious problem—that is to say, into a crisis of identity (for such is certainly the ultimate sense of Orthodoxy for the Byzantines).”18 The groups who reentered the empire considered themselves orthodox, catholic, apostolic Christians, but theologians in the great capital on the Bosphoros tended to label some of them as heretics. Others, considered orthodox, were not quite “Roman.” Armenians and Syrians, for example, might be neither “Roman” nor “orthodox” (meaning, to a Constantinopolitan, Chalcedonian in their theology). Then again, they might be “orthodox” but not Roman. Some Armenians had even become both “Roman” and “orthodox,” although this group probably did not include the recent immigrants. Only time would answer questions about the identity of these people—“foreign” or “Roman,” “orthodox” or “heretical,” “us” or “them.” Meanwhile, fierce struggles ensued. Most importantly for the evolution of Byzantine views of Latins, three of these questionable groups in the empire raised the issue of unleavened bread. The Armenians were the most important of the three. Armenia had been under Muslim rule until the ninth century. Then the decline of Abbasid power had allowed a period of independence. Then, in the second half of the tenth century, as Byzantium expanded eastward, Armenia was annexed to the empire, becoming the theme of Iberia in the early eleventh century. From 1045 to 1071 (battle of Manzikert), Armenia was ruled by the Byzantine Empire. Initially, the emperors involved in the annexation and integration of Armenia and Armenians into the empire were fairly tolerant of religious differences. Because Nikephoros II Phokas (963–969) and John Tzimiskes (969–976) wanted to repopulate eastern regions of Anatolia, they welcomed Armenian noble families who migrated into Cappadocia and southeastern Anatolia. These Armenians settled themselves and their ecclesiastical hierarchy within the empire.19 But this sort of tolerance would not last. After the annexation of Ani in 1045, when the last independent Armenian area fell to the armies of Constantine IX Monomachos (1042–55), relations between Armenian communities and their Greek Chalcedonian neighbors worsened. Very soon after the conquest, Monomachos began to crack down on Armenian heterodoxy. In 1048 Peter I, katholikos of the Armenian church, traveled to Constantinople for discussions. Discussions were friendly enough (at least the katholikos managed to stay out of prison), but in general both the emperor and his patriarch, Michael Keroularios, were 18 G. Dagron, “Minorite´s ethniques et religieuses dans l’orient byzantin a` la fin du Xe et au XIe sie`cle: L’immigration syrienne,” TM 6 (1976): 177–79. 19 G. De´de´yan, “L’immigration arme´nienne en Cappadoce au XIe sie`cle,” Byzantion 45 (1975): 41–116.

Tia M. Kolbaba [ 123 ] determined to wipe out the Armenian species of Monophysitism (as they saw it). They would soon begin to act on that determination.20 Thus, in the decade before the more famous events of 1054, a group of antiChalcedonian, “azymite” Christians debated with Chalcedonian, leavened-bread Christians. From these debates came some of the first treatises against azymes.21 This battle with the Armenians had a negative impact on discussions with Latins, for when the Greeks discovered that Latins were using unleavened bread, they “often seem[ed] too preoccupied with contemporary Armenian and Jewish polemics to evaluate properly the Latin position.”22 The same series of tenth-century conquests that made Armenia part of the empire also reintegrated Syria and its capital, Antioch, a competitor with Constantinople for ecclesiastical and even imperial preeminence.23 As they had encouraged the Armenians, the emperors also encouraged the Syrian Monophysites to repopulate imperial territories, especially northern Syria.24 This influx of foreign heretics was decried by Chalcedonian churchmen, and conflict ensued between those who advocated or at least practiced tolerance and coexistence and those who would not tolerate the “heretics.” The history of competition between Chalcedonians and Monophysites in these territories was ancient and bloody. As it had with the Armenians, imperial tolerance dissolved after the death of Basil II (976–1025). In 1029 the non-Chalcedonian patriarch John VIII Bar Abdoun was summoned to Constantinople. After a chance to state his views, he was condemned, excommunicated, and exiled by the synod. But worries about heterodoxy in the region of Melitene continued for some years.25 In Antioch in the 1050s, there were some fearful fights, including the burning of Orthodox churches.26 The link between Syrian Monophysites and the azyme controversy is not direct, for they use leavened bread in the eucharist. Still, their presence in the empire influenced the eleventh-century azyme controversy in two ways. First, the conflict with these heterodox Christians added to the general crisis of identity within the empire. Indeed, the documents regarding their status open for us one of the few windows onto such a crisis, through which we get not only a clear view of those whose definitions of “orthodox” and “Roman” won in the end, but also a fleeting glimpse of their opponents. Those opponents seem to have acted more than they spoke. We can only guess at their motives. 20

J. Hussey, The Orthodox Church in the Byzantine Empire (Oxford, 1986), 131. Knowledge of later events has often led to the conclusion that these treatises originate with the LatinByzantine conflicts of the 1050s. As Mahlon Smith, Jean Darrouze`s, and John Erickson have pointed out, however, the earliest anti-azyme treatises were ammunition in the debates with the Armenians. J. Darrouze`s, “Trois documents de la controverse gre´co-arme´nienne,” REB 48 (1990): 89–153; idem, “Notes: Un faux Peri` tw'n ajzu´ mwn de Michel Ce´rulaire,” REB 25 (1967): 288–90; Smith, And Taking Bread, 128 ff, 173; Erickson, “Leavened and Unleavened,” 175. 22 Erickson, “Leavened and Unleavened,” 175. 23 Dagron, “Minorite´s,” 205–7. 24 Ibid., passim. 25 V. Grumel, Les regestes des actes du patriarcat de Constantinople, rev. ed. J. Darrouze`s, vol. 1, fasc. 2–3 (Paris, 1989), nos. 838–40, 846. Dagron, “Minorite´s,” 200–204. 26 The Chronicle of Matthew of Edessa, trans. A. E. Dostourian, Armenia and the Crusades, Tenth to Twelfth Centuries (Lanham, Md., 1993), 2.2, 84–86; Dagron, “Minorite´s,” 208; Smith, And Taking Bread, 111. 21

[ 124 ] Byzantine Perceptions of Latin Religious “Errors” Some emperors, for example, apparently conceived of the empire as an ecumenical body, capable of integrating heterodox Christians in the short run and of converting them to Chalcedonian orthodoxy later. On this side of the debate, too, were the bishops and imperial officials around Melitene whom the patriarchal synod reprimanded for excessive tolerance of the “Jacobites,” as they called the Syrians. Among other things, these officials were accused of tolerating marriages between orthodox people and heretics and of accepting the testimony of heretics in court.27 One would like to know more about this largely unrecorded segment of the population for whom, it seems, the lines between “orthodox” and “heretic” were less clear or less important than they were for the members of the synod. On the other side of the debate were those whose voices have come to us in a multitude of texts. These men thought that the heretics would never convert. It was a self-fulfilling prophecy, especially since, as time passed, they tended to give heresy “a definition more geographic and ethnic than dogmatic.”28 For these men, the definition of orthodox Romans included not only a Chalcedonian dyophysite creed, but also a set of rituals and customs that were, in fact, the rituals and customs of only part of the empire. Latins, Armenians, Syrians, and many others who might consider themselves both orthodox and Roman were excluded.29 The second link between Syrian Monophysites and the azyme controversy was in the minds of these same orthodox adherents of the Council of Chalcedon, for they did not always distinguish Armenians from “Jacobites.” They were encouraged in this conflation by the actions of the groups themselves, who sometimes forgot their differences in their common hatred of the imperial-orthodox establishment and its attempts to enforce conformity.30 So, for example, Syrians and Armenians did occasionally collaborate in violent opposition to imperial attempts to shut down their churches. Thus, although a direct link between Jacobites, who use leavened bread, and polemic against users of unleavened bread is questionable, it is significant that the first eleventh-century figure to write a treatise against azymes was Patriarch Peter III of Antioch, a city where clashes between non-Chalcedonians and Chalcedonians had recently resulted in the burning of several Chalcedonian churches. Meanwhile, Jews, the group with the longest history of challenging Christian identity and self-definition, had not disappeared either. The number of Jews within the empire was increased by the return of areas of southern Italy to imperial control in the ninth century. Bari and Oria, for example, had substantial Jewish communities.31 After sporadic persecutions in the same century, renewed imperial tolerance for Jews encouraged many to migrate into the cities of the empire, especially into Constantinople, from further east.32 The status of these Jews in the empire remained ambiguous. On the one hand, 27

Grumel, Regestes, no. 846. Dagron, “Minorite´s,” 213. 29 Dagron, “Minorite´s,” 204: “The synod is alarmed, and we sense that there is a complete divorce between the orthodox, centralizing ideas of Constantinople and the political, social, and economic life of a region that is perhaps also ‘Byzantine’ but in a way different from the capital.” 30 Smith, And Taking Bread, 110–11. 31 A. Sharf, Byzantine Jewry from Justinian to the Fourth Crusade (London, 1971), 2. 32 Ibid., 107–16. 28

Tia M. Kolbaba [ 125 ] imperial laws from the period continued the Byzantine tradition by which Jews were second-class citizens. For example, the laws spelled out the penalties a Jew should suffer if he should manage—by influence or bribery, for he could not do it legally—to attain a civil or military office in the government. On the other hand, such laws indicate that some Jews had sufficient influence and wealth to circumvent the laws and that some Christians were willing to help them do so.33 In the years leading up to the quarrel between Cardinal Humbert and Keroularios in 1054, Jews had also come to the attention of imperial authorities in more negative ways. In 1042 they had participated, with Armenians and other “foreigners,” in the riots that accompanied an attempt to depose the empresses, Zoe and Theodora.34 In 1051 the Jews of Bari revolted, and the Christian citizens of the town retaliated by burning down the Jewish quarter.35 All of this is relevant to the azyme issue because Byzantines associated unleavened bread with the Jewish commemoration of Passover. Here Byzantines made all sorts of connections that modern historians find unconvincing, but our skepticism does not mean that the Byzantines themselves were not honestly convinced. Byzantine polemicists argued that using unleavened bread was in itself a “Judaizing” practice, indicating a lack of recognition that the New Testament had, in all ways, superseded the Old. From this perspective, Christians who used unleavened bread in the eucharist revealed that they were too attached to the Old Testament world of shadows and types, and not convinced of the grace of the new dispensation.36 Byzantine conviction on this point was reinforced by their belief that Armenians “Judaized” in other ways as well. They maintained a hereditary priesthood and they offered sacrificed meat within the sanctuary of the church—both practices that the Council in Trullo had condemned as “Jewish customs.”37 Although these arguments are not accepted by modern historians and were not accepted by Jews, Armenians, and Latins at the time, they were nonetheless strongly felt by some Byzantines. So we see that the early eleventh century had been a period of debate about orthodox identity, especially about who was to be excluded from the category of “orthodoxy.” In that debate, unleavened bread had been used as a marker—the symbol that distinguished nonorthodox “them” from orthodox “us.” When some of the same men who had excluded Armenians for this reason became aware that Latins, too, used unleavened bread, they concluded that Latins, too, were heretics. But other orthodox churchmen did not agree. Thus Leo of Ohrid’s letter against azymes, which is usually seen as the first volley in the war between Michael Keroularios and Humbert of Silva Candida, was addressed not to the pope or his cardinal, but rather to one of Leo’s acquaintances, John, bishop of Trani. Trani is in southern Italy and was at that time under Byzantine authority. John 33

Ibid., 112–13. Ibid., 116–17. 35 Ibid., 123. 36 Erickson, “Leavened and Unleavened,” 165–69. 37 Trullo 33 and Trullo 99—P.-P. Joannou, Fonti. Fascicolo IX: Discipline ge´ne´rale antique (IIe–IXe s.), vol. 1.1, Les canons des conciles oecume´niques (Rome, 1962), 166–67, 235–36 [⫽ G. A. Rhalles and M. Potles, Su´ ntagma tw'n qei´wn kai` iJerw'n kano´ nwn (Athens, 1852–59), 2:379, 2:543]. See also G. Dagron, “Judaı¨ser,” TM 11 (1991): 365. 34

[ 126 ] Byzantine Perceptions of Latin Religious “Errors” was “a Byzantine sympathizer” and “an honorary member of the hierarchy of the Great Church in Constantinople.” Leo wrote to John because he had heard that this otherwise orthodox bishop was accepting azymites in his church.38 In sum, the quarrel about azymes and Latins became fierce because it was internal— not a simple matter of “us” versus “them,” but a debate about the very definition of “us.” The Twelfth Century In the early years of the Komnenoi, the main concern remained azymes, even for those whose lives were disrupted by the Crusades.39 Patriarch John IV of Antioch (1089– 98) saw the First Crusade capture his city. Initially he stayed in Antioch, where he presided over both Greek and Latin clergy, but he later quarreled with the Latin rulers, fled to Constantinople, and abdicated. Around 1112 he wrote a tract on azymes in which he explicitly stated that he saw azymes as the most important error of the Latins: “The principal cause of the division between them and us is in the matter of azymes. . . . The matter of azymes involves in summary form the whole question of true piety; if it is not cured, the disease of the church cannot be cured.”40 John represents Byzantine churchmen who were convinced that the use of azymes was itself heretical. Other Orthodox theologians disagreed. Around 1090, Theophylact, archbishop of Ohrid (1088/ 89–post 1126), reproached those who raised trivial issues, including azymes, against the Latins: “It seems to me,” he wrote, “that a man versed in church tradition and aware that no custom is important enough to divide the churches, except for that which leads to the destruction of dogma, will not” agree that the Westerners “commit unpardonable sins” in matters such as azymes.41 This issue, then, was still debated because it was still not settled. Taking second place after azymes in the twelfth century was the issue of papal primacy.42 For example, in one of his texts written for debates with papal envoys in the capital in 1112, Niketas Seides (fl. first half of the 12th century) named twelve Latin errors, but insisted that only three were truly important: the procession of the Holy Spirit, azymes, and not calling Mary Theotokos. The importance of primacy is demonstrated by his first treatise, for although he began by saying that his concern was the three doctrinal issues, he ended up writing a refutation of the claims of the papal legates that Rome is the Mother of the churches.43 From that refutation, he moved to doctrinal matters by arguing that even if Rome were the Mother of the churches, mothers deserve to be followed only if they are faithful to God. His example of how Rome had not been faithful, and the subject of his second discourse, was azymes. The emphasis on papal 38

Smith, And Taking Bread, 91, 91 n. 47, 114–18. Beck, Kirche und theologische Literatur, 610. 40 B. Leib, Deux ine´dits byzantins sur les azymes au de´but du XIIe sie`cle (Rome, 1924), 245 [113], cited and trans. in Pelikan, Spirit of Eastern Christendom, 177. 41 Theophylact of Bulgaria, Proslali´a tini tw'n aujtou' oJmilitw'n peri` w»n ejgkalou'ntai Lati'noi, ed. P. Gautier, The´ophylacte d’Achrida: Discours, traite´s, poe´sies. Introduction, texte, traduction et notes (Thessaloniki, 1980), 279. 42 Darrouze`s, “Documents,” passim. 43 Ibid., 55. 39

Tia M. Kolbaba [ 127 ] primacy in Seides’ work, and in that of his contemporary, Theodore Smyrnaios (fl. 1080–1112), had two roots. First, it was a direct response to papal pretensions. In a letter to Emperor Alexios I, Pope Pascal II (1099–1118) had indicated that acceptance of Rome’s primacy, in matters of doctrine as in all else, was a prerequisite for ecclesiastical peace.44 Seides, Smyrnaios, and others had been assigned by the emperor to refute such claims. So, too, Patriarch John X Kamateros’ (1198–1206) later refutation of papal primacy was a direct response to Innocent III’s (1198–1216) assertion of that primacy.45 The second reason for Byzantine interest in papal primacy in the twelfth century relates directly to the Crusades. When Latin Crusaders conquered Antioch (1098) and Jerusalem (1099), they installed Latin patriarchs in both places. Those patriarchs owed at least nominal allegiance to the pope; at most, as was often the case, they were actually appointed by the pope. Constantinopolitans had not exactly been enthusiastic about the independence of the other Eastern patriarchs before the Crusades, and the patriarchs of Constantinople had been known to interfere in the other patriarchates. Nevertheless, they were quick to denounce Rome’s attempts to control them. Still, Joseph Gill’s assessment of Byzantine denials of papal primacy before 1204 rings true: they lack heat. John Kamateros’ debate with Innocent III is “largely academic” in tone, with “little sense of urgency.”46 Lists of Latin errors, the lowest and most rabid kind of polemic, do not raise the issue of papal authority until after 1204.47 The question is crucial in high-level negotiations with Rome, but it is not contested within the Byzantine church. Debates about papal primacy have Greek-speaking, Orthodox people on one side, Latins on the other. Even if the Latins score points in a debate, papal primacy is not going to be applied to the East. There is no identity crisis here. “They” believe in papal primacy; “we” do not. As with other issues, it is only when papal primacy becomes an issue within Greek circles that it generates some heat, and that happens only after 1204. Finally, the Filioque reemerges in the Komnenian period. It scarcely seemed important in the furor about azymes around 1054, but by the late eleventh century Theophylact of Ohrid and others returned to Photios’ claim that this was the truly horrible error. A century later, Innocent III called for the return of the Greek “daughter” church to her “mother,” the Roman church. Patriarch John Kamateros responded that it was the Roman church, in fact, that left, by teaching a heresy and adding to the creed.48 In sum, Byzantines in the Komnenian period worried about many of the same issues as in the time of Keroularios, especially azymes. They were also increasingly troubled by papal claims to plenitudo potestatis and all that that meant. Among other things, the

44

Letter 437, PL 163:588–89. Darrouze`s, “Documents,” 51–54, 57. A. Papadakis and A.-M. Talbot, “John X Camaterus Confronts Innocent III: An Unpublished Correspondence,” BSl 33 (1972): 26–41. Also published in part in PL 214:325–29, 756–72, and in J. Spiteris, La critica bizantina del Primato Romano nel secolo XII (Rome, 1979), 324–31. Discussion of these letters, with further bibliography: T. M. Kolbaba, “Barlaam the Calabrian: Three Treatises on Papal Primacy,” REB 53 (1995): 43–44. 46 Gill, Byzantium and the Papacy, 12. 47 The first reference to papal primacy in such texts is in Constantine Stilbes’ list, compiled after 1204: Darrouze`s, “Me´moire,” para. 4, 44. 48 Papadakis and Talbot, “John X Camaterus,” 34–35. 45

[ 128 ] Byzantine Perceptions of Latin Religious “Errors” Byzantines were beginning to realize that they could not openly discuss differences with the Latins if the Latins were not willing to give up the idea that the pope could do things alone, without the approval of the other patriarchs. 1204–1261 The next major development in Greco-Latin relations was traumatic and unlikely to endear Latins to theologians or any other Byzantines. The Fourth Crusade ended in the Latin army’s sack of Constantinople on 12 April 1204. Pope Innocent III hoped that the establishment of a Latin emperor in Constantinople might lead to reunion of the churches. On the contrary, it stiffened resistance to Latins within the Greek-speaking churches of Nicaea and Epiros. The Latin conquest did, however, change the priorities of Byzantines who criticized Latin doctrines and practices. Although azymes remained important and the Filioque was growing in importance,49 the dominant issue was now papal primacy.50 Texts from this period emphasize the role of the pope in the church and do so in ways that are not at all “academic.” For example, in discussions held between Greeks and Latins in the capital after the appointment of Thomas Morosini as Latin patriarch in 1204, the issue of papal primacy was central. In December 1204, the papal legate Peter Capuano held discussions with Greek clergy. He asked them to submit to the pope, but most of them refused. In 1206 most of the Greek clergy in the capital were still refusing to accept Morosini as their patriarch. In August, September, and October of that year, Patriarch Morosini and the papal legate Benedict, Cardinal of St. Susanna, debated with Greek clergy. The discussion laid out the arguments for and against papal primacy in what had, by then, become a formula. The Greeks remained adamant. They wanted their own, Greek patriarch.51 These events from early in the period of Latin rule reveal the first reason for the centrality of papal primacy in this period: in the areas they controlled after 1204, the Latins insisted that Greek clergy and bishops take an oath of obedience to the pope or be deposed from their churches. Moreover, they were quite open about the implications of that oath, for they refused to separate theological issues from papal primacy. If the pope was indeed the head of the church, if Rome was the mother of all the other churches—if, in short, all the Western claims for papal primacy in law and doctrine were true—then the only solution to the schism was for the daughter church to return to the mother, the schismatics to return to the catholic church. All other issues were subsumed 49 For example, in discussions between Greek theologians and papal legates in 1234, the Greek representatives insisted that the Filioque was the most important issue, while the Latins condemned the Greek refusal to accept unleavened bread in the eucharist. Gill, Byzantium and the Papacy, 65–72. 50 See, e.g., the letter of Patriarch Germanos II to the clergy on Cyprus (1229), PG 140:613–21; summary in Gill, Byzantium and the Papacy, 60. It might be possible to count up the number of references to any given topic in all the extant texts of the era, but it would not be particularly useful. Most texts that survive do so because later eras are interested in their content. Based on my reading of surviving theological texts and on the accounts of historians, I have reached the conclusion that papal primacy is the central issue in most debates; that judgment, while defensible, remains subjective. 51 Gill, Byzantium and the Papacy, 32–34. For typical Greek arguments against papal primacy, the best summaries are Darrouze`s, “Documents,” 42–88, and Spiteris, La critica bizantina.

Tia M. Kolbaba [ 129 ] under the issue of papal primacy.52 So the Latins gave the Greeks good reasons to think that this issue must be settled before any others could be. The Greeks learned the lesson well. When, in 1253, John III Vatatzes (1221–54) sent an embassy to the pope, seeking reunion of the churches, his proposals began with an acknowledgment of papal primacy, including the right of appeal to Rome in all church matters.53 But some Greek clergy and people were less adamant than the debaters of 1204 or 1206. Around 1205 the bishops of Rodosto and Negroponte submitted to papal authority.54 Papal letters from the pontificates of Innocent III and Honorius III (1216–27) reveal a number of monasteries that submitted to the pope and received, in return, papal protection for their rights and properties.55 In 1214 Patriarch Theodore Irenikos of Nicaea (1214–16) wrote to the people of Constantinople, exhorting them to remain true to their faith and not to vow obedience to the pope: “For how would your faith be preserved and safe-guarded, if you should agree to be one of the pope’s faithful?”56 Often cited as an example of Greek resistance, this letter is equally an indication that some of the people were wavering, possibly because of the “conciliatory policy towards the Greeks” that the second Latin emperor of Constantinople followed.57 Later, the case of the clergy of Cyprus reveals a similar ambivalence within the Orthodox community outside the City. Cyprus had been under Latin rule since 1191, and the Greek clergy there could not agree among themselves about the best way to coexist with the Latins. To what extent should they compromise? Could they take an oath of obedience to the pope and/or to the Latin bishop without compromising their orthodoxy? They quarreled about this issue for years, arguing about the limits of oikonomia. Asked for guidance, Patriarch Germanos II (1223–40) and his synod in Nicaea also failed to agree. They first ruled that the clergy of Cyprus could compromise with the Latin archbishop in various ways without betraying their faith, but later, under pressure from a more radical element from Constantinople, they modified this decision. Thus they added confusion to the situation instead of alleviating it.58 In light of this evidence, it is fair to say that past scholarship has often overemphasized the Greek clergy and bishops who fled to Nicaea and Epiros rather than take an oath of obedience to the pope. Most Greeks did resist the Latins, but some did so passively, while others compromised. The important point is that the compromisers existed; they must have had reasons for their actions. When their opponents admit that they exist, they

52 Examples of this Latin emphasis on Roman primacy can be found in nearly every piece of papal correspondence from the period. For examples, see Gill, Byzantium and the Papacy, 65, 67, 89, 93. 53 Ibid., 92–95. 54 J. Richard, “The Establishment of the Latin Church in the Empire of Constantinople (1204–1227),” in Latins and Greeks in the Eastern Mediterranean after 1204, ed. B. Arbel, B. Hamilton, and D. Jacoby (London, 1989), 47–48. 55 Gill, Byzantium and the Papacy, 52–53; Richard, “Establishment of the Latin Church,” 54. 56 Cited in Gill, Byzantium and the Papacy, 40. 57 Henry of Hainault reigned from 1206 to 1216 in Constantinople. See G. Ostrogorsky, History of the Byzantine State, trans. J. Hussey (New Brunswick, N.J., 1957), 381. 58 J. Gill, “The Tribulations of the Greek Church in Cyprus, 1196–c. 1280,” ByzF 5 (1977): 78–80; Hussey, Orthodox Church, 201–6; M. Angold, “Greeks and Latins after 1204: The Perspective of Exile,” in Latins and Greeks in the Eastern Mediterranean after 1204 (as in note 54), 72–75.

[ 130 ] Byzantine Perceptions of Latin Religious “Errors” claim that they were merely weak or evil, willing to sell their souls for safety or political preference. Maybe some were craven traitors, but we need not take their opponents’ word for it. It is equally likely that some men honestly believed that an oath of obedience to the pope was no stain on their orthodoxy. That belief, however, made their definition of orthodoxy quite different from that of the anti-Latin, anti-papal rigorists. Their compromise goes a long way toward explaining the heat with which other men attacked papal pretensions. Those who opposed Western ideas and Western authority—whether we call them intransigent or steadfast—were so fierce in their opposition because their definition of the boundaries of orthodoxy was not universally accepted. They were trying either to convince the compromisers that they were wrong or, failing that, to convince the rest of their contemporaries that the compromisers should be anathematized. The Palaiologan Period In 1261, in a serendipitous accident that many considered miraculous, a small army from Nicaea recaptured the city of Constantinople. No longer in exile, Michael VIII (1259–82) and the other leaders from Nicaea proceeded to reestablish the Roman Empire of Constantinople. But the question of reunion of the churches of Rome and Constantinople would not evaporate along with the Latin Empire. In fact, for Michael VIII, the question was perhaps more urgent than for any of his predecessors because various Western enemies proposed a “crusade” against the “schismatic Greeks” to recover the empire for catholic Christendom. To fend off these attackers, Michael opened and maintained negotiations with the papacy for reunion of the churches. For these negotiations, even more than for those during the Empire of Nicaea, papal primacy was the dominant theme, and for many of the same reasons. The popes still insisted that this was the fundamental issue,59 and Byzantines still disagreed among themselves about compromise. In the collection of documents related to the Second Council of Lyons (1274) published by Vitalien Laurent and Jean Darrouze`s, the dominant issues remain papal primacy, the right of appeal to the papacy, and the commemoration of the pope in the Byzantine liturgy.60 But the Filioque continued to grow in importance. It became the central issue sometime around the Second Council of Lyons. Many will challenge the idea that it was around 1274—and only then—that the Filioque became the crucial issue. After all, it was the most important issue for Photios, for Theophylact of Ohrid, for Greek theologians at Nicaea in 1234, and for theologians of the Palaiologan period. It remains the most important issue for many theologians today. It is quite natural to conclude that it has always been the most important issue, at least for thoughtful Christians. But it has not. The disputants of the 1050s hardly mentioned it. Treatises on the topic, including the statement composed by the Nicene synod in 1234,61 appear in the twelfth and thirteenth centuries, but they are far less common than discussions of azymes and papal prim59

See Kolbaba, “Barlaam the Calabrian,” 43–48, esp. the letter of Clement IV quoted there. V. Laurent and J. Darrouze`s, Dossier grec de l’Union de Lyon (1273–1277), Archives de l’Orient chre´tien 15 (Paris, 1976). 61 Gill, Byzantium and the Papacy, 72. 60

Tia M. Kolbaba [ 131 ] acy.62 This does not mean that the Filioque was not important to many people and at some times before 1274. Certain men of a philosophical bent seem always to have been troubled by the implications of double procession, while men who were concerned with authority within the church often challenged the unilateral character of the addition to the creed. But the Filioque was not the subject of more treatises, more debates, or more invective than papal primacy or azymes before 1274. So we should be more surprised than we are when we see how the issue dominated the Palaiologan period. This dominance can be seen in a number of ways. For example, a rough count of the authored works listed in the Greek Index Project reveals that about 70 percent of all the polemical works written in the Palaiologan period were against the Latins. Of those, about half were about the procession of the Holy Spirit.63 For another example, Barlaam the Calabrian wrote twenty-one anti-Latin treatises. Fifteen of these are concerned in some way with the relations of the persons in the Trinity; ten explicitly mention the procession of the Holy Spirit in their titles.64 This dominance needs explaining. We cannot simply claim that the Filioque is the most important issue in an absolute, philosophical sense; such a claim cannot be proven or verified. More to the point, even if it is the most important issue, it was not always seen and treated as such in Byzantium. As was true of azymes and papal primacy, the Filioque became a burning issue only when it became an issue within the Eastern church. There were no Byzantine defenders of the Filioque before the 1270s. It was not necessary to write treatises to convince other Byzantines that the addition to the creed was illegitimate and possibly heretical. Within Byzantium—and therefore within most theological discussions in Byzantium—the belief that the Spirit proceeded from the Father alone could be assumed; it did not have to be defended. Moreover, in general, when the question did come up in arguments with the Latins, its theology was seldom discussed systematically. The error of the Latins in this matter came primarily from their unilateral addition to the creed, as Theophylact of Ohrid expressed so clearly: “But the Symbol of the faithful must be the Symbol freed from all alteration . . . for not even the axe-wielders of Ezekiel spared those marked with the sign if they did not observe that their sign was not counterfeit.”65 Perhaps the Latins had simply not thought through or were not capable of thinking through the theological implications of this novelty. Theophylact surmised that the Latin language had no way of distinguishing the “procession” of the Holy Spirit from the Father from his “having been sent” by the Son. He assumed that if he simply showed the Latins their misunder-

62 I base this statement on a close study of the polemical works catalogued in H.-G. Beck’s Kirche und theologische Literatur, as well as on more recent studies of the theological debates of the period. 63 R. E. Sinkewicz and W. M. Hayes, Manuscript Listings for the Authored Works of the Palaeologan Period (Toronto, 1989). The other half is divided among the following, in roughly descending order: treatises entitled generally “Contra Latinos” or “De unione” or something similar; treatises against purgatory, against azymes, against papal primacy; and miscellaneous single occurrences, such as a treatise “Against Thomas [Aquinas].” 64 For a list of Barlaam’s works, see R. E. Sinkewicz, “The Solutions Addressed to George Lapithes by Barlaam the Calabrian and Their Philosophical Context,” MedSt 43 (1981): 185–94. 65 Theophylact, Peri` w»n ejgkalou'ntai Lati'noi, 251.

[ 132 ] Byzantine Perceptions of Latin Religious “Errors” standing and explained how dangerous the theological implications of this addition were, they would concede the point.66 During the discussions held at Nicaea and Nymphaeum in the period of the Latin Empire of Constantinople, that condescension was less evident. Latins came to discussions armed with both a knowledge of Greek and manuscripts of the writings of the Greek fathers of the church. Still, the Greek theologians involved in those debates stressed the Filioque precisely because it was the area where they felt the firmest ground beneath their feet. They were utterly convinced of the rightness of their position, and nothing the Latins said changed any of their minds. But the firm ground began to shake in the period around the Council of Lyons. In late 1273 or early 1274, John Bekkos (born ca. 1230, patriarch of Constantinople 1275– 82), an important Constantinopolitan churchman, became convinced that the Latin position on the Filioque was theologically defensible. From his time on, there was a debate within Byzantium about the Latin position. The unionist defenders of the Filioque adduced quite sophisticated arguments, based not only on logic but also on the writings of Greek and Latin church fathers. The anti-unionist, anti-Filioque people were caught off-guard by this, at first, and did not always do a good job of defending their position. Although some partisan historians still dismiss him with a few scathing words, Bekkos was not obviously and self-evidently wrong. He may not have been the most subtle theologian in history, but he convinced many other men of his position. He also became patriarch after the Council of Lyons and the union manufactured there. Later, when that union was repudiated, the first synod convened to condemn Bekkos and the other unionists was more a lynch-mob than a thoughtful discussion. The second synod, convened years later, was unable to convert Bekkos and his supporters to the anti-Filioque opinion.67 Patriarch Gregory II of Cyprus (1283–89) produced strong, reasoned refutations,68 but Bekkos had his supporters, both at the time and down to the end of the empire.

Changing Tone John Bekkos did not convert to Roman Catholicism; he merely believed that the theologians who argued in favor of the double procession of the Holy Spirit were correct and in agreement with the fathers of the church. In the centuries after his death, several prominent Byzantine intellectuals would reach a broader conclusion: that Latin theologians and philosophers were right about many things. Some of these intellectuals would convert; the most famous example is Cardinal Bessarion (ca. 1399–1472).69 Demetrios Kydones, a fourteenth-century convert to Catholicism, put it best when describing how, 66

Ibid., 253–55. Hussey, Orthodox Church, 247. 68 Hussey, Orthodox Church, 247–49; A. Papadakis, Crisis in Byzantium: The ‘Filioque’ Controversy in the Patriarchate of Gregory II of Cyprus (1283–1289) (New York, 1984). 69 The most thorough study of his life and work is L. Mohler, Kardinal Bessarion als Theologe, Humanist und Staatsmann, 3 vols. (Paderborn, 1923–42). A more accessible and up-to-date survey is J. Gill, Personalities of the Council of Florence (New York, 1964), 45–54. 67

Tia M. Kolbaba [ 133 ] in his youth, he began to study Thomas Aquinas and other Latin theologians: “Now it would become apparent that the Latins too had people capable of the highest intellectual attainments—something that had not been widely known in the past among the Byzantines. . . . For too long, my Byzantine countrymen had been content to hold on to the staid old notion that mankind was divided into two groups: Greeks and Barbarians. . . . The Latins could not be credited as being capable of anything worthy of human beings.”70 This awareness of Latin theological sophistication brings us to the second major point of this paper: between Photios and Bekkos there had been a fundamental shift in how Byzantine intellectuals perceived their Western European brethren, a shift reflected in the changing tone of anti-Latin texts. The Ninth Century Photios and others after him manifested the classic middle Byzantine attitude toward Western “barbarians.” Photios claimed that Pope Leo (III?) had made Christians in Rome say the creed in Greek because Latin was such an inferior language that it “often render[s] false notions of the doctrines of the faith.”71 Photios was willing to blame most Western errors on ignorance and lack of education. Even when he descended to namecalling and aspersions, his epithets did not resemble later polemic. Rather, he used the classic terminology of heresy and heretics: arrogance, rashness, insolence, impudence, pride. In The Mystagogy of the Holy Spirit, for example, he described the advocates of the Filioque in ways that echo talk about heretics throughout history. For example, he wrote of “the arrogance of those contentious men who hold fast to unrighteousness and strive against the truth.”72 He referred to their “rash impudence,” “brutal and insolent attacks,” and “lawlessness.”73 “When all is said and done,” he wrote, “it comes down to the same unending pride.”74 We would err if we put too much weight on these descriptions as indications of what Photios thought of “Latins.” If we compare these epithets to Photios’ synopsis of the ecumenical synods in his letter to the Khan of Bulgaria, we see striking parallels. Arios was also proud; he had “an overweening attitude” and refused to “see something that is true of everything and self-evident.”75 Makedonios, too, ignored the obvious and was “arrogant” and “insolent.”76 This is an old story: heretics are proud and devil-inspired, refusing to see what any honest, humble, praying man would see. They are mad, arrogant, insolent, blasphemous, and willfully blind. When Photios described 70 Demetrios Kydones, Apologia 1, ed. G. Mercati, in Notizie di Procoro e Demetrio Cidone, Manuele Caleca e Teodoro Meliteniota ed altri appunti per la storia della teologia e della letteratura bizantina del secolo XIV, ST 56 (Vatican City, 1931), 364; trans. J. Likoudis, Ending the Byzantine Greek Schism (New Rochelle, N.Y., 1983), 26. See also F. Kianka, “The Apology of Demetrius Cydones: A Fourteenth-Century Autobiographical Source,” ByzSt 7.1 (1980), 57–71; eadem, “Demetrius Cydones and Thomas Aquinas,” Byzantion 52 (1982): 264–86. 71 Photius, Lo´ go" peri` th'" tou' Agi J ´ou Pneu´ mato" mustagwgi´a", PG 102:376, trans. J. P. Farrell, Saint Photios: The Mystagogy of the Holy Spirit (Brookline, Mass., 1987), 103. 72 PG 120:280; trans. Farrell, 59. 73 E.g., see PG 120:297–301; trans. Farrell, 66–68. 74 PG 120:324; trans. Farrell, 80. 75 Photius, Epistulae et Amphilochia, 1:5; trans. D. S. White and J. R. Berrigan Jr., The Patriarch and the Prince (Brookline, Mass., 1982), 42. 76 Photius, Epistulae et Amphilochia, 1:6; trans. White and Berrigan, 43–44.

[ 134 ] Byzantine Perceptions of Latin Religious “Errors” Frankish missionaries in Bulgaria in such terms, he was not commenting on the ethnic or racial characteristics of Westerners; he was describing heretics. The Eleventh Century The tone of the arguments in 1054 was a bit worse than at the time of Photios. Yet it was nowhere near as acrimonious as generally assumed. That general assumption rests on a reading of only what Humbert of Silva Candida and Michael Keroularios wrote. In the balance against these writings we need to put not only the oft-noted irenic position of Peter of Antioch, but also a multitude of other texts from the period. In these texts, Byzantine writers still condescended to their Western brothers. Both Peter of Antioch and Leo of Ohrid assumed that the Latins had wandered from the true path out of ignorance, and that if they were corrected by their more learned, wiser Eastern brethren, they would return to the straight and narrow.77 Latins were barbarians, ignorant of doctrine. The superior orthodox Christians must be patient with them. Byzantine disputants also limited their instruction to a part of the Western church; they did not maintain that the whole Western church had fallen into error. Keroularios, for example, although he was inconsistent on this point, usually insisted that the pope was not to blame for the errors of the West or for his dispute with Humbert. He distinguished between the pope, with whom he wanted an alliance, and the “Franks,” including his archenemy and the Byzantine governor in southern Italy, the Lombard Argyros. Keroularios’ synod in 1054 did not condemn the pope or Westerners in general, but claimed that Humbert and the other legates were impostors bearing forged letters altered by Argyros.78 Peter of Antioch defended Westerners on the basis of his knowledge of them, and insisted that if some Westerners were violating canon law (by eating strangled things or marrying within forbidden degrees), they must be doing so without the knowledge of the pope.79 Behind this last comment, and fundamental to our understanding of the events of 1054, was an awareness that the West was not a monolith. Peter of Antioch, Leo of Ohrid, and Michael Keroularios did not live in a world where the division of East from West was clear and all-important. Instead of that bipolar world—Rome facing Constantinople—that dominates modern accounts of 1054, we find different groups, with different interests, involved in ecclesiastical negotiations and disputes. In Italy alone the actors included the pope, the German emperor, the Normans in southern Italy, the Lombards in southern and central Italy, and the indigenous Italians of the same region, some of whom still considered themselves subjects of the emperor in Constantinople. Sometimes a single individual embodied this complex world: Argyros, whom Keroularios blames for the whole fiasco, was a Lombard of the Latin rite. He had lived in Constantinople. In 1054 he was the Byzantine imperial governor in southern Italy. Other examples 77 E.g., Peter of Antioch, Letter to Michael Keroularios, PG 120:805: “For they are our brothers, even if it happens that, through rusticity and lack of education, they have often fallen from what is proper.” 78 Synodal Judgment, PG 120:741. 79 Letter to Keroularios, PG 120:808.

Tia M. Kolbaba [ 135 ] of this multilateral world include the recipients of two of the earliest and most important anti-azyme texts: Dominic of Grado and John of Trani. Peter of Antioch wrote to Dominic, bishop of Grado, probably in the spring of 1054.80 No one who studies the events of 1054 overlooks this text, for it is one of the earliest.81 One aspect of its context, however, is seldom explicitly noted. As a result, it is generally presented as a straightforward example of a letter from an “Eastern” patriarch to a “Western” bishop, as if the distinction were as clear-cut as it would be in the thirteenth or fourteenth century. But history had made Grado an odd sort of liminal place. For reasons too complicated to discuss here, the metropolitans (or patriarchs, as they came to call themselves) of Grado sat on a cathedra not on the island of Grado but in Venice. Now, in hindsight, we see clearly that Venice had gained political independence from Constantinople in the course of the tenth century. But its cultural independence could not have been so clear. Dominic of Grado, the recipient of Peter’s letter, built the “new” church of San Marco, the one we see today. He imported architects and skilled craftsmen from the East to do so, which explains the fundamentally Byzantine character of San Marco.82 So the geography of Venice, between Byzantium and the West, was reflected in its culture, as it always has been. Leo of Ohrid’s treatise on azymes illuminates another section of this multicultural world. Leo’s addressee, John, bishop of Trani, was a representative of the Byzantine church in southern Italy, and he was “asked to call these matters to the pope’s attention only after he [had] ‘corrected himself.’”83 Aware of the ethnic and religious diversity of Western Europe, eleventh-century authors of Byzantine religious texts did not yet engage in the kind of name-calling that would characterize later anti-Latin polemic. For them (if not for the contemporary historians),84 Westerners were not barbarian “Franks” or “Kelts.” Usually they were called “Romans,” even when they erred. Thus Keroularios told Peter of Antioch about Roman errors ( RJ wmai¨kw'n sfalma´ twn).85 Peter replied, speaking also of “Romans,” whom he distinguished from “Vandals,” although he feared that the Romans might have been influenced by the Vandals.86 When tribal names of barbarians appear, it tends to be in what we would call secular contexts. Keroularios, for example, wrote about his desire to form an alliance with the pope against the “Franks,” by which he meant the men we call “Normans.”87 Put simply, Byzantines were not yet constructing a world in which the “Latins” or “Franks” from the West were a monolithic, threatening group. Keroularios was the only person to imply that the Western church as a whole was in a state of schism. Although

80

PG 120:755–82. Smith, And Taking Bread, 54–59, 134, 157, 173, 178–79. 82 DHGE 21 (1986), s.v. “Grado.” 83 Leo of Ohrid, Letter to John of Trani, PG 120:835–44. Quote from Smith, And Taking Bread, 114. For other information regarding Leo’s letter, see Smith, 106–8, 156–57, 173, 174. 84 See Alexander Kazhdan’s contribution to this volume. 85 PG 120:789. 86 PG 120:805. 87 Letter to Peter of Antioch, PG 120:784. 81

[ 136 ] Byzantine Perceptions of Latin Religious “Errors” he usually maintained that they had nothing against the popes, and that his only quarrel was with the “false” envoys and their “forged” papers, Keroularios did tell Peter of Antioch that: “From the sixth holy and ecumenical council to the present, the commemoration of the pope has been excised from the sacred diptychs of our holy churches. [This is] because the pope of that time, Vigilius, did not want to come to that council, nor to anathematize what Theodoritos wrote against the orthodox faith and against the twelve chapters of St. Cyril, or the letter of Ibas. And from that time to the present the pope has been cut off from our holy and catholic church.”88 But Peter rebuked Keroularios for this statement, pointing out that it was wrong both in its central point and in its knowledge of history: I was ashamed of these latter things contained in the letter of Your Honor, nor do I know what to say, believe me. . . . For before examination and complete understanding, from vain rumor you have set forth that which never happened as if it had happened. . . . For Vigilius was at the fifth council . . . , but he was not at the sixth council. The interval between these two synods was 139 years. It did happen, for a brief while, that commemoration was cut off on account of [Vigilius] contending with the most holy patriarch Menas and subjecting him to demotion. [This schism lasted] until the archbishops made peace and were reconciled with one another. At the sixth holy synod, the pope was the priest Agathon, a worthy and divine man, wise in divine things. Read the acts of the sixth council, as it is customary to do on the Sunday after the Exaltation of the Venerable Cross. For you will find there that the aforementioned Agathon was gloriously acclaimed in that holy council.89 Even later writers who copied and expanded Keroularios’ list of Latin errors tended to leave off his erroneous introduction. The idea that the popes were heretics who had been in error for centuries was not commonly accepted in 1054. Keroularios’ claim that they were was idiosyncratic. Two other pieces of evidence are often adduced in support of the idea that something radically different and more hostile took place around 1054. First, it is often asserted that Keroularios closed the Latin churches in Constantinople.90 However, as Mahlon Smith has noted, this statement is based on slim evidence. Humbert of Silva Candida alleged on several occasions that Keroularios persecuted Latin churches. This persecution seems mostly to have taken the form of “mocking” the Latins by calling them “azymites.”91 Humbert claimed only once that Keroularios actually closed Latin churches in the capital, and even then he qualified his statement as hearsay. Later, when he was in Constantinople, he reformed the practices of certain churches there. These must have been churches founded for the Westerners in the city; neither Keroularios nor any other 88

PG 120:788–89. PG 120:797–800. 90 For example, it is stated in passing as a fact by Hussey, Orthodox Church, 132. Grumel, Regestes, no. 863, “Ordre de fermer les e´glises Latines de la capitale,” assumes that such a document once existed but notes that no such document is extant. 91 PL 143:759. 89

Tia M. Kolbaba [ 137 ] Greek in the city would have let him “reform” a Greek church. If Humbert “reformed” Latin churches in Constantinople, those churches must have been open.92 Second, it is often asserted that Keroularios stirred the common people up to join his conflict with the emperor and Western envoys, and that he found it easy to do so because the people harbored a xenophobic hatred for Westerners.93 A closer look makes this assertion even more questionable than the first. Evidence for popular anti-Latin sentiment in this period is meager. Keroularios did indeed raise the rabble on more than one occasion, but the people’s rage seems to have had other roots: unhappiness with Constantine IX Monomachos, as well as the general malaise of this period of instability in Byzantium.94 In sum, the events of 1054 were insignificant for the short run. They received little attention in the empire at the time. The first known references to a “schism” between Keroularios and Humbert date from the early twelfth century.95 As in the time of Photios, much of the conflict was more individual than general—Humbert versus Keroularios, rather than Rome versus Constantinople. Many of the features of later relations between Byzantium and Latins were not yet evident. Perception of the West as a unity and a threat; anxiety about Latin theological sophistication; popular antipathy—all of these would emerge later. The Twelfth Century During the Komnenian period (1081–1204), Latin penetration of the empire grew exponentially. This has been discussed too often and too well by other scholars to need elaboration here.96 But this was not the impact of an active, vibrant, potent, masculine Western force on a passive, decadent, impotent, effeminate Eastern despotism. Byzantine emperors and their subjects reacted to Western pressures and exerted their own pressures on Westerners. To read Paul Magdalino’s account of Manuel I’s empire (1143–80) or Ralph-Johannes Lilie’s account of Byzantine relations with the Crusader kingdoms of Syria and Palestine is to see Byzantine emperors both exerting great influence on Westerners and adopting Western ideas and methods.97 The central dichotomy of Alexios I’s (1081–1118) or his grandson’s world was not as much between “Rhomaioi” and “Latinoi” as between “those who are for me” and “those who are against me”; not between “Roman” ways of doing things and “Latin” ones, but between “what works” and “what does not.” The borders were permeable, and both sides were changed by extended contact. 92 For a fuller statement of this argument, with citations of the primary texts, see Smith, And Taking Bread, 119–21. 93 Ostrogorsky, History of the Byzantine State, 297; Hussey, Orthodox Church, 134. 94 He´le´ne Ahrweiler concurs: “Recherches sur la socie´te´ byzantine au XIe sie`cle: Nouvelles hie´rarchies et nouvelles solidarite´s,” TM 6 (1976): 121. 95 A. Michel, Humbert und Kerullarios, Studien, vol. 1 (Paderborn, 1924), 30–33. 96 Even to survey the bibliography of this topic would be a monumental task. See the bibliography in P. Magdalino, The Empire of Manuel I Komnenos, 1143–1180 (Cambridge, 1993), or in M. Angold, Church and Society under the Comneni, 1081–1261 (Cambridge, 1995). 97 Magdalino, Manuel I; R.-J. Lilie, Byzantium and the Crusader States, 1096–1204, trans. J. C. Morris and J. E. Ridings (Oxford, 1993).

[ 138 ] Byzantine Perceptions of Latin Religious “Errors” This two-way flow of men and ideas is a recurring theme in recent histories of the politics, economy, and warfare of the empire. In contrast, the theological sphere in this period is generally portrayed as impermeable. Adopting the view our Byzantine sources would like us to adopt, we tend to see Orthodoxy as a kind of fortress with outer walls of slippery-smooth marble. Latin theological ideas, recognized as novelties and dangerous heresies, bounced off this marble without so much as leaving a smudge. In the end, however, the same sources inadvertently show us a rather different Byzantium, and we cannot quite believe their explicit message of a faith untouched by Western ideas. Perhaps Byzantine theologians remained truly unaffected by Latin ideas in the time of Photios, for at that time their theological and philosophical training was superior to that of anyone in the West. By the twelfth century, however, such superiority was melting away. The Latins were catching up, and some Greeks knew they were. Given the acrimonious ad hominem attacks of 1054, it is unlikely that Humbert’s logic impressed them. In contrast, when Alexios I listened to debates regarding the procession of the Holy Spirit between a Latin bishop and some Greek theologians (1112), he was convinced by the Latin arguments and sent his own theologians back to the drawing board. The extant account of this debate was written by the Latin bishop, so we should be skeptical. Yet it is true that various Greek theologians worked very hard at refuting Latin arguments around 1112. One senses that their confident condescension had been shaken.98 Still more interesting is the controversy, in the time of Manuel I, over the Gospel phrase, “The Father is greater than I.” The trouble began when an imperial ambassador, Demetrios of Lampe, returning from the Latin West, brought with him ideas about the relations of the persons in the Trinity. He reported that he had heard Latins say that the Son was at the same time inferior to and equal to the Father, and he proclaimed that opinion ridiculous. Emperor Manuel differed with Demetrios and deputed Hugo Eteriano, a Pisan theologian and friend, to argue against him. Probably most of the churchmen of Constantinople supported Demetrios, in part because they believed that his opinion was the traditional, native one, while Hugo’s was a Latin innovation.99 However that may be, the emperor dominated the council that decided the matter in 1166, and so Demetrios’ position was anathematized in the Synodikon of Orthodoxy and Hugo’s inscribed on enormous marble tablets attached to the walls of the Great Church.100 The doctrine that had been criticized by a Greek and defended by a Latin became official dogma within the Greek church.101 98

Beck, Kirche und theologische Literatur, 616; ODB, s.v. “Grossolano, Peter,” 2:885. Magdalino, Manuel I, 287–90. 100 On the tablets, see C. Mango, “The Conciliar Edict of 1166,” DOP 17 (1963): 317–30. 101 The sources for this controversy have been studied very carefully, and much has been learned from rather scanty material. Still, it may be worthy of another study; ideas about the import of Latin influence in the Komnenian period have changed a great deal since the last careful study of the origins and meaning of this conflict. My analysis of the events is based on a quick reading of the primary sources and on reflections inspired by Magdalino’s version of the events in P. Magdalino, “The Phenomenon of Manuel I Komnenos,” in Byzantium and the West, c. 850–c. 1200, ed. J. D. Howard-Johnston (⫽ ByzF 13; Amsterdam, 1988); repr. in Tradition and Transformation in Medieval Byzantium (Aldershot, 1991), no. , 196–98. The most thorough article on the controversy, with publication of texts, is P. Classen, “Das Konzil von Konstantinopel und die Lateiner,” BZ 48 (1955): 338–68. See also G. Thetford, “The Christological Councils of 1166 and 1170 in Constantinople,” 99

Tia M. Kolbaba [ 139 ] Although many details of the case elude historians, its broad outlines reveal three important features of Byzantine relations with Latins in the twelfth century. First, the intellectual boundaries between East and West were permeable. Demetrios traveled west and picked up a controversy there. Hugo Eteriano, an Italian, lived in Constantinople and entered the lists as the emperor’s champion. This permeability complicates historical reconstruction of the boundaries between East and West, Greek and Latin, Orthodox and Catholic. Even what initially seems like the simplest question has no answer: Which was the “Latin” position here? At first glance, it seems clear that Hugo’s was the “Latin” opinion. But the answer is probably more complex than that. According to the historian John Kinnamos, Demetrios’ position was imported, for he had “returned from [the West] full of drivel.”102 We cannot unquestioningly believe Kinnamos, however, who was ever-supportive of the emperor and ever-contemptuous of Latins; he omitted any mention of Hugo’s role and presented the emperor’s opinion as wholly self-generated. In the end, it seems that both sides of the debate were inspired by Western concerns, but that their accounts after the fact were anxious to hide Western influences.103 When we read between the lines, Orthodox Constantinople’s walls look more like a cellular membrane than a marble castle. Second, Byzantines continued to disagree among themselves about the definition of orthodoxy and specifically about whether that definition included Latins. On the one hand, around this time a list of Latin “errors,” which had probably been circulating for a while, surfaces in the historical record. This, the most scurrilous kind of anti-Latin polemic, reveals the existence of a virulently anti-Latin contingent in Constantinople.104 On the other hand, there is obviously a problem with concentrating too much on the anti-Latin crowd. At least in the controversy surrounding Demetrios of Lampe, they seem to have lost. Hugo and Manuel won; it was their position that became the orthodox position. The marble tablets in the Great Church remained there until 1567.105 Besides, that list of Latin errors was written to convince fellow Byzantines, not to convince Latins. In other words, the list reveals the existence of Byzantines who were not convinced that Latins were filthy heretics, as well as the existence of those who were. Third, it was necessary for each side to downplay the role of Western influence in these events. Demetrios and his supporters saw and portrayed themselves as diligent polishers of the smooth marble walls of Orthodoxy, trying to rebuff all things Latin, while

SVThQ 31 (1987): 143–61; A. Dondaine, “Hugues Ethe´rien et la concile de CP de 1166,” HJ 77 (1958): 473–83; J. Gouillard, “Le Synodikon de l’Orthodoxie,” TM 2 (1967): 76–80, 216–26; L. Petit, “Documents ine´dits sur le concile de 1166 et ses derniers adversaires,” VizVrem 11 (1904): 465–93. 102 John Kinnamos, Deeds of John and Manuel Comnenus, trans. C. M. Brand (New York, 1976), 189, book 6.2; Ioannes Kinnamos, Epitome rerum ab Ioanne et Alexio Comnenis gestarum, ed. A. Meineke, CSHB (Bonn, 1836), 251. 103 Magdalino, “Phenomenon,” 196–97. 104 Hugo Eteriano translated such a list into Latin sometime before 1178. Hergenroether published both Hugo’s translation and a Greek list that mostly corresponded to it, entitled: Peri` tw'n Fra´ ggwn kai` tw'n loipw'n Lati´nwn: J. Hergenroether, Monumenta graeca ad Photium eiusque Historiam Pertinentia (Ratisbon, 1869), 62–71. See also A. Argyriou, “Remarques sur quelques listes grecques e´nume´rant les he´re´sies latines,” ByzF 4 (1972): 13–15; Darrouze`s, “Me´moire,” 54–55. 105 Mango, “Conciliar Edict,” 317–21.

[ 140 ] Byzantine Perceptions of Latin Religious “Errors” Kinnamos did not mention Hugo Eteriano’s role and portrayed Demetrios’ heresy as an imported product. A summary of these mixed messages is difficult. Given that it was desirable for both sides to hide any “Latin” connection in their teaching, it seems that anti-Latin sentiment was strong and getting stronger. Yet the presence of Latins in the empire and frequent embassies of Greeks to the West allowed both friendly and unfriendly interaction, as well as mutual influences that usually went unrecorded in contemporary sources. This was still the empire of the Komnenoi, with their ties of blood and friendship to Latins, and with subjects who had not yet experienced the traumatic events of 1204. 1204–1261 After 1204 the opinion that Latins are Christian brothers was modified, even in the most pro-Latin circles. Everyone took the separation of the churches for granted. The period was characterized by one attempt after another to reunite the churches. It also saw both violent resistance to these efforts and a kind of openness to Western ideas. Paradoxically, the openness was often a result of the division. Some Byzantine scholars thought they had to learn more about the Latins before they could refute them. In the end, these trends led to a wide spectrum of opinion. At one end of the spectrum is the most scurrilous kind of polemic against the Latins, which became popular. Lists of Latin “customs” or “errors” circulated widely. These included not only theological issues such as the Filioque and liturgical issues such as azymes, but also disgust at the things Latins ate and the clothes they wore.106 In this way, opponents of church union sought to undermine the advocates of union by associating them with filthy heretics. For example, the anti-unionists could not acknowledge that John Bekkos might have reached a conviction of the orthodoxy of the Filioque by reading patristic texts. He must, rather, be a servant of the pope. One of the most infamous anti-Latin texts, the “Dialogue of Panagiotes with an Azymite,” dates to around the Second Council of Lyons. It begins with a description of the arrival of a papal envoy in Constantinople. He is met by John Bekkos, who wears a miter and a ring, which, the author assures us, are symbols of the pope. The papal envoy is leading a mule with a basket on its back. In the basket is an image of the pope. Both Bekkos and Emperor Michael VIII perform acts of submission in front of this mule. Michael actually leads it by its bridle, an idea familiar to Western medievalists and an allusion to Michael’s earlier, hypocritical submission to Patriarch Arsenios.107 An even more striking example of this sort of condemnation-by-association is the case of Patriarch Gregory II of Cyprus. In spite of being the person who finally defeated Bekkos, in spite of his sophisticated elucidation of the Greek doctrine of the procession of the Holy Spirit, he ran into trouble. His enemies accused him of heresy in that very elucidation, and more than one of them 106

See, e.g., Darrouze`s, “Me´moire,” para. 30, 63, 65, 66, 75. D. J. Geanakoplos, Interaction of the ‘Sibling’ Byzantine and Western Cultures in the Middle Ages and Italian Renaissance, 330–1600 (New Haven, Conn., 1976), 158. Greek text ed. A. Vasiliev, Anecdota graeco-byzantina (Moscow, 1893), 179. 107

Tia M. Kolbaba [ 141 ] implied that his error came from his origins on Latin-dominated Cyprus.108 He may have been deposed from the patriarchate for entirely political reasons, but a good way to justify such a deposition was to portray him as a “Latin.” At the other end of the spectrum, some intellectuals of the Palaiologan period admired Latin learning and considered it superior to Greek. This shift is symbolized most vividly by the year 1274, for 1274 is not only the date of the Second Council of Lyons but also the date of the death of St. Thomas Aquinas. In other words, Latin theologians were reaching their peak. Popular opposition to the Latins was not significantly changed by this, but the opposition of intellectuals took on a different tone. Unlike Photios, Theophylact of Ohrid, or Peter of Antioch, intellectuals of the Palaiologan period were not certain of their own superior theological reasoning. Some among them even decided that the Latins had surpassed them.109 The Greek delegation at the Council of Florence (1438–39) was not condescending; it was defensive. The complaint used to be Latin barbarism; for the opponents of reunion at Florence, the complaint was that the Latins were oversubtle.110

Conclusion Byzantine identity was not a simple matter. Modern historians struggle for ways to describe what made Byzantine people Byzantine. Was theirs an “ethnic” identity? a religious identity? an imperial identity? Perhaps, instead of searching for a single definition that works once and for all, we need to acknowledge that our confusion is justified. A group of people defines itself and is defined as much by whom it excludes as by whom it includes. In both senses, the people we misleadingly call Byzantines did not always agree on a definition of themselves. Even where they did agree, the definition had to be refined more than once in their thousand-year history. That refinement of definition was always contested. Despite histories written by the victors, which often pretend that

108 Nikephoros Gregoras, Byzantina Historia, ed. L. Schopen, CSHB 19, vol. 1 (Bonn, 1829), 165. German trans. and comm., Rhoma¨ische Geschichte, J. L. van Dieten, Bibliothek der griechischen Literatur 4, vol. 1 (Stuttgart, 1973), 148. Hussey, Orthodox Church, 248–49. 109 I. Sˇevcˇenko, “Intellectual Repercussions of the Council of Florence,” ChHist 24 (1955): 294; repr. in idem, Ideology, Letters and Culture of the Byzantine World (London, 1982), no. , with new pagination, 6. Since Sˇevcˇenko’s seminal article, a great deal has been published regarding Greek reception of Latin theology in the Palaiologan period. A full bibliography is not possible here, but the interested reader might begin with the fol¨ bersetzung von acht Briefen des Hl. Augustinus (Vienna, 1984); F. Kianka’s lowing: H. Hunger, Prochoros Kydones, U articles on Demetrios Kydones (note 70 above); R. Flogaus, “Der heimliche Blick nach Westen: Zur Rezep¨ B 46 (1996): 275–97. tion von Augustins De trinitate durch Gregorios Palamas,” JO 110 ˇ Sevcˇenko, “Repercussions,” 298 (repr. 10). Demetrios Kydones complained that his fellow Greeks, rather than learn Latin positions and refute them intelligently, said, “‘The Latins are sophists. They attack us with sophistry, and when one refutes their sophisms, then there is nothing left but blasphemy and absurdity. We, however, stand by the folly of the evangelical message and the simplicity of fishermen. We did not receive Divine Revelation clad in worldy wisdom and we do not intend to surrender it to such wisdom, lest we strip the Cross of its Christ’” (ed. Mercati, 388; trans. Likoudis, 53).

[ 142 ] Byzantine Perceptions of Latin Religious “Errors” the challenge came from without, the crucial problems were debates within the empire, among those who considered themselves heirs of Rome and children of Christ. The iconoclast controversy, for example, despite all efforts to attribute it to Jewish or Islamic influences, seems most likely to have originated within the Christian Roman Empire. Doubts about the use of icons were certainly related to external influences, especially to the Arab invasions that threatened the very existence of the empire in the early eighth century. Still, the earliest iconoclasts were Christian bishops in Asia Minor. The history written a century later by the victorious iconodules portrays a true church, staunchly in favor of icons, oppressed by a minority of evil men who tried to force an alien doctrine and practice upon the orthodox, but modern historians doubt such an account, pointing to evidence that the iconoclasts were popular and that their beliefs were sincerely held and theologically justified. If iconoclasm had been, as its opponents claimed, an alien and obviously heinous belief, it would not have caused more than a century of strife within the empire. Nor would icon veneration have been enshrined at the center of the Orthodox definition of themselves if it had been uncontested.111 Western European Christians presented another sort of question of identity, for they had been citizens of the Roman Empire as recently as the time of Justinian I (527–565) and they were Christians. Nominally, then, they were included in the Byzantine definition of “us.” Minimal contact with Westerners in the eighth to tenth centuries enabled this status to stand. It was recognized that Westerners were not quite up to Constantinopolitan standards. They were rustic cousins, baptized barbarians—but, then, so were people in other provinces of the empire. Contact with mercenary soldiers from Scandinavia simply reinforced this comfortable semi-inclusion and superiority. But then these rustic cousins began to penetrate the empire in other ways—not just as mercenary soldiers overawed by the empire’s wealth and as occasional papal or imperial legates, who had contact with few Byzantines. The Crusades and the ambitious ventures of Italian merchants revealed to the Byzantines that their country cousins were strong and self-confident (“arrogant,” the Byzantines tended to say). Sometimes they acted like enemies. Even when they did not, they had ideas above their station: they actually claimed to have their own emperor of the Romans! This awareness of Latin difference brought Latins into the center of the ongoing debates about the boundaries of Byzantine society. It did so at a time when other groups were presenting similar challenges, especially the Armenians and Syrians who reentered the empire with the conquests of the tenth century. If all Byzantines had agreed that Westerners were excluded from the ecumenical empire and church, there would have been relatively little Greek literature about Latin beliefs and customs. Probably for the average Constantinopolitan going about his daily business, it was taken for granted that Westerners were different simply because they spoke different languages and wore different clothes. For the hierarchy of the church, however, the problem was more profound. They could not merely adduce the obvious ethnic differences. The church, as Christ’s body, is not supposed to make distinctions 111

For a recent account of the iconoclast controversy that involves this sort of analysis, see J. Herrin, The Formation of Christendom (Princeton, 1987), 306–43, esp. 331–43.

Tia M. Kolbaba [ 143 ] based on ethnicity, “for you are all one in Christ Jesus.”112 Within the church, then, the status of Latins needed clarification. Were they included in the orthodox, catholic Christian church, in spite of their strange customs and odd clothing? Or were they excluded by their own beliefs? Were they, in a word, heretics? It was not an easy question to answer. As a result, the literature for and against Latin heresy began to pile up. Not surprisingly, that literature began with issues and definitions that had been used earlier on other groups, hence the centrality of azymes in the 1050s. The impact of the Crusades on the evolution of Byzantine attitudes toward Latins was indirect but important. The Crusades made Western Europeans and one’s attitude toward them crucial issues for all Byzantines, from highly educated Constantinopolitan theologians to peasants in the Morea. Eastern hostility toward the Latins, including their status as the favorite target of religious polemic, is a result of the Crusades, a “radicalization” caused by the behavior of their Latin “brothers.” For most people, the Latins became a threat to body before they were a threat to soul; religious aversion followed violent conflict. It is a paradoxical conclusion, but it seems to fit the facts: anti-Latin polemic in the period of the Crusades is intimately linked to the Crusades, and yet it hardly ever mentions them. Princeton University

112

Gal. 3:28.

Funduq, Fondaco, and Kha¯n in the Wake of Christian Commerce and Crusade Olivia Remie Constable The arrival of the Crusaders in the Near East brought warfare to the region, but their coming also encouraged an increased presence of Western merchants in Egypt, Syria, and the Crusader states. To gain access to commercial goods and local markets, these Western traders needed institutional and political support for their activities. In many respects, their needs paralleled those of contemporary Muslim and Jewish traders operating in the same commercial sphere. On the other hand, the presence of Western Crusaders and merchants also put new pressures and demands on the structure of trade in the Near East during the twelfth and thirteenth centuries. One of the most critical areas of demand was the need for lodging and security. A chronological analysis of the terminology and function of certain types of merchant hostelry reveals shifts in the commercial infrastructure that shaped long-distance trade in the Near East during and after the era of the Crusades. Christian, Jewish, and Muslim merchants required short- and long-term lodging for themselves and their animals as well as secure storage for their commercial goods. In the medieval Islamic world, these needs could be accommodated in a number of ways. Among the most common types of merchant hostelry throughout much of the medieval period were the funduq (a word adopted from Greek) and the kha¯n (adopted from Persian). These terms—funduq (pl. fana¯diq) and kha¯n (pl. kha¯na¯t)—had been introduced into Arabic during the first centuries after the Muslim conquest, and both were well attested by the tenth century. Although sometimes used synonymously by medieval travelers as well as modern scholars, the two terms were not identical, either in origin or purpose. This distinction is made clear by references in which both institutions are mentioned together. The tenth-century Muslim traveler Ibn H . awqal, for instance, noted both fana¯diq and kha¯na¯t in Nishapur and elsewhere listed “fana¯diq, houses, baths, and kha¯na¯t” as different elements in a charitable foundation (waqf ).1 Architecturally, buildings for the two institutions were often indistinguishable, a fact that has led scholars to emphasize their similarities in form and function rather than their differences.2 Ibn H . awqal, Kita¯b .su¯rat al-’ard, ed. M. J. de Goeje and J. H. Kramers, Bibliotheca geographorum arabicorum, 3d ed. (Leiden, 1967), 2:418, 184. 2 See, e.g., the remarks of R. Hillenbrand, Islamic Architecture: Form, Function, and Meaning (New York, 1994), 331–32, 340. 1

[ 146 ] Funduq, Fondaco, and Kha¯n Indeed, observation of the differences, overlaps, and shifts in the medieval understanding of these two institutions sheds light on the changing climate of trade and travel in the Near East during and after the Crusades. For much of the earlier medieval period, until the twelfth century, there seems to have been considerable functional overlap between the funduq and kha¯n. Both served as hostelries providing either free or paid lodging for merchants, pilgrims, and other travelers. There was a degree of geographical distinction between them, however, since the funduq was more common in lands bordering the Mediterranean (especially in regions that had been subject to Byzantine influence or control), while the kha¯n predominated in the central and eastern Islamic world. During the twelfth and thirteenth centuries, the relationship between the funduq and the kha¯n began to change. The catalyst for this shift may lie with the arrival of European merchants and Crusaders in Islamic cities along the Mediterranean, together with the subsequent introduction of another cognate hostelry, the fondaco. The presence and commercial needs of these Westerners led to a new understanding of the funduq, fondaco, and kha¯n in the minds of both Christians and Muslims.3 The evolution and meaning of each term rested on the shifting roles of these three institutions. The Arabic word funduq derives from the Greek pandocheion, a word used for an inn, a tavern, or even a brothel.4 In the Gospel of Luke (10:35), for example, the good Samaritan brought the man whom he had rescued to a pandocheion. The parable makes clear that this pandocheion, as in other early Greek citations, was a “for profit” hostelry, since the Samaritan left money with the innkeeper to pay for the care of the invalid guest. Texts and archaeological data suggest that such hostelries were quite common along roads in the late Roman world. Remains of a somewhat later pandocheion lie along the route between Batnae and Edessa, though no record survives of its fees. Greek and Latin inscriptions by its door note that in the late third century “Aurelius Dassius . . . prefect and governor of Osrhoene . . . made in this place a pandocheion . . . so that travellers may enjoy refreshment and repose.” 5 The term pandocheion continued to be used later in the Byzantine world, though it appears less frequently than the word xenodocheion (or xenon), which referred to a charitable hostelry for lodging strangers and the poor. Some of the xenodocheion’s association with good works seems to have been incorporated into the understanding of the Byzantine pandocheion, since, unlike those from antiquity, later pandocheia sometimes provided charitable hospitality. Others, however, continued to function as ordinary inns. Many Byzantine pandocheia were privately owned, though some were built in conjunction with churches or monasteries to lodge needy travelers. John Chrysostom, writing in the late 3 Medieval Jewish merchants and travelers also made use of fana¯diq, as is shown by documents in the Cairo Geniza. See S. D. Goitein, A Mediterranean Society (Berkeley–Los Angeles, 1967), 1:349–50 ff. See also note 11, and O. R. Constable, “The Funduq and Fondaco in the Context of Cross-Cultural Relations in the Medieval Mediterranean World,” proceedings of the conference “Muslim Arab Civilization: The Non-Muslim Dimension,” Royal Institute for Interfaith Studies, Amman, Jordan, 16–21 August 1997, forthcoming. 4 See, e.g., Strabo, Geography, 5.3.9 and 12.8.17, Loeb (Cambridge, Mass., 1988), 410; Polybius, The Histories, 2.15.5–6, Loeb (Cambridge, Mass., 1922), 276; Epictetus, Enchiridion, ed. H. Schenkl (Leipzig, 1916), 11:11. 5 C. Mango, “A Late Roman Inn in Eastern Turkey,” Oxford Journal of Archeology 5 (1986): 223–31.

Olivia Remie Constable [ 147 ] fourth century, reported that pandocheia were found everywhere along the roads so that travelers and their animals could stop and rest. He likewise urged that a person should allow his house to be “Christ’s pandocheion,” where guests might stay free of charge.6 These admonitions apparently carried weight, since according to the later hagiographer Symeon Metaphrastes (fl. ca. 960), as soon as Chrysostom had spoken, a man named Theodorichos collected “all of his wealth, except a remainder from which he and his children might live, and gave it to the pandocheion of the church.” 7 Arabic texts make clear that there were fana¯diq throughout the Islamic world by the tenth century. Presumably these hostelries also existed earlier, either as new foundations or adaptations of existing pandocheia. The Muslim historian al-T . abarı¯ (d. 923) wrote that in 900 the caliph al-Mu‘tad. id passed through a place called Funduq al-H . usayn, near Alexandretta (modern Iskenderun, Turkey), on his way to Aleppo. The exact location is unclear, but this may have been the site of an earlier pandocheion, though the second half of its name suggests a Muslim foundation.8 Tenth-century geographers, including . awqal and al-Muqaddası¯, likewise mentioned fana¯diq in Muslim cities from Spain Ibn H to Khurasan. Extant documents from waqf foundations indicate that the Muslim funduq preserved and even augmented the charitable intent of its Byzantine predecessor. A very early waqf inscription from Ramlah, dated 913, recorded a “funduq with all its boundaries and rights, its land and building, its lower and upper floor . . . [as] a pious foundation.” 9 Later . awqal reported that pious foundations devoted to good works had in the century, Ibn H once been established in all regions of the Islamic world, including Syria, Iran, Yemen, Egypt, and the Maghrib, with their activities supported by fana¯diq, kha¯na¯t, and other . awqal, these foundations had all disappeared sources of revenue.10 According to Ibn H by his day, but later records contradict this assertion and show a link between the medieval Muslim funduq and good works. Some fana¯diq provided free lodging for needy travelers, while others were ordinary “for profit” hostelries dedicated to worthy causes. In Egypt, fragments from the Cairo Geniza from the twelfth century and later mention several fana¯diq owned by the Jewish community. These buildings were maintained by the community and used to house Jewish refugees or to raise revenues for the community’s coffers (qo¯desh) through rent.11 The Mamluk historian al-Maqrı¯zı¯ (d. 1442) likewise recorded that in the late twelfth century a servant of Saladin named Masru¯r “established a waqf [in Cairo] for the benefit of prisoners of war and the poor. . . . [This large funduq] had ninety-nine rooms and a mosque for Friday services. . . . Masru¯r [also] converted his own house into a school 6

PG 56:111.50–53; PG 60:320.22. PG 114:1129. I am grateful to Stamatina McGrath for her help in finding other references to Byzantine hostelries in the Dumbarton Oaks Hagiography Database. 8 Al-T . abarı¯, Ta’rı¯kh al-rusul wa al-mulu¯k (Beirut, 1995), 5:635. Hillenbrand (Islamic Architecture, 334) notes several buildings in Jordan and Syria that may have been earlier pandocheia. 9 M. Sharon, “A Waqf Inscription from Ramlah,” Arabica 13 (1966): 78–79. 10 Ibn H . awqal, Kita¯b .su¯rat al-’ard, 184. 11 Many of these texts have been collected and edited by M. Gil, Documents of the Jewish Pious Foundations from the Cairo Geniza, (Leiden, 1976), docs. 65, 67–69, 72, 80–81, 83–87, 89, 91, 98–99, 101–3, 106, 131–32, 134, 138, 142. 7

[ 148 ] Funduq, Fondaco, and Kha¯n [madrasa], and the revenues of a small funduq went to support this school.”12 In the 1180s, the Andalusian geographer and traveler Ibn Jubair observed a similar situation in Syria, where a vizier to the ruler of Mosul had “built fana¯diq in the cities between Iraq and Syria in order to lodge poor travelers who could not pay the fees.”13 Architecturally, fana¯diq were usually built around a square central courtyard, sometimes with one or two stories. Goods and animals were accommodated on the ground floor, while merchants and other inhabitants lodged in the rooms around the courtyard or on the upper floors. Some buildings were fortified, especially in rural areas.14 During the twelfth century, changing conditions of trade, travel, and warfare in the Near East began to alter the institution of the funduq. While Muslim and Jewish travelers continued to patronize fana¯diq, new cognate forms of the institution evolved to meet the needs of recently arrived European Christian merchants in Egypt, Syria, and Palestine during the era of the Crusades. By the middle of the twelfth century, three versions of the institution flourished in the southeastern Mediterranean: first, an older JudaeoMuslim version (the funduq); second, a new Christian version in Muslim-ruled Egypt and Syria (called fondaco in Italian);15 and third, another new Christian version (also fondaco) in the Crusader states.16 The first version, the funduq, primarily served as an urban hostelry for Muslim and Jewish traders and as a storage facility. The funduq was often still linked to charitable intent and could be founded as a good end in itself, providing free or inexpensive lodging for merchants and other travelers. More frequently, however, it charged fees for lodging and storage, and these revenues were dedicated to worthy causes, such as mosques, synagogues, schools, or poor relief. The second version was the fondaco as it took root in Muslim cities, especially Alexandria, where it became a specialized enclave for the lodging and business of foreign Christian merchants. Although closely related to the funduq (and still called by that name in Arabic), the fondaco was quite different in its administration and purpose. In general terms, a fondaco housed foreign travelers and provided a safe place for the storage of their goods. Venetian or Genoese merchants doing business in Muslim Alexandria, for example, were expected to reside in the Venetian or Genoese fondaco. From their point of view, the institution provided a familiar haven in a foreign city, where they could feel safe among their own countrymen. From the perspective of Muslim urban authorities, 12 Al-Maqrı¯zı¯, Khit.a¯.t (Bulaq, 1853), 2:92. See also al-Maqrı¯zı¯, Les marche´s du Caire: Traduction annote´e du texte de Maqrı¯zı¯, ed. A. Raymond and G. Wiet in Textes arabes et e´tudes islamiques 14 (1979): 133–35. 13 Ibn Jubair, Rih. la, ed. W. Wright and M. J. de Goeje (London, 1907), 126. 14 There are a number of architectural studies of the funduq and other related hostelries, and it is clear that building types could vary by region, period, and urban or rural setting. See Hillenbrand, “The Caravansarai,” Islamic Architecture, 331–76, and bibliography, 606; also N. Elisse´eff, “Kha¯n,” in EI 2 (Leiden, 1990), 4:1010–17; J. Sauvaget, “Caravanse´rails syriens du moyen-aˆge,” AI 6 (1939): 48–55, and ibid., 7 (1940): 1–19; K. Erdmann, Das anatolische Karavansaray des 13. Jahrhunderts (Berlin, 1961); and M. Kia¯ni and W. Kleiss, Ka¯rva¯nsara¯ha¯-ye Ira¯n (Iranian Caravansarais) (Tehran, 1995). 15 Although many different cognate forms appear in the documents ( fundico, fontico, etc.), this study uses the Italian version fondaco throughout. 16 This does not take into account contemporary versions of the pandocheion and cognates in Byzantium.

Olivia Remie Constable [ 149 ] the institution provided a place in which strangers to the city could be contained and monitored, and where their goods could be assessed for taxation. Local Muslim governments often granted special privileges to the Western fondaci, whose inhabitants were usually allowed to live according to their own religion, law, and customs. By the early thirteenth century, each of the important Western trading cities had acquired the right to a fondaco in Egypt for its merchants. This version of the fondaco has received considerable scholarly attention, with questions generally focusing on the commercial privileges granted to the Venetian, Genoese, Pisan, and other national fondaci and their internal administrative structure.17 Less attention has been paid to the origins of this form of the institution, since it is assumed that Western merchants arriving in the Near East simply adapted the local institution of the funduq to suit their needs. This is a reasonable supposition, particularly given the fact that Muslim writers continued to refer to these specialized Western enclaves by the Arabic word. On the other hand, we should also consider the possibility that the idea of the fondaco as a building dedicated to lodging a particular community of foreign merchants and their goods, with certain negotiated rights and restrictions (including ovens, churches, the use of bathhouses on designated days, permission to consume wine, tax privileges, and a nocturnal curfew), arrived with the European merchants from the western Mediterranean. Certainly the conception of the fondaco as a restricted residential and commercial enclave appears more akin to earlier institutions documented in Muslim and Christian Spain than to the version of the funduq that Westerners would have encountered in Egypt. Italian merchants traded in Egypt as early as the tenth century, but the earliest surviving reference to a fondaco appears in an 1154 treaty between Pisa and the Fatimid caliph al-Z . a¯fir.18 Although this fondaco may have had earlier roots in Egypt (a possibility that is suggested in the 1154 treaty), there are better-documented prior examples from the western Mediterranean. Four years previously, in 1150, Pisa had already drawn up a similar treaty at the other end of the Mediterranean, arranging for a Pisan funduq in Muslim Valencia.19 In 1146, both the king of Castile and the count of Barcelona had promised the Genoese a fondaco, bath, oven, and garden in Almerı´a in return for their naval help in capturing that city.20 Earlier in the twelfth century, Ibn ‘Abdu¯n, a Muslim market inspector in Seville, had instructed that foreign merchants be confined within a “funduq where they will be under the care of their fellow residents until morning.”21 Nocturnal confinement would become standard practice in the fondaci in Egypt, where it was noted and resented by many later residents. 17 For an excellent recent discussion, see articles by D. Jacoby, “Les Italiens en Egypte aux XIIe et XIIIe sie`cles: Du comptoir a` la colonie?” and M. T. Mansouri, “Les communaute´s marchandes occidentales dans l’espace mamlouk (XIIIe–XVe sie`cle),” in Coloniser au moyen aˆge, ed. M. Balard and A. Ducellier (Paris, 1995), 76–111. Also, P. Racine, “Les de´buts des consulats italiens outre-mer,” in Etat et colonisation au moyen aˆge et a` la Renaissance, ed. M. Balard (Lyons, 1989), 267–76. 18 M. Amari, I diplomi arabi del R. Archivio fiorentino (Florence, 1863), 243. 19 Ibid., 240. 20 C. Imperiale, Codice diplomatico della Repubblica di Genova, 3 vols. (Rome, 1936–42), 1: docs. 167, 169. 21 Ibn ‘Abdu¯n, Risa¯la, ed. E. Le´vi-Provenc¸al, in Documents arabes ine´dits sur la vie sociale et e´conomique en occident musulman au moyen aˆge: Trois traite´s hispaniques de H . isba (Cairo, 1955), 18.

[ 150 ] Funduq, Fondaco, and Kha¯n Whether or not the European fondaci in Egypt were based on an Eastern or Western prototype, they were clearly distinct from local Muslim fana¯diq. In contrast to the older Eastern funduq, the new Christian fondaco evinced no hint of charitable purpose. The latter appears to have been a purely “for profit” venture, benefitting both the Italian communes and the Egyptian administration through commercial taxes and revenues. Both parties received security and financial advantages. The Europeans were assured a safe place where they could live among their compatriots, enjoy certain privileges, store their goods, and transact business. At the same time, the Egyptian government had the assurance that Christian merchants would remain segregated from the local population and closely supervised by urban authorities. The fondaco buildings were owned and maintained by the Muslim government, and their doors were locked from the outside at night and during times of Muslim prayer. This emphasis on regulation and segregation set the fondaco apart from the average Egyptian funduq. Earlier fana¯diq were patronized by particular groups of Muslim or Jewish merchants, who perhaps shared a geographical origin or branch of trade, but this division seems to have been based on personal preference and commercial convenience rather than on governmental requirement.22 The institution of the fondaco appears to have grown rapidly in Egypt in the second half of the twelfth century. A treaty of 1173 between Saladin and Pisa reconfirmed Pisan access to a funduq, and in 1182 Saladin wrote to Baghdad justifying his support of Italian merchants as suppliers of arms and other critical goods to Egypt.23 It is noteworthy that he felt the need to explain and justify the situation, a circumstance that further suggests that the fondaci represented a new venture in Muslim-Christian economic relations, even in the midst of military and political tensions. In fact, the warfare of the Crusades may have been a catalyst to the desire, common to both Muslims and Christians, to regularize and institutionalize their commercial affairs. The fondaco provided both parties with security and segregation in a time of ongoing hostilities.24 Venice obtained a fondaco in Alexandria by 1184, and Genoa certainly claimed one by the turn of the thirteenth century.25 Other Western cities, including Barcelona, Florence, and Marseille, later obtained fondaci in Alexandria. Several less important Western fondaci also existed in Muslim Syrian cities.26 This system, established during the early crusading period, survived through the Ottoman conquest of Egypt in the early sixteenth century. Differences between the fondaci in Egypt and those in the Crusader Levant suggest that the institutions may have developed independently. The fondaci in Crusader cities

22 It has been noted that the segregation of the Christian fondaci in Egypt was not unlike the regulation of the Byzantine mitata, but I know of no documentation for a direct link between the two. 23 Amari, Diplomi arabi, 258. Saladin’s letter is reproduced by both Ibn Wa¯.sil and al-Qalqashandi; cf. Mansouri, “Les communaute´s marchandes occidentales,” 90. 24 In this regard, the fondaco fits into a much broader spectrum of institutions promoting and protecting cross-cultural trade. These have been noted by P. Curtin, Cross-Cultural Trade in World History (Cambridge, 1984), 78. 25 R. Morozzo della Rocca and A. Lombardo, Documenti del commercio veneziano nei secoli XI–XIII (Rome, 1940), 1:341; Jacoby, “Les Italiens en Egypte,” 81. 26 There were also similar fondaci in cities in North Africa and Spain.

Olivia Remie Constable [ 151 ] appeared somewhat earlier than their Egyptian counterparts and had a different administrative structure and residential function.27 Unlike the Egyptian fondaci, which represented, in their very walls, the physical and legal boundaries of the Christian merchant communities in Muslim cities, the fondaci in Christian Acre, Tyre, and Antioch were merely buildings within larger urban quarters acquired by the Venetians, Genoese, and other foreign communities. It is striking that while Western commercial treaties with Egypt almost invariably mentioned arrangements for a fondaco, treaties with Christian rulers in Crusader cities only occasionally granted fondaci, more often simply citing houses, streets, and squares. Owing to differences in the religious and political situation, the Crusader fondaci were not regulated to the same degree as those in Egypt. They merely provided rented lodging for merchants during the sailing season, and a place to store, assess, and even trade goods. In contrast to the restricted living situation in Egypt, most longer-term foreign residents in Crusader cities owned or rented lodgings outside of their national fondaco. Although some Crusader fondaci may have taken over established hostelries, any charitable heritage seems to have disappeared with the transfer. Instead, as with Western fondaci elsewhere, the institution was devoted to mercantile convenience, residential needs, and commercial profit.28 Fondaci appeared in Crusader cities immediately in the wake of military conquest. Already in 1098, Bohemond, the prince of Antioch, granted a fondaco to the Genoese, together with a church, a well, and thirty neighboring houses. This grant was reconfirmed by Bohemond II in 1127, and again by Bohemond III in 1169.29 Later Venetian treaties with the same city in 1140, 1153, 1167, and 1183 regularly renewed their rights to a Venetian fondaco.30 The Pactum Warmundi in 1123 referred to a Venetian funda in Tyre and its revenues.31 Pisa also had fondaci in both Tyre and Acre by the late twelfth century, as documented by references to a fundacum Pisanorum (granted together with 27 The foreign colonies in crusader cities have been discussed in a number of works. See especially J. Prawer, “I Veneziani e le colonie veneziane nel regno latino di Gerusalemme,” in Venezia e il Levante fino al secolo XV, ed. A. Pertusi, Atti del I convegno internazionale di storia della civilta` veneziana (Venezia, 1–5 giugno 1968) (Florence, 1973), 1.2:625–56; D. Jacoby, “L’expansion occidentale dans le Levant: Les Ve´netiens a` Acre dans la seconde moitie´ du treizie`me sie`cle,” Journal of Medieval History 3 (1977): 225–64; and V. Slessarev, “Ecclesiae Mercatorum and the Rise of Merchant Colonies,” Business History Review 41 (1967): 177–97. 28 The combination here of fondaci with houses suggests that fondaci often came to serve nonresidential functions in the crusader cities. This may be the origin of the usage of fondaco to mean “warehouse” in late medieval Italy. 29 Imperiale, Codice diplomatico della repubblica di Genova, 1: docs. 7, 47; 2: doc. 49. It is not clear whether the first fundaco was a preexisting hostelry or an imported institution, but its link with a church raises the possibility that the latter was an established religious house that already had a hostelry annexed to it. Unlike Muslim Egypt, where it would become common to grant rights to a church in connection with a fondaco, there was no necessity for this privilege in crusader cities under Christian rule. 30 G. L. F. Tafel and G. M. Thomas, Urkunden zur a¨lteren Handels- und Staatsgeschichte der Republik Venedig mit besonderer Beziehung auf Byzanz und die Levante, 3 vols. (Vienna, 1856–57), 1:102, 134, 149, 176 (hereafter Tafel and Thomas). 31 Ibid., 1:86. Funda and fondaco were cognates; the former held more the sense of a market, while the latter signified a residence (though trade might be conducted within its walls). This distinction is discussed at length by J. Riley-Smith, “Government in Latin Syria and the Commercial Privileges of Foreign Merchants,” in Relations between East and West in the Middle Ages, ed. D. Baker (Edinburgh, 1973), 109–32.

[ 152 ] Funduq, Fondaco, and Kha¯n ovens, baths, mills, and houses) in privileges from 1187 and 1189.32 A long list of rent revenues from the Venetian fondaco in Acre in 1244 makes clear the continuing residential nature of the institution.33 The Crusader fondaci were a less long lasting phenomenon than their counterparts in Egypt, and none survived the final conquest of Acre in 1291 in their original form. Several, however, continued to have a residential and commercial function under Muslim rule, though they were later called kha¯na¯t. The Kha¯n al-Ifranj in Acre, for example, seems to have occupied the same spot as the earlier Venetian fondaco, and part of the building’s structure dates to the Crusader period. Its Arabic name also suggests a Frankish heritage.34 Likewise, the Pisan fondaco in Acre probably stood at the same location as the existing Kha¯n al-Shu¯na.35 This name change signified more than just a switch from Christian to Muslim rule. Instead, the change from fondaco to kha¯n was part of a much broader contemporary shift in the terminology of trade and travel in the Near East. By the late twelfth century, a mixture of names and types of hostelries already existed in Crusader cities. In the 1180s, for example, when Ibn Jubair arrived in Acre, he was “taken to the custom-house, which is a kha¯n prepared to accommodate the caravan. . . . The merchants deposited their baggage there and lodged in the upper story.”36 His description points to an evolving distinction between the fondaci, which housed Christian merchants from specific national communities, and the kha¯n, an all-purpose hostelry for merchants and other travelers, including Muslims. These changes and overlaps in word usage and institutional function make it difficult, even impossible, to lay out definitive rules to differentiate among the funduq, the fondaco, and the kha¯n. Nevertheless, certain trends distinguish the development of the kha¯n. The Persian institution of the kha¯n had always been present in the Islamic world, though fana¯diq were more common around the Mediterranean until the thirteenth century. From early on, kha¯na¯t, like fana¯diq, had been founded by rulers and local officials, often for the public good. Al-T . abarı¯ recorded that in 720 the caliph ‘Umar b. ‘Abd al‘Azı¯z wrote to the governor of Samarqand instructing him to “establish kha¯na¯t in your lands so that whenever a Muslim passes by, you will put him up for a day and a night and take care of his animals; if he is sick, provide him with hospitality for two days and two nights; and if he has used up all of his provisions and is unable to continue, supply him with whatever he needs to reach his hometown.”37 Other kha¯na¯t were more obviously commercial, including one in Wa¯sit described by the eleventh-century writer alHarı¯rı¯ al-Basrı¯.38 32 G. Mu¨ller, Documenti sulle relazioni delle citta` toscane coll’Oriente cristiano e coi Turchi fino all’anno MDXXXI (Florence, 1879), docs. 23, 31. 33 Tafel and Thomas, 2:393–96. 34 D. Jacoby, “Crusader Acre in the Thirteenth Century: Urban Layout and Topography,” in idem, Studies on the Crusader States and on Venetian Expansion (Northampton, 1989), 32. 35 Ibid., 24. 36 Ibn Jubair, Rih. la, 317–18. 37 Al-T . abarı¯, Ta’rı¯kh, 4:69 (D. S. Powers, trans., The History of al-Tabarıˆ [Albany, 1989], 24:94). 38 Al-Harı¯rı¯, Maqa¯ma¯t (Beirut, 1958), 228. This text provides a base for S. Guthrie’s discussion of the kha¯n in her Arab Social Life in the Middle Ages (London, 1995), 94–100.

Olivia Remie Constable [ 153 ] According to Ibn H . awqal, kha¯na¯t were connected with waqf foundations before the 39 tenth century. Nevertheless, they began to appear in large numbers of waqf grants only in the twelfth and thirteenth centuries, and they became even more numerous in later centuries. Some kha¯na¯t were the objects of endowments, as in the case of a waqf inscription dating from 1115 referring to a pious foundation created to support a kha¯n in Kazwin.40 More often, however, the revenues of a kha¯n were devoted to a mosque, school, or other worthy purpose. Al-Maqrı¯zı¯ reported that the Kha¯n Manku¯rush in Cairo had been founded by a Mamluk of Saladin who died in 1182, and its revenues were dedicated to good works.41 The same author likewise noted another late twelfth-century hostelry with a charitable purpose, a public kha¯n founded in Cairo to lodge “sons of the road and [other] travellers [who were] received without charge.”42 These foundations date from the period in which the kha¯n was beginning to replace the funduq, and to take over a more exclusive role in lodging travelers on the road. During the twelfth and thirteenth centuries, the institution of the kha¯n gradually took over the essential function of the earlier funduq as it evolved to meet new needs of Muslim travelers. This shift is demonstrated in the example of a hostelry near Damascus that bears an inscription recording that the “building of this blessed funduq [had] been ordered by . . . [Saladin], Sultan of Islam and the Muslims” in 1181.43 Despite this official designation as a funduq, the hostelry soon became known as Kha¯n al-‘Aru¯s. Only a couple of years after its dedication, Ibn Jubair described it as “the Kha¯n of the Sultan, which was built by Saladin, the lord of Syria. It is the zenith of strength and handsomeness, with iron doors after the fashion in the building of kha¯na¯t. . . . The road from Homs to Damascus is little populated, except for three or four places where there are these kha¯na¯t.”44 Beginning in the later twelfth century, a network of new kha¯na¯t were established by Ayyubid, Seljuq, Mamluk, and Ottoman rulers along the caravan routes of Egypt, Syria, Iraq, and southern Anatolia. Contemporary inscriptions on surviving buildings confirm the proliferation of new kha¯na¯t, usually at the behest of the sultan, government officials, and other wealthy individuals. Ibn Jubair also noted many recently built kha¯na¯t on his travels, such as the “large new kha¯n” he encountered in a village after leaving Mosul, further adding that “in all the stages of the road there are kha¯na¯t.”45 He passed another night in a “large new kha¯n” near Harran, and a few days later, leaving Qinnasrin, he “halted to rest . . . in a large kha¯n, strongly fortified, called the Kha¯n of the Turkomans. The kha¯na¯t on this road are like fortresses. . . . Their doors are of iron, and they present the utmost strength.”46 Data on kha¯na¯t in the twelfth and thirteenth centuries show not only the rapid growth of the institution at this time, but also the ways in which late medieval kha¯na¯t had begun . awqal, Kita¯b .su¯rat al-’ard, 184. Ibn H Re´pertoire chronologique d’e´pigraphie arabe, ed. E. Combe, J. Sauvaget, and G. Wiet, Publications de l’Institut franc¸ais d’arche´ologie orientale du Caire, 18 vols. (Cairo, 1937), 8: no. 2967, 108–12. 41 Al-Maqrı¯zı¯, Khit.a¯.t, 2:93. 42 Ibid. For a list of waqf endowments mentioning kha¯na¯t in Mamluk cities, see I. Lapidus, Muslim Cities in the Later Middle Ages (Cambridge, Mass., 1967), app. A, 195–98. 43 Re´pertoire, 9: no. 3368, 115. 44 Ibn Jubair, Rih. la, 269; and Sauvaget, “Caravanse´rails syriens du moyen-aˆge,” AI 6 (1939): 50–52. 45 Ibn Jubair, Rih. la, 247. 46 Ibid., 257, 264. 39 40

[ 154 ] Funduq, Fondaco, and Kha¯n to differ from contemporary fana¯diq and fondaci. Descriptions by Ibn Jubair and other travelers indicate that kha¯na¯t were strongly built, even fortified, but many maintained only a minimal staff or none at all. This lack of close supervision contrasts with the strict regulation of the fondaci for Christian merchants in cities such as Alexandria and Damascus. Likewise, it is clear that numerous kha¯na¯t were founded systematically along welltraveled inland rural routes, especially in Syria, Iraq, Iran, and Anatolia, for the benefit of Muslim merchants, state employees, post riders, and other travelers. There were also kha¯na¯t in many late medieval Muslim cities, including Damascus, Aleppo, Homs, Mosul, Bursa, and Isfahan. Unlike the urban fana¯diq and fondaci, however, which were usually located near the heart of a city’s commercial district, kha¯na¯t were more often found on the outskirts of the city center. Meanwhile, Muslim fana¯diq were losing their primary character as hostelries for travelers. Increasingly devoted to urban commerce and manufacture, they often contained shops, warehouses, and workshops as well as living space for long-term residents. By the fifteenth century, al-Maqrı¯zı¯ described how the kha¯na¯t of Cairo were filled with travelers passing through, while the fana¯diq were crammed with residents.47 At the same time, the European fondaci in Egypt continued to function as strictly regulated enclaves for Christian merchants and other Western travelers. The appearance of the fondaco in the twelfth and thirteenth centuries, in the wake of the Crusades, was in part responsible for this shift, as the new demands of Christian traffic influenced the patterns of Muslim trade and travel. The rapid evolution of the fondaco, and its success as an enclave for Christian traders, may have blurred the identity of the funduq and encouraged the growth of other types of lodging for Muslims. Unlike the fana¯diq and other hostelries, most late medieval kha¯na¯t were not individual establishments but part of larger regional networks. The building of kha¯na¯t coincided with a revival of Muslim overland travel, commerce, and communications beginning in the twelfth century, as well as a new institutional emphasis. The revival of the barı¯d, or postal system, by Sultan Baybars after 1260 was part of the same phenomenon, and later Mamluk and Seljuq rulers continued to augment the infrastructure for overland communications through the foundation of kha¯na¯t and other facilities. The kha¯n became the dominant institution for lodging travelers throughout Syria and the Eastern Islamic world in the wake of the Crusades. In Egypt, fana¯diq were converted into kha¯na¯t, yet fana¯diq still remained common in Cairo. Because of this, the terms are often more confused with regard to Egypt than Syria, although the two institutions were generally distinct. In the late fourteenth century, Ibn Duqma¯q (d. 1406) listed fortythree fana¯diq and one kha¯n in his description of Cairo; in the first half of the next century, al-Maqrı¯zı¯ noted nineteen fana¯diq and eleven kha¯na¯t in the same city.48 One of these was Kha¯n al-Masru¯r, which had been founded as Funduq al-Masru¯r in the late twelfth century.49 Although these tallies indicate the continued fluidity of the terms and the transfer

Al-Maqrı¯zı¯, Les marche´s, 2. Ibid., 23–24. 49 Al-Maqrı¯zı¯, Khit.a¯.t, 2:92. 47 48

Olivia Remie Constable [ 155 ] of certain buildings from one designation to another, both authors considered the institutions sufficiently distinct to count them separately.50 Contemporary travelers’ reports likewise support the impression of hostelries with overlapping but not identical functions. For example, the fourteenth-century Maghribi traveler Ibn Bat..tu¯.ta drew attention to the distinction between urban and rural hostelries in Egypt. He described a roadside hostelry near Cairo as a “funduq, which they call a kha¯n, where travelers alight with their beasts.” His comment makes clear that this nonurban facility for passing travelers went by the title kha¯n in Egypt, although it was similar to what he would have recognized as a funduq in the Western Islamic world.51 Christian travelers in Egypt also commented on the terms fondaco and kha¯n, which were made even more confusing by the translation of Arabic words into Western languages. Although Europeans called the fondaci in Egypt by that name, the term, like funduq, was often blurred with kha¯n in Arabic. A linguistically muddled description of the fondaci in Alexandria by the traveler Lionardo di Niccolo` Frescobaldi in 1384 testifies to this confusion: “All the Christian Franks are locked in a building called a cane [kha¯n], and the keeper of the cane locks them in, and this name comes from the fact that we are [considered to be] cani [dogs]” by the Muslims.52 A century later, however, the Flemish traveler Joos van Ghistele drew attention to clear distinctions between the fondaci and kha¯na¯t, noting architectural and administrative differences, including nocturnal curfews imposed on the Christian inhabitants of the fondaci.53 By the fourteenth century, the funduq had become rare in Syria and Iraq, where its decline may have been connected with the rise of the kha¯n and the increasing focus of Syrian traffic on long-distance interurban overland routes traveled by Muslims. Meanwhile, the funduq continued to flourish in the Western Islamic world and to a lesser degree in Egypt, though in both regions it had become more of an urban entrepoˆt, warehouse, or atelier than a hostelry for travelers. Many of the fana¯diq in Mamluk Cairo mentioned by Ibn Duqma¯q and al-Maqrı¯zı¯ were associated with particular trades or types of commercial goods. For example, Funduq Da¯r al-Tuffa¯h. , near Ba¯b al-Zuwaila, handled fruit arriving from the Egyptian countryside, and the Funduq of T . urunt.a¯y 54 served as a depot for olive oil coming from Syria. To further complicate the situation, another similar institution, the waka¯la, also began to develop in this period. Al-Maqrı¯zı¯ thought it necessary to explain that a waka¯la was similar to a funduq or kha¯n, noting that merchants from Syria arrived at the Waka¯la Qaws.u¯n with olive and sesame oil, soap, walnuts, and other goods.55 Some late medieval Egyptian waqf endowments continued 50 N. D. MacKenzie discusses the overlap of the two terms in his Ayyubid Cairo: A Topographical Study (Cairo, 1992), 165–77. 51 Ibn Bat..tu¯.ta, The Travels of Ibn Bat..tu¯.ta, .. 1325–1354, ed. C. Defre´mery and B. R. Sanguinetti, trans. H. A. R. Gibb (Cambridge, 1958), 1:110, 71–72 (Arabic text, 111–12). 52 L. Frescobaldi, Visit to the Holy Places of Egypt, Sinai, Palestine, and Syria in 1384, trans. T. Bellorini and E. Hoade, Publications of the Studium Biblicum Franciscanum 6 ( Jerusalem, 1948), 42. 53 Joos van Ghistele, Le Voyage en Egypte de Joos van Ghistele, 1482–1483, trans. R. Bauwens-Pre´aux (Cairo, 1976), 113–14. 54 Al-Maqrı¯zı¯, Khit.a¯.t, 2:93–94. 55 Ibid., 93. On the waka¯la, see also A. Raymond, The Great Arab Cities in the 16th–18th Centuries (New York, 1984), 44–54; and N. Hanna, An Urban History of Bu¯la¯q in the Mamluk and Ottoman Periods (Cairo, 1983).

[ 156 ] Funduq, Fondaco, and Kha¯n to mention fana¯diq, but their sponsorship only remained common in the Maghrib. Elsewhere, new endowments rarely noted fana¯diq, although earlier foundations may have continued to contribute funds to good causes. After the fall of Acre in 1291, the fondaco, like the funduq, became a phenomenon limited almost exclusively to Egyptian and Maghribi port cities. Following the Crusades, Muslim rulers attempted to limit Western merchant activity and funnel their business toward certain markets. Thus the fondaci survived in Mamluk Egypt because the structure of Egyptian trade concentrated traffic through Alexandria and Cairo and because the Mamluk administration closely monitored and taxed this traffic. By the fourteenth century, Western merchants no longer enjoyed a free choice of ports or style of residence as they had in Crusader cities. Instead, the concentration of fondaci in Alexandria channeled most Christian traffic in the eastern Mediterranean through that port, where Western traders could be closely supervised from Cairo and where profits could be more efficiently directed to the sultan’s treasury. Although fondaci survived in a few Syrian cities, most notably Damascus, they never ranked with their Egyptian counterparts in terms of volume of trade. The highly regulated nature of the institution also increased over time, strictly controlled by the edicts of custom and diplomacy. By the fifteenth century, many of the European consuls in charge of the fondaci actually received a stipend from the Egyptian government for their services.56 The continued presence of European traders and the institutionalization of the fondaco also stimulated the evolution of the kha¯n in a new and distinctly Muslim form. The development of a system of kha¯na¯t in Syria, Iraq, Anatolia, and elsewhere demonstrates the pattern of center and periphery in Near Eastern trade in the later Middle Ages. Within their lands, the Mamluk government sought to channel commerce through Egyptian markets, routing communications and revenues toward Cairo, its political and economic center. The long-distance caravan routes through Syria that connected with routes to Egypt became critical to this mission, as did the kha¯na¯t established at regular intervals along these routes to facilitate travel by government officials and private merchants. This system of inland routes was differentiated by language, religion, and function from the maritime trading system of the eastern Mediterranean, which was now dominated by Western Christian merchants. By means of the strictly regulated urban fondaci, these Christian traders had access to certain Muslim markets but not, generally, to the Muslim inland network that gained renewed importance during the thirteenth and fourteenth centuries. This vital network of Muslim communication and trade depended on the caravan routes and kha¯na¯t linking Syria and Egypt, the Red Sea, the Hijaz, the Persian Gulf, and points east. Differences in the terminology and function of institutions for trade and travel in the Near East—including the shifting usage of funduq, fondaco, and kha¯n—reflected these new Christian and Muslim commercial interests and routes in the wake of the Crusades. University of Notre Dame 56

Amari, Diplomi arabi, 339.

Byzantine Trade with Christians and Muslims and the Crusades Angeliki E. Laiou, with an Appendix by Ce´cile Morrisson In May 1192, at a time when the Third Crusade was still in progress, the Byzantine emperor Isaac II sent one of his virtually annual embassies to Saladin, seeking, among other things, an offensive and defensive alliance directed, inevitably, against Western Europeans. In late summer or early autumn of the same year, a Venetian ship carrying the Byzantine ambassadors, Saladin’s envoys and gifts to Isaac, and goods and merchandise belonging to Isaac, his brother and future emperor Alexios, an imperial official, and “Greek and Syrian merchants,” set sail from Egypt toward Constantinople. Near Rhodes, it was attacked by Pisan and Genoese ships led by the Genoese corsair Guglielmo Grasso. The goods were seized, and the ambassadors and the merchants were killed, or so Isaac said in his letter of complaint to Genoa.1 Nor were the sums involved negligible. Isaac claimed that the merchandise was valued at 96,000 hyperpyra and 566 nomismata,2 of which 39,000 hyperpyra and 193 nomismata belonged to merchants of Constantinople. This affair initiated a series of diplomatic and not-so-diplomatic negotiations that lasted until September 1195. Almost immediately after the attack, in November 1192, Isaac wrote to the Commune of Genoa describing what had happened and seeking satisfaction (iJka´ nwsi") for the property lost. Otherwise, he said, the Genoese merchants in I should like to thank my research assistant, Charles Dibble, for his help. 1 The main information about this incident is contained in G. Bertolotto, Nuova serie di documenti sulle relazioni di Genova coll’Impero bizantino, Atti della Societa` Ligure di storia patria 28.2 (Genoa, 1898), doc. , pp. 448–53 (⫽ F. Miklosich and J. Mu¨ller, Acta et diplomata graeca medii aevi sacra et profana, 6 vols. [Vienna, 1860– 90], 3:37–40), and doc. , pp. 454–64, and J. Mu¨ller, Documenti sulle relazioni delle citta` Toscane coll’oriente cristiano e coi Turchi fino all’anno , Documenti degli archivi toscani (Florence, 1879; repr. Rome, 1966), nos.  (pp. 61–64) and  (pp. 66–67). Brief discussions of the affair in C. M. Brand, “The Byzantines and Saladin,” Speculum 37 (1962): 173–78, and W. Heyd, Histoire du commerce du Levant au Moyen Age, 2d ed., 2 vols. (Leipzig, 1936; repr. Amsterdam, 1967), 1:233–35. On piracy, cf. H. Ahrweiler, Byzance et la mer (Paris, 1966), passim, esp. 288–92, and M. Balard, La Romanie ge´noise, vol. 1 (Rome, 1978), 35. 2 cilia´ da" uJperpu´ rwn ejnnenh´ konta e’x, nomi´smata o”moia pentako´ sia eJxh´ konta e“x. The two different terms may refer to two different coins: the gold hyperpyron and the electrum “nomisma.” The small number of the electrum coins would reflect their limited usefulness for international trade. I owe this observation to C. Morrisson. But see also M. F. Hendy, Coinage and Money in the Byzantine Empire, 1081–1204, DOS 12 (Washington, D.C., 1969), 35–37.

[ 158 ] Byzantine Trade Constantinople must sell their goods and deposit the money as a guarantee of the eventual restitution to be made by the Commune.3 Similar letters, with a similar content, must have been sent to Pisa: although the early phases of negotiations with Pisa elude us, the surviving documentation makes reference to previous embassies. Very soon thereafter, indeed in the same month of November 1192,4 Isaac did seize the property of Genoese and, undoubtedly, Pisan merchants. The name of a Genoese merchant, Enrico Novitella, who had sailed into Constantinople in November, is specifically mentioned. In a sigillion given to the city of Genoa in October 1193, Isaac explained why he had taken this extraordinary measure. Once again, he recalled Grasso’s attack, which had taken place, he reminded the Genoese, shortly after the conclusion of a treaty with Genoa.5 He was most emphatic on the damages suffered by his merchants, many of whom, he said, were from Constantinople itself, and among the most important merchants in the City (kai` ta` prw'ta tw'n ejn aujth' pragmateutw'n ferome´ nou"). He, Isaac, could not take this lightly, especially since the merchants (or, one assumes, the heirs or creditors of those who were killed) vociferously demanded justice, revenge, and reparation. They had asked to be allowed to get satisfaction from the property of Genoese merchants in Constantinople; the Genoese in question had not responded to the emperor’s request that they make reparations, which upset their Byzantine counterparts even more. Isaac held the city of Genoa responsible, even though it had tried to claim that the corsairs were outlaws—expelled from the city—and that therefore the Commune bore no responsibility for their actions. In the second instance, Isaac held responsible the Genoese merchants active in Constantinople. Under heavy pressure from the people,6 and fearing a riot,7 but not wanting to take extreme measures such as rescinding the privileges of the Genoese, he had seized a portion of the money and goods of some Genoese merchants and given them to Byzantine “guarantors” (ejgguhta´ ") to hold in deposit (parakataqh´ kh). The sum so deposited was 20,000 hyperpyra, just over half what the Byzantine merchants claimed to have lost; it is legitimate to suppose that the goods of Pisan merchants, which we also know to have been seized,8 made up the other half. This property was meant to be returned if the Genoese (and Pisan) Commune made reparations. If not, it would be given outright to the Byzantine merchants in reparation (iJka´ nwsi"). Between November 1192 and October 1193, the Genoese (and the Pisans, one assumes) had sent envoys, promising to pursue the corsairs and deliver them into Isaac’s hands, and asking that the Commune not be made to suffer for the actions of individuals. Isaac chose to believe that the Commune would make restitution, or else that the prom-

Bertolotto, Nuova serie, doc. . For the date, see Bertolotto, Nuova serie, p. 457. 5 For the treaty (April 1192), see F. Do¨lger, Regesten der Kaiserurkunden des ostro¨mischen Reiches, vol. 1.2 (Munich, 1925), no. 1610; Miklosich and Mu¨ller, 3:25–37; Bertolotto, Nuova serie, pp. 413–33. 6 The text uses the word demos; must one assume general popular discontent, or the discontent of merchants? 7 On riots on the part of “the men of the marketplace” during the reign of Alexios III, cf. Niketas Choniates, Historia, ed. J. L. van Dieten (Berlin-New York, 1975), 523–26. 8 Mu¨ller, Documenti, no. , p. 67. 3 4

Angeliki E. Laiou [ 159 ] ise itself was sufficient. In any event, although the Genoese did not send the money seized by the corsairs,9 the Byzantines returned the 20,000 hyperpyra held in deposit. The man who returned it, and who perhaps had held it in deposit, was John Oxeobaphopoulos.10 The transaction was effected with all the Byzantine legal forms: the Genoese, for example, gave formal assurance that they had, indeed, received the money (ajnarguri´an probale´ sqai oujk e“comen). The Genoese Commune agreed not to raise any further claims for damages suffered by its merchants whose goods had been confiscated, not to refer to the matter again, nor to seek revenge. And in return, the emperor issued his sigillion, closing the matter, and renewing Genoese privileges. That the accused neatly turned into accusers is only one of the interesting aspects of the story. Clearly, we are at a crossroads in the development of the law of reprisals, and the principle prevailed that reprisals should not be sought of innocent parties; but in the process, the Byzantine merchants did not receive restitution. It is, I think, unlikely that in closing the matter in this way Isaac II was swayed by arguments regarding the respective responsibilities of individuals and collectivities. For the dispute with Pisa dragged on until 1195. It was, eventually, resolved in a way parallel to the settlement with Genoa, but only because Pisan ships kept harassing the environs of Constantinople, attacking Byzantine ships, seizing goods, and killing people.11 One wonders whether in the Genoese case as well there was not an element of military or political persuasion. The origins of the events of 1192–95 were embedded in political affairs: the last effort on the part of Isaac II to ally himself with Saladin against the Crusaders. It has been argued that the corsair attack was also in some degree political, the corsairs being in part bent on creating trouble for Saladin, the great enemy of the Crusader states.12 The conclusion of the affair was also, it would seem, political, for the capture and murder of the ambassadors ended the close relationship between Isaac and Saladin. That having proved unproductive or even counterproductive, Isaac now made a full turn toward Genoa, Pisa, the pope, and the Normans13 —which may serve as another explanation of the ease with which Pisan and Genoese goods were returned. Despite the political aspects, however, the role of the merchants is central to the story, and the importance of Constantinopolitan merchants is especially noteworthy. They appear as a large, influential, rich, and dangerous group, and it is no accident that Isaac began by demanding restitution for everyone’s goods (including his own, his brother’s, and his official’s), then very quickly limited his demands to reparations for the goods of the merchants, even though in the end he got nothing. Modern scholars have given this story scant attention. Yet it cannot be equaled as a

9 Heyd, Commerce, 1:234, says that restitution of the money was, in fact, made, but the documentation nowhere suggests that; he may have misunderstood the text in Bertolotto, Nuova serie, p. 457. Brand, “The Byzantines and Saladin,” 178, follows Heyd. 10 Cf. below, 177. 11 Mu¨ller, Documenti, no. . 12 Heyd, Commerce, 1:233. Heyd presents this as a hypothesis, suggesting that Isaac’s alliance with Saladin and the presence of Egyptian ambassadors aboard this ship made it doubly interesting to the Genoese pirates. 13 Brand, “The Byzantines and Saladin,” 178: “Saladin was disillusioned with Isaac’s military capabilities, while Isaac finally realized that Saladin was too distant to protect him from the Latins.”

[ 160 ] Byzantine Trade snapshot of conditions in the eastern Mediterranean in the late twelfth century, for it illustrates a number of important developments. For one thing, it makes evident the close connections between the Crusades and the increasingly strong presence of Italian merchants in the eastern Mediterranean. It brings to the fore the Byzantine bankers and merchants, here engaged in international money transactions. It allows us more than a glance into the developing law of the sea regarding issues of paramount importance to merchants, here the question of reparations and reprisals. And it hints at the nature of political and commercial connections between the Byzantine Empire and the Muslims during the time of the Crusades. These topics will be discussed in what follows. For the Byzantine Empire, the question of the economic influence—if any—of the Crusades is almost inextricably connected with the question of the influence of the Italian merchants on the Byzantine economy, which both predates the crusading movement and becomes closely tied to it, certainly by the time of the Fourth Crusade. Indeed, the presence of Italian merchants on Byzantine soil eventually became dominant, reducing the degrees of freedom of the native merchants, although possibly increasing their opportunities. This is a topic of significance, touching primarily the economy of exchange, and it has been treated by a number of scholars.14 I will not offer a reconsideration of the question; I simply note the double presence of Italian merchants and Crusaders in many geographic areas, constant but in small numbers in the case of the first; somewhat more sporadic but sometimes in huge numbers in the case of the Crusaders. The “Crusades” were a frequent phenomenon of the twelfth and thirteenth centuries. We are accustomed to taking account of the major crusades, the ones with numbers, but crusading expeditions of one kind or another took place often, and certainly the Christians and Muslims of the area were aware of the fact. The First Crusade, insofar as the Byzantine Empire was concerned, did not end in 1099. It extended from 1096 until 1108 and the Treaty of Devol, which marked the end of Bohemond’s quasi-crusade. And the 1120s were punctuated by crusading expeditions undertaken by Pisans and Genoese by sea, while in 1122 a Venetian Crusader fleet on its way to Palestine attacked Corfu in retaliation for the attempt of John II Komnenos to reduce Venice’s commercial privileges; it pillaged Byzantine lands on the way to and from Palestine and extracted the confirmation and expansion of Venetian commercial privileges in the Byzantine Empire.15 Thus large armies and small and large fleets often traversed Byzantine lands and waters, presenting the Byzantines with the new problem of dealing, in economic ways too, with theoretical friends in large numbers. The other general point that should be remembered is the existence of Crusader states

14 See, most recently, D. Jacoby, “Italian Privileges and Trade in Byzantium before the Fourth Crusade: A Reconsideration,” Anuario de estudios medievales 24 (1994): 349–68; A. Laiou, “Byzantium and the Commercial Revolution,” in G. Arnaldi and G. Cavallo, eds., Europa medievale e mondo bizantino: Contatti effetivi e possibilita` di studi comparati, Istituto Storico Italiano per il Medio Evo, Nuovi Studi Storici 40 (Rome, 1997), 239–53; M. F. Hendy, “Byzantium, 1081–1204: ‘The Economy Revisited,’ Twenty Years On,” in idem, The Economy, Fiscal Administration, and Coinage of Byzantium (Northampton, 1989), no. . 15 J. Riley-Smith, “The Venetian Crusade of 1122–1124,” in G. Airaldi and B. Kedar, eds., I comuni italiani nel regno crociato di Gerusalemme, Jerusalem, 24–28 May 1984, Collana Storica di Fonti e Studi 48 (Genoa, 1986), 337–50.

Angeliki E. Laiou [ 161 ] after 1099, which meant that there were three sets of Christians in the eastern Mediterranean: Byzantines, Crusader states, and Italian merchants (not to mention the native Christians of the East, who will not enter this discussion); the interconnections between them will be seen to have been of importance. That having been said, I should like to pose a few questions that are somewhat different from the questions scholars have been asking. The focus, to the extent possible, and bearing in mind the sometimes inextricable interconnections between Crusaders and merchants, will be on the effects of the Crusades themselves, and of the Crusader states, on Byzantine commercial relations. What new and specific problems and challenges did the Crusaders and the existence of the Crusader states pose to the Byzantine Empire? Did they influence the mechanisms and methods of trade? Did they help bring about any structural changes? Or were they irrelevant, and do the only questions continue to be those associated with the Italian presence and eventual dominance over the commerce of these areas? I will concentrate here not on the overall canvas (the bird’s-eye view), but rather on the worm’s-eye view to start with, and then on the point of view of the flying-fish— the middle distance, the structural and institutional developments that took place in Byzantine trade with Christians and Muslims especially during the twelfth century, but also in the later period—connected with conditions in which the Crusades played a role. I will take for granted the very large changes after 1204, insofar as Byzantine trade with Western Christians is concerned.

I. The Byzantine Economy and the Crusades This is a time when armies and navies passed through the territories of a still relatively intact and prosperous Byzantine Empire. The discussion of the First Crusade will include the expedition of 1100–1101 and end in 1108.16 There were two major relevant problems: that of provisioning the armies as they crossed Byzantine lands and the related problem of currency exchange. A. Provisioning Major challenges were posed by the large size of the crusading armies. According to the most recent estimates, the Peasant Crusade had around 20,000 participants; the main armies at Nicaea counted approximately 50,000 to 60,000 members including noncombatants. Many had already died on the way to Constantinople. The Crusade of 1101, for which there are huge contemporary estimates, must have been larger.17 No figures are 16 J. France, Victory in the East: A Military History of the First Crusade (Cambridge–New York, 1994), 142, and J. Flori, “Un proble`me de me´thodologie: La valeur des nombres chez les chroniqueurs du Moyen Age: A propos des effectifs de la premie`re Croisade,” Moyen Age 99 (1995): 399–422. J. Heers, Libe´rer Je´rusalem: La premie`re Croisade (1095–1107) (Paris, 1995), does not give estimates. On the Crusade of 1100–1101, see J. L. Cate, “The Crusade of 1101,” in K. M. Setton, ed., History of the Crusades, vol. 1, The First Hundred Years, ed. M. W. Baldwin (Madison, Wisc., 1969), 343–67. 17 Cate, “The Crusade of 1101,” 351.

[ 162 ] Byzantine Trade given for the French and German armies of the Second Crusade, but combined they must have been as large as those of the First Crusade. For the army of Frederick I Barbarossa, we have the figure of 100,000, which includes 20,000 mounted troops.18 In terms of the problems of provisioning, we might compare these figures with the most recent and persuasive estimates of the size of the Komnenian army on campaign, which is approximately 15,000 to 20,000 men, thus much smaller than the Crusaders’ armies.19 One must also take into account the length of time the Crusaders spent on Byzantine soil; thinking only of the Balkans, I have estimated an average of two to two and a half months for most of the armies of the First Crusade, except for those of Raymond of Toulouse and Bohemond, which took five and a half to six months. The passage of most of the armies, including the Peasant Crusade, was swift: the comparison to locusts, found in both Anna Komnene and Orderic Vitalis,20 seems apposite. The armies of the Second Crusade took two and a half to three months. But that of Frederick Barbarossa spent almost nine months on Byzantine soil, mostly in southern Bulgaria and Thrace, heavily taxing the resources of the area. As these numbers suggest, the problems of provisioning were ubiquitous; as for questions of exchange, those arose on the ground, and, during the Second and Third Crusades, formed a part of the negotiations for safe passage. As is well known, the First Crusade was too amorphous and disorganized for formal arrangements to have been made beforehand to assure provisioning, although by 1101 agreements were indeed concluded as the Lombard army was about to enter Bulgaria.21 Provisions do not seem to have been brought as far as the Byzantine Empire, although one of the chroniclers mentions that the Crusaders carried some provisions with them;22 these undoubtedly were exhausted before the Crusaders left Western Europe. The Crusaders also brought money and marks of silver with them, presumably to be used to buy provisions.23 The problem of provisioning was recurrent during the passage of the armies of the First Crusade and of the subsequent two. The army of Raymond of St. Gilles, passing through Dalmatia, could get neither safe-conduct nor commercium until it reached Skoutari, where, in January 1097, Raymond made a pact with the local ruler (Vodin), giving him much money so that the army would be allowed to buy provisions, although Vodin’s promises seem to have remained a dead letter.24 The Crusaders were certainly expected, by the Byzantine authorities, to buy their food: Alexios I, as soon as he heard of the Crusade, had sent generals to Durazzo and Avlona to ensure that there would be panegyreis (the Latin term is usually mercatum or necessarium negotium)25 in all the lands along the 18

France, Victory in the East, 136. J.-C. Cheynet, “Les effectifs de l’arme´e byzantine aux Xe–XIe sie`cle,” CahCM 38 (1995): 319–35, at 331–32. 20 Alexiade 10.5.7, ed. B. Leib (with P. Gautier), 4 vols. (Paris, 1939–76), 2:208; Orderic Vitalis, The Ecclesiastical History, ed. and trans. M. Chibnall, 6 vols. (Oxford, 1969–80), 5.30. 21 Albert of Aix, Historia Hierosolymitana, RHC, HOcc 4 (Paris, 1879), 559; Orderic Vitalis, 5:327. 22 Roberti Monachi Historia Iherosolimitana, RHC, HOcc 3 (Paris, 1866), 744, on Bohemond. 23 See below, 168–69. 24 J. Hugh and L. Hill, eds., Le “Liber” de Raymond d’Aguilers (Paris, 1969), 37–38; Guillaume de Tyr, Chronique, ed. R. Huygens, CC continuatio mediaevalis 63–63 (Turnhout, 1986), 2.17 (1:182 ff ). 25 E.g., Orderic Vitalis, 5:31, 33, 43, 49, 69. 19

Angeliki E. Laiou [ 163 ] way.26 Some Crusaders, especially the excitable Tancred,27 thought it too bad that they should have to buy food—and they took different courses of action.28 The emperor, perhaps in response, seems to have made the availability of provisions contingent on the Crusaders’ oath of fealty or friendship, as also happened during the Second and Third Crusades.29 The Byzantine authorities seem to have been confident that they could provide adequate markets, so much so that Niketas Choniates thought that Manuel Komnenos’ failure to supply adequate markets to the participants of the Second Crusade was deliberate. Byzantine armies on campaign normally brought some of their provisions with them, or bought them along the way, as long as they were on friendly territory.30 But the Crusader armies, especially those of the Second and Third Crusades, which crossed into the Balkans all together, were much larger than a Byzantine army on campaign. In fact, the passage of the crusading armies taxed the productive resources of the area and most particularly the mechanisms of distribution. It is surprising that the system did not significantly break down before the time of Frederick Barbarossa. The provisioning of the armies of the First Crusade (including that of 1101) on Byzantine territory was carried out in three ways. First, the Crusaders bought food, and a recurrent complaint is that there was no food to be bought or, as in Kastoria, that the inhabitants did not wish to sell, being suspicious of the Crusaders’ motives.31 In the Balkans, in Asia Minor, and also in the Holy Land, much of the provisioning of the crusading army depended on the purchase of food (and fodder for the horses), as William of Tyre makes most clearly evident. Indeed, the fact that various crusading chronicles quote prices, mostly at times of famine, proves that food was bought.32 The second way they found food was from imperial donations—either at times of joint victory or at times of considerable hardship. Alexios I perhaps fed some of the army of Peter the Hermit in Constantinople and promised Hugh of Vermandois not only an adequate market (copiosum mercatum) but also alms for the poor.33 Before the capture of Nicaea, 26 Alexiade 10.5.9 (ed. Leib, 2:209). According to Anna, when Alexios heard of the crusade, he told his officials to panhgu´ rei" te dayilei'" ejx aJpasw'n tw'n cwrw'n kata` th` n oJdo` n ejxa´ gein. Later, Alexios heard about the arrival of more Crusaders, and again he sent word to ta` zwarkh' tou´ toi" ejrcome´ noi" ejpicorhgeiˆn ejp∆ aujtv' tou´ tv tetagme´ noi", wJ" mh` labei'n aujtou` " to` para´ pan ejschke´ nai. Alexiade 10.10.3 (ed. Leib, 2:228). Orderic Vitalis says that during the passage of the crusaders of Walter Sans-Avoir, the doux and the officials of Philippopolis allowed them to enter the city “et mercatum concesserunt,” which they had not done before: Ecclesiastical History, 5.31 (this information is found nowhere else, according to the editor’s note). 27 R. Hill, ed., Gesta Francorum et aliorum Hierosolimitanorum (London, 1962), 10. 28 See below, 164. 29 E.g., Robert the Monk, 744, 748, 749. Also William of Tyre, 2.6, 2.12, 2.14 (1:168–69, 175–77, 178–79). 30 R. Va´ri, Leonis imperatoris Tactica, 2 vols. (Budapest, 1917–22), 1:225. 31 Robert the Monk, 745: Bohemond tries to get “forum rerum venalium”; cf. Hill, Gesta Francorum 8: “quesivimus mercatum.” Cf. Orderic Vitalis, 5.45, 49. 32 Among the prices, see Alexiade 11.4.3 (ed. Leib, 3:20): in Antioch, hJ kefalh` tou' boo` " ejpi` trisi` crusi´noi" stath'rsi ajpempolei'to. On the fact that food was normally purchased as opposed to donated or acquired by raids, see William of Tyre, 1:142, 143, 148, 164–65, 173, 176, 178, 184. 33 “[I]mperator iusserat dari mercatum, sicuti erat in civitate”: Hill, Gesta Francorum 2; but he uses mercatum also to mean a market; cf. Robert the Monk, 732: “mercatum eis habere concedebat, quod etiam in civitate erat.” Orderic Vitalis, 5.43; cf. p. 327 on ingentia dona to the leaders of the Crusade of 1101 and p. 335 on ship-

[ 164 ] Byzantine Trade Alexios allowed a ship to bring provisions for purchase; but after the fall of that city, he gave to the poor of the army alms, which Stephen of Blois defines as distributions of food.34 The third way in which the Crusaders found food was plunder. Relatively limited during the First Crusade, at least on Byzantine territory, despite the depredations of Tancred and Bohemond and those of the Lombard participants in the Crusade of 1101,35 it became a way of life when the Crusaders entered Turkish territory. Spoils of war, not quite plunder, were also available, as also were large donations by the rulers of various Muslim towns once the Crusaders had entered Syria and Palestine. As for the Second and Third Crusades, the matter of provisioning was paramount in the eyes of the Crusader leaders, and in the eyes of the Byzantine emperors it had become, along with safe conduct, the most powerful negotiating tool. In his letter to the pope regarding the crusade of Louis VII, Manuel I Komnenos promised safe passage and markets (and suitable exchange: “forum idoneum, concambium competens”), in exchange for an oath of security for his realm.36 He then sent orders everywhere to have provisions brought to the roads through which the Crusaders would pass.37 The German army too was promised “hospitality” (uJpodoch'" te ajpolau´ onta" eij" to` eijko` " kai` filofrosu´ nh" th'" a“llh") in exchange for an oath of friendship.38 The governor of Nish, Michael Branas, was ordered to provide “necessities” (proujnoeiˆto h“dh tw'n ajnagkai´wn aujtoiˆ", ou”tw prostetagme´ non aujtv'),39 and similarly in Sofia. In Sofia, at least, it is certain that its governor made sure that Louis VII had the right to buy provisions.40 Manuel forced the French to reconfirm their oaths in exchange for market privileges several times: most importantly in Constantinople, where he exacted the oath of homage in return for a promise of guides, fair exchange, and markets everywhere,41 and, later, in Attaleia.42 When arrangements broke down, either by design on the part of the Byzantines or by chance, or because there was not enough food to feed the Crusaders (as in Attaleia), the Crusaders plundered.43 The arrangements with Frederick Barbarossa, as reported by Niketas Choniates, Ansbert and other sources, were almost a point-by-point copy of those made earlier with Conrad and Louis VII—which means that the imperial bureaucracy had learned more

loads of tetartera given by Alexios to the Crusaders in 1101 to persuade them to cross into Asia Minor. Ekkehard of Aura (Hierosolymita, ed. H. Hagenmeyer [Tu¨bingen, 1877], 233–34), also says that in 1100 (1101?) Alexios promised to provide markets and give alms to the poor. 34 Robert the Monk, 778; First letter of Stephen of Blois to his wife, RHC, HOcc 3:886. According to Stephen, Alexios distributed food to the poor even before the fall of Nicaea, daily. After the capture of the city, “omnia vero victualia peditibus distribuantur.” 35 Albert of Aix, 559–60. Cf. the attack on Adrianople in 1101: Ekkehard of Aura, 234–35. 36 Ne´ o" JEll. 11 (1914): 111–12; Odo of Deuil, De profectione Ludovici VII in orientem, ed. V. G. Berry (New York, 1948; repr. 1965), 29. 37 Choniates, ed. van Dieten, 61 ff. 38 Ioannis Cinnami Epitome rerum ab Ioanne et Alexio Comnenis gestarum, ed. A. Meineke (Bonn, 1836), 2.2, p. 68. 39 Kinnamos (Bonn ed.), 2.13; p. 70. 40 Odo of Deuil, 45. 41 Odo of Deuil, 77–83. 42 Odo of Deuil, 129. 43 Odo of Deuil, 41, 97; Kinnamos (Bonn ed.), 2.15 ff, pp. 75 ff.

Angeliki E. Laiou [ 165 ] or less what to do, unless it simply means that Choniates is repeating himself. The second eventuality, however, is unlikely, since he had firsthand knowledge of the later arrangements, and the statements of Odo of Deuil (and Ansbert) and Choniates corroborate each other at several points.44 Once again the quid pro quo was safety for imperial lands in return for markets, although, as is well known, the problem deteriorated into a conflict over imperial titles. Once again the emperor appealed to the provincial governors to “transfer goods from the various regions to where the king would pass.”45 In the case of the Third Crusade, promises of safe-conduct and optimi fori were made in Nuremberg in November 1186, and repeated in Nish and Branitsevo. The question of iustum concambium arose in October 1189.46 But arrangements broke down very soon indeed, and the German army—100,000 of them—found its food by plunder, which must have been very painful for the population. Permission to plunder was solemnized by the Treaty of Adrianople in February 1190. According to Choniates, at the time of the Third Crusade, the Turks of Konya also promised safe-passage and provisions, but broke their promises.47 How was this food marketed, and who sold it? In all cases, the emperors gave the original orders, which mandated two courses of action: that markets should be provided to the armies and that provisions should be collected by the governors of the regions.48 William of Tyre, perhaps not entirely trustworthy on this point, adds that Alexios’ edict regarding the army of Bohemond envisaged the death penalty for anyone who disregarded his orders to “iusto precio et equo pondere ducis exercitui quelibet mercimonia venderentur.” His narrative often reiterates that Alexios not only gave the Crusaders the right to buy and sell, but also ordered that the sale be carried out with correct weights and measures and at a just price.49 During the Crusade of 1100–1101, permission was given the Crusaders to buy and sell (“emendi et vendendi”) in Rosa (the Rusa of the First Crusade: Xanthe or Komotini), Panidos, Rodosto, Didymoteichon, “castello . . . de Natura,” Selymbria, Adrianople, Philippopolis;50 some of these towns, like Panidos and Rodosto, were major grain markets. What content we should give to these statements is another story. In some cases, for example, Alexios’ gratis provisioning of the poor at Nicaea, it is probable that imperial stores were implicated. When it is a matter of markets, panegyreis, mercatum, fora, we have to imagine one-time markets, for the most part situated in towns along the way, which 44 See, for example, the story about the inhabitants of cities carrying out their transactions with the Crusaders by means of ropes thrown down from the walls: Odo of Deuil, 40; Choniates, ed. van Dieten, 66. A further and amazing corroboration: According to both Odo of Deuil (76, 82) and Ansbert (Historia de expeditione Friderici, ed. A. Chroust, MGH, ScriptRerGerm, n.s., 5 [Berlin, 1928], 65), Manuel and Isaac II respectively agreed that if the army was not able to buy provisions, it had the right to plunder, though not to occupy the territory in question—or, in the case of Frederick I, not to give it to the Turks. This underlines the problems posed by provisioning at politically difficult times: Roger of Sicily was attacking Greece at the time of the Second Crusade. 45 Choniates, ed. van Dieten, 402. 46 Ansbert, 15–16, 29, 33, 48. 47 Ansbert, 25, 26, 37, 39, 44, 59, 66, 71, 73. On the Turks, see Choniates, ed. van Dieten, 412; Ansbert, 69. 48 Cf. Choniates, ed. van Dieten, 402–3, for the Third Crusade. 49 William of Tyre, 2.12 (1:176); cf. 1.18.54 (1:142), 1.19.12 (1:143); 2.14.15 (1:179). 50 Albert of Aix, 559.

[ 166 ] Byzantine Trade in any case would have known the institution of the periodic market or fair. The imperial orders would have been to agree to sell to the Crusaders, a somewhat frightening proposition considering the vast numbers of people going through. Imperial officials would do their best to persuade the inhabitants to sell commodities: this was the role given to the pansebastos Eustathios Philokales, who was to accompany the army of Frederick I as far as Philadelphia, for exactly that purpose.51 What the Crusaders “sold” would have been jewels, silver plate, their arms and horses in times of dire need, and money.52 There was, undoubtedly, a combination of factors that created a market. Despite the fact that the emperor issued orders regarding provisioning, generally speaking the markets were not, I think, state controlled, but were, rather, composed of producers (both landlords and peasants) and merchants.53 In other words, as far as we can see, food was not requisitioned by the Byzantine state, nor were imperial or military stores opened. Occasional references allow us a glimpse at the various groups of people who sold food to the Crusaders. Fulcher of Chartres mentions “citizens” of Constantinople, who, by order of the emperor, brought food to sell outside the walls of the city.54 These were, presumably, merchants. In Attaleia, during the Second Crusade, the food supply depended entirely on maritime trade, for the surrounding territory was in Turkish hands. The grain must have been brought in by merchants.55 Direct sale by producers may be deduced from other texts. William of Tyre ascribes the absence of markets at Nish, during the First Crusade, partly to the actions of the Byzantine governor and partly to the fact that the peasants had fled.56 As the army crossed Macedonia, its escort told the “inhabitants of the land” to bring provisions, the Crusaders not being allowed to enter the cities.57 These “inhabitants” could be either merchants or direct producers. When Bohemond reached Nicaea, he ordered “maximum mercatum conduci per mare, et pariter utrinque veniebant, ille per terram et ille per mare, et fuit maxima ubertas in tota Christi militia.”58 Since 50,000 to 60,000 people had to be fed, the sellers can hardly have been peasants: they must have been either merchants or landlords. Food seems to have been concentrated mostly in cities and towns; the process would have been something like the one that brought to Halmyros, along a west-east axis, Greeks with grain to sell, as reported by Edrisi, writing at approximately the time of the Second Crusade.59 Prices seem to have been formed on the ground, although the Crusaders clearly expected state control of both prices and weights and measures. An indication of this, at a 51

Ansbert, 65. Cf. below, 167–68, and Odo of Deuil, 75. 53 See Kinnamos (Bonn ed.), 2.13, pp. 70–71: the Germans apply unjust force “on those offering them food for sale in the market,” in the plains that come after Dacia, i.e., in Bulgaria: toi'" te kat∆ ejmpori´an ta` w“nia sfi´sin ajpodidou'si cei'ra ejpe´ balon a“dikon. 54 Fulcher of Chartres, Historia Hierosolymitana (1095–1127), ed. H. Hagenmayer (Heidelberg, 1913), 331. 55 Odo of Deuil, 129; William of Tyre, 16.26 (2:753). 56 William of Tyre, 1.19.26–36 (1:143) (army of Walter Sans-Avoir). 57 Gesta Francorum 10. 58 Gesta Francorum 14. 59 P. A. Jaubert, La ge´ographie d’Edrisi, 2 vols. (Paris, 1840), 2:291, 196. 52

Angeliki E. Laiou [ 167 ] later period to be sure, is offered by the reports of the final accord between Isaac II and Frederick I (the Treaty of Adrianople, February 1190)—namely, that the emperor was obliged to ensure that there would be “good” (i.e., fair) markets (bona mercata), selling at a just price (iusto pretio), which here seems to be understood as the price that the emperor would have had to pay if he were buying the food.60 An anonymous letter says that the sale should be at half the price the Crusaders were paying until then.61 So one job of the imperial officials would have been to guarantee fair prices and stop speculative sales at famine prices. Such, at least, seems to have been the case in 1190, and William of Tyre’s narrative suggests that it was also the case during the First Crusade; but this historian’s insistence on “just prices” may well be anachronistic. Indeed, the great fluctuation in food prices because of scarcity, or because the armies were so large, suggests that fair prices, even if promised, were not delivered.62 Choniates wrote scathingly about the inhabitants of cities who did not provide adequate markets and seized the opportunity to sell at extravagant prices. They closed their gates, he says, and threw down ropes, to collect the money of the Crusaders and then send down whatever victuals they saw fit— while the worst of them simply took the money and gave nothing in return. They used unjust weights and measures and even tampered with the quality of the food.63 Finally, during the Third Crusade also, the inhabitants of some cities, such as Philippopolis, refused to sell to the Crusaders.64 All this shows the limits of imperial control of the situation. There was an interplay, then, of free-market forces and imperial orders; this was free exchange in which the state was expected to intervene and did intervene to a limited extent. There was some pressure toward increased state control, but it was not very successful. Freedom for foreigners to buy and sell within the frontiers of the empire was a specific privilege. It was originally granted to the Venetians and was then extended to Crusaders; there are, therefore, formal connections between commercial institutions and the provisioning of the crusading armies. Also, the distribution mechanisms must have been influenced by the problems of provisioning. However, the acute and episodic nature of arrangements for the provisioning of the armies in all probability did not have longlasting effects on the Byzantine economy. Rather, these arrangements are inscribed in the larger framework of developing commercial mechanisms. It may be otherwise with questions regarding currency exchange. It must first be stressed that provisioning and currency exchange were closely connected. Even when 60 The reference is undoubtedly to the ejxw´ nhsi", the obligatory sale to the state of various products, at fixed prices, on which see N. Oikonomide`s, Fiscalite´ et exemption fiscale a` Byzance (IXe–XIe sie`cle) (Athens, 1996), 97–99. 61 Ansbert, 65–66; K. Hampe, “Ein ungedruckter Bericht u¨ber den Vertrag von Adrianopel zwischen Friedrich I. und Isaak Angelos vom Febr. 1190,” Neues Archiv der Gesellschaft fu¨r deutsche Geschichtskunde 23 (1898): 400. The editor of the letter thinks that this was impossible and simply the stuff of rumor. 62 Cf. Robert the Monk, 749: the price of bread falls. See also Odo of Deuil, 135, who complains of extremely high prices exacted by the Greeks in Attaleia. Cf. idem, p. 97, on the high price of food outside Nicaea, where the sellers demanded cuirasses and swords in payment. 63 Choniates, ed. van Dieten, 66–67; Odo of Deuil, 41. 64 Choniates, ed. van Dieten, 403.

[ 168 ] Byzantine Trade sources such as William of Tyre refer to the buying and “selling” by the members of the First Crusade, they may actually be referring to the “sale” of coins, that is, to exchange transactions. It should also be noted that Crusader sources normally quote prices in Western coins, which again raises the problem of currency transactions. B. Currency Exchange During the First Crusade, through 1101, not a word is spoken about the problem of currency exchange. The Crusaders brought with them a multiplicity of coins (Raymond of Aguilers mentions seven types, but we know there were many more)—all billon deniers of different intrinsic values (some quite good silver)—and marks of silver. The idea floated by M. Matzke, that perhaps they took a limited and deliberately chosen type of coin does not seem plausible.65 Some of the great barons, like Godfrey of Bouillon, brought with them considerable quantities of marks of silver. How did the Crusaders fend for themselves, once they had reached the frontiers of the Byzantine Empire? While the sources of the First Crusade remain mute on the problem of currency exchange, it must have been posed brutally by reality. The problem of “markets” and high prices, omnipresent in the sources, undoubtedly but tacitly includes that of exchange—not only a fair exchange, but, to start with, exchange itself. Who, in the towns along the way, would have been willing to change the Western deniers against Byzantine coins, and at what equivalence? If there was no exchange, how many people would have been willing to accept this unlikely money in payment and, again, at what value? Ingots of silver may have been easier for the locals to accept. While marks of silver were too valuable for small-scale transactions, smaller ingots (not attested in our sources, but possible) may have been used. In this case, one must assume that the Byzantine money changer or merchant would have had the opportunity to give an exchange rate lower than the intrinsic value of the ingot. The exiguous number of coins datable to the First Crusade that has been found in the Balkans,66 Asia Minor, and Syria-Palestine could be interpreted in a number of ways: as evidence that the Crusaders did not so much purchase their food as live off the land, which contradicts the rest of the evidence, at least for the Balkans or, more plausibly, as 65 M. Matzke, “Die sieben Kreuzfahrermu¨nzen und das Papsttum,” SM 44 (1994): 13–19, mentioned in D. M. Metcalf, Coinage of the Crusades and the Latin East in the Ashmolean Museum, Oxford, 2d ed. (London, 1995), 13–14. As Metcalf says, this proposition depends on the accuracy and completeness of Raymond of Aguilers’ list of coins, which is doubtful. Godfrey of Bouillon got from the bishop of Lie`ge 1,300 or 1,500 marks of silver: History of the Crusades, 1.267. At Antioch, Tancred asked for and was promised 400 marks of silver for guarding a castle: Gesta Francorum 43. The list of coins appears in Raymond of Aguilers, ed. Hill, 111–12. 66 The only significant find in terms of numbers is a hoard of 83 coins in Constantinople possibly associated with the First Crusade: Metcalf, Coinage of the Crusades, 6. A. M. Stahl (“The Circulation of European Coinage in the Crusader States,” in V. Goss, ed., The Meeting of Two Worlds: Cultural Exchange between East and West during the Period of the Crusades, Studies in Medieval Culture 21 [Kalamazoo, Mich., 1986], 85–102, at 86) mentions the hoard of 1,600 coins found in 1884 in Zombor, Croatia, as belonging to the First Crusade, but the hoard has been dated to the Second Crusade. Another 3,000 coins found in Hungary may be datable to the Second Crusade. A very large hoard of 7,700 coins and ingots (with, unfortunately, an unknown find-spot) has been associated with a member of Frederick Barbarossa’s army during the Third Crusade: see Metcalf, Coinage of the Crusades, 7–10.

Angeliki E. Laiou [ 169 ] an indication that the Crusaders’ coins were routinely and quickly melted down by the Byzantines, perhaps to be reminted; if so, there must have been some sort of government service that bought the coins from the citizens.67 Similarly, the fact that coins were minted very soon after the conquest of Syria and Palestine suggests that Western coins were not originally well received in the East.68 After 1140, the Crusader states began to mint silver pennies. Before and after, there was a heavy flow of silver to the East.69 Once the members of the First Crusade reached Constantinople, their money supply began to become replenished through imperial gifts of cash, and the problem of exchange was eased to some extent, though only partially. Large gifts of gold and silver on the part of Alexios to the Crusader leaders put Byzantine money into circulation: Fulcher of Chartres reports that after the Crusader leaders had taken the oath of fealty, Alexios gave them “de numismatibus suis,” clearly coined gold and perhaps silver.70 Anna Komnene speaks of huge gifts to Bohemond, including gold and silver coins.71 There were also gifts of smaller coins. After the fall of Nicaea, the emperor distributed to the leaders gold, silver (“de auro suo et argento”),72 as well as other precious things, both from the spoils and from his treasury. The rank and file—the foot soldiers—received distributions “de sumis suis aeneis quos vocant tartarones.”73 The tetarteron in question here is Alexios’ tetarteron, made first of lead, then of billon, then of copper.74 Smallscale transactions, at a time when markets were pretty well established outside Nicaea, were made possible with these coins. In 1101 Alexios distributed boatloads of tetartera, according to Orderic Vitalis.75 When the Crusaders began to be victorious over the Muslims, gold came their way; some undoubtedly was minted gold, as we know for the case when the governors of the towns of Syria and Palestine offered the Crusaders tribute—or gifts with which they bought immunity—not to mention the ghoulish recovery of bezants from the inside of the corpses of slain Turks.76 It is telling that the leaders, when they were in Syria and Palestine, still had substantial sums of money, in marks of silver, even after the long trip.77 Only two rates of exchange are mentioned in the sources of the First Crusade, and

67

Metcalf, Coinage of the Crusades, 3–11, gives a list of the coins that belong to the first three crusades. The princes of Antioch and the counts of Edessa seem to have struck almost immediately copper coins imitating the Byzantine follis of the late 11th century. See Metcalf, Coinage of the Crusades, 22–23, 31 ff. 69 Stahl estimates a total of ca. 1 million kg of silver ⫽ 1,000 tons throughout the crusading period: “Circulation of European Coinage,” 97. 70 According to William of Tyre, after his accord with Godfrey of Bouillon, Alexios sent him every week, between Epiphany and Ascension, as many gold coins (“auree monete”) as two strong men could carry, and 10 modioi of copper coins (“de ereis vero denariis X modii”), to be distributed among his nobles and the plain soldiers: William of Tyre, 2.12 (1:176 ff). 71 Alexiade 10.11.5 (ed. Leib, 2:233). 72 Fulcher of Chartres, 333–34. Is the “silver” coin the electrum trachy? 73 Fulcher of Chartres, 333–34; Stephen of Blois boasted to his wife that he doubled his worth in gold and silver: RHC, HOcc 3:886. 74 M. F. Hendy, Studies in the Byzantine Monetary Economy c. 300–1450 (Cambridge, 1985), 515. 75 Orderic Vitalis, 5.334, says that the tetartera were as current as bezants in transactions in Thrace and Bithynia. 76 E.g., Raymond d’Aguilers, ed. Hill, 111: 15,000 aureos from the governor of Tripoli; Gesta Francorum 80. 77 See, e.g., Gesta Francorum 97; Raymond d’Aguilers, ed. Hill, 111–12. 68

[ 170 ] Byzantine Trade they are not for small coins, but rather for gold hyperpyra or dinars exchanged for silver (billon) deniers. One is the rate of 8 hyperpyra for 120 soldi (1 soldo, a unit of account ⫽ 12 deniers), outside Antioch, at a time of famine: “vendebant [the Syrians and Armenians] onus unius asini octo purpuratis, qui appreciabantur centum viginti solidis denariorum,” that is, 1 hyperpyron for 15 soldi, or 1 hyperpyron for 180 deniers.78 This is high in terms of the approximate intrinsic value of the coins in question.79 It represents a rate of exchange for poor Western coins in times of famine. The other rate of exchange is not with the hyperpyron but rather with the gold coin of Tripoli: “Valebat quippe unus aureus eo tempore octo vel novem solidos monetae nostri exercitus,” says Raymond of Aguilers, and continues with the famous list of seven coins that he considered to be the army’s money (or at least the money of the army of Raymond of St. Gilles).80 Of these coins, the denier of Le Puy is said by Raymond of Aguilers to have been worth only half the value of the others.81 One wonders whether the difference in fineness was as obvious to the locals as it was to Raymond of Aguilers and is to modern numismatists. In any case, if the gold coin in question was the normal dinar, the Crusaders got approximately double what they had received outside Antioch. The rate of exchange with the dinar seems normal, and therefore the Byzantine one was very detrimental to the Crusaders. Gold and silver were apparently exchanged on the basis of intrinsic value. Outside Antioch, the exchange was between hyperpyra and billon deniers, and was heavily influenced by the Crusaders’ need. More interesting phenomena start with the Second Crusade. The Second and Third Crusades were different from the First in the matter of exchange as in other matters. All the kings who planned to go through the Byzantine Empire asked not only for markets but for fair exchange; and undoubtedly Manuel I was as interested as they were in some kind of regularization of the currency exchange, without which the problem of provisioning was bound to become exacerbated. Also, the participants of these betterorganized expeditions seem to have carried considerable amounts of money with them. It is also with the Second Crusade that Western sources begin to show an evident concern with problems of currency exchange, and there are strident complaints from the pen of Odo of Deuil.82 The complaints, and the situation to which they refer, repay close examination. The members of the French army had brought both silver deniers (presumably the denier Parisis, which I will take as the basis for calculations) and marks of silver (it is not clear which mark is meant). The money, as far as we can tell from the complaints, was not accepted as means of payment in the local markets. The insistence of Louis VII on concambium competens also suggests that coins were exchanged, as does Choniates’ state78

Gesta Francorum 33. I am very grateful to Philip Grierson and Ce´cile Morrisson for their valuable help with this section. According to Michael Hendy, the “purpuratus” is the nomisma trachy, 20.5 carats fine, introduced by Alexios I: Hendy, Coinage and Money, 34–35. Throughout this section, the reader should consult the table and Ce´cile Morrisson’s appendix. 79 See the appendix below. 80 Raymond of Aguilers, ed. Hill, 111–12. 81 Metcalf, Coinage of the Crusades, 13. 82 Odo of Deuil, 40, 66.

Angeliki E. Laiou [ 171 ] TABLE 1 E R Early 1098 1 hyperpyron ⫽ 15 soldi ⫽ 180 deniers1 Second Crusade In the Balkans 1 stamenon ⫽ 5 deniers2 [1 hyperpyron ⫽ 240 deniers] 144 stamena ⫽ 1 mark of silver 3 [3 hyperpyra ⫽ 1 mark of silver]4 [Official value of the billon trachy in 1136: 1/48 hyperpyron]4 In Constantinople 1 stamenon ⫽ less than 2 deniers 396 stamena ⫽ 1 mark of silver 5 Third Crusade (Treaty of Adrianople, 1190) 5.5 hyperpyra ⫽ 1 mark of (pure?) silver 6 3 hyperpyra ⫽ 1 mark of (impure?) silver7 120 stamena ⫽ 1 hyperpyron8 1

Gesta Francorum 33. Odo de Deuil, 41. 3 Ibid. 4 For the official value, see Dmitrievskii, Opisanie, 689 (as in note 86). The equivalences 1 hyperpyron ⫽ 240 deniers and 3 hyperpyra ⫽ 1 mark of silver are not mentioned in the sources for 1147–1148. They are theoretical and derive from the official value of the billon trachy. 5 Odo de Deuil, 89–90. 6 Ansbert, 66. 7 Hampe, “Ungedruckter Bericht,” 400. 8 Ansbert, 66. 2

ment—unless it is rhetorical—that people cheated the participants of the Second Crusade of their gold and silver.83 On what basis did the exchange take place? We note, first, that Odo of Deuil quotes the exchange rates as between the denier and the mark of silver and the “stamenon,” that is, the petty Byzantine coin, not the gold hyperpyron. The stamenon, a corruption of the rare Byzantine term histamenon, is the billon trachy (“aspron trachy”), a billon coin, mostly copper, with, at that time, 6.3 83 Choniates, ed. van Dieten, 66. See also P. Grierson, “A German Crusaders’ Hoard of 1147 from Side (Turkey),” in LAGOM: Festschrift fu¨r Peter Berghaus (Munster, 1981), 195–203. The West had no gold coinage yet; Choniates may be referring to either unminted gold or Byzantine gold coins acquired by exchange. C. Morrisson and Marc Bompaire suggest that by “concambium competens” Louis VII may have meant that the French coins should be exchanged on the basis of their intrinsic value plus a corresponding premium. According to M. Bompaire, such a practice is attested in French documents of the late 12th century for coins circulating outside the territory where they were legal tender and where they had a nominal value higher than their real value. Cf. M. Castaing-Sicard, Monnaies fe´odales et circulation en Languedoc, XIIe au XIIIe sie`cle (Toulouse, 1961). I am grateful to M. Bompaire for this information and for the reference.

[ 172 ] Byzantine Trade percent silver.84 The Crusaders refer to it as a copper coin, as, to all intents and purposes, it was. Thus they suffered a first, psychological, shock, at the exchange of silver for copper. Within the Byzantine Empire, the billon trachy functioned as a virtual token or quasitoken coin.85 Its equivalence to the hyperpyron was legislated, and, in 1136, it was worth 1/48 of an hyperpyron, that is to say, one gold coin was worth 48 billon trachea or stamena.86 The intrinsic value of the billon trachy (based on its silver content) would have been much lower. It was, then, against this token coin that the denier and the mark were exchanged. The question arises whether the exchange was based on the intrinsic value of the coins or not. M. Hendy has suggested that the Crusaders would have expected an exchange based on the intrinsic value of each denomination, while the Byzantines would have insisted on the official (partly token) value of the billon trachy.87 This is a highly plausible hypothesis and may well help interpret some of the complaints of the Western sources; but by itself it does not suffice to explain the actual exchange rates. Odo of Deuil mentions three exchange rates, two in the provinces and one in Constantinople. The rates in the Balkans and in Asia Minor were very close to each other, and very different from that in Constantinople. According to Odo of Deuil, when the Crusaders first entered the Byzantine Empire, they bought 1 stamenon for 5 deniers, “et pro duodecim solidis [earum] marcam triste dabamus vel potius perdebamus,” that is, they bought 144 billon trachea for one mark of silver.88 In Asia Minor, three days’ march away from Constantinople, the exchange rate was almost exactly the same.89 This exchange rate was very unfavorable to the Crusaders, whether the exchange was 84 Hendy, Coinage and Money, 21, 31. The billon aspron trachy, with a “pronounced silvery surface” (on which see M. F. Hendy and J. A. Charles, “The Production Techniques, Silver Content, and Circulation History of the Twelfth-Century Byzantine Trachy,” in Hendy, The Economy, no. ), would not have been mistaken for a silver coin. 85 Copper money was almost always a token coin in Byzantium. The billon coin is called by C. Morrisson “‘fausse monnaie’ ou monnaie fiduciaire au sens large”: “La monnaie fiduciaire a` Byzance, ou ‘vraie monnaie,’ ‘monnaie fiduciaire’ et ‘fausse monnaie’ a` Byzance,” Bulletin de la Socie´te´ franc¸aise de Numismatique 34 (1979): 615–16. 86 The rate is given in the typikon of the monastery of Pantokrator in Constantinople: A. Dmitrievskii, Opisanie liturgicˇeskih rukopisei, hranyasˇˇcihsya v bibliotekah pravoslavnogo Vostoka, vol. 1 (Kiev, 1895), 689. See the Table, above. 87 Hendy, Coinage and Money, 21. Cf. above, note 82. 88 Odo of Deuil, 41. The part about the exchange rate of the mark has been misunderstood by Chalandon, Les Comne`nes, 2 vols. (Paris, 1900–1912), vol. 2, Jean II Comne`ne (1118–1143) et Manuel I Comne`ne (1143–1180), 298–99 n. 2, and, following him, by the most recent editor of Odo of Deuil. The correct interpretation was suggested to me by Philip Grierson and hinges on the fact that the “duodecim solidi” means 12 units of 12 stamena, i.e., 144 stamena. There is a further uncertainty concerning the mark, which had different weights in different parts of Europe. I assume that, unless the sources explicitly state otherwise, when they mention marks they refer to pure silver weighing one mark. On the circulation of unminted silver, traveling in bars of ingots of a standard fineness and “frequently of a standard weight,” see P. Spufford, Money and Its Uses in Mediaeval Europe (Cambridge, 1988), 209–24. I owe this reference to C. Morrisson. On the ingots in the “Barbarossa hoard,” see below, note 100. 89 Odo of Deuil, 66: they bought one stamenon for 5 or 6 deniers and paid one mark of silver for 144 deniers.

Angeliki E. Laiou [ 173 ] based on the nominal value of the billon trachy or on the intrinsic value of the coins.90 If, indeed, the equivalence in the Byzantine Empire was 1 hyperpyron ⫽ 48 billon trachea, the Crusaders paid 240 deniers for 1 hyperpyron (and 1 mark of silver for 3 hyperpyra), much more than in the First Crusade or in the negotiated settlement of 1190.91 This shows that, if the calculations are made on the basis of the hyperpyron, the Crusaders would have lost even more money than would be expected by an exchange on the nominal value of the billon trachy. They lost considerable money also if the exchange was in terms of the intrinsic value of the coins. It is very difficult to figure the intrinsic value of such coins, but an exercise might be interesting. The denier Parisis at the time of Philip Augustus weighed approximately 1 gram, and contained 36–37 percent silver, thus having about 0.4 g silver. The billon trachy weighed around 4.30 g and its fineness was 6.3 percent silver, therefore its silver content was about 0.27 g.92 The difference was visible to the naked eye. The silver content of the billon trachy was less than that of the denier Parisis both absolutely and relatively, and any exchange rate that quoted the billon trachy in multiples of the denier Parisis was bound to be excessive. A fair exchange, in the eyes of the Crusaders, was what they got in Constantinople, after an agreement with the emperor: 1 stamenon for less than 2 deniers,93 and 396 stamena for 1 mark of silver.94 If we translate that into hyperpyra, and assume a nominal value of 1/48 hyperpyron for the stamenon, the Crusaders would have bought 8.25 hyperpyra for 1 mark of silver in Constantinople, a rate much more favorable to the mark than at the time of Frederick Barbarossa. The implied gold-silver ratio would be 1:7.86, very weak for Byzantium. Certain observations are in order. The exchange rate between the stamenon and the denier, as that between the stamenon and the mark of silver, is consistent, grosso modo, with an exchange based on the nominal value of the stamenon, which normally was overvalued by something like 2.5 times its intrinsic value.95 If the Crusaders then bought hyperpyra with their stamena, the gold-silver ratio would have been 1:8. It is, however, much more likely that the conversion of stamena to hyperpyra was either impossible or extremely limited, since both the fiduciary nature of the stamenon and the cheap rate that would result for the hyperpyron argue against it. The difference in exchange rates between Constantinople and the provinces is striking indeed. In part, it reflects the real problems of exchanging money in the provinces. The presence of large Crusader armies would mean a sharp rise in demand and a shortage of Byzantine coins, thus making them expensive. Add to this the cost of metallic exchange to the money changer. The result was high rates and a psychological shock to the Crusaders, who exchanged silver for what looked to them like copper. In Constantinople, Manuel seems to have insisted on the nominal value of the sta90

See the Table, above. The difficulty with the mark, mentioned in note 88, must be kept in mind. For 1190, see below, 175. 92 See the appendix below. 93 Odo of Deuil, 66; cf. Chalandon, Jean II Comne`ne, 298–99 n. 2. 94 “[E]arum triginta tres solidos propter marcam,” i.e., 33 units of 12 stamena for one mark of silver. 95 Cf. the appendix below. 91

[ 174 ] Byzantine Trade menon, a deal that the Crusaders also considered to be fair. He may also, however, at the same time, have devalued the stamenon given to the Crusaders, as Choniates charges, taking with one hand what he had given with the other. Hendy discounts the possibility that Manuel had minted debased billon trachea specifically for the Crusaders, as charged by Choniates: ajrgu´ rion ajdo´ kimon eij" no´ misma ko´ ptetai kai` tou'to proba´ lletai toi'" ejk tou' tw'n Ij talw'n strateu´ mato" ajpodo´ sqai ti qe´ lousi.96 Numismatic evidence, it seems, indicates that the debasement of the billon trachy (the stamenon) took place much later, and, it would appear, very fast. Yet it is not impossible that Manuel may have done exactly what Choniates accuses him of, that is, minted a debased billon trachy (with more copper) precisely to defraud the Crusaders and for use in Constantinople. One other point is worth noting. None of the sources for the Second Crusade mentions an exchange rate for the hyperpyron, although we are told that in Constantinople the Crusaders did, in fact, exchange silver, both minted and unminted, for gold.97 This suggests that the negotiated rates involved only the petty Byzantine coin and did not extend to gold-silver exchanges. In this case, the rate of exchange for gold and silver would have been negotiated on the ground, leaving open the possibility that the Crusaders overpaid for Byzantine gold, according to their need. This hypothesis finds some corroboration in a wonderful story, related by Odo of Deuil, which also shows how the exchange took place. In Constantinople the Crusaders could exchange money on board the food ships and also “ante palatium vel etiam in tentoribus,” that is, the money changers came to them at the site of the camp; or they could change money in the money changer’s shops.98 Once the Crusaders had left the city to cross over to Asia, they were attended by food ships, on which there were also money changers (cum cambitoribus). The money changers disembarked and set up their benches, which fulgent auro. Here came the Crusaders to exchange money as they needed to. Odo of Deuil speaks of the silver vessels that the money changers had bought from the Crusaders; so they also exchanged unminted metal for coin. A fairly orderly procedure, except for the fact that there were some heroic Crusaders who thought this a good occasion for plunder. A riot ensued, where the money changers lost their money and fled to Constantinople. Eventually Louis VII made restitution, and the money changers returned. But this is a rare picture of how operations took place: the money changers were right at the place where the food was, and the currency exchange went hand in hand with the purchase of necessities.99 How such transactions were carried out in the provinces is less clear; in fact, we have no information. One can imagine, however, that part of the business of creating a market included setting up facilities of exchange. Indeed, it is almost necessary to posit this, and it is not at all a far-fetched supposition, since we know that at Byzantine fairs credit transactions also took place; so the money changers were used to being at fairs. Nor

96

Choniates, ed. van Dieten, 67; cf. Hendy, Coinage and Money, 22, 170–71. Odo of Deuil, 66. 98 Odo of Deuil, 66. 99 Odo of Deuil, 75. 97

Angeliki E. Laiou [ 175 ] would the money changers have been averse to setting up business on the route of the Crusaders: even if Manuel had not debased the coinage (indeed, especially if he had not debased the coinage), they stood to gain quite a lot in their transactions with people in need who had money of which to be relieved. By the time of the Third Crusade, the balance of blackmail had changed, so that Frederick Barbarossa was able to exert great political and military pressure on the Byzantines.100 The exchange agreements between the German emperor and Isaac II, in 1190, were highly sophisticated. The Treaty of Adrianople quotes both the exchange rate of the mark of silver, in hyperpyra, and the equivalence between the hyperpyron and the billon trachy. One mark of (pure) silver would sell at 5.5 hyperpyra, each hyperpyron to be counted as being worth 120 stamena, whether these were old or new stamena.101 Thus no one could play games by giving different rates for hyperpyra and stamena. According to another Western source, the exchange rate between the mark and the hyperpyron varied according to whether the mark was non examinata or examinata, that is, non-assayed or assayed, with a control stamp, in other words, impure or pure. In the first place, the exchange rate would be 3 hyperpyra to the mark, in the second, 5 hyperpyra to the mark.102 The distinction may well have been made on the insistence, not of the Germans, but of Isaac II or, more likely, the money changers of Constantinople, who would otherwise have assumed the risk involved in changing impure silver ingots into gold hyperpyra. The rate offered for the impure mark represented considerable compensation for the risk of the money changers. The rate of exchange between the hyperpyron and the pure mark of silver seems to be within the normal boundaries of the gold-silver ratio. The growing sophistication of exchange transactions is indicated by the fact that the treaty states the equivalences between the hyperpyron and the billon trachy. The equivalence 120 new (debased) stamena to the hyperpyron implies the overvaluation of the stamenon and thus its fiduciary nature. When the same equivalence is applied to the old stamenon, however, then the fiduciary nature of the coin is undermined, and the emperor abandons part of his seignorage, that is, the state’s premium, the buying power that is over and above the intrinsic value of the coin. One may suppose that this latter equivalence was a special arrangement extracted by Frederick I and did not obtain in normal transactions in Constantinople. Later in the century, in 1199, Pisan notarial records give the rate of 184 stamena to the hyperpyron.103 This was undoubtedly the market (exchange) value current in Constantinople. There may have been a further devaluation of the billon trachy between

100 Barbarossa’s army must have brought quite a lot of money. The “Barbarossa hoard” contains 7,700 coins as well as ingots: Metcalf, Coinage of the Crusades, 8–10. The Byzantines, too, had learned that their coins were much sought after: as Frederick Barbarossa attacked the area of the monastery of Bacˇkovo, in the fall of 1189, its treasurer buried part of the annual budget: M. F. Hendy, “The Gornoslav Hoard, the Emperor Frederick I, and the Monastery of Bachkovo,” in idem, The Economy, no. . 101 Ansbert, 66; Hendy, Coinage and Money, 21–22. Hendy posits a devaluation of the billon trachy, which would have been the reason for this last provision. 102 Hampe, “Ungedruckter Bericht,” 400. 103 Mu¨ller, Documenti, no. , p. 77; Hendy, Coinage and Money, 22.

[ 176 ] Byzantine Trade 1190 and 1199, although the coins of Alexios III do not show lower silver content than those of Isaac II.104 But devaluation is not really what makes the difference between the negotiated rates of 1190 and the attested rates of 1199. Rather, the difference is due to a revaluation down of the billon trachy, to reflect a demand that would bring it closer to its intrinsic value, probably as a result of the double pressure of foreign merchants and Crusaders. Thus the highly overvalued equivalences mandated by the state in the 1130s were no longer effective, even though the face value of the stamenon remained lower than the nominal value; that is to say, the overvaluation of the billon trachy continued, but at lower levels than had obtained earlier. A hundred and fifty years later, Pegolotti wrote of the stamenon: “ma a questi stammini non si fa nullo pagamento se none in passagio di Gostantinopoli per lo paese, e per erbe e cose minute.”105 In their exchange transactions with the Crusaders, the Byzantines seem to have made good money and shown considerable knowledge of monetary matters. The merchants must have played an important role in transactions on the ground, if only because exchange was so closely tied to the purchase of merchandise on the part of the Crusaders. Those merchants who had activities both with the countryside and with Constantinople would have profited very considerably, especially given the unequal exchange rates obtaining in the capital and in the provinces. The merchants would have acted also as money changers. There are indications of such a development in the exceptional importance assumed by merchants and money changers in late twelfth-century Constantinople. The Treaty of Adrianople provides a first glimpse into this development, but it is a striking one. The treaty was guaranteed not only by the oaths of representatives of the two emperors and the confirmation of the patriarch, but also by the oaths of large numbers of their subjects. In the case of the crusading army, the oath was taken in Adrianople by 15,000 milites, in the presence of Byzantine ambassadors.106 This was given in exchange for an oath taken in Constantinople, in the solemn precincts of Hagia Sophia, by “500 men of the marketplace and the court” (ajpo` tw'n ajgorai´wn kai` th'" basilei´ou aujlh'"),107 who swore “that the emperor would keep the treaties inviolate and would give the leaders of the Germans safe passage and provisions.”108 This is an unprecedented event in Byzantine international relations, with obvious constitutional overtones. The oath was undoubtedly demanded by Frederick Barbarossa, as Ansbert’s text suggests; and the men who took it, whom another source calls “quingenti de melioribus terrae,” were understood to have sworn “quod ipse rex Grecie concanbium bonum et forum rectum et omnium suorum tutelam sine mala fraude ineundo et redeundo postaret.”109 The latter source is a letter, by an unknown author, that reports on the terms of the treaty as he heard them

104 Hendy explains the drop in the value of the billon trachy between the time of John II and the end of the 12th century by the decline in fineness: Studies, 518. However, his own figures show that the rate of decline in fineness is less than the rate of decline in value. 105 Francesco Balducci Pegolotti, La pratica della mercatura, ed. A. Evans, Medieval Academy of America Publication 24 (Cambridge, Mass., 1936), 40. 106 Ansbert, 66. 107 Ansbert, 66, calls them “quingenti homines meliores civitatis et imperii.” 108 Choniates, ed. van Dieten, 411. 109 Hampe, “Ungedruckter Bericht,” 399.

Angeliki E. Laiou [ 177 ] from “cives noti qui . . . Constantinopoli fuerunt”; the editor of the letter suggests, very plausibly, that the cives noti were (Italian) merchants.110 These Italian merchants, then, understood that their equivalent Byzantines, the very men who would have been engaged in trade and money changing, were made to guarantee the Treaty of Adrianople. Was their oath (along with that of some of Isaac’s aristocrats) demanded by the German emperor because he was used to dealing with the “best men” of cities? Or were the “best” businessmen of Constantinople playing a new role in relations with Westerners, merchants, and states? Here we must recall the near-contemporary affair of the piratical attack of summer 1192 and the subsequent negotiations, with which we began, and Isaac II’s efforts to get reparations for the goods of “the first among the merchants of Constantinople,” in whom it is legitimate to see a subgroup of the meliores terrae of the Treaty of Adrianople.111 It is pertinent to our discussion that it was these merchants (and money changers too, undoubtedly) who functioned as guarantors (ejgguhtai´) of the goods of Genoese and Pisan merchants held in deposit. Thus they engaged in money business as well, a phenomenon already mentioned above in connection with the probable role of merchants in money transactions during the Crusades.112 John Oxeobaphopoulos, who returned the deposit to the Genoese, bears a name that clearly connects him to the marketplace, for it means the “‘red’ purple dyer”; whether he was a silk manufacturer or a silk merchant, or both, or whether his was a family name, cannot, of course, be determined. Kalomodios, the one great merchant and money changer whose name and profession are explicitly attested, lived in the same period, during the reign of Alexios III. We also know that during Alexios III’s reign, cloth merchants and money changers and other tradesmen were able to buy honorific titles: oiJ ejn trio´ doi" kai` ajgorai'" kai` kollubistai` kai` pratai` tw'n ojqonw'n sebastoi` ejtimh´ qhsan.113 The important and complex role of the Constantinopolitan merchant and banker is the result of many developments: the extensive commercial activity of the Byzantines in the twelfth century; the development of Byzantine-Italian relations, which, on the institutional front (and I mean legal and economic institutions), was proceeding at a rapid pace and creating new situations; and undoubtedly to some extent it was also due to the active role of the merchants and money changers in the negotiations with Crusaders and in the implementation of agreements. The Treaty of Adrianople did not create this group. But it gave its existence a certain solemnity; and it may have contributed something to this group’s apparently significant power in Constantinople, and so to the fact that its interests were very much taken into account in the state relations between the empire, Genoa, and Pisa—and in the discussion of the terms in which both trade and piracy were to be conducted. The expanded role of the merchants and money changers, which recalls the earlier developments of the eleventh century, was cut short by the events of 1204. Would these developments have been different in degree or in kind without the pas-

110

Ibid., 400 n. 5. Above, 157–59. 112 Above, 176. 113 Choniates, ed. van Dieten, 523–24, 483–84. 111

[ 178 ] Byzantine Trade sage of the Crusader armies? One does not wish to engage in counterfactual argumentation. What may be stressed is the fact that relations between Crusaders and Byzantines, especially during the Second and Third Crusades, were arrangements between states. Thus they ratified and confirmed developments that may already have been taking place. In sum, the Crusades (always speaking of the first three major movements) had a number of influences on the Byzantine terms of trade with Westerners. Sometimes the connection is obvious: thus the first Pisan commercial privileges incorporate a clause that anyone who travels on Pisan ships to Jerusalem “against the pagans” will suffer no impediment in terms of passage, or of dapanai (stipendia in the Latin text), which is not really the equivalent of mercatus, although it does refer to provisions, or in terms of their military equipment.114 This guarantee of safe-conduct applied only to those Crusaders who would swear an oath to the Pisan ship captain not to harm the Byzantine Empire— hence R.-J. Lilie’s correct interpretation that those who did not swear not to harm the Byzantine Empire would not be guaranteed safe passage, and thus that there was a partial Byzantine boycott of the transport of Crusaders. These provisions bear considerable substantive similarity to the terms on which Bohemond had been allowed to recruit soldiers from Western Europe to Antioch, terms included in the Treaty of Devol, three years earlier.115 The connection between the two events is real at the political level. Since 1099 Alexios had been in conflict with Pisan and Genoese ships sailing to the Holy Land, absolutely parallel to the skirmishes and sometimes overt hostilities that took place with the passage of Crusader armies over land.116 After Bohemond went west to raise troops for an antiByzantine crusade, Alexios tried to include the Pisans (as well as the Genoese and the Venetians) in his system of defense against Bohemond; the Pisans had engaged in hostile piratical action in the wake of the war with Bohemond (in 1111).117 There is also a diplomatic connection, or a formal, institutional connection, since similar provisions found their way into commercial privileges very early indeed, and in treaties with the Crusaders (and Bohemond), almost from the beginning. Such connections, however, I have not discussed, preferring instead to concentrate on other issues. Sometimes the crusading arrangements and those with Italian city-states were based on different principles: a combination of statism and free trade in the case of some arrangements with the Crusaders regarding markets and exchange rates, versus the free trade agreements that characterized, to a significant extent, the commercial privileges, and that created a market response to currency exchange rates in Constantinople. Sometimes the two form part of a developing global pattern, as was the case with repara-

114 Mu¨ller, Documenti, no. , pp. 44 and 53. This is the privilege of 1111, inserted into the treaty of 1192. Cf. R.-J. Lilie, Handel und Politik: Zwischen dem byzantinischen Reich und den italienischen Kommunen Venedig, Pisa und Genua in der Epoche der Komnenen und der Angeloi (1081–1204) (Amsterdam, 1984), 69 ff. 115 Alexiade 13.12.17 (ed. Leib, 3:133): Bohemond agrees that all who cross the Adriatic with him will swear doulei´an to the empire, in the person of a man whom Alexios will send to Italy (the equivalent of swearing an oath to the ship captain—in both cases, before departure): eij d∆ ajpophdw'si to` n o”rkon, mh` a“llw" eja'sai diapera'n wJ" ta` aujta` fronei'n hJmi'n ajpanainome´ nou". 116 Alexiade 11.10, 11 (ed. Leib, 3:42–48); Heyd, Histoire du commerce, 1:190–91. 117 Alexiade 12.1.2, 14.3 (ed. Leib, 3:54, 154 ff ). The Genoese had also participated in the piratical attacks.

Angeliki E. Laiou [ 179 ] tions and reprisal arrangements. Thus there were new ways of dealing with the Westerners in commercial matters; and this went in tandem with dealings with the Crusaders, one set of arrangements reinforcing the other. Similarly, the bilateral agreements, confirmed by oaths, which we see most clearly in the Treaty of Adrianople, are a part of an evolving pattern of relationships with Western powers, primarily the crusading sovereigns and the maritime cities of Italy. Here, too, the Crusades are inscribed in a larger pattern of evolution, which includes the first bilateral state agreements (with Venice in 1186–87, with Pisa in the treaty of 1192, and with Genoa in 1155).118 Scholars have pointed out that the chrysobull issued by Isaac II in February 1187 differs in form from earlier privileges, for it is clearly bilateral, incorporating, on the one hand, the obligations of the Venetians, and on the other hand, and in exchange, the concessions given by the Byzantine emperor.119 This document preserves the traditional form in one important way: that the Venetians swear by oath to keep their promises, whereas the emperor’s word is his chrysobull. The Treaty of Adrianople, on the other hand, is a bilateral convention in this also: that both parties swear an oath through their representatives and an exchange of presents, bilateral in Choniates,120 unilateral in Ansbert.121 Of constitutional significance for Byzantium, this conjunction of bilateral arrangements also suggests a pattern of international agreements between states, whether crusading or mercantile, agreements that undoubtedly reinforced each other.

II. “International” Commercial Institutions In political terms, we can detect in the Byzantine Empire a certain dissolution of native institutions regarding aspects of trade and the exercise of the merchant’s profession. These then reappear in more general, Mediterranean, forms. In the course of the twelfth and thirteenth centuries, one may see the elaboration of similar general provisions and institutions throughout the eastern Mediterranean, in the trade relations between Byzantines, other Christians, and Muslims.122 In some instances, the first stages of

118

Lilie, Handel und Politik, 82. M. Pozza and G. Ravegnani, eds., I trattati con Bisanzio, 992–1198, Pacta veneta (Venice, 1993), p. 80, and doc. 8, pp. 90 ff. 120 Ed. van Dieten, 411. 121 Ansbert, 64, 66; Choniates (ed. van Dieten, 411) says the “emperor and the king renewed their oaths”; Frederick’s magnates gave the oath, and Isaac gave hostages. He then moves into the discussion of the oath taken in Constantinople. 122 S. D. Goitein (“Mediterranean Trade in the Eleventh Century: Some Facts and Problems,” in Studies in the Economic History of the Middle East, ed. M. A. Cook [London, 1970], 51–62) has argued that already in the 11th century, and before the Crusades, the “Mediterranean area gave the impression of a free-trade area.” Here I argue that the Crusades and the Crusader states were instrumental in the development and spread of institutions that facilitated Mediterranean trade. Many important topics, for example, the institutional impact of Western trading stations in Egypt on the development of Italian trading colonies in Constantinople, are necessarily left out of this discussion. I also do not discuss the Amalfitan and Venetian trading stations in Constantinople. These predate the crusading period, but neither their existence nor their evolution is critical to the topics elaborated below. 119

[ 180 ] Byzantine Trade development took place in the Byzantine Empire; in other cases in the Crusader states; in all cases the Crusader states played a pivotal role in this development. In general terms, it may be argued that states give up some prerogatives in the general trend toward the free and safe movement of men and merchandise—as was also the case with the freedom from, or the reduction of, the commercium. This is a very rich topic, with many aspects to it. I should like to examine one example only: the law of salvage, which is connected with the law of reprisals and seems to be developing together with laws regarding the disposition of the goods of merchants dying in a host country. The first two have to do not only with the normal risks run by merchants who carried out their business at the mercy of the elements, but also with piracy, increasingly a problem in the Mediterranean in the twelfth century and after.123 I will not deal with the evolving laws on reprisals here. Salvage In the case of salvage, one must begin with the Byzantine law, which derived from but became stricter than the Roman (i.e., Justinianic) law of salvage. There are two aspects to the question: who has the right to salvaged goods, and what penalties there are for those who pilfer them. As to the first point, Justinianic law, repeated in the Basilics, is quite clear: the goods belong to their owner; there is no time limitation on his rights; and what he can salvage he may keep.124 The Rhodian Sea Law adds some practical provisions that give a reward to those who help salvage the ship or the goods it carries, presumably in an effort to avoid looting: the people who help salvage a ship or parts of a ship are entitled to 1/5 of its value; if they find goods on land and return them, they get 1/10 of the value; and if they dive to salvage gold or silver or anything else, they are entitled to 1/3 or 1/2, depending on the depth.125 As for penalties for those who seize the goods of shipwrecks, while the Rhodian Sea Law does not mention any, in the Basilics there were very high financial penalties: within the first year, the pilferers had to restore to the owners the goods at their value in quadruple; after the first year, there was simple restitution.126 This was the Roman, Justinianic, provision. But the Procheiros Nomos stated that after the first year the goods should be restored at double their value.127 Custom, apparently, was even stricter than the law. A novel of Leo VI mentions the custom that mandated the death penalty for those who hid goods lost in shipwreck, for it was deemed a sin to deprive individuals of their wealth. 123 The connection has been made by H. Ahrweiler, “Course et piraterie dans la Me´diterrane´e orientale aux IV e`me–XV e`me sie`cles (empire byzantin),” in Course et piraterie: Etudes pre´sente´es a` la Commission Internationale d’Histoire Maritime a` l’occasion de son XVe colloque internationale pendant le XIVe Congre`s International des Sciences Historiques, San Francisco, 1975, 2 vols. (Paris, 1975), 1:10, 16, 17–19. Salvage is connected with jettison, on which see O. R. Constable, “The Problem of Jettison in Medieval Mediterranean Maritime Law,” Journal of Medieval History 20 (1994): 207–20. 124 Bas. 53.3.23 ⫽ D 41.2.21, §§ 1.22 jo. D 16.3.18; Bas. 53.3.15 ⫽ D 14.2.7. 125 Bas. 53, appendix, pt. 2, nos. 45, 46, 47. The Rhodian Sea Law has many other provisions regarding shipwreck, but they deal with the respective rights of the captain and the merchants aboard the ship. See, e.g., nos. 27, 29, 37, 40. 126 Bas. 60.20.1 ⫽ D 47.9.1; cf. Cod. 4.2.18. 127 Pr. 39.25 (⫽ Eisagoge 40.28), in Zepos, Jus 2:218, 361.

Angeliki E. Laiou [ 181 ] The emperor delivered himself of a somewhat rhetorical argument by analogy, to show that the death penalty was too heavy a punishment for a sin that is, after all, a crime of property: it is not just, he argues, to take an immaterial and immortal thing, one’s soul, in compensation for material and perishable things.128 Instead, the emperor restored the Roman penalty of a fine at four times the value of the objects in question, payable to the owner, although he gave no limitation of time, that is, the fine was not reduced after one year had elapsed. However, it would seem that all of this legislation applied only to Byzantine subjects, not to foreigners. What happened in the case of foreigners in the early and middle Byzantine period is not clear. The question of salvage came up in the Russo-Byzantine treaty of 911, where the Rus and the Byzantines engaged themselves to help each other’s ships during a tempest; if there was a shipwreck, and the goods were pilfered or the sailors/ merchants were murdered, the penalties for theft and murder would apply.129 In the twelfth century, it would seem that the goods aboard the ships of foreigners not covered by treaties and wrecked in Byzantine waters were considered to belong to the fisc: indirect evidence to that effect is to be found in the chrysobull for Pisa (1111) and in an incident ca. 1200.130 The Seljuks and the Latin rulers of Cyprus in the late twelfth and early thirteenth century apparently also considered that such goods belonged to the fisc. Quite early in Byzantine relations with the Westerners, clauses referring to shipwreck and salvage were incorporated in Imperial privileges. The treaties with Venice did not incorporate any such clauses, presumably because the Venetians, treated as Byzantine subjects, were covered by Byzantine law. Pisa was another matter. The chrysobull of 1111 was the first commercial privilege granted to an Italian power whose friendship was not a given, and with which, indeed, there had already been hostilities at sea. The Pisans requested and received a privilege regarding both piracy and shipwreck.131 It constitutes a key text in some respects. The emperor promised that if there was shipwreck in Byzantine waters, the Pisans could have without impediment their salvaged goods (the implication is that until then they could not). If some Byzantines helped salvage goods, the Pisans were allowed to keep those too, after making payment “according to the custom of the place, or according to any agreement they may have made.” While this was less than what had been agreed with the Rus, since no help was promised, it did lift the impediments on the recovery of shipwrecked goods. The Pisans retained the ownership of salvaged goods and had to pay those who helped recover them. But the emperor did not guarantee the safety of these goods. This is a highly simplified and attenuated form of the Byzantine law on salvage. We next find legislation on salvage in the privileges granted to Italian merchants in the variP. Noailles and A. Dain, Les Novelles de Le´on VI le Sage (Paris, 1944), novel 64. I. Sorlin, “Les traite´s de Byzance avec la Russie au Xe sie`cle,” Cahiers du monde russe et sovie´tique 3 (1961): 334, 357; in 944 the Rus engage themselves not to harm shipwrecked Byzantine ships; those who steal goods will be judged according to Byzantine law: ibid., 4 (1961): 450, 461. 130 Below, 184. 131 Mu¨ller, Documenti, no. , p. 44. The clause on piracy says that if a Pisan ship is seized in Byzantine waters and goods are taken by Byzantine subjects, the emperor will examine the matter and do the right thing, at the right time. 128 129

[ 182 ] Byzantine Trade ous Crusader states, where they appear in a simple but significantly stronger form. The first mention is in the privilege granted to the Venetians by the king of Jerusalem in 1123, as repayment for their help in the siege of Tyre: “Si vero aliquis Veneticorum naufragium passus fuerit, nullum de suis rebus patiatur dampnum.” Significantly, this is followed by the clause “si naufragio mortuus fuerit, suis heredibus aut aliis Veneticis res sue remanentes reddantur,” that is, the heirs of the shipwrecked person or other Venetians retained their rights on his property.132 Exactly the same clause was inserted in the privilege given by Baldwin II, king of Jerusalem (1118–31) to the Venetians; Renault of Chaˆtillon, regent of Antioch, and Bohemond III gave similar privileges in 1153 and 1167, respectively.133 There was an implicit guarantee given in these treaties, certainly greater than the mere permission to recover their goods that Alexios I Komnenos had given to the Pisans. Although no penalties are mentioned for those who seized goods, the state did promise their safety, a clause almost certain to eventually lead to reprisals. Some years later, the Byzantines also gave new guarantees to Italian merchants; in substance they are similar to the implicit guarantees given by the Crusader states, but in form they are more developed. In 1169 Manuel I was trying to bring Genoa into an anti-German alliance and was also negotiating a commercial treaty. He stated that in case of shipwreck, if goods were seized by someone, he, the emperor, charged himself with avenging the ill done and restoring the goods to the Genoese: “fiat de his vindicta a majestate mea et restauratio hujusmodi rerum” (“My majesty will give satisfaction and restore these goods”).134 This clause was incorporated in the treaty of 1192, which states: kai` eja` n ploiˆon Genoui¨tiko` n ajf∆ oiJoudh´ tino" to´ pou ejrco´ menon eij" RJ wmani´an kinduneu´ sh kai` sumbh' tina` tw'n ejn aujtv' pragma´ twn ajfaireqh'nai uJpo´ tinwn, i”na ge´ nhtai pro` " tau'ta th'" basilei´a" mou ejkdi´khsi" kai` ejpana´ swsi" tw'n toiou´ twn pragma´ twn.135 There seems to be a development that would bring the treatment of shipwrecked Italian ships closer to that of Byzantine subjects; in 1169 Manuel, unlike Alexios in 1111, guarantees the return of the merchandise of the Genoese, and if the penalty of the quadruple does not appear, the word vindicta in the text suggests punishment. Like the privileges granted earlier by the Crusader states, this is, I believe, partly an effort to preclude

132 The privilege also is the first one to state that if a Venetian dies in Tyre, his property goes to the Venetian authorities: William of Tyre, 12.25 (1:577 ff); cf. G. L. F. Tafel and G. Thomas, Urkunden zur a¨lteren Handelsund Staatsgeschichte der Republik Venedig, Fontes rerum Austriacarum, Diplomataria et acta 12–14, 3 vols. (Vienna, 1856–57), 1:79 ff. 133 Tafel and Thomas, Urkunden, 1:92, 134, 148–49. Cf. also the privilege of Renault of Chaˆtillon to Pisa in 1154, with return of goods to heirs: Mu¨ller, Documenti, no. ; privilege to Pisa by Bohemond III in 1170, with return of goods to heirs: Mu¨ller, Documenti, no. ; cf. Venice and the Seljuks, Tafel and Thomas, Urkunden, 2:484–86 (1254). On the early privileges granted to Italian merchants in crusader Syria and Palestine, cf. M. Balard, “Les re´publiques maritimes italiennes et le commerce en Syrie-Palestine (XIe–XIIIe sie`cles),” Anuario de estudios medievales 24 (1994): 313–48, and J. Riley-Smith, “Government in Latin Syria and the Commercial Privileges of Foreign Merchants,” in Relations between East and West in the Middle Ages, ed. D. Baker (Edinburgh, 1973), 109–32. 134 Bertolotto, Nuova serie, 351; Balard, La Romanie ge´noise, 1:28 ff. 135 Miklosich and Mu¨ller, 3:36.

Angeliki E. Laiou [ 183 ] reprisals. But one may also suggest that the Italians, having received a greater guarantee from the Crusader states, returned to Byzantium asking here, too, for better guarantees. In the 1180s, Andronikos I passed a measure of exceptional harshness toward those who pilfered shipwrecks. Choniates, who reports the measure, informs us that from time to time Byzantine emperors passed laws on salvage, although, after the Rhodian Sea Law, the only extant legislation, to my knowledge, is the Macedonian one, including Leo VI’s novel. But he also says that the laws were not applied and complains that there was a long-standing custom, found only among the Byzantines, that ships that foundered or were cast ashore were not helped by the inhabitants; rather, the natives seized anything that the sea did not.136 Andronikos’ measure is not transmitted in the form of a novel, but is rather in that of an entole, an order addressed to his relatives and the imperial officials.137 The emperor begins by saying that he believes his predecessors were incapable of exacting obedience to their laws because they only pretended to wish to put a stop to this evil; if they had really wanted to, they would have punished it with the death penalty. He himself wants to stop practices that harm the commonweal, especially the custom of seizing the cargo of shipwrecks and sometimes even demolishing the ship. The penalty he decrees is harsh: those who do not heed his orders will be hanged from the ship’s mast or will be impaled on the coast, so that they will be visible from afar and become a lesson to those who see them. Andronikos holds responsible not only those who physically plunder the ships, but also the imperial officials who govern the area, as well as the landlords whose tenants engage in such activities. Given Andronikos’ reputation as a man of his word, Choniates tells us that the officials wrote to those who ruled the provinces in their names, and the landlords to their agents, not to harm the shipwrecks. As a result, nothing was stolen, no ships were destroyed, no anchors taken away, and the people of the coastlands as well as the officials helped the shipwrecked sailors and merchants to recover all their possessions.138 Choniates relates this imperial order at some length, using it as the single detailed illustration of his general statement that Andronikos tried to ease the lot of the poorer people, partly by insisting on fair and predictable taxation (no multiplication of dues), and partly by stopping the sale of offices and by appointing officials who would not use their office to enrich themselves. It is thus presented as part of the populist policy of this otherwise hateful (at least to Choniates) emperor. But it is, in fact, much more than that. It is a stage in the development of a law of the sea, established in a harsh and inelegant way, to be sure, that aims to protect merchants and to minimize claims. Although noth136 Choniates, ed. van Dieten, 326: e“qou" . . . ajlogwta´ tou. Interestingly, Hugh I of Lusignan, in his treaty with the sultan of Konya (1216), also mentions that such an unjust custom prevailed in Cyprus (kata` th` n ejpikrath´ sasan a“dikon sunh´ qeian)—from Byzantine days? See S. Lambros, “ JH JEllhnikh` wJ" ejpi´shmo" glw'ssa tw'n Soulta´ nwn,” Ne´ o" JEll. 5 (1908): 49. A much later text, from the 1320s, gives details about people who cause shipwrecks in order to steal the goods aboard ship—a simple form of piracy: Theodoulos Monachos, Peri` basilei´a", PG 145:481–84; cf. Ahrweiler, “Course et piraterie,” 17 ff. 137 Do¨lger, Regesten, no. 1566, s.a. 138 Choniates, ed. van Dieten, 326–39; Skoutareiotes in K. Sathas, Mesaiwnikh` Biblioqh´ kh, vol. 7 (Paris, 1894), 350; cf. A. Laiou, “Byzantine Traders and Seafarers,” in The Greeks and the Sea, ed. S. Vryonis Jr. (New Rochelle, N.Y., 1993), 89–90.

[ 184 ] Byzantine Trade ing in the text suggests that this was prompted by foreign merchants or was to apply to them, the international implications are clear. Even Choniates’ opening statement, that this was “an extremely irrational custom that is practiced only among the Romans” suggests that international concerns were present. Andronikos’ measure was made necessary both because of the increased volume of maritime trade and because the increased presence of foreign merchants created a situation where the emperor was responding to pressures for safety in sailing, pressures coming not only from his own subjects but also from foreigners. The measure carries imperial orders one step further: people are to help salvage goods. And it is general enough to be applicable to both Byzantine and foreign shipping, despite Andronikos’ well-known hostility to the Latins. Indeed, a number of incidents in the Black Sea at the turn of the century show both the need for a proper law of salvage and the fact that the powers of the area still considered that, in the absence of treaties to the contrary, salvaged goods belonged to the fisc. One incident implicated merchants both Byzantine and Turkish. Around 1200,139 Alexios III sent to the Black Sea a certain Constantine Frangopoulos, ostensibly to examine the cargo of a ship sailing from Phasis and shipwrecked near Kerasous—a clear indication that the fisc was claiming goods salvaged from foreign ships.140 The man, perhaps on imperial orders, perhaps (as Alexios later claimed) not, plundered merchantmen sailing to and from Constantinople, near Samsous, it seems,141 and killed some of the merchants. The loot—or part of it—came to Alexios who added it to his treasury, oblivious to the complaints of the merchants. But some of them were subjects of the sultan of Konya. Since this happened at a time when a treaty (or the renewal of a treaty)142 was being discussed, the sultan tied the treaty negotiations with the reparation demands of the merchants. Eventually, the sultan got, along with other things, “50 pounds of silver” as damages for his merchants.143 The other incident happened in 1223 and involved Russian merchants fleeing Sudak and shipwrecked near Sinope. The Turks claimed the right to keep salvaged goods, “according to a custom with regard to a place where there existed no concessions.”144 The laws on salvage, which began in the Byzantine Empire, developed significantly in the Crusader states and returned to Byzantium to develop further. They spread all over the eastern Mediterranean in the thirteenth century, in one form or another, depending on the model that was being followed. The treaties between the sultan of Konya Kaikaus I and Hugh I of Lusignan of Cyprus in 1216 contain clauses on both piracy and salvage, as does the Byzantine-Genoese treaty 139 Do¨lger, Regesten, no. 1658 (1201); cf. C. Cahen, La Turquie pre´-ottomane, Varia turcica 7 (Istanbul-Paris, 1988), 64, 122–23. 140 Similar rights were claimed by Kaikaus in 1216; cf. below, 184–85. 141 Samsous had fallen to the Turks shortly before 1194: Cahen, La Turquie pre´-ottomane, 64. 142 There were earlier treaties with Kaikaus I of Konya, which Alexios broke (perhaps in 1198), again by attacking the properties (merchandise and animals—uJpozu´ gia) of Greek and Turkish merchants, subjects of the sultan: Choniates, ed. van Dieten, 493–94. 143 Choniates, ed. van Dieten, 529: eij" ajpo´ doma w»n oiJ e“mporoi ajfhre´ qhsan. Do¨lger, Regesten, no. 1658 (the sultan was Rukn al-Din). Reparations for goods seized by the fisc were also given to the Venetians, after the seizure of their goods by Manuel I in 1171—again, a parallel development. 144 Cahen, La Turquie pre´-ottomane, 125. The quotation is from Cahen.

Angeliki E. Laiou [ 185 ] of 1192. In both states it seems that salvaged goods of foreigners were seized; the treaties put an end to that, promising the return of the surviving men and their merchandise.145 Hugh I promised that if pirates attack and pillage a ship and its merchandise and go to Cyprus, the spoils will be returned to the rightful owners, if the king captures the pirates. And in cases of shipwreck, “the surviving men as well as the salvaged goods shall be kept safe and restored, and not seized as per the prevailing unjust custom”; similar provisions, although somewhat differently worded, were included in the sultan’s letter to Hugh, which repeats and confirms the terms of the treaty.146 A few years later, in 1220, Alaeddin Kaikobad, sultan of Konya, signed a treaty with Jacopo Tiepolo, podesta` of the Venetians in Constantinople. A lengthy clause addresses the issue of salvage and, in vaguer terms, of piracy (or rather, of Venetian ships being chased by ships of other powers). On the issue of shipwreck, the sultan promises not only that the goods of Venetians would be restored, but also that his subjects would help the Venetians recover them—a clause that appears only in the Russian-Byzantine treaty of 911 and implicitly in Andronikos’ measure.147 The same privilege was given to the Venetians by the lord of Rhodes, the Caesar Leo Gabalas, in 1234: “et omnes a gente mea habebunt subsidium et favorem.”148 The privilege granted by Theodore I Laskaris to the Venetians (Jacopo Tiepolo again) in 1219 contains two clauses that make reciprocal arrangements for ships and merchants of the two powers, to return their property to the men who survive the shipwreck. The Venetians are given the added privilege, contained already in the Byzantine privilege of 1198, but appearing for the first time in the Pactum Warmundi in 1123, that the goods of Venetians who died in the Empire of Nicaea would revert to their heirs.149 Italian treaties with Egypt, after the middle of the thirteenth century, also incorporate clauses promising safety for men and goods in case of shipwreck.150 What all of this indicates is that a law of the sea was developing in the twelfth century and into the thirteenth, which regulated certain important matters and which eventually applied to all merchants active in the area and to the states whose subjects they were.

145 Lambros, “ JH JEllhnikh´ ,” 48–50; A. Savvides, Byzantium in the Near East: Its Relations with the Seljuk Sultanate of Rum in Asia Minor, the Armenians of Cilicia and the Mongols, .. c. 1192–1237, Byzantina keimena kai meletai 17 (Thessaloniki, 1981), 141–42; C. Cahen, “Le commerce anatolien au de´but du XIIIe sie`cle,” in idem, Turcobyzantina et Oriens Christianus (London, 1974), no. , 93. 146 Lambros, “ JH JEllhnikh´ ,” 52. 147 Tafel and Thomas, Urkunden, 2:223; the reciprocal privilege to the sultan’s men is found on pp. 223–34; it includes a clause that, if the man dies, his property will be given to his heirs, which harks back to the Venetian privileges in the Pactum Warmundi and the Byzantine treaty of 1198 (Pozza and Ravegnani, I trattati, doc. 11, p. 136). The sultan also promises that any Crusaders found on Venetian ships will not be imprisoned but will be freed. There do not seem to be extant privileges to the Genoese: Cahen, “Le commerce anatolien,” 98–99. 148 Tafel and Thomas, Urkunden, 2:321–22; cf. M. Angold, A Byzantine Government in Exile: Government and Society under the Laskarids of Nicaea, 1204–1261 (London, 1975), 114. Again, no privileges to the Genoese seem to have been issued. 149 Tafel and Thomas, Urkunden, 2:206–7. 150 See, for example, the treaty with Venice, in 1254, Tafel and Thomas, Urkunden, 2:484: “et omnes Veneti sint salui et securi in personis et hauere et toto suo navigio.” Cf. the treaty of 1290 with Genoa, and of 1281 with Byzantium (below, 189–91).

[ 186 ] Byzantine Trade Italian merchants, Byzantines, and Turks were all covered by the same protections, at least in the law of salvage. Crusader states, the sultanate of Konya, the Byzantine Empire, and then the Empire of Nicaea, and Egypt, incorporated in their agreements with each other the same or similar arrangements. There are intriguing connecting lines between some of these agreements: for example, the occasional appearance in this connection of the guarantee of the inheritance of merchants dying abroad; the similarities in phraseology between the treaties of Hugh of Cyprus and Kaikaus I on the one hand, and Manuel I’s agreements with the Genoese on the other. Most important, not in any intrinsic way, but for our topic, is the similar concern expressed by Andronikos I and Hugh I of Cyprus regarding the “unjust custom” (ejpikrath´ sasan a“dikon sunh´ qeian) (Hugh I) or the irrational custom (e“qou" ajlogwta´ tou) (Andronikos I) of plundering shipwrecks. Both men opposed this custom, and treaties incorporated clauses that forbade the seizure of goods and sometimes pledged help for their recovery. There was, thus, an effort throughout the eastern Mediterranean to improve the conditions of trade and travel, an effort that almost certainly originated with the concerns of Western merchants and Crusaders, who plied these seas, but that led to measures that became generally adopted. Political considerations connected with the politics of crusading and the Crusader states were important indeed for the adoption and spread of such institutions, but the motive force behind them was commercial. The old Byzantine laws regulating these issues became simplified as their substance was generally adopted. Gone is the distinction between the first year of the seizure of goods and subsequent years; gone also the obligation to return the goods in quadruple for the first year; these elements had disappeared already at the time of Andronikos I, but equally, his extremely severe measures were unique to him. Complex arrangements would have been impossible to implement in this mobile world, hence the simplicity. The simplification is evident as early as the Pisan privilege of 1111, which, along with the Pactum Warmundi, is the basic document. The law of the sea and mercantile law were also constantly evolving, as may be seen from the provisions regarding the goods of men dying outside their own country, whether with a testament or intestate. While this issue cannot be examined here, it may be noted that the process of its development is very similar to that of the law of salvage. The principle of the devolution of property to a man’s heirs or his fellow nationals, rather than to the fisc of the country in which he died, a very important principle for merchants, appears first in the privilege granted to the Genoese by Baldwin I of Jerusalem (1105)151 and the Pactum Warmundi in 1123. It then took the form of the return of the goods to Venetian authorities, but by the middle of the twelfth century, in the Crusader states, the right devolved to the heirs.152 In the Byzantine Empire such certainly was not yet the practice in 1165–66, when the considerable property of the Pisan merchant Signoretto, who was also a burgensis, was claimed by the fisc.153 It appears for the first time in the treaty of 1198 with Venice. There can be no doubt that this principle devel-

151

C. Imperiale di Sant’Angelo, ed., Codice diplomatico della repubblica di Genova, 3 vols., Fonti per la storia d’Italia 77, 79, 89 (Rome, 1936–42), 1: no. 15, pp. 20 ff. 152 See above, note 129. 153 Mu¨ller, Documenti, no. , pp. 11–13.

Angeliki E. Laiou [ 187 ] oped in the Crusader states, traveled to Byzantium some ninety years later, and became generalized in the eastern Mediterranean, including Egypt, in the course of the thirteenth century.

III. Trading with Muslims Regarding Byzantine trade with the Muslims during the period of the Crusades, the available information is sporadic and can perhaps best be understood in the light of what has already been said regarding the institutional and structural changes in the conditions of trade in the Byzantine Empire and in the eastern Mediterranean generally.154 Quite as was the case with Western merchants, the ideology of Holy War did not interrupt trade between the Byzantines and the Muslims; the opposite in fact was the case with Western merchants,155 whose presence and business in Egypt increased in the course of the twelfth and thirteenth centuries, with a short interruption for a few years after 1187. Byzantine trade with Egypt continued in the course of the twelfth century, as we know from the testimony of Benjamin of Tudela,156 as well as from a few narrative and documentary sources such as the affair regarding the piratical activities of Guglielmo Grasso and his fellow corsairs. However, S. Goitein’s statement that the Geniza documents show a shift of importance, suggesting an increased presence of Western merchants relative to the Byzantines, must stand. Some other observations may also be made. The relations of the Byzantines with Muslim powers had always had a strong political and military component, which overwhelmed private and commercial relations, existent but difficult to discern in the sources. The statement has been made by M. Canard with regard to the tenth century.157 It remains true in the twelfth century and then again in the second half of the thirteenth and the fourteenth century, for the Byzantines not only had political and ceremonial relations with the Muslim rulers of Egypt, but also considered themselves and were considered by others as having a special role to play in the protection of the Christian populations of the Egyptian state. As a result, virtually all of the commercial contacts that are visible to us from Byzantine or Muslim sources appear in the context of political relations. It is primarily the Geniza documents that show an ongoing commercial activity of the Byzantines in the eleventh and twelfth centuries that is quite independent of official exchanges of embassies. 154 On the topic generally, cf. S. Labib, “Egyptian Commercial Policy in the Middle Ages,” in Cook, Studies ¨ gyptens im Spa¨tin the Economic History of the Middle East (as above, note 122), 63–77, and idem, Handelsgeschichte A mittelalter, 1171–1517, Vierteljahrschrift fu¨r Sozial- und Wirtschaftsgeschichte Beihefte 46 (Wiesbaden, 1965). 155 D. Jacoby, “Les Italiens en E´gypte aux XIIe et XIIIe sie`cles: Du comptoir a` la colonie?” in Coloniser au Moyen Aˆge, ed. M. Balard and A. Ducellier (Paris, 1995), 78 ff. 156 Benjamin bar Jonah of Tudela, The Itinerary of Benjamin of Tudela, trans. M. Adler, A. Asher, ed. M. Singer (Malibu, Calif., 1993), 70. 157 M. Canard, “Deux e´pisodes des relations diplomatiques arabo-byzantines au Xe sie`cle,” in idem, Byzance et les musulmans du Proche Orient (London, 1973), no. , p. 51. On trade see, among others, P. von Sivers, “Taxes and Trade in the Arab Thugur, 750–962/133–351,” Journal of the Economic and Social History of the Orient 25 (1982): 71–99.

[ 188 ] Byzantine Trade An interesting mention of Byzantine merchants in the early twelfth century shows them having quite a prominent position in Alexandria. The information comes from Orderic Vitalis, who notes, in 1102, the presence in Cairo of merchants from Constantinople, multimodis mercimoniis. He says that, according to the laws of the people (or of nations—leges gentium), they paid the required taxes and stayed there for some time. They were very rich, and during their stay they visited churches, the poor Christians, and the Latin captives who had been brought to Cairo after the fall of Ramleh. They are credited with persuading Alexios I to blackmail the sultan into releasing Arpin of Bourges, under the threat that otherwise he, Alexios, would “have all the Egyptian factors and mercenaries in the whole Empire of Constantinople arrested” (“omnes Babilonicos institores et stipendiarios per totum imperium Constantinopolitanum comprehendi iuberet”).158 According to this account, Byzantine merchants seem to have been visiting Cairo in a routine fashion, staying for certain periods of time; they paid commercial duties, and their movements, at least within the city, were not restricted. Anna Komnene mentions that the emperor heard of the imprisonment of many famous Crusader knights and sent “a certain Vardales” and gave him letters to the sultan along with a great deal of money.159 The phrase tina` Bardalh'n suggests that this was not a prominent courtier and may, indeed, have been a merchant. In the course of the twelfth century, trade with Egypt seems to have been connected, to some extent, with political relations intimately related to the Crusades: the clearest example is the combined ambassadorial and trade mission to Saladin in 1192, which had the unhappy end that we have mentioned.160 Its main purpose had been diplomatic, while trade (and gift exchange) played an important but secondary role. There were also changes in the commercial relations between Byzantines and Muslims during the period in question. The Fourth Crusade reoriented the trade of the Byzantines (the Empire of Nicaea) with the Muslims. Relations with Egypt were interrupted. Very few embassies (I count two) are attested from the Empire of Nicaea to Egypt, and their object was either ceremonial or unknown to us.161 There are no indications anywhere of continuing trade relations with Egypt; the inward-looking policy of the Nicene emperors and the sumptuary laws passed by John III Vatatzes162 would suggest that the absence of information is not a matter of chance but reflects the reality of very limited relations. On the other hand, political contacts with the sultanate of Konya were frequent, and, although there is again very limited information about trade relations, one may assume that they existed; their extent is impossible to recover. In the late twelfth century, there had been an active trade between the sultanate of Rum and Constantinople, through the Black Sea.163 But we have little information regarding Black Sea commerce between 1204 and 1261. We find here Venetians and Turks, with the 158 Orderic Vitalis, 5.351–52. Arpin may have been the prior of Charite´-sur-Loire: ibid., p. 350 n. 1. Albert of Aix attributes Arpin’s release to a Byzantine knight: RHC, HOcc 4:594. 159 Alexiade 11.7.3 (ed. Leib, 3:33); cf. Do¨lger, Regesten, no. 1216. 160 Ibn al-Athı¯r also mentions Greeks, along with “Franks,” in Acre in 1187: RHC, HOrient, vol. 1 (Paris, 1872), 689. 161 Do¨lger, Regesten, nos. 1713a (1226) and 1763a (after 13 Aug. 1238). 162 Do¨lger, Regesten, no. 1777 (early 1243). 163 See above, 184, during the reigns of Andronikos I and Alexios III.

Angeliki E. Laiou [ 189 ] Empire of Trebizond fighting to retain some control. Trade with Constantinople would have bypassed the Nicene Empire. The sultans of Rum had developed a very significant trade network including central Anatolia and the ports of the southern coasts and the Black Sea coast, cemented by the capture of Sinope (1214) and Attaleia (1207) and the establishment of a Seljukid protectorate in Sudak in 1225.164 How much this commercial flourishing involved the Empire of Nicaea is not certain; it has been suggested that commercial relations between Nicaea and Konya increased after the Mongol invasion of Asia Minor in 1243.165 What is certain is that at least at one point, at a time of famine in the Seljuk state, there were massive exports of grain, sheep, goats, oxen, and other foodstuffs from Nicaea to the Seljuks; according to Gregoras, all the wealth of the Turks, in gold, silver, precious textiles, and other luxury objects, was drained into the coffers of both the state and private individuals.166 The export of wheat, forbidden in Byzantium for centuries,167 seems to have been entirely free in this instance. With the recovery of Constantinople, there was, once again, a political and to some degree economic reorientation toward Egypt. Of primary importance to the Mamluks, and also important for the nexus of relationships between the Byzantines and the Muslims, was the slave trade, which brought to Egypt slaves for its armies from the Crimea through Constantinople. Almost immediately after the recovery of the capital, Emperor Michael VIII and the Egyptian sultan Baibars exchanged embassies regarding the importation of slaves from the Black Sea.168 Relations with Egypt went through ups and downs after 1264–65, but in 1281 there was, once again, a treaty between Michael VIII and the new sultan, Qalawun. This was the time when Michael VIII was engaged in fullscale hostilities with Charles of Anjou, and the original plan of the treaty incorporated a clause that would have guaranteed that Michael VIII would have denied free passage to anyone who wanted to go through his state to attack the Egyptians.169 The treaty of 1281, which has been published and analyzed by Canard, embodies some elements important to our topic. It should be read in conjunction with the treaty signed in 1290 between Qalawun and the Genoese,170 and against the background of 164

Cahen, “Le commerce anatolien,” 94–95. Angold, A Byzantine Government in Exile, 115–16. 166 Nikephoros Gregoras, Byzantina historia, ed. L. Schopen and I. Bekker, 3 vols. (Bonn, 1829–55), 1:42–43. 167 Cf. Leo VI’s novel 63, Noailles and Dain, pp. 231–33 (but it does not mention grain specifically; it deals with punishment of export of kekolymena generally). 168 George Pachymeres, Relations historiques, ed. A. Failler (Paris, 1984), 234; cf. 243, relations with Nogai. Gregoras (Bonn ed.), 1:101–2, makes the explicit connection with the Egyptian need to sail to the northern coast of the Black Sea once a year, to procure slaves. Cf. Do¨lger, Regesten, nos. 1902–4 (Nov. 1261–Nov. 1262); M. Canard, “Le traite´ de 1281 entre Michel Pale´ologue et le Sultan Qala’un,” Byzantion 10 (1935): 669–80. 169 Canard, “Le traite´,” 679. Canard thinks that this clause was dropped because Charles of Anjou reached an agreement with Qalawun. It may be noted that Michael VIII, despite a treaty with Baibars in 1261, renewed in 1267, had undertaken the obligation of joining a crusade against Egypt in 1274–76: D. J. Geanakopolos, Emperor Michael Palaeologus and the West: A Study in Byzantine-Latin Relations (Cambridge, Mass., 1959), 285–94. 170 Latin text published by Antoine Isaac Sylvestre de Sacy, Notices et extraits des manuscrits de la Bibliothe`que du Roi (Paris, 1827), 33–52 (p. 42); 33, 40: reciprocal vs. simple guarantee of shipwrecked goods; p. 38 on the goods of merchants dying abroad. The Arabic text is translated by P. M. Holt, “Qalawun’s Treaty with Genoa in 1290,” Der Islam 57 (1980): 101–8. 165

[ 190 ] Byzantine Trade other treaties signed between the Egyptians and the maritime cities of Italy in the course of the thirteenth century. Such a reading reveals the following. The treaty of 1281 includes reciprocal clauses guaranteeing the free access of the merchants of both countries to the markets of both countries, against payment of the appropriate dues. Such a clause is usual in all treaties of the period.171 Other clauses, which had been general in the Mediterranean since the thirteenth century, include the statement that neither state would take reprisals against the merchants of the other for piratical activities undertaken in its territorial waters or by people who claimed to be its subjects; reprisals could only be taken against the individuals guilty of the act of piracy.172 A clause that is unique to this document has to do with the right of the Byzantines to buy off Christian slaves and the right of freed Christian slaves to sail to the Byzantine Empire. The treaty allows the Egyptians to export Byzantine wheat: an interesting clause because the Palaiologan emperors tried mightily to restrict the export of Byzantine wheat by Western merchants: in the treaty of 1265 with Venice, export was permitted only after its price in Constantinople was under 50 hyperpyra per kentenarion, which rose to 100 hyperpyra in the treaty of 1277, while treaties with the Genoese allowed export only after specific imperial permission.173 The reciprocal privilege allows Byzantine merchants to buy thoroughbred horses in the Egyptian possessions. The principle of free access to markets and merchandise extends even to the controlled wheat trade. On the other hand, this treaty lacks the very detailed provisions regarding residence, the payment of duties, exemption from forced purchases, relations with customs officials, relations with the fisc (what happens to the property of foreign merchants dying in the state), and relations between individuals (resolution of private disputes and debts) that appear in the treaties between the Italian maritime states and Egypt or Byzantium. What this suggests is that the commercial relations between Egypt and the Byzantines, which we know existed,174 were nowhere nearly as extensive as those of the Italian merchants and these two states. Thus the same general principles of trade existed between the Byzantines, the Christian merchants, and the Muslims (and similarly between the Egyptians and Byzantine and Western merchants), but the realities of trade were that Byzantine relations with Egypt were much more heavily political than economic/commercial in nature. What distinguishes Michael VIII’s treaty with Qalawun from most others is the insistence of the Egyptians to have freedom of access to the Black Sea area (“le pays de 171 Cf., e.g., Tafel and Thomas, Urkunden, 3:68 (treaty between Michael VIII and Venice, 1265); cf. ibid., 141–43, 146 (treaty of 1277). Similar though not identical provisions in the Arab text of the treaty between Genoa and Qalawun in 1290: Holt, “Qalawun’s Treaty,” 106. 172 Canard, “Le traite´,” 677–80. There is, however, nothing on shipwreck, which, on the contrary, appears in both Venetian-Byzantine treaties and treaties between the Egyptians and Western maritime cities. See, for example, Tafel and Thomas, Urkunden, 3:73, 144 (with Byzantium, 1265, 1277), and 2:338–39, 484–85 (Venetian-Egyptian treaties of 1238 and 1254). Cf. Holt, “Qalawun’s Treaty,” 102. 173 A. Laiou, Constantinople and the Latins: The Foreign Policy of Andronicus II, 1282–1328 (Cambridge, Mass., 1972), 65, 73; eadem, “The Byzantine Economy in the Mediterranean Trade System,” in eadem, Gender, Society and Economic Life in Byzantium (Hampshire, 1992), no. , 213. On the grain trade see also J. Chrysostomides, “Venetian Commercial Privileges under the Palaeologi,” StVen 12 (1970): 267–356. 174 See, for example, Bertolotto, Nuova serie, 521 (ca. 1290), and Theodore Metochites, AASS, Nov. 4: 672.

Angeliki E. Laiou [ 191 ] Sudaq”), for the specific reason of buying slaves, with the proviso that they not be Christian slaves. This was, in fact, the major interest of the Egyptians in the Byzantine Empire and the principal reason why this treaty was concluded. And it was intimately connected with the politics surrounding the crusading movement as it developed after the Fourth Crusade, that is, in this specific case, with the plans of Charles of Anjou and the fate of the last remaining outposts of the kingdom of Acre. In their formal aspects, that is, in the matter of treaties and official agreements, Byzantine trade relations with Egypt remained connected with political concerns and with the concerns of the Byzantine emperors for the non-Latin Christians of the East. This was so in the late thirteenth century and remained so in 1348/49, in the negotiations of John VI Kantakouzenos with Malik Nasir Hasan, in which the security of Byzantine merchants in Egypt is embedded in discussions regarding the Christians of the Mamluk empire.175 While the formal relations were embedded in politics, the general conditions and mechanisms were similar to those between Byzantine and Western Christians, following the general lines elaborated in the twelfth century. After 1291 and the fall of Acre, we know that there were repeated calls in the West for a boycott of trade with Egypt. This did not involve Byzantium; on the contrary, some of the plans for a boycott of Egyptian trade included, as a side effect of the eventual rearrangements in the eastern Mediterranean, the reconquest of Byzantium by the Crusaders.176 Interestingly, however, there were, in the early fourteenth century, some vague and some not-so-vague mutterings in Byzantium against the continuation of good relations with Egypt: vague in the comments of Theodore Metochites, who, in his Oration to the neo-martyr St. Michael, somewhat sheepishly and defensively explains that the Byzantine emperor was in the habit of sending frequent embassies and friendly messages to the “impious ruler . . . not because of some need,” since they were not contiguous neighbors and thus did not have the close contact of contiguous neighbors, but because the emperor always did everything in his power for the protection and well-being of Christians in the Mamluk possessions.177 In a much more explicit manner, George Pachymeres criticized in no uncertain terms the policy of Michael VIII that “opened the way” to the Egyptians toward the “Scythians,” that is, the Cumans of the Black Sea. He 175 Ioannis Cantacuzeni eximperatoris historiarum libri quattuor, ed. L. Schopen, 3 vols. (Bonn, 1828–29), 3:90– 104; cf. M. Canard, “Une lettre du Sultan Malik Naˆsir Hasan a` Jean VI Cantacuze`ne (760/1349),” in idem, Byzance et les musulmans du Proche Orient (as above, note 157), no. , 29–52. 176 This includes the early plans of Marino Sanudo, ca. 1306–12: A. Laiou, “Marino Sanudo Torsello, Byzantium and the Turks: The Background to the Anti-Turkish League of 1332–1334,” Speculum 45 (1970): 374–92. On the crusading plans of this period, see S. Schein, Fideles Crucis: The Papacy, the West, and the Recovery of the Holy Land (1274–1314) (Oxford, 1991). 177 AASS, Nov. 4: 672–73. Do¨lger thinks the embassy in question may have been the one dated 1311–13: Regesten, no. 2326. This text also provides important information regarding Byzantine trade relations with Egypt. According to Metochites, Alexandria is always full of people—among them, Byzantines—who go there for trade and other reasons (676). At the time of the martyrdom of St. Michael, there were, in the city, both Byzantine ambassadors and Byzantine merchants (673, 676). The ambassadors seem to have sailed on a Byzantine ship, for Michael, a captive in Alexandria, tried to escape on this ship by disguising himself as one of the passengers or crew who were to sail back to Constantinople (673). This incidental information suggests that even in the early 14th century there were Byzantine ships sailing these seas and that the Italian control of shipping in the eastern Mediterranean was not absolute.

[ 192 ] Byzantine Trade connected that with the Egyptian offensive against the last remaining Crusader possessions in Palestine (1268–91); he lamented the fall of these possessions; and he ended this narrative with the statement: “This is the profit to Christendom wrought by our imprudence and our ungoverned impulsive actions and appetites.”178 Pachymeres, clearly, would much rather have seen a Christian economic and political alliance that would have kept the Egyptians out: something like the boycott advocated in Western Europe. Thus, after a very long time, the Byzantines themselves connected the Crusades and the fate of the Crusader states with the trading policies of the Byzantines and the Muslims. To the modern scholar, the connections between the crusading movement and the Byzantine commercial relations with Christians and Muslims appear much more complex. The complexity is partly the result of the fact that with the Crusades the entire political scene of the Middle East changed; and with the expansion of Italian merchants the economics of the area changed as well. These are long-term phenomena, and there is a broad and long-term connection between political and economic affairs. There are also short-term and immediate connections between specific crusading efforts and matters affecting commerce, for example the grant of specific charters of privilege. I have tried to focus on the phenomena that lie between the long term and the short term and to point out some of the structural and institutional developments that may be considered to have occurred because of the Crusades wholly or, more often, in part. The development of exchange mechanisms and mechanisms of negotiation in Byzantium were connected both to specific events, the Second and Third Crusades, and to the evolving presence of Western merchants. The terms of trade with Christians and Muslims, in the exemplar case of the law of salvage, have been seen to have evolved over the entire eastern Mediterranean, in response to both political events and economic necessities, the evolution starting in Byzantium and making the rounds of the Crusader states and Egypt, becoming reinforced on the way. In the case of Byzantine trade relations with the Muslims, the Crusades appear to have imposed two successive reorientations—and political and economic affairs were closely intertwined. In sum, in my view, the Crusades and the existence of the Crusader states played a much greater role than modern scholars tend to allow, not only in the general patterns of trade in the eastern Mediterranean but also in the conditions and mechanisms of trade between the Byzantines and the Muslims, and especially between the Byzantines and the Western Christians. Harvard University and Academy of Athens

178 Ed. Failler, 241, 243; cf. Georgii Pachymeris de Michaele et Andronico Palaeologis libri tredecem, ed. I. Bekker, 2 vols. (Bonn, 1835), 2:87, 456–58. Cf. A. Laiou, “On Political Geography: The Black Sea of Pachymeres,” in The Making of Byzantine History, ed. R. Beaton and C. Roueche´ (London, 1993), 110–11, 119.

Appendix by Ce´cile Morrisson

1. Comparison of the Known Exchange Rates from the First Crusade (comments on pp. 169–70) The exchange rate deduced from the Gesta Francorum (1 hyperpyron ⫽ 15 soldi ⫽ 180 deniers) is evidently one obtaining in an exceptional context (famine and the presence of a huge army). The intended coins are obviously, on the one hand, the hyperpyron, although the Gesta provides, to my knowledge, the only occurrence of the form purpurati.1 The Crusaders may have coined this name by identifying hyperpyra with “imperial” coins, as purpuratus had primarily this meaning. On the other hand, which deniers the author of the Gesta may have had in mind is not clear. His South Italian origin could have made him allude to the deniers from Pavia, Lucca, Venice, and Rouen,2 which are attested there by textual and archaeological evidence in the late eleventh century. Or he might simply have alluded to the various French and Italian deniers that the Crusaders brought with them. Little is known of the silver fineness of these,3 and we have to be content with mentioning figures for royal French deniers of the late eleventh century. The deniers Parisis

The tentative estimates of coins’ real values given here are, it must be stressed, based on average values of weights and metal contents and should not be taken as absolute values but as indications of order of magnitude. The reliability of the figures is even less for Western denominations, namely deniers, than it is for Byzantine coins because of the great variety of issuing authorities, the scattered nature of the data, the virtual absence in the 12th century of surviving monetary specifications, and the small number of available analyses. I am most grateful to Marc Bompaire for advice on Western European numismatics and for communication of his and Maria Guerra’s forthcoming article (M. Bompaire and M. Guerra, “Analyse de monnaies franc¸aises du XIe sie`cle: Le proble`me du zinc,” in Actes du XIIe Congre`s international de numismatique, Berlin, 1997, ed. B. Kluge). The results obtained for silver fineness are not commented upon there but show great variety in time and space. 1 The only reference in Du Cange, Glossarium mediae et infimae Latinitatis (Paris, 1937–38), is to the Gesta and its various versions or compilations. This source and coin name are not mentioned by B. Koutava-Delivoria, “Les chichata, les protocharaga et la re´forme mone´taire d’Alexis I Comne`ne,” RBN 141 (1995): 13–36. 2 See L. Travaini, La monetazione nell’Italia normanna (Rome, 1995), 362–94, esp. 394; J.-M. Martin, “Le monete d’argento nell’Italia meridionale del secolo XII secondo i documenti di archivio,” Bollettino di numismatica 6–7 ( Jan.–Dec. 1986): 85–96, at 86; F. Dumas and J. Pilet-Lemie`re, “La monnaie normande, Xe–XIIe sie`cle. Le point de la recherche en 1987,” in Les Mondes Normands (VIIIe–XIIe s.): Actes du IIe Congre`s International d’Arche´ologie Me´die´vale, Caen 2–4 octobre 1987, ed. H. Galinie (Caen, 1989), 125–31. 3 The denier of Pavia in the early 12th century (ca. 1102) contained some 0.5 g fine silver, according to C. Brambilla quoted by C. Cipolla, “Currency Depreciation in Medieval Europe,” EcHistR 15 (1963): 413–22, at 24. Such a poids de fin is very close to that of the contemporary denier Parisis that I use in the following estimates. One denier of Rouen of the late 11th century with a fineness of some 40–49% contained ca. 0.3–0.7 g (F. Dumas, “Les monnaies normandes (Xe–XIIe s.) avec un re´pertoire des trouvailles,” RN 21 [1979]: 84–140, at 103).

[ 194 ] Byzantine Trade of Philip I (1060–1108) were struck at 1/312 or 1/324 to the pound (1.30–1.26 g) and had a fineness of 40 percent.4 Alexios I’s hyperpyra, weighing ca. 4.30 g, are, according to the latest analyses,5 some 82 percent fine. The comparison of these average intrinsic values, ⫾ 3.74 g fine gold ⫽ 90.7–93.6 g fine silver,6 entails a gold-silver ratio of 1:24 to 1:25, apparently excessive but understandable in the circumstances. The rate of 1 “aureus” to 8 or 9 soldi (96–108 deniers) appears to be a more “normal” one. In fact, the contemporary Tripoli dinar had, like the other Fatimid dinars, a very high purity (ca. 97%).7 Reckoning a denier with some 0.5 g silver fine, the comparison becomes: ⫾ 4.07 g fine gold ⫽ 48–54 g fine silver, entailing a gold-silver ratio of 1:11.8 or 1:13.2, more usual in an eastern Mediterranean context.8 Michael Metcalf is therefore right when he comments on this famous statement of Raymond of Aguilers and concludes that “eight or nine shillings does not sound in any way extortionate as an exchange-rate against the Islamic gold dinar.”9

2. Intrinsic Value of Coins Exchanged during the Second Crusade (comments on pp. 170–75) The exchange rate in the Balkans of 1 stamenon to 5 deniers, in terms of the intrinsic values involved, implies equating ca. 0.27 g fine silver in the Byzantine stamenon10 to the 2 or 1.9 g represented by 5 deniers Parisis of the time.11 The denier’s real value was thus grossly underestimated (by about 7 to 8 times). As A. Laiou points out (p. 173), “any exchange rate that quoted the billon trachy in multiples of the denier Parisis was bound to be excessive.” The other equivalence between 1 mark of silver and 3 hyperpyra (144 stamena rated

4 J. Lafaurie, “Numismatique: Des Carolingiens aux Cape´tiens,” CahCM 13 (1970): 117–36, at 136. The figures given by Lafaurie imply some 0.5–0.52 g fine silver in a denier. 5 C. Morrisson, et al., L’or monnaye´, vol. 1, Purification et alte´rations. De Rome a` Byzance, Cahiers ErnestBabelon 2, CNRS (Paris, 1985), 154 and 232. 6 Estimating at 87.2% the fine content of a hyperpyron weighing ca. 4.3 g (containing 86.4% Au and 10% Ag reckoned as equivalent to 0.84% Au at a 1:12 gold-silver ratio). 7 According to reliable specific gravity measurements. See A. S. Ehrenkreutz, “Studies in the Monetary History of the Near East in the Middle Ages—the Standard of Fineness of Some Types of Dinars,” Journal of the Economic and Social History of the Orient 2 (1959): 128–61, and W. A. Oddy, “The Gold Contents of Fatimid Coins Reconsidered,” in Metallurgy in Numismatics, vol. 1, ed. D. M. Metcalf and W. A. Oddy (London, 1980), 98–118. 8 The calculations of the gold-silver ratio and the identification of its nature (ratio for unminted metal, market ratio for coins, etc.) are, as is well known, tricky. See A. M. Watson, “Back to Gold—and Silver,” EcHistR 20.1 (1967): 1–34, at 33. In what follows I have relied on Watson’s results (23–29). Neither the details of his calculations nor their sources could be given in his article. 9 Metcalf, Coinage of the Crusades (above, note 65), 12. 10 A “billon trachy” of some 4.45–4.30 g with a 6.3% silver content; see Hendy and Charles, above, note 84. The average fineness is that of Manuel’s first and second coinages. 11 Louis VII’s (1137–80) deniers were struck at 1/240 or 1/250 to the mark (1.01–0.97 g) and had a fineness of ca. 39.8%, according to Lafaurie, “Numismatique,” 136 n. 6.

Ce´cile Morrisson [ 195 ] at 48 to the hyperpyron) implied the following approximate intrinsic values: 234.4 g fine silver12 ⫽ 10.86 g fine gold.13 This entails an excessive gold-silver ratio of 1:26. The rate in Constantinople of 1 stamenon to less than 2 deniers amounts to equating the Byzantine billon coin containing 0.28–0.26 g fine silver with two Western silver ones containing 0.8–0.76 g silver. It implies overvaluating the stamenon nearly three times (2.8–2.9) and agrees with M. Hendy’s conclusion, on a different basis, that “the overvaluation will have been something in the order of 2 1/2 times the bullion value of the trachy.”14 The other Constantinopolitan exchange rate of 396 stamena for 1 mark of silver leads to roughly the same result: 396 ⫻ 0.27 g ⫽ 106.9 g, being considered equivalent to the 234.4 g fine silver in a mark, which implies an overvaluation of the stamenon of some 2.1 times. The more favorable rating of the mark may be attributed to its guaranteed, or easily assayable, fineness and to its mere bulk (important absolute value) compared to retail exchanges dealing with small sums in all sorts of deniers of varying fineness.

3. Intrinsic Value of Coins Exchanged during the Third Crusade and the Exchange Rates (comments on pp. 175–76) The 1190 treaty probably involved the Cologne mark, which was slightly lighter and contained some 231 g of highly pure silver. The rate of 5.5 hyperpyra for an assayed (examinatum)15 mark amounts to equating some 19.7 g16 fine gold to 231 g fine silver and implies a gold-silver ratio of 1:11.8.17 The difference between examinata and non examinata (the latter rated 1.8 times less than the former, at 3 hyperpyra) seems at first very high, but must have incorporated an important premium for risk. Moreover, it is known that “all of the surviving German

12 A Troyes mark with a weight of 244.7 g and a fineness of 12d in argent-le-roi (95.8%). If the mark were one of pure or nearly pure silver, the silver weight would be ca. 240 g and the ratio 1:21.87. The order of magnitude is the same. 13 Manuel’s hyperpyron with a weight of 4.3 g and a fineness of ca. 85% (84.2% Au and 12% Ag) would amount to 3.65 g fine (see above, note 5). 14 M. F. Hendy and J. A. Charles, “The Production Techniques, Silver Content, and Circulation History of the Twelfth-Century Byzantine Trachy,” in M. F. Hendy, The Economy, Fiscal Administration, and Coinage of Byzantium (Northampton, 1989), 18. 15 “Examinatum” is clearly a technical word for assay. In the late 12th century, the mint of Melgueil, for instance, had two gardes: one in charge of the custody of dies and the other in charge of issag, i.e., the punch used to certify the testing of the ingots’ fineness. 16 Isaac II’s hyperpyron with a weight of 4.3 g and a fineness of ca. 83.3% (82.2% Au and 14% Ag) would amount to 3.58 g fine (see above, note 5). 17 Watson, “Back to Gold,” 23: “before Europe returned to gold the ratio in most parts of Europe was generally between 9 and 10.” See P. Spufford, Money and Its Use in Medieval Europe (Cambridge, 1988), table 2, p. 272.

[ 196 ] Byzantine Trade ingots that have been analyzed have proved to be of various poor qualities of argentum usuale.”18 The official value of 120 stamena to the hyperpyron equates the ca. 12 g silver contained in the now debased billon coin19 with ca. 3.58 g fine gold. The implicit goldsilver ratio is three times less than the one derived from the exchange rate of the assayed mark and three times less as well than the one assumed to have prevailed in the eastern Mediterranean. This is a good index of the persisting overvaluation (partly token nature) of the billon denomination. In 1199 the decreasing value of 184 stamena to the hyperpyron deduced from Pisan documents equates ca. 14.7 g silver with 3.6 g fine gold. The implicit gold-silver ratio is 21 percent higher than the 1190 one and confirms the downward trend of the stamenon overvaluation emphasized above (p. 175). It should be pointed out that comparisons between the metal contents of the stamenon and the hyperpyron as well as the implicit derived ratios are very unreliable for two reasons: first, the uncertainty of metrological figures, but above all the limits certainly fixed to the exchange of stamena into hyperpyra, as is always the case with token coinages.20 CNRS-Colle`ge de France, Paris, and Dumbarton Oaks

18

Spufford, Money, 221, quoting A. [von] Loehr, “Probleme der Silberbarren,” NZ 64 (1931): 101–9, with illustrations. The figures given by von Loehr (p. 106) for ingots of the early 13th century vary between 85.4% and 79.8% fine. 19 Assuming 120 stamena of 4–3 g with a fineness of 2.5% (0.1 to 0.07 g fine silver), equating a hyperpyron of 3.58 g. Hendy and Charles, “Byzantine Trachy”; T. Bertele`, Numismatique byzantine (Wetteren, 1978), 77. 20 To the observation of Pegolotti mentioned above (176 n. 105) might be added the fact that the Nea Logarike, according to my interpretation at least, only accepted the (electrum) aspron trachy and not the stamenon in payment of taxes (C. Morrisson, Monnaie et finances a` Byzance [Aldershot, 1994], no. , 460–64).

Changing Economic Patterns in Latin Romania: The Impact of the West David Jacoby The dramatic fall of Constantinople in 1204 and the Latin conquest of the empire’s provinces in the following decade resulted in the dismemberment of Romania. The Latins established a fairly large number of new political entities in the region, most of which remained under their rule for more than two centuries. These long-term political and territorial developments also generated profound economic changes.1 The present survey is not aimed at providing an overall picture of the complex economic evolution of Latin Romania until around the mid-fifteenth century. Rather, it attempts to determine, as far as possible, the nature, extent, and pace of the Western impact on some of its aspects and trends. To this effect, it will be necessary time and again to look back at conditions and patterns existing before the Fourth Crusade. It would be tedious to review the complex history of the political entities created by the Latins on Byzantine soil after 1204. However, two of them warrant our attention, since they were of particular importance for the economic evolution of Latin Romania and happen to be fairly well documented: the Venetian maritime empire and the Frankish principality of the Morea in the Peloponnesos. Venice laid the foundations of its centralized maritime empire in 1207, when it began the conquest of Crete and occupied Coron and Modon, two ports of southern Messenia. In 1211 it obtained a quarter in the main city of Euboea, Euripos, called Negroponte by the Latins, a name also used for the island itself. The second stage of Venetian expansion in Romania took place in the 1380s and 1390s, when Venice extended its domination over the island of Corfu, the whole of Euboea, and several cities and lordships in the Peloponnesos and the Aegean.2 The principality of the Morea was the largest among the lordships created in non1 I shall not deal here with the short-lived Kingdom of Thessalonica and Latin Empire of Constantinople, nor with the few islands of the Aegean occupied by the Genoese in the 14th century. They differed markedly from most territories of Latin Romania in their economic evolution. 2 General historical background and detailed treatment of various issues by S. Borsari, Il dominio veneziano a Creta nel XIII secolo (Naples, 1963); idem, Studi sulle colonie veneziane in Romania nel XIII secolo (Naples, 1966); F. Thiriet, La Romanie ve´nitienne au Moyen Age: Le de´veloppement et l’exploitation du domaine colonial ve´nitien (XIIe– XVe sie`cles), Bibliothe`que des Ecoles franc¸aises d’Athe`nes et de Rome 193, 2d ed. (Paris, 1975); D. Jacoby, La fe´odalite´ en Gre`ce me´die´vale: Les “Assises de Romanie.” Sources, application et diffusion (Paris–The Hague, 1971), 185– 308; J. Koder, Negroponte: Untersuchungen zur Topographie und Siedlungsgeschichte der Insel Euboia wa¨hrend der Zeit ¨ sterreichische Akademie der Wissenschaften, Philosophisch-historische Klasse, der Venezianerherrschaft, O Denkschriften 112 (Vienna, 1973); short survey by P. Lock, The Franks in the Aegean, 1204–1500 (London–

[ 198 ] Changing Economic Patterns in Latin Romania Venetian territories, most of which adopted a Western-type feudal regime. The return of Byzantium to the Peloponnesos in 1262 compelled Prince William II of Villehardouin to seek the support of King Charles I of Sicily, who extended his rule over Frankish Morea after the prince’s death in 1278. The direct and indirect domination of the king’s successors over the principality continued for more than a whole century.3 Among other topics, it will be useful to investigate whether differences in political conditions between the Venetian colonies and feudalized areas, primarily Frankish Morea, had any impact on their respective economic evolution. Two intertwined processes promptly following the conquest had a marked impact on the economy of Latin Romania: the confiscation and redistribution of urban and especially rural resources, the most important components of which were land, the peasantry, and public rights of taxation; and Latin settlement in the conquered territories. The extent to which resources were confiscated varied from one area to another. It was largely determined by the circumstances leading to the submission of the local population, either based on agreements with the Latin leaders or imposed by force, by the size of lordless property, and by the amount of assets required for the needs of the conquerors. Large estates belonging to the Byzantine crown, to members and relatives of the imperial family, as well as to magnates, dignitaries, and ecclesiastical institutions closely associated with the imperial court and based in Constantinople were clearly among the first to be seized by the Latins. Such was the fate, for instance, of estates in the western Peloponnesos held before 1204 by Irene, daughter of Emperor Alexios III Angelos, members of the Kantakouzenos and Branas families, and Constantinopolitan monasteries.4 There were also estates belonging to the crown and to Constantinopolitan landlords in Crete and in other territories occupied by the Latins.5 In addition, the conquerors took hold New York, 1995), 142–60; also D. Jacoby, “Byzantium after the Fourth Crusade: The Latin Empire of Constantinople and the Frankish States in Greece,” in The New Cambridge Medieval History, vol. 5, c. 1198–c. 1300, ed. D. Abulafia (Cambridge, 1999), 525–42. 3 See J. Longnon, L’Empire latin de Constantinople et la principaute´ de More´e (Paris, 1949); A. Bon, La More´e franque: Recherches historiques, topographiques et arche´ologiques sur la principaute´ d’Achaı¨e (1205–1430) (Paris, 1969); Jacoby, Fe´odalite´, 17–91, 179–83; idem, “The Encounter of Two Societies: Western Conquerors and Byzantines in the Peloponnesus after the Fourth Crusade,” AHR 78 (1973): 873–906, repr. in idem, Recherches sur la Me´diterrane´e orientale du XIIe au XVe sie`cle: Peuples, socie´te´s, e´conomies (London, 1979), no. ; K. M. Setton, The Papacy and the Levant (1204–1571), vol. 1, The Thirteenth and Fourteenth Centuries (Philadelphia, 1976), 1–162, 405–73; Lock, Franks, 1–134, 193–239, 266–309; D. A. Zakythinos, Le Despotat grec de More´e, 2d ed. (London, 1975). On Latin Romania in general, see D. Jacoby, “Les e´tats latins en Romanie: Phe´nome`nes sociaux et e´conomiques (1204–1350 environ),” in XVe Congre`s international d’e´tudes byzantines (Athe`nes, 1976), Rapports et corapports, vol. 1.3 (Athens, 1976), 1–51, repr. in idem, Recherches, no. ; idem, “Social Evolution in Latin Greece,” in A History of the Crusades, ed. K. M. Setton, 2d ed. (Madison, Wisc., 1969–89), 6:175–221. 4 See D. Jacoby, “Les archontes grecs et la fe´odalite´ en More´e franque,” TM 2 (1967): 422–27, repr. in idem, Socie´te´ et de´mographie a` Byzance et en Romanie latine (London, 1975), no. . The church of the Theotokos at the Blachernae in Constantinople presumably also had land in the western Peloponnesos prior to 1204. This is suggested by the monastery bearing its name, recorded in the 15th century and later, which stood to the east of Chiarenza; its construction goes back to the Byzantine period: Bon, More´e, 325, 561–74. 5 For a general view of Constantinopolitan estates throughout the empire, see J.-C. Cheynet, Pouvoir et contestations a` Byzance (963–1210) (Paris, 1990), 237–45; P. Magdalino, The Empire of Manuel I Komnenos, 1143– 1180 (Cambridge, 1993), 160–71.

David Jacoby [ 199 ] of assets previously owned or held as pronoiai by local landlords who had fled or had opposed the Latin armies. Finally, whenever convenient they curtailed the property of Greek lay landlords, churches, and monasteries remaining under their rule.6 While varying in density, the number and especially the distribution of knightly fiefs in the Peloponnesos suggest large-scale confiscations.7 In Crete the state’s seizure of property belonging to various archontes who had submitted to Venetian rule triggered a Greek rebellion in 1212.8 Confiscated lands and peasants were partitioned and partly allotted, both in feudalized and Venetian territories, in return for specific military and fiscal obligations that do not concern us here. In Crete, however, Venice retained under its direct authority the city and district of Candia, as well as the latter’s rural workforce, and appears to have acted similarly in the territories of Coron and Modon.9 As for the rights of the Byzantine crown, they were privatized in feudalized areas and taken over by the state in Venetian territories. Some Greek archontes, though, benefited from the collapse of Byzantine authority, since they retained in their hands usurped imperial land and rights. The judicial and fiscal prerogatives exercised by Alexios Kallergis in Crete in the late thirteenth century, the most conspicuous and best documented such case, suggest that his ancestors too had enjoyed them since 1204, if not earlier.10 It has recently been argued that a free peasantry directly accountable to the state still existed in the empire in the early thirteenth century.11 If this indeed was the case in the territories occupied by the Latins, free peasants too would have been affected by the processes just described. These processes did not alter the nature of the predominantly agrarian economy of the conquered territories, although they had a definite impact on various phases of its operation that will be examined below. In the short term, however, the remaining Greek and the new Latin landlords as well as the peasantry had a common interest, regardless of the political regime and the changes in lordship imposed by the conquerors in their respective territories. They were eager to ensure the preservation of the Byzantine economic infrastructure in the countryside and the continuity of the latter’s exploitation. 6

See Jacoby, “Archontes,” 441–42. R. Hiestand, “Nova Francia—nova Graecia: Morea zwischen Franken, Venezianern und Griechen,” in R. Lauer and P. Schreiner, eds., Die Kultur Griechenlands in Mittelalter und Neuzeit, Abhandlungen der Akademie der Wissenschaften in Go¨ttingen, Philologisch-historische Klasse, Dritte Folge 212 (Go¨ttingen, 1996), 59, argues that the Greeks of the Peloponnesos who cooperated with the conquerors did not suffer any losses. If this had indeed been the case, no land would have been available for distribution in large areas. 8 See Borsari, Dominio, 32–33, and for an overview of Greek rebellions in the 13th century, ibid., 27–66. 9 For Crete, see Borsari, Dominio, 27–28; for Coron and Modon, Jacoby, Fe´odalite´, 225–26. 10 See Jacoby, “Etats,” 11, 26–28; idem, “Evolution,” 184–85, 201; C. Maltezou, “Byzantine ‘consuetudines’ in Venetian Crete,” DOP 49 (1995): 270–71; further evidence on the standing of Alexios Kallergis in eadem, “Creta fra la Serenissima e la Superba,” in Oriente e Occidente tra medioevo ed eta` moderna: Studi in onore di Geo Pistarino, ed. L. Balletto (Genoa, 1997), 768–69. 11 See J. Lefort, “Rural Economy and Social Relations in the Countryside,” DOP 47 (1993): 101–13, esp. 111 ff. Archaeological evidence suggests a dwindling free peasantry in southwestern Boeotia already earlier, in the 11th–12th century: see J. Bintliff, “The Frankish Countryside in Central Greece: The Evidence from Archaeological Field Survey,” in The Archaeology of Medieval Greece, ed. P. Lock and G. Sanders (Oxford, 1996), 4–5; for a more general view, see A. Harvey, “Peasant Categories in the Tenth and Eleventh Centuries,” BMGS 14 (1990): 250–56. For Crete, see M. Gallina, Una societa` coloniale del Trecento: Creta fra Venezia e Bisanzio, Deputazione di storia patria per le Venezie, Miscellanea di studi e memorie 28 (Venice, 1989), 85–88. 7

[ 200 ] Changing Economic Patterns in Latin Romania Indeed, there was continuity in the patterns of land and water uses, crops and types of cultivation, animal breeding and pastoral activity, as well as in the production of wine, cheese, hides, wool, and silk, to name the most important products of the rural economy.12 Continuity also prevailed in the forms of exploitation, either by the peasants themselves in their small household holdings or on domain land held by the landlords, who for that purpose relied on a variety of means. These included corve´e or compulsory labor services owed by the peasants, the ajggarei´a of the Byzantine period,13 hired labor,14 and various contracts associating peasant and landlord and entailing a division of produce between the two parties or the payment of rents by the peasant. Continuity in these fields is well illustrated by the survival of Byzantine institutions and practices, whether administrative, fiscal, or legal, by the structure of the large estates in Frankish Morea, reflected in fourteenth-century surveys, as well as by numerous agricultural and pastoral contracts drafted in Crete, which make abundant use of Greek terms.15 It remains to determine the extent of continuity on the ground. The Latin conquest was largely conducted in swift campaigns that do not seem to have affected the operation of the rural economy. Later, however, intermittent warfare in the 1270s between Frankish and Byzantine forces in Euboea must have inflicted some damage upon the countryside.16 More severe disruptions in the exploitation of rural resources took place in Crete. The first instance occurred in the years 1207–11, during the struggle between Venetian forces and those of the Genoese Enrico Pescatore, who for several years ruled large sections of the island.17 A letter written by some Cretan Greeks to the Venetian doge Pietro Ziani in 1224 or 1225 offers convincing evidence to this effect.18 The same 12

For details, see below. Originally, the Byzantine ajggarei´a constituted a public labor service owed to the state, which occasionally transferred it to landlords and was then used for the cultivation of their domain land: see A. StauridouZaphraka, “ JH ajggarei´a sto` Buza´ ntio,” Byzantina 11 (1982): 23–54. Its public nature was preserved under Venetian rule in the districts of Coron and Modon: see D. Jacoby, “Un aspect de la fiscalite´ ve´nitienne dans le Pe´loponne`se aux XIVe et XVe sie`cles: Le ‘zovaticum,’” TM 1 (1965): 408, repr. in idem, Socie´te´, no. . About 1270 labor services existed on a moderate-sized Byzantine estate with about a dozen peasant households: see J. Lefort, “Une exploitation de taille moyenne au XIIIe sie`cle en Chalcidique,” in Afie j ´ rwma sto` n Ni'ko Sboro´ no, vol. 1 (Rethymno, 1986), 362–72, esp. 366. For Crete, see Gallina, Societa`, 79–80, 86. 14 For Crete, see Gallina, Societa`, 49–50, 88–89; C. Gaspares, H gh kai oi jagro´ te" sth mesaionikh´ Krh´ th, 13o"–14o" ai., National Hellenic Research Foundation, Institute for Byzantine Research, Monographs 4 (Athens, 1997), 175–78. 15 See Jacoby, “Evolution,” 216–18; idem, review of A. Carile, La rendita feudale nella Morea latina del XIV secolo (Bologna, 1974), in BZ 73 (1980): 359–61; idem, “From Byzantium to Latin Romania: Continuity and Change,” Mediterranean Historical Review 4 (1989): 10–23, repr. in Latins and Greeks in the Eastern Mediterranean after 1204, ed. B. Arbel, B. Hamilton, and D. Jacoby (London, 1989); Maltezou, “Byzantine ‘consuetudines,’” 270–74; eadem, “ JO o“ro" ‘metacherissi’ sti`" ajgrotike` " misqw´ sei" th'" benetokratoume´ nh" Krh´ th",” Byzantina 13 (1985): 1142–46; Gallina, Societa`, 31–58, 71–72, 78–79; Gaspares, H gh, 143–75; P. Topping, “Viticulture in Venetian Crete (XIII C.),” in Pepragme´ na tou' D´ Dieqnou'" Krhtologikou' Sunedri´ou, vol. 2 (Athens, 1981), 509–20. 16 On the military operations, see D. J. Geanakoplos, Emperor Michael Palaeologus and the West, 1258–1282: A Study in Byzantine-Latin Relations (Cambridge, Mass., 1959), 235–37, 296–99. 17 On this struggle, see below, pp. 207–8. 18 Ed. G. B. Cervellini, Documento inedito veneto-cretese del Dugento (Padua, 1906), 13–18, esp. 14–16. For its dating, see Borsari, Dominio, 32–33, esp. n. 17. 13

David Jacoby [ 201 ] letter refers to the harsh measures implemented during the Greek rebellion of 1222–24 by the duke of Crete, Paolo Querini, which led to the death or flight of numerous peasants.19 During the Greek rebellions of the thirteenth century, Venetian military operations and forays by the insurgents caused at times severe destruction.20 In addition, the Venetian government prohibited on several occasions cultivation and grazing in areas likely to provide supplies to the insurgents. Though not continuous, the most serious interruptions in land use resulting from such cases occurred between 1287 and 1299, the year in which the rebellion led by Alexios Kallergis came to an end.21 In 1307 Andrea Corner requested the Venetian government to compensate him for revenue losses incurred in that period in Lombaro, a village located in the Lassithi area that belonged to the militia or fief he had obtained from the Commune.22 He referred to the period in which the village had been partly deserted, “quando e lomefo deshabitadho,” indicating that of the thirty-three villein households settled earlier, only seven remained. He also mentioned damage caused by fire and the neglect of vineyards cultivated by peasants under lease.23 Incidentally, the detailed document he submitted is of particular interest, since it offers a unique insight into the components and sources of income of a Cretan fief. Several feudatories other than Andrea Corner had suffered similar losses.24 It should be noted that in all these instances the disturbances were circumscribed to specific localities or areas of Crete, peasants either fleeing or being moved to new locations in which they contributed to the extension of cultivation.25 For instance, Gabriele Querini was allowed between 1234 and 1236 to transfer his villeins to the new military tenement the Commune had granted him, since he could not exploit the one he held in the Lassithi plain.26 Land appears to have been only temporarily abandoned by peasants and herdsmen, generally for less than five years, considering the cases adduced earlier. 19

For the dating of Querini’s action, see Borsari, Dominio, 32, 39–40, 128. On Greek rebellions in the first half of the 14th century, see Thiriet, Romanie, 164. The one of 1332–33 was short-lived and localized; the one begun in 1342 appears to have inflicted more damage: see F. Thiriet, “La condition paysanne et les proble`mes de l’exploitation rurale en Romanie gre´co-ve´nitienne,” StVen 9 (1967): 59, repr. in idem, Etudes sur la Romanie gre´co-ve´nitienne (Xe–XVe s.) (London, 1977), no. . 21 Borsari, Dominio, 82–83; particular instances are recorded in documents drafted in 1307: S. M. Theo´a AQhnw' tokes, ed., Qespi´smata th'" Benetikh'" Gerousi´a", 1281–1385, Akadhmi j j n, Mnhmei'a th'" JEllhnikh'" JIstori´a", to´ mo" B, vol. 1 (Athens, 1936–37), 41–47, 48–56, esp. 43, nos. 9–11, 13–16. 22 Partial edition by Borsari, Dominio, 81–82 n. 76, and full one by Theotokes, Qespi´smata, 1:47–48, no. 12, yet both are marred by several mistakes. The legal nature of the Cretan militie and serventarie differed from that of fiefs and sergeantries, respectively, in feudalized territories: see Jacoby, “Evolution,” 192–93. 23 Note the important share of newly planted vines in the damage estimate: Theotokes, Qespi´smata, 1:48, lines 1–3. A contract for the planting and cultivation of a vineyard precisely at Lombaro, dated 1279, appears in M. Chiaudano and A. Lombardo, eds., Leonardo Marcello, notaio in Candia, 1278–1281, Fonti per la storia di Venezia, Sez. 3, Archivi notarili (Venice, 1960), no. 102 (hereafter Marcello). 24 See above, note 21, and esp. Theotokes, Qespi´smata, 1:53, lines 91–106; Gaspares, H gh, 201–8 and 299–330, nos. 1–6. 25 On the status and transfer of villeins, see Jacoby, “Etats,” 35–39; M. Gallina, Vicende demografiche a Creta nel corso del XIII secolo, Quaderni della Rivista di studi bizantini e slavi, Studi bizantini e slavi 2 (Rome, 1984), 12–21. 26 C. Maltezou, ed., Venetiae quasi alterum Byzantium. “Oyei" th'" JIstori´a" tou' Benetokratoume´ nou JEllh` Tekmh´ ria (Athens, 1993), 152, no. 9. Similar moves of peasants from the Lassithi area ocnismou', Arceiaka j curred during the rebellion of Alexios Kallergis: see above, note 24. 20

[ 202 ] Changing Economic Patterns in Latin Romania In 1222 and 1252 the Venetian government expected the military settlers sent to Crete to ensure the resumption of land cultivation in the area granted to them within two years after their arrival, state subsidies ensuring their livelihood in that period.27 Exceptionally, though, land exploitation was discontinued for some eight years.28 In addition to general circumstances, spontaneous peasant mobility also determined the degree of continuity in land use.29 In Crete peasants mainly moved within the island itself, although some of them attempted to flee. The authorities sought to prevent their escape and to attract immigrants.30 Peasant mobility appears to have been greater in the Peloponnesos. The unclear definition of boundaries between Frankish Morea and the Venetian enclaves of Coron and Modon, on the one hand, the existence of a Byzantine province in the Peloponnesos since 1262 and its subsequent expansion, on the other, prompted peasants to cross the common borders of these territories, whether in one direction or the other.31 However, in the late thirteenth and in the first half of the fourteenth century this movement may have been somewhat restricted by the peaceful coexistence of Latin and Greek lords in specific areas along the Frankish-Byzantine borders. These lords jointly exploited several villages of Frankish Morea and shared their revenues and, therefore, had a vested interest in the stability of the local peasantry.32 While generating individual mobility, agreed exchanges of peasants between landlords did not disrupt rural work. In Crete the state occasionally granted villeins to military settlers or authorized these to settle a number of them on their tenements.33 On the whole, then, it would seem that the thirteenth-century mobility of the rural workforce in Latin Romania was neither general nor continuous, but rather a local or regional phenomenon, limited in both extent and time. Recent research points to demographic growth in eleventh- and twelfth-century Byzantium.34 This trend appears to have been sustained in the territories occupied by the Latins for another century, as suggested by evidence regarding the peasantry, the demand for land, and rising yields.35 Conditions changed for the worse in several coastal areas and islands of the Aegean from the early fourteenth century, once the Turks of Asia Minor began their forays, 27 D. Jacoby, “La colonisation militaire ve´nitienne de la Cre`te au XIIIe sie`cle: Une nouvelle approche,” in Le partage du monde: Echanges et colonisation dans la Me´diterrane´e me´die´vale, ed. M. Balard and A. Ducellier (Paris, 1998), 304, 309, 311. The land granted in 1222 had been hit by the rebellion of 1217–19. 28 Theotokes, Qespi´smata, 1:52, lines 36–37. 29 For an overview of peasant mobility in Latin Romania, see D. Jacoby, “Une classe fiscale a` Byzance et en Romanie latine: Les inconnus du fisc, e´leuthe`res ou e´trangers,” in Actes du XIVe Congre`s international des e´tudes byzantines (Bucarest, 1971), vol. 2 (Bucharest, 1975), 139–52, repr. in idem, Recherches, no. ; idem, “Etats,” 36–39. 30 See previous note; Gallina, Vicende, 12–14; and below, p. 230. 31 On Frankish-Venetian boundaries, see Jacoby, Fe´odalite´, 223–25, 229–30; C. Hodgetts and P. Lock, “Some Village Fortifications in the Venetian Peloponnese,” in Lock and Sanders, Medieval Greece (as in note 11), 77–80. 32 See D. Jacoby, “Un re´gime de coseigneurie gre´co-franque en More´e: Les ‘casaux de parc¸on,’” Me´lRome 75 (1963): 111–25; repr. in idem, Socie´te´, no. . 33 See above, note 29; for Crete, see also Gallina, Vicende, 28–47. 34 See A. Harvey, Economic Expansion in the Byzantine Empire, 900–1200 (Cambridge, 1989), 47–67, 245–48, 250–55. 35 These last two aspects are discussed below. For Crete, see Gallina, Vicende, 9–47, an essentially positive assessment. I no longer maintain the view to the contrary, expressed in Jacoby, “Classe fiscale,” 142.

David Jacoby [ 203 ] which occasionally resulted in severe damage and depopulation.36 On the other hand, territories shielded from their activity, such as Crete, benefited from an influx of refugees.37 The Black Death appears to have inflicted serious demographic losses, aggravated by subsequent bouts of plague, upon the peasantry of Latin Romania as a whole.38 The comparison of two surveys of the same villages in Frankish Morea, one carried out in 1338 and the other in 1354, thus before and after the Black Death, respectively, enhances this assessment.39 It has been argued that in Crete demographic losses were partly offset by the import of slaves. To be sure, slaves were occasionally put to work in the countryside, yet compared with villeins they surely remained a marginal factor in its exploitation, since most of them, whether male or female, lived in urban households.40 Losses of rural labor resulting from the plague were compounded in the Peloponnesos by unstable political conditions in the following period and the Turkish, Byzantine, and Albanian incursions related to them. In 1396 the prince of Morea requested the return of peasants who had moved from the principality to the Venetian territories of Coron and Modon. Three years later Venice took measures to induce the villeins of Argos to return to their land, while in 1407 Latin landlords of Coron requested manpower from the Commune in order to bring their abandoned land under cultivation.41 Although many Albanians 36 On their activity, see D. Jacoby, “Catalans, Turcs et Ve´nitiens en Romanie (1305–32): Un nouveau te´moignage de Marino Sanudo Torsello,” StMed, 3d ser., 15 (1974): 246–47, 251–53, 257–61, repr. in idem, Recherches, no. ; E. A. Zachariadou, “The Catalans of Athens and the Beginning of the Turkish Expansion in the Aegean Area,” StMed, 3d ser., 21 (1980): 821–38, repr. in eadem, Romania and the Turks (c. 1300–c. 1500) (London, 1985), no. ; eadem, “Holy War in the Aegean during the Fourteenth Century,” Mediterranean Historical Review 4 (1989): 212–18, repr. in Arbel, Hamilton, and Jacoby, Latins (as in note 15); documents in R.-J. Loenertz, ed., Les Ghisi, dynastes dans l’Archipel, 1207–1390 (Florence, 1975), 236 (lines 63–68), 241 (no. 72), 251 (no. 76): attacks on the area of Corinth and other regions of the Peloponnesos in the 1340s and movement of villeins between Aegean islands and Crete in the 1350s and 1360s. 37 See Jacoby, “Classe fiscale,” 140 n. 6, 149 n. 52; also above, note 35. 38 M.-H. Congourdeau, “Pour une e´tude de la Peste Noire a` Byzance,” in Euyuci´a: Me´langes offerts a` He´le`ne Ahrweiler, Byzantina Sorbonensia 16 (Paris, 1998), 149–63, also includes information about the plague in Latin Romania until the 1460s. More specifically, for Crete, see Thiriet, “Condition,” 59–60, 62–63; Jacoby, “Classe fiscale,” 151; Gaspares, H gh, 75, 77, 80; Gallina, Societa`, 37, refers to heavy losses, yet on 89–90 relies on the high numbers of villeins residing in 1356 in three Cretan fiefs to suggest that they were lower than commonly believed. One may wonder, though, whether these pieces of evidence do not reflect a concentration of villeins in specific tenements, also attested in other periods (see below, note 47), rather than the general picture. 39 The two surveys appear in J. Longnon and P. Topping, eds., Documents sur le re´gime des terres dans la principaute´ de More´e au XIVe sie`cle (Paris–The Hague, 1969), 55–115, nos.  and . These and other surveys drafted from 1336 to 1379 record peasant households in feudal estates of Frankish Morea or offer indirect evidence on demographic trends. They still await a thorough and balanced examination in this respect. Their analysis by Carile, Rendita, 80–183, is unsatisfactory: see my review (cited above, note 15), 358–59. 40 See Thiriet, Romanie, 314–15, 335, 413, on slaves in agriculture; C. Verlinden, L’esclavage dans l’Europe me´die´vale, vol. 2, Italie-Colonies italiennes du Levant-Levant latin-Empire byzantin (Ghent, 1977), 876–78, is less emphatic. Thiriet relies on rather slim evidence and remains unconvincing. Inter alia he refers to the loan of 3,000 hyperpers offered in 1393 by the Venetian authorities to encourage the import of male slaves for settlement in villages: document ed. by Verlinden, ibid., 877 n. 553. However, given their average price in Crete around that time, on which see ibid., 875–76, the sum would have sufficed for only about eighty slaves. For the sake of calculation, I assume that sale prices in Crete were double the original purchase prices elsewhere. S. McKee, “Households in Fourteenth-Century Venetian Crete,” Speculum 70 (1995): 58–65, has found only two references to slaves settled in the countryside in 785 wills of the 14th century. 41 For evidence, see above, notes 29, 36, and 39; also D. Jacoby, “Italian Migration and Settlement in Latin Greece: The Impact on the Economy,” in H. E. Mayer, ed., Die Kreuzfahrerstaaten als multikulturelle Gesellschaft:

[ 204 ] Changing Economic Patterns in Latin Romania raised cattle and horses, a number of them tilled the land. Their settlement in continental Greece in the late fourteenth and the early fifteenth century clearly points to partial depopulation.42 On balance, despite the Greek rebellions of the thirteenth and fourteenth centuries, Crete appears to have enjoyed a greater stability of rural labor than the Aegean islands or the Peloponnesos. It is in the nature of written sources bearing on populations to emphasize mobility and disruption, rather than to record or reflect stability and continuity.43 One should remember, though, that unless on a massive scale in a given area peasant mobility did not necessarily result in overall interruption of land exploitation. It is impossible to arrive at quantitative estimates of demographic trends for Latin Romania, except for some villages of the Peloponnesos in a limited period.44 We have to rely, therefore, on evaluations partly based on circumstantial evidence, as for instance the recent surface prospection of southwestern Boeotia. The pottery, structures, and settlement pattern discovered in this region point to stable conditions, high population and demographic rise, as well as to prosperity and economic expansion throughout the Frankish period up to the Turkish conquest.45 To be sure, this region willfully submitted to the Frankish conquerors in 1204 and to the Catalan Company in 1311. In addition, it was less exposed than the Aegean islands and coastal areas to foreign incursions. Future field research in other regions of the Greek mainland may well yield similar results. The general impression, then, is one of a fairly high degree of continuity in rural exploitation in many areas of Latin Romania in the thirteenth and fourteenth centuries. These trends are confirmed by the eagerness of Latins, including merchants and bankers, and indigenous Greeks to obtain landed estates, from feudal lords in feudalized areas and either from the Commune or from the holders of tenements in Crete.46 The value of Cretan fiefs varied substantially according to their nature, the type and quality of their land, and the number of households they contained, which differed widely.47 The circumstances of their transfer and the purchasers’ expectations were also relevant in this Einwanderer und Minderheiten im 12. und 13. Jahrhundert, Schriften des Historischen Kollegs, Kolloquien 37 (Munich, 1997), 123–24; J. Chrysostomides, ed., Monumenta Peloponnesiaca: Documents for the History of the Peloponnese in the 14th and 15th Centuries (Camberley, 1995), 373, 406–7, 506, 571–72, 583–85, 589. 42 See P. Topping, “Albanian Settlements in Medieval Greece: Some Venetian Testimonies,” in Charanis Studies: Essays in Honor of Peter Charanis, ed. A. E. Laiou-Thomadakis (New Brunswick, N.J., 1980), 261–71; A. Ducellier, “Les Albanais dans les colonies ve´nitiennes au XVe sie`cle,” StVen 10 (1968): 405–20, reprinted in idem, L’Albanie entre Byzance et Venise, Xe–XVe sie`cles (London, 1987), no. . The settlement of some 2,000 Armenians in Crete in 1363 and some 4,000 Greeks from Tenedos in Crete and Euboea in the 1380s also points to rural depopulation, yet it is impossible to evaluate the contribution of these immigrants to rural production. See Thiriet, Romanie, 264–65, and further studies cited by Gaspares, H gh, 80. 43 Thiriet, “Condition,” 35–69, and Carile, Rendita, 80–183, have failed to take this into consideration and paint excessively bleak pictures of the state of the rural workforce and land exploitation in Latin Romania. 44 See above, p. 203. 45 See Bintliff, “Frankish Countryside,” 1–18, esp. 4–7. For other regions, see A. Harvey, “The Middle Byzantine Economy: Growth or Stagnation?” BMGS 19 (1995): 254–55. 46 On the bankers, see below, p. 211. On the Moreot archontes, see Jacoby, “Encounter,” 891–96; on those of Crete, Borsari, Dominio, 35–66, passim. 47 On the number of villein households per militia or fief in Crete, from six to twenty-five, see Gallina, Vicende, 12–40, esp. 21–22, yet higher figures appear in 1356, on which see above, note 38. Lombaro had thirtythree villein households before it was hit by the rebellion of Alexios Kallergis: see above, note 22. On higher figures of households in 1414, including free ones, see Gaspares, H gh, 282–84 and 288–90, tables 31 and 37.

David Jacoby [ 205 ] respect. On the whole, though, prices appear to have been on the rise,48 except in areas hit by rebellions.49 The acquisition of these tenements, whether by residents or by newcomers to the island, was considered a good investment.50 In Crete some fiefs provided in the late thirteenth century yearly revenues of more than 1,000 or even 1,500 hyperpers.51 In the Morea and other feudalized territories, the knight’s fief was supposed to yield a yearly income of 1,000 hyperpers.52 The four known figures of the fourteenth century reflecting the actual revenue of entire knightly fiefs differ widely, two being substantially lower.53 Both in feudalized areas and in Venetian Crete, military tenements remained for several generations within the same families. In the island, though, they changed hands more often as a result of greater mobility among military settlers, some of whom in the thirteenth century returned to Venice or other localities in Italy after residing in Crete for a number of years.54 The redistribution of resources in the conquered lands, examined above, was coupled with Latin settlement proceeding on a scale and along patterns unknown earlier in Romania. The geographic distribution of the Latin settlers also differed substantially from what it had been before 1204. These features were bound to have a strong impact on the economic development of Latin Romania. To begin with, there was a swift and significant growth in Latin settlement, encouraged by the new lords, including the Venetian government. To be sure, these lords were eager to strengthen the small Latin nuclei in their respective territories in order to increase their military power and enhance their rule, yet economic considerations were also of major importance in this respect. The collapse of centralized imperial control over specific branches of trade and manufacture, among them silk,55 brought about a striking departure from traditional Byzan-

48 Borsari, Dominio, 83–84, and table facing 84; Gallina, Societa`, 106–11. In 1216 Giovanni Longo bought two militie for 300 Venetian pounds or around 250 hyperpers: R. Morozzo della Rocca and A. Lombardo, eds., Documenti del commercio veneziano nei secoli XI–XIII (Turin, 1940) (hereafter DCV ), no. 574; see Jacoby, “Colonisation,” 307 n. 46, for the calculation of this exceptionally low price in Cretan currency, and 304–5. Prices failed to rise around the mid-14th century, presumably in connection with the sequels of the Black Death. 49 In 1307 the value of a specific sergeantry in the region of Canea was estimated at 300 hyperpers and its annual revenue at 18 hyperpers, in addition to 4 hyperpers paid by each of the ten villein households it contained, ´ sei" Mei´zono" Sumbouli´ou Beneti´a", 1255–1669, thus a total of 58 hyperpers: S. M. Theotokes, ed., Apofa j ´a Aqhnw' Akadhmi j j n, Mhmei'a th'" JEllhnikh'" JIstori´a", to´ mo" A.2 (Athens, 1933), 49, no. 39. Thiriet, Romanie, 273, has mistakenly multiplied by ten the first two figures appearing in the document. Since the sergeantry was equivalent to one-sixth of a fief, the latter would have yielded around 350 hyperpers. Both its low value and low revenue may be explained by severe damage inflicted during the rebellion of Alexios Kallergis, as in the case of Lombaro documented for the same year, on which see above, note 22. 50 The evidence in this respect contradicts the assumptions of Thiriet, Romanie, 137, and Gallina, Societa`, 10, that in the 13th century “normal” economic exploitation in Crete was excluded, and that rather than being concerned with it, the military settlers focused on the establishment and strengthening of military and institutional structures. Incidentally, this was the task of the Venetian authorities. 51 Theotokes, Qespi´smata, 1:53, lines 91–106. On revenues, see also Gaspares, H gh, 264–65, table 13. 52 See Jacoby, “Archontes,” 449 and n. 156. 53 See Carile, Rendita, 118, 126–27, 140–41, 172: revenues of 768 hyperpers in return for full yearly military service, 1,150 for nine months only, 1,165 for three months only, around 743 hyperpers for a whole year. The reduction of military service in the second and third cases was clearly granted as a favor on a personal basis. 54 On the fate of some Cretan fiefs, see Gallina, Vicende, 28 and 55–57, appendix, tables –. On nonVenetians and the return to Italy, see Jacoby, “Colonisation,” 307, 313. 55 On which see D. Jacoby, “Silk in Western Byzantium before the Fourth Crusade,” BZ 84–85 (1991–92): 452–500; repr. in idem, Trade, Commodities and Shipping in the Medieval Mediterranean (Aldershot, 1997), no. .

[ 206 ] Changing Economic Patterns in Latin Romania tine attitudes, policies, and practices. The political and territorial fragmentation of Latin Romania created a climate of competition between Latin lords, which induced the latter to attract settlers, merchants, and ship operators, ensure them of safe conditions in their respective territories and ports, enable direct access to local producers, markets, and fairs, and enhance foreign purchases of local goods. In addition, these lords sought ways to stimulate investments, as well as transit trade and shipping through their ports. Their primary purpose was to increase thereby their own financial gains and fiscal revenue. However, there were some marked differences between the initial settlement patterns affecting non-Venetian and Venetian territories, respectively. Among the latter this was especially true with respect to Crete, due to the particular background and circumstances of its conquest. It is commonly believed that in the early thirteenth century Venice’s interest in Crete was exclusively or primarily stimulated by strategic considerations, namely, the island’s location at the crossing of important sea lanes, as well as the havens and logistic support it could offer for their control. However, there were also other weighty factors, largely overlooked until now, that prompted Venice to invest considerable means in the purchase of Crete from Boniface of Montferrat in 1204 and in the island’s conquest in the following years. Venice displayed substantial interest in the economic and fiscal exploitation of the island. This interest had been especially stimulated from the second half of the twelfth century by the acquaintance of Venetian merchants with the agricultural and pastoral resources of Crete, its growing production and exports, its trade networks, and its increasing role in trans-Mediterranean navigation between Italy and the Levant.56 The interplay between private and governmental factors with respect to Crete is further illustrated around the time of the island’s conquest. Venetian trade with Crete appears to have continued unabated, regardless of the political vacuum in the island resulting from the collapse of imperial power, partial Genoese occupation from 1206 to 1211, and Venice’s efforts to enlarge its rule beyond Candia, captured in 1207.57 Giovanni Corner was about to leave Venice for Crete in March 1205, after receiving 100 Venetian pounds in collegantia.58 Merchants accompanied the military expedition of 1209 to Crete, as suggested by two collegantia contracts concluded in Venice. According to one of them, Maria, wife of Doge Pietro Ziani, entrusted 120 Venetian pounds to Tommaso Viadro, an experienced merchant about to sail for Candia.59 At that time Venetian merchants must have traded in the territory already held by Venice and were presumably also involved in the provisioning of the military forces stationed in Crete. In the autumn of the same year, Venetians were engaging in trade between Candia and Alexandria, and 56

On Crete in Venetian trade and shipping in that period, see D. Jacoby, “Italian Privileges and Trade in Byzantium before the Fourth Crusade: A Reconsideration,” Anuario de estudios medievales 24 (1994): 349–56, repr. in idem, Trade, no. ; idem, “Byzantine Crete in the Navigation and Trade Networks of Venice and Genoa,” in Balletto, Oriente (as in note 10), 517–30, 537–40. 57 See below, p. 207. 58 DCV, no. 469. 59 A. Lombardo and R. Morozzo della Rocca, eds., Nuovi documenti del commercio veneto dei sec. XI–XIII (Venice, 1953) (hereafter NDCV ), no. 73: in presenti venturo stolo; no. 74: ire debebas in exercitu Veneciarum.

David Jacoby [ 207 ] in the following spring Giovanni Corner, already encountered five years earlier, invested in Constantinople in a business venture with Crete and Venice.60 Not surprisingly, therefore, the first wave of military settlers leaving Venice for Crete in the autumn of 1211 included merchants such as Tommaso Viadro, who had previously traded in Crete, and others who had surely obtained information about the economic potential of the island. In addition, merchants arriving on their own in Crete also settled on the island in that period.61 It is noteworthy that the charter delivered by Doge Pietro Ziani to the military contingent of 1211 contains several detailed provisions regarding the future settlers’ trade in the island’s products. One of them deals with state control over wheat exports, a sensitive issue connected with Venice’s food supply.62 Military settlers became involved in internal trade and in the export of local commodities shortly after their arrival in the island. Some of them also invested in maritime ventures and from 1218, or 1222 at the latest, even directly undertook trade journeys overseas.63 A similar conjunction of commercial, fiscal, and strategic factors and interplay between private and state initiatives provided the background for the invasion of Crete by Enrico Pescatore in 1206 and for his efforts to secure his rule over the island, which eventually collapsed in 1211. Although Pescatore’s expedition was a private enterprise, Genoa displayed vivid interest in it and provided large-scale naval, military, and financial assistance to ensure its success. In 1208 the Commune responded to Pescatore’s request by sending him ships, men-at-arms, and supplies, as well as money for the purchase of horses, obviously in Crete itself. In 1210 he requested further financial assistance and, in return, promised Genoa a quarter with its facilities in each Cretan city, jurisdiction in the whole island, as well as full exemption from taxes to Genoese merchants. The charter he delivered was obviously modeled after those obtained by Genoa in the Crusader Levant. Pescatore further promised the reimbursement in three yearly installments of the financial assistance provided by the Commune, to the amount of 18,000 Genoese pounds, beginning two years after the total subjection of the island to his rule. Finally, he stated that the Commune would inherit Crete should he die without legitimate heirs.64 Genoa’s interest in the economic and fiscal exploitation of the island was clearly con60

DCV, nos. 516, 518. See Jacoby, “Colonisation,” 299–300, 303–8. 62 G. L. F. Tafel and G. M. Thomas, eds., Urkunden zur a¨lteren Handels- und Staatsgeschichte der Republik Venedig (Vienna, 1856–57) (hereafter TTh), 2:132 (victualia, here clearly wheat), 140. 63 See Jacoby, “Colonisation,” 305–6, 308; M. Gallina, “Finanza, credito e commercio a Candia fra la fine del XIII secolo e l’inizio del XIV,” Memorie della Accademia delle Scienze di Torino, II. Classe di Scienze Morali, Storiche e Filologiche, 5th ser., 7–8 (1986): 13–21, 23–31, 41–68, passim. Borsari, Dominio, 85–87, suggests that those who settled in Crete opted for security with less income, rather for the high risk involved in maritime trade. This assumption is not convincing, nor is it plausible that the prestige of armed service acted as an additional inducement to settlement in Crete, considering the attitudes, values, and interests of the mercantile milieu. The case of Doge Ranieri Zeno to which Borsari refers cannot be considered typical. 64 See G. Gerola, “La dominazione genovese in Creta,” Atti dell’I. R. Accademia di Scienze e Lettere ed Arti degli Agiati in Roverete, 3d ser., 8.2 (1902): 134–75; Borsari, Dominio, 21–25, 27; D. Abulafia, “Henry Count of Malta and His Mediterranean Activities, 1203–1230,” in A. T. Luttrell, ed., Medieval Malta: Studies on Malta before the Knights, Supplementary Monographs of the British School at Rome (London, 1975), 113–19, repr. in D. Abulafia, Italy, Sicily and the Mediterranean, 1100–1140 (London, 1987), no. ; Maltezou, “Creta,” 763–67. 61

[ 208 ] Changing Economic Patterns in Latin Romania nected with the expectations of Genoese merchants. They had traded there since around the mid-twelfth century.65 Pescatore’s occupation of Crete in 1206 and his power over large sections of it in the following five years prompted private Genoese interest and investments in his Cretan venture, as in 1210.66 It presumably also stimulated an increase in trade with the island, some merchants personally sailing that same year to Crete on three vessels leaving with reinforcements and provisions for Pescatore, one of them the Glauca.67 Pescatore’s eviction from Crete in 1211 was followed by a sharp reduction in the volume of Genoese trade with the island, which remained minimal until the Genoese began to export Cretan wine in the fifteenth century.68 The territorial extent and the large population of Crete, presumably also the fierce struggle of 1206–11 with Genoa for the possession of the island, induced Venice to implement an original and unique, highly structured and institutionalized immigration and settlement policy. Combined with strict control over land and peasantry, it was aimed at promoting the state’s political and economic interests in the island. The settlement of Venetians liable to serve in a military capacity in return for land granted by the state, initiated in 1211, was partly based on somewhat earlier precedents in Venice’s portion of the Latin Empire of Constantinople.69 The novelty in Crete, in which no Venetian settlers had resided prior to 1204,70 was that both the immigration and settlement of these individuals were state-sponsored, state-organized, and supported by state subsidies. From 1211 to 1252 Venice established several small military contingents numbering a total of 189 men granted fiefs, who were accompanied by one or two horsemen, and 60 foot soldiers offered sergeantries. In this framework the repopulation and reconstruction of Canea after 1252 was part of Venice’s endeavor to consolidate its rule over the western section of the island. Originally the fiefs or sergeantries were to be granted exclusively to Venetian citizens, each of them to a single settler. However, already a few years after the beginning of the military colonization process, we find some Venetians holding several military tenements simultaneously, as well as non-Venetians originating from northern and central Italy among the holders of such tenements.71 Venice resorted 65

See Jacoby, “Byzantine Crete,” 530–40. A contract to this effect in Gerola, “Dominazione,” 158; Abulafia, “Henry Count of Malta,” 116–17. There is reason to believe that there were additional ones. M. Balard, “Les Ge´nois en Romanie entre 1204 et 1261: Recherches dans les minutiers notariaux ge´nois,” Me´lRome 78 (1966): 474–75, repr. in idem, La Mer Noire et la Romanie ge´noise, XIIIe–XVe sie`cles (London, 1989), no. , suggests that in 1210 Guglielmo Porco, admiral of the kingdom of Sicily and a close associate of Pescatore, obtained a loan to help him in Crete. 67 H. C. Krueger and R. L. Reynolds, eds., Notai liguri del sec. XII e del sec. XIII, vol. 6, Lanfranco (1202– 1226) (Genoa, 1953), 1: no. 638: a young man from Rapallo leaving for the island, whether to fight or trade, leases a piece of land until his return; no. 652: accommendatio of 12 Genoese pounds for trade in Creti in nave Glauca et inde quo iero causa negotiandi; Balard, “Ge´nois en Romanie,” 474. 68 On which see D. Jacoby, “Creta e Venezia nel contesto economico del Mediterraneo orientale fin alla meta` del Quattrocento,” in Venezia e Creta, ed. G. Ortalli (Venice, 1998), 86, 93–94. 69 See D. Jacoby, “The Venetian Presence in the Latin Empire of Constantinople (1204–1261): The Chal¨ B 43 (1993): 144–45, 154–61. lenge of Feudalism and the Byzantine Inheritance,” JO 70 See Jacoby, “Italian Privileges,” 365–66. 71 See Jacoby, “Colonisation,” 306–11, 313. The figures of military settlers are based on name lists. F. Thiriet, “Recherches sur le nombre des ‘Latins’ immigre´s en Romanie gre´co-ve´nitienne aux XIIIe–XIVe sie`cles,” in Byzance et les Slaves: Etudes de civilisation, Me´langes Ivan Dujcˇev (Paris, 1979), 427, suggests higher figures which, however, are based on the numbers of expected settlers, far larger than those of the actual ones. In 1294 66

David Jacoby [ 209 ] again to state intervention from 1301 to 1324 by settling in its colonies of Coron and Modon 124 men-at-arms and craftsmen, among them woodworkers, metalworkers, and stonemasons, whose activities would support specific aspects of local defense or the servicing of naval forces. All these settlers received an annual salary that ensured their service in the local militias, yet were allowed to work for a daily wage. With their respective families, their total number reached around four hundred people.72 In the fourteenth century Venice sought to encourage foreign Latins to settle in its colonies and outposts of Latin Romania by granting those who undertook to reside there for at least ten years the privileges of Venetian nationality in Romania, the status enjoyed by the subjects of Venice’s colonies. In addition to diplomatic protection, this status ensured them the same fiscal privileges as Venetian citizens in Romania, yet not in Venice itself; nor did it confer upon them the commercial advantages enjoyed by citizens. After the Black Death, Venice offered Latins settling in its colonies full Venetian citizenship. These measures had only very limited success.73 It should be stressed that in Venetian territories the ongoing process of spontaneous nonmilitary immigration of Venetians and other Latins, primarily motivated by private economic considerations, contributed far more than statesponsored immigration to Latin demographic growth and resulted in a more diversified population.74 This is clearly attested for Coron and Modon.75 Latins settling in Crete displayed an obvious preference for urban residence. Their heaviest concentration occurred in Candia, while others resided in Canea, Rethymno, Sitia, and some inland cities. The overwhelming majority of military settlers lived in a house in Candia or Canea that was included in their tenement. Many of them were involved in trade.76 Nevertheless, some settlers, like Baldovino Lombardo, apparently and 1299 military equipment for a total of some 500 men was sent to Crete: R. Cessi, ed., Deliberazioni del Maggior Consiglio di Venezia (Bologna, 1931–50), 3:354, no. 136, and 447, no. 45. Since it was to be distributed among Cretan fiefholders as well as burgenses, it does not provide any indication about the number of the former, contrary to Thiriet, ibid., 430. For the decades following the Black Death, Thiriet, ibid., 430, postulates that each member in the feudatories’ council had an average of five family dependents, without taking into account that several members of the same household served on that council. On the territorial aspects of military settlement, see C. Maltezou, “Concessio Crete. Parathrh´ sei" sta` e“ggrafa dianomh'" feou´ dwn stou` " ´ a G. Kalokairi´nou (Irakleion, prw´ tou" Benetou` " ajpoi´kou" th'" Krh´ th",” in Loibh` eij" mnh´ mhn Andre j 1994), 107–31. 72 See A. C. Hodgetts, “The Colonies of Coron and Modon under Venetian Administration, 1204–1400” (Ph.D. diss., University of London, 1974), 151–52, 156, 355. 73 Jacoby, “Etats,” 20. On the distinction between Venetian citizenship and nationality in Romania, see D. Jacoby, “Les Ve´nitiens naturalise´s dans l’Empire byzantin: Un aspect de l’expansion de Venise en Romanie du XIIIe au milieu du XVe sie`cle,” TM 8 (1981): 217–35, esp. 219–20, repr. in idem, Studies on the Crusader States and on Venetian Expansion (Northampton, 1989), no. . On the practical implications, see Thiriet, Romanie, 279–82. 74 Both the demographic and military importance of state-sponsored settlement in Crete has been grossly overrated: see Jacoby, “Colonisation,” 312–13. The estimates of Latin population in Romania suggested by Thiriet, “Nombres,” 428, are purely hypothetical. They are based on the cumulative number of merchants attested over a long period, which does obviously not reflect the number of Latins at any given moment. On variety of origin among the settlers, see above, p. 208. 75 See below, p. 222. 76 Jacoby, “Etats,” 19. Note that in 1299 the Commune ordered that 110 crossbows out of a total of 430 should be sent to Canea, which clearly points to a much smaller number of Latin settlers in this city: Cessi, Deliberazioni, 3:447, no. 45. According to the rule, two Venetian brothers should have resided in the city of Canea

[ 210 ] Changing Economic Patterns in Latin Romania resided or at least spent a few months a year on their estates. In 1285, during a Greek rebellion, Lombardo requested state funding for the building of a defensive tower to enhance his security, implying that his residence in the countryside would strengthen Venice’s military position in the area.77 In the autumn of 1363, shortly after the great rebellion of St. Tito had begun, a number of military settlers were staying in villages included in their respective tenements, presumably in order to inspect them or to collect taxes and payments, yet not as permanent residents.78 Latins who were not military settlers seem to have been established in larger numbers in the countryside in order to engage in land cultivation and the raising of animals or to practice crafts using local raw materials, such as the dressing of hides and tanning.79 However, at times Latin surnames in rural areas do not necessarily point to Latin identity, since it was customary for the illegitimate offspring of Latin fathers and Greek mothers to adopt the former’s surname, which did not prevent them from remaining in the latter’s community.80 In any event, the number of Latins residing in the Cretan countryside may have increased in the fourteenth century, during which security conditions improved on the whole. On the other hand, it is unlikely that Latins would have settled outside Coron or Modon in the small Venetian enclaves of southern Messenia. Surprisingly, before 1204 Latin temporary or permanent settlement in the territories later ruled by feudal lords was limited to Thebes and Corinth, with the possible addition of Sparta. Venetians apparently resided in Thebes almost continuously from the 1150s at the latest until the Fourth Crusade and had two churches there. They also had one church in Corinth and another in Sparta.81 Latin settlement after the Fourth Crusade in the same non-Venetian territories became substantially larger, far more varied in nature, and more dispersed. Fairly small Latin nuclei of Western knights and sergeants owing military service established themselves in castles and fortified mansions in the countryside or in the acropolis of the main cities of major lordships, as in Thebes and Acrocorinth. Shortly after the conquest, some of these nuclei counted less than twelve individuals, yet must have subsequently grown with the arrival from the West of the latter’s relatives and additional immigrants.82 Other knights permanently resided in the major after having inherited sergeantries in this city’s district in 1334. In 1339, however, they were allowed to remain in Candia, where for a long time they had been conducting trade: Theotokes, Qespi´smata, 1:185, no. 14. 77 Excerpt from his request ed. by Borsari, Dominio, 83. 78 Lorenzo de Monacis, Chronicon de rebus venetis, ed. F. Cornelius (Corner) (Venice, 1758), 179: “omnes nobiles qui extra civitatem diffusi erant per sua casalia.” They were there on their own since they, yet not their family members, were murdered by the insurgents who intended to kill all Latins. 79 A. Lombardo, ed., Imbreviature di Pietro Scardon (1271), Documenti della colonia veneziana di Creta 1 (Turin, 1942), nos. 24, 33, 164, 266, 294, 440; Marcello, nos. 256–57, in 1280; E. Santschi, “Contrats de travail et d’apprentissage en Cre`te ve´nitienne au XIVe sie`cle d’apre`s quelques notaires,” Revue suisse d’histoire 19 (1969): 65, for the 14th century. 80 See Jacoby, “Etats,” 29–30; S. McKee, “Greek Women in Latin Households of Fourteenth-Century Crete,” JMedHist 19 (1993): 229–30. 81 DCV, no. 166: a Venetian having served for two years as commercial agent in Thebes prior to May 1165; see also Jacoby, “Silk in Western Byzantium,” 494–96. On the churches, see also S. Borsari, Venezia e Bisanzio nel XII secolo: I rapporti economici, Deputazione di storia patria per le Venezie, Miscellanea di studi e memorie 26 (Venice, 1988), 41; R.-J. Lilie, “Die lateinische Kirche in der Romania vor dem vierten Kreuzzug: Versuch einer Bestandaufnahme,” BZ 82 (1989): 202–6, 209–11. 82 The figure twelve is implied by a letter of Pope Innocent III, Epist., XIII.16, PL 216:216. See Jacoby, “Etats,” 19.

David Jacoby [ 211 ] city located in the vicinity of their fiefs, such as Negroponte, Naxos, or Patras. Some had a secondary residence in Andravida, Chiarenza, Modon, or Coron.83 The Sienese and Florentine merchants and bankers operating in Chiarenza who obtained fiefs from the princes of the Morea in the second half of the thirteenth and in the fourteenth century maintained their residence in that city.84 In the principality there may have been 170 knight-fiefs around 1225, with a total of some 450 horsemen. In the 1320s or 1330s the Venetian Marino Sanudo estimated their number at between seven hundred and one thousand. In 1338 or somewhat later the number of fiefs held by the vassals and rearvassals of the prince of the Morea was calculated at more than one thousand.85 It is unclear whether these numbers included horsemen endowed with money-fiefs instead of estates, once the land available for distribution had been exhausted. The number of these horsemen at any given time is unknown.86 The initial settlement pattern of the knights in non-Venetian territories, concurrent with the territorial fragmentation typical of feudal landholding, as well as military pressure in some regions, led to the repair of existing castles and the construction of new ones. Castles served as secure bases from which neighboring areas could be controlled, defended, or attacked, yet also functioned as administrative and consumption centers. In addition, numerous mansions and towers were erected in the countryside. The distribution, location, and architectural variation of the eighty or so massive towers in Attica and Boeotia and the other fifty in Euboea suggest that most of them were not planned to fulfill a strategic purpose, but rather meant to be status symbols or to serve as administrative bases and provide storage for resources.87 By contrast, a new phase of construction began in several areas in the late fourteenth century. It was carried out or sponsored by Venice for military purposes in response to the Turkish threat.88 Some of these buildings contributed indirectly to a growth in rural output and improved marketing by providing protection and enabling a more efficient collection and storage of products. On the other hand, their construction and maintenance diverted substantial means in labor and cash from investments in the expansion of the rural infrastructure. Seignorial and state con83

Ibid.; also Jacoby, “Encounter,” 901 n. 134; idem, “Migration,” 105–6. See below, p. 224. 85 See Jacoby, “Etats,” 20–21; A. Ilieva, Frankish Morea (1205–1262): Socio-Cultural Interaction between the Franks and the Local Population (Athens, 1991), 165–68; Marino Sanudo Torsello, Istoria del Regno di Romania, ed. C. Hopf, Chroniques gre´co-romanes ine´dites ou peu connues (Berlin, 1873), 102–3, obtained information from Marco II Sanudo, duke of Naxos. 86 On money-fiefs, see Jacoby, Fe´odalite´, 135; idem, “Encounter,” 887. 87 Livre de la conqueste de la prince´e de l’Amore´e, Chronique de More´e (1204–1305), ed. J. Longnon (Paris, 1911), para. 218: “li baron dou pays et li autre gentil homme si commencerent a faire fortresses et habitacions, quy chastel, quy maisons sur sa terre.” The passage refers to the period after the capture of Monemvasia by Prince William II in 1248, a dating contested by H. A. Kalligas, Byzantine Monemvasia: The Sources (Monemvasia, 1990), 86–94. In any event, construction began much earlier: see A. Bon, “Forteresses me´die´vales de la Gre`ce centrale,” BCH 61 (1937): 136–208; Bon, More´e, 601–84; P. Lock, “The Frankish Towers of Central Greece,” BSA 81 (1986): 101–23; idem, “The Medieval Towers of Greece: A Problem in Chronology and Function,” Mediterranean Historical Review 4 (1989): 129–45, repr. in Arbel, Hamilton, and Jacoby, Latins (as in note 15); P. Lock, “The Towers of Euboea: Lombard or Venetian; Agrarian or Strategic,” in Lock and Sanders, Medieval Greece (as in note 11), 107–26; Lock, Franks in the Aegean, 75–80, 82. See also P. Lock, “The Military Orders in Mainland Greece,” in The Military Orders: Fighting for the Faith and Caring for the Sick, ed. M. Barber (Aldershot, 1994), 333–39. 88 See Lock’s studies in previous note, and Hodgetts and Lock, “Fortifications,” 77–90. 84

[ 212 ] Changing Economic Patterns in Latin Romania struction, the latter in Venetian territories, was largely achieved by compulsory peasant labor, although it occasionally also relied on salaried work.89 Latin commoners spontaneously immigrated to non-Venetian territories in Latin Romania in numbers exceeding by far those of the knights. They established themselves exclusively in cities, some of which the early settlers had visited or inhabited before 1204. These immigrants displayed a marked preference for ports such as Negroponte and Corinth, which, like Candia, Coron, and Modon, had already previously functioned as collection and distribution centers, as outlets for their respective hinterland and neighboring islands, or as regular stopovers for vessels sailing between Italy and eastern Mediterranean ports. Chiarenza, a new port founded on the western coast of the Peloponnesos after the Latin conquest, prospered thanks to favorable geopolitical and economic conditions, attracted settlers, and became the main emporium of Frankish Morea by the second half of the thirteenth century. Latin commoners also settled in cities manufacturing silk textiles, such as Thebes and Negroponte, as well as in old and new political and administrative centers, namely, Andravida, Thebes, Athens, Corinth, and Naxos.90 These cities also attracted Greek immigrants from the countryside wishing to take advantage of new opportunities in the exercise of crafts and in the service sector, especially in trade and transportation.91 The most spectacular population growth occurred in Candia and resulted in the development of a new suburb beginning in the second half of the thirteenth century.92 Generally speaking, Latin immigration and related demographic and economic developments in Latin Romania generated important shifts in the relative importance of cities, as well as in the geographic distribution and the hierarchy of consumption, industrial, and trade centers. Despite various manifestations of continuity in the countryside of Latin Romania, noted above, a partial restructuring at the basic level of management and exploitation was unavoidable after the Latin conquest. This appears to have been especially the case in confiscated crown and other extensive estates of Constantinopolitan landlords. Their fragmentation into smaller units and the division of their workforce among new landlords must have often prevented the upholding of large-scale compulsory labor services. Yet there also was a tendency to replace services and so-called gifts owed by the peasants with cash payments, in particular in Crete where the state-granted military tenements were rather small and at best moderate-sized. At times commutation was applied even when a fairly large labor force was available, as in the late thirteenth century at Lombaro, 89

On the economic implications of the latter, see below, p. 229. See Jacoby, “Migration,” 103–8, 112–13; idem, “Etats,” 19–22. 91 For Crete, see Santschi, “Contrats,” 34–74, esp. 59, 65; C. Maltezou, “Me´tiers et salaires en Cre`te ve´nitienne (XVe sie`cle),” ByzF 12 (1987): 322–23, 326, 330. 92 Gallina, “Finanza,” 8, 10. Candia’s expansion is illustrated by the inclusion of rural churches apparently constructed in the Byzantine period within the newly built urban territory. See M. Georgopoulou, “The Topography of Chandax, Capital of Crete in the Second Byzantine Period,” Cretan Studies 4 (1994): 91–136, esp. 116–23. Incidentally, the author (ibid., 102 and n. 41) cites a 13th-century Venetian text that she wrongly interprets as ascribing the building of Candia’s fortifications to Enrico Pescatore, whose activity in Crete extended from 1206 to 1211. See above, pp. 200, 207. However, the text refers to Candia as being nondum (instead of nundum [sic]) muribus circumdata, “not yet surrounded by walls” at that time. This seems to contradict the archaeological evidence, on which see ibid., 102–3. 90

David Jacoby [ 213 ] the village held by Andrea Corner, which included thirty-three peasant households before it was hit by the rebellion of Alexios Kallergis.93 In other cases commutation was implemented only when the peasants could not carry out their labor service, evaluated in 1281 at 4 hyperpers per year in the village of Apano Trifora.94 In Frankish Morea villeins were required to perform their servicium personale. The latter’s evaluation at 5 hyperpers per peasant unit appears only in official documents computing the income of fiefs and, therefore, does not necessarily imply that it was replaced by cash payments.95 Commutation was a convenient device saving the cost of supervision and the need to coerce peasants to fulfill their obligations. Landlords applying it relied more heavily upon the lease of domain land to peasants and upon profit-sharing ventures with them in land cultivation and animal husbandry.96 Far more important and wide ranging, at both the local and regional levels, were changes in the channeling and destinations of agricultural and pastoral surpluses produced by the confiscated estates of large absentee landlords. Before 1204 these surpluses were partly, if not entirely, intended for self-supply, mainly in the capital, or else for gifts to ecclesiastical institutions. Thus, for instance, until 1171 the monastery of St. John of Patmos and other institutions benefited from yearly allowances of wheat from crown lands (basilikai` ejpiske´ yei") in Crete. These surpluses remained outside the commercial circuit and were conveyed to their destinations within noncommercial networks. On the other hand, it may sometimes have been more convenient and profitable for absentee landlords residing in Constantinople to sell them at rural markets and fairs or in urban centers located in the vicinity of their estates, and save thereby the cost and nuisance involved in their transportation to the capital.97 They could then use the proceeds of sales for the purchase of supplies and consumption elsewhere. The Latin conquest and the fragmentation of their extensive estates after 1204 severed the link between the latter and Constantinople. A portion of the surpluses had to be redirected toward the new Latin landlords and the retinues established on their land or in nearby cities, while the rest was transferred to markets and fairs in the region. The volume of produce affected by these changes cannot be assessed, yet may have been quite substantial in areas in which there had been a heavy concentration of Constantinopolitan estates. 93 See above, note 22. For a contemporary case, see above, note 49. Although situated outside the region upon which this study focuses, the case of Lampsakos in western Asia Minor may prove instructive. In 1219 cash equivalents were established for labor services and gifts owed by seventy-two peasant households, commutation being possibly adopted afterwards: see Jacoby, “Presence,” 175–77. 94 Partial edition by Borsari, Dominio, 90 n. 105, who reads fourteen days per year, and full edition by Z. N. Tsirpanles, “Kata´ stico ejkklhsiw'n kai` monasthriw'n tou' Koinou'” (1248–1548). Sumbolh` sth` mele´ th tw'n sce´ sewn Politei´a" kai` jEkklhsi´a" sth` benetokratou´ menh Krh´ th (Ioannina, 1985), 261, no. 187, who reads forty-five days. Both figures are problematic, since the number should either be one day per week or else one or several days per month and thus twelve or a multiple of twelve: see Jacoby, “Presence,” 176–77. 95 See Carile, Rendita, 95–98; my review of the latter (cited above, note 15), 359; Longnon and Topping, Documents, 271–72; Jacoby, “Migration,” 123. 96 On which see above, note 15. 97 D. Tsougarakis, ed., The Life of Leontios, Patriarch of Jerusalem: Text, Translation, Commentary (Leiden, 1993), 102, chap. 61, and 190–91, commentary. See also P. Magdalino, “The Grain Supply of Constantinople, NinthTwelfth Centuries,” in Constantinople and Its Hinterland, ed. C. Mango and G. Dagron (Aldershot, 1995), 37–39, 45–46, who, however, does not deal with sales by lay landlords.

[ 214 ] Changing Economic Patterns in Latin Romania On a more general level, the patterns of Latin settlement and economic activity generated some significant changes in the siting and relative importance of markets and fairs. As a result, existing commercial routes were partly deflected to new courses and destinations. Numerous markets were presumably held in the territories included in Latin Romania both before and after the conquest, yet only few are documented, namely, in Corinth, Cosmina, Vasilicata, and Androusa in the Peloponnesos.98 The term ejmpo´ rion, used for several places in that region, indicated the existence of a market or a settlement serving as marketplace.99 As elsewhere in the empire, fairs too must have been quite common before the Fourth Crusade in the territories conquered by the Latins. However, only three of them are documented in the Peloponnesos in the thirteenth and fourteenth centuries, none of them being attested either earlier or later.100 The small village fair at Macona in southern Messenia, recorded in 1338, yielded an annual revenue of only 8 hyperpers. It is not clear whether it was an annual event or was held several times a year, in connection with the rural calendar.101 The annual fair at Vervena, attested in 1296, was undoubtedly far more important. It took place in the Frankish part of the Skorta region some 10 to 15 km south of Andritsaina, in the vicinity of the Frankish-Byzantine border newly delineated in the late 1270s, in the wake of the empire’s northward expansion in the central Peloponnesos.102 If this fair perpetuated a Byzantine institution, it certainly gained in importance after the 1270s, in view of its location. Yet it may also have been established around that time in order to attract subjects of both the Frankish

98 Longnon and Topping, Documents, 139, line 36; 146, line 20; 162, line 8 (lo comerchio del merchato); 168, line 2; see also 275–76. Silk cocoons were traded at the market of Androusa in 1328: C. Hodgetts, “Venetian Officials and Greek Peasantry in the Fourteenth Century,” in KAQHGHTRIAÚ Essays Presented to Joan Hussey for Her 80th Birthday, ed. J. Chrysostomides (Camberley, 1988), 493, doc. , para. 1. 99 See ODB 1:694, s.v. “Emporion” and references in A. Ilieva, “Images of Towns in Frankish Morea: The Evidence of the ‘Chronicles’ of the Morea and of the Tocco,” BMGS 19 (1995): 106–7. 100 On Byzantine fairs, see S. Vryonis Jr., “The Panegyris of the Byzantine Saint: A Study in the Nature of a Medieval Institution, Its Origins and Fate,” in S. Hackel, ed., The Byzantine Saint, Studies supplementary to Sobornost 5 (1981), 196–226; A. E. Laiou, “Ha¨ndler und Kaufleute auf dem Jahrmarkt,” in Fest und Alltag in Byzanz, ed. G. Prinzing and D. Simon (Munich, 1990), 53–70, 189–94 (notes), repr. in eadem, Gender, Society ˇ ivojinovic´, “The Trade of Mount Athos Monasand Economic Life in Byzantium (Aldershot, 1992), no. ; M. Z ¨ ffentlichkeit,” in Mentalita¨t und teries,” ZRVI 29–30 (1991): 112–14; K.-P. Matschke, “Die spa¨tbyzantinische O Gesellschaft im Mittelalter: Gedenkschrift fu¨r Ernst Werner, ed. S. Tanz (Frankfurt am Main, 1993), 159–64. On fairs in the 14th-century Greek Despotate of the Morea and the 17th-century Peloponnesos, see Zakythinos, Despotat grec, 2:253–54. 101 Longnon and Topping, Documents, 64, line 12: “Panegerii de Amachonu reddunt in porcione yperpera quatuor.” Note the plural, also used in the following case which clearly refers to annual fairs. The sum mentioned represented half of the total revenue, the fief and its revenue being divided in equal shares between two lords: ibid., 165, lines 22–23, and 166, lines 8–26. For the location, see ibid., 250. On contemporary village fairs in the empire on the land of monasteries, also yielding small annual revenues, see Laiou, “Ha¨ndler,” ˇ ivojinovic´, “Mount Athos Monasteries,” 112 n. 49. 62–63; Z 102 Livre de la conqueste, para. 802–3: “les foires que on claime panejours, lesquelles se font au jour de huy au demie juyn.” The sale of silk is explicitly mentioned: see D. Jacoby, “Silk Production in the Frankish Peloponnese: The Evidence of Fourteenth-Century Surveys and Reports,” in Travellers and Officials in the Peloponnese: Descriptions-Reports-Statistics, in Honour of Sir Steven Runciman, ed. H. A. Kalligas (Monemvasia, 1994), 45, 59–60, repr. in idem, Trade, no. . I correct here the location of Vervena mentioned in that study by relying on Bon, More´e, 169–70, 380–89, esp. 387–88, 512–15, who presents convincing geographic and archaeological arguments against a proposed siting of Vervena in Arkadia. On the Byzantine advance, which went beyond Kalavryta, see Bon, More´e, 144–46. It excluded the holding of a Frankish fair in 1296 in Arkadia, which by then had already been in Byzantine hands for some twenty years.

David Jacoby [ 215 ] principality and the Greek despotate and thereby promote exchanges between these territories.103 It should be noted that the region of Vervena was fairly close to the borderland further west in which later, in 1322, Frankish and Greek lords shared the revenue of several villages of the principality.104 The annual fair of St. Demetrius, attested in 1338, was presumably the most important of the three fairs of the Peloponnesos documented in the Latin period. It was held inland at some distance from Chiarenza. Venetian merchants attending it were supposed to return with their goods to that city within eight days after its conclusion.105 The timing of this fair in late October coincided with the marketing of agricultural and pastoral produce for export and the passing of Venetian state galleys and other ships returning from the eastern Mediterranean to the Adriatic.106 In view of the Byzantine connotation of its name, one would assume that this fair was inherited from the Byzantine period, yet this is far from certain, in the same way as the use of the Greek term panhgu´ ri" in its Latin, French, or Italian versions does not offer any clues with respect to continuity or the creation of new fairs in the Latin period.107 The St. Demetrius fair mentioned here may have developed from a local or regional event into a more important gathering after the foundation of Chiarenza, yet it is not excluded that it was established by a Frankish lord in order to take advantage of that port. As in the Byzantine period, changing circumstances induced landholders to transfer existing fairs to new locations or to establish new fairs on their estates.108 In the Venetian territories, developments regarding markets and fairs appear to have been quite different from those occurring in the Peloponnesos. Fiscal expediency prompted the state to impose the channeling of all commercialized rural products to urban markets by land and by sea, regardless of their ultimate destination, in order to supervise and tax their sale. In Crete the maritime transportation of these products was directed toward the main ports of the island, namely, Candia, Canea, Rethymno, and Sitia.109 Two late thirteenth-century customs lists record taxes levied at the land gate and in the harbor of Candia.110 Goods brought for sale to the city had to be weighed or measured at the Commune’s official station located at the marketplace. This rule applied 103 The proximity of the border is further illustrated by the events that followed the fair: Livre de la conqueste, para. 804–25. Note the chronicler’s remark that “nowadays [the fairs] are held in mid-June”: see text in previous note. It is impossible to determine whether this remark already appeared in the original version of the chronicle, in which case it may reflect a change in timing related to the events of the 1270s, or in the abridged version of 1320–24. On the French versions of the chronicle, see D. Jacoby, “Quelques conside´rations sur les versions de la ‘Chronique de More´e,’” JSav (1968): 133–50, 181–89, repr. in idem, Socie´te´, no. . 104 See Jacoby, “Coseigneurie,” 114–15. 105 Baron Blanc, ed., Le flotte mercantili dei Veneziani (Venice, 1896), 59. See also Jacoby, “Silk Production,” 60. 106 On state galleys and timing, see below, pp. 222, 228. 107 For Latin and French, see above, notes 101 and 102. The 14th-century Francesco Balducci Pegolotti, La pratica della mercatura, ed. A. Evans (Cambridge, Mass., 1936), 17, lists several equivalents of Tuscan mercato and fiera, among them panichiero in grechesco. 108 See Laiou, “Ha¨ndler,” 54–57. 109 Note the provisions of 1316–17 prohibiting the loading or unloading of goods in the bay of Dermata, to the west of Candia’s harbor, only fishing boats being allowed to anchor there: P. Ratti Vidulich, ed., Duca di Candia: Bandi, Fonti per la storia di Venezia, Sez. I, Archivi pubblici (Venice, 1965) (hereafter Bandi), nos. 144, 174. 110 Ed. E. Gerland, Das Archiv des Herzogs von Kandia (Strasbourg, 1899), 108–9, and for their dating to 1298–99, see 107 n. 1 and 135.

[ 216 ] Changing Economic Patterns in Latin Romania regardless of whether they would be sold or delivered there to a specific customer, according to an earlier agreement, or be transported afterwards to a private home, a private warehouse, or to the harbor for export. The weighing or measuring was carried out by state-appointed officials in the presence of state assessors, who also acted as official brokers on behalf of the Commune. We may safely assume that similar practices existed in the other ports and inland cities of Crete. In any event, they appear in Coron and Modon.111 In these circumstances, it is rather unlikely that rural fairs should have survived in Crete or that any subsisted outside the cities of Coron and Modon in the exiguous territories of Venetian Messenia.112 As before 1204,113 rural surpluses in Latin Romania were partly conveyed to nearby rural or urban markets and fairs by producers, whether peasants, herdsmen, or craftsmen, who sold their own products or else delivered them to a specific customer, in conformity with a contract between them.114 Some producers also acted as middlemen between their peers and markets. According to a Venetian document of 1328, a number of Greek peasants from Venetian Messenia sold in Coron their own silk cocoons, as well as others bought in neighboring Frankish Morea.115 Individuals wishing to avoid the costly and time-consuming transportation and marketing of small amounts of products relied on professional merchants and carriers, such as the traveling Venetian merchants visiting the fair of St. Demetrius in the area of Chiarenza in 1338.116 Professional merchants settled in Latin Romania took advantage of their greater ability to offer transportation services. In Crete a number of them combined trade with the holding of military tenements.117 Large landholders also served occasionally as middlemen between rural producers on the one hand and urban markets and exporters on the other, concentrating large amounts of goods in their hands. Unless delivering their surpluses to a specific customer, many peasants were left with fairly small amounts of them after paying rents and taxes in kind. They may have found it convenient, therefore, to sell them to large landholders rather than to market them on their own. Some Cretan documents refer to such deals.118 A similar pattern may be assumed for the produce of small estates. From 1341 to 1344 the serventaria of Castri in the region of Milopotamo produced for its holder 791 misure of wheat or an annual average of about 264 misure, around 3.380 metric tons, from which a certain amount had to be deducted for personal consumption.119 At times, however, 111 See D. Jacoby, “Cretan Cheese: A Neglected Aspect of Venetian Medieval Trade,” in Medieval and Renaissance Venice, ed. E. E. Kittel and T. F. Madden (Urbana-Chicago, 1999), 54–55; Hodgetts, “Colonies,” 448–53. 112 In 1473 the Venetian Senate deplored that for some time Cretan merchants were buying and storing local products in the countryside, thereby depriving the state of revenue and Candia of its economic activity: H. Noiret, ed., Documents ine´dits pour servir a` l’histoire de la domination ve´nitienne en Cre`te de 1380 a` 1485 (Paris, 1892), 526–27. 113 For the Byzantine period, see Laiou, as above, note 100. 114 On the latter, see below, p. 218 and note 129. 115 Hodgetts, “Officials,” 493, para. 1; see also Jacoby, “Silk Production,” 44–45, 60–61. 116 See above, p. 215. 117 See above, pp. 207, 209 and note 76. 118 See Jacoby, “Cretan Cheese,” 52–53. 119 E. Santschi, ed., Re´gestes des arreˆts civils et des me´moriaux (1363–1399) des archives du duc de Cre`te, Bibliothe`que de l’Institut helle´nique d’e´tudes byzantines et post-byzantines de Venise 9 (Venice, 1976), Memoriali, no. 735; for the location, see ibid., no. 1100. The wheat misura of Crete is estimated at 12.8 kg: see E. Schilbach, Byzantinische Metrologie (Munich, 1970), 94–96, 149–50.

David Jacoby [ 217 ] large landholders must have applied pressure to acquire from their own as well as from other villeins the products they wished to market. Such a practice may have been fairly common when they had agreed to deliver specific amounts of products to a private customer or a merchant and had to meet a deadline.120 A similar pattern was even more imperative in Crete with respect to wheat. From the late thirteenth century, Latin military settlers and the powerful Cretan archontes of the Kallergis clan undertook to deliver each year specific amounts of wheat to the Commune, an obligation they had to fulfill at all cost.121 In these circumstances, pressure on peasants appears all the more likely on the hereditary estates owned by the Kallergis, in which they had preserved and even reinforced their traditional standing and authority.122 The function of large landowners as middlemen in the marketing of rural produce is also documented for Frankish Greece. Large landowners did not directly engage in silk growing, which was exclusively carried out by peasants. Moreover, they apparently obtained only small amounts of silk as payment in kind for the use of their processing facilities by the peasants, less than 15 light pounds per village as illustrated by some cases in the 1370s which, admittedly, do not necessarily reflect an average.123 In any event, we may safely assume that the Greek archon of Frankish Morea who in 1296 came to the fair of Vervena would not have bothered to attend it, had he not assembled a sizable quantity of silk for sale.124 In addition to the small amounts he had collected as payment from silk growers, he must have purchased silk from his and other villeins willing to sell it or compelled them to do so under pressure. It is not excluded, though, that the archon also acted as middleman between other landholders of his area, who would have collected silk in a similar way, and merchants attending the fair. The same functions of middleman may be assumed with respect to John Laskaris Kalopheros, a Byzantine adventurer who had wedded the daughter of Erard III Le Maure, one of the most powerful barons of Frankish Morea, and had himself become a fiefholder in the principality. In 1381 Kalopheros sold in Modon 2,773 light pounds or around 950 kg of raw silk, quite a sizable amount.125 A document of 1328 regarding Munista in Venetian Messenia reveals that individual peasant households produced fairly small amounts of cocoons, between 10 and 25 light pounds, which in turn yielded between ca. 2 and ca. 5 light pounds of raw silk.126 Judging by these figures, the silk sold by Kalopheros would have represented the production of more than 550 peasant households. Only a few landlords of Frankish

120 See Jacoby, “Cretan Cheese,” 52–53; Marcello, no. 125, a contract of 1279 specifying that the price of cheese and wool would be determined “ad eam videlicet racionem quam villani de Sythea vendiderint eorum dominis.” The use of pressure is hinted by the following: “ita tamen quod ipsi villani vendidere non debeant dictum caseum neque lanam,” the reference being to free marketing by the peasants themselves. See also ibid., no. 213, drafted in 1280: a feudatory promises to deliver his own cheese and wool as well as those of his villeins. 121 On which see below, p. 223. 122 See above, p. 199. 123 On such payments in Frankish Morea, in all likelihood perpetuating a Byzantine practice, see Jacoby, “Silk Production,” 51–53; on amounts, see ibid., 57. 124 See below, note 214. 125 Jacoby, “Silk Production,” 55, 60. The light pound of Modon weighed 343 g: see ibid., 55 n. 52. 126 For these and other production figures from Frankish Morea, see ibid., 57. The ratio between the weight of cocoons and silk, respectively, is around 5:1.

[ 218 ] Changing Economic Patterns in Latin Romania Morea would ever have had so many households on their estates.127 It follows that Kalopheros was not only selling silk produced by his own villeins. Despite the evidence regarding the archon attending the fair of Vervena, adduced above, most Greek archontes of Latin Romania must have been weakened in their capacity to act as middlemen. Their resources or those they inherited had been reduced after the conquest, and they had lost the high social status and official functions they or their ancestors had enjoyed in the Byzantine period. In addition, they had to face competition from Latin landlords and from Latin merchants in particular. There were nevertheless some individual archontes who were in a better position than their peers, such as the Kallergis in Crete and the few archontes who in Frankish Morea had attained prominent positions within the knightly class or were serving as high-ranking officials in the baronial or princely administrations.128 As for Kalopheros, a latecomer to the principality, although a Greek he had rapidly integrated into the Frankish feudal elite, which was quite exceptional. The number of Latin settlers engaging in trade in western Romania appears to have been constantly growing since the early thirteenth century and, in any event, was far larger than before the Fourth Crusade. The continuous presence of these merchants in the region enabled them to monitor trade and shipping and prospect the area around their city of residence. They thereby enhanced trade, banking, credit, and entrepreneurship ventures, whether their own or those of other settlers and traveling merchants. Especially merchants involved in wholesale and export operated with more liquid capital than their peers or than landlords. Their contacts with producers and landlords on the one hand and with customers and other merchants, on the other, was easier and far more extensive than before 1204. The bulk of evidence regarding these practices comes from Crete. Direct transactions regarding agricultural and pastoral products between producers and customers or merchants concentrated in ports enabled both sides to dispense with middlemen. These transactions were often based on sale credit, in the form of anticipated payment for the delivery of an agreed amount of produce at a specific date or within a specific period. There is good reason to believe that sale credit was far more common after the conquest than in the preceding Byzantine period, although the absence of quantitative data prevents any solid assessment. The use of sale credit was greatly furthered by the increasing external demand for the products of Crete, to which I shall soon return, by the infusion of capital it promoted, as well as by the growing monetization of the economy. On the whole this system, with its concealed loan, favored the creditor who was in a position to exert pressure on the producer in need of cash and obtain favorable prices from him.129 It also eliminated any possibility of collective bargaining by producers. In this respect, after 1204 Latin landlords fared no better than the Greeks. In short, these developments generated a shift from a sellers’ market to a buyers’ market. 127 See the small numbers of households in villages surveyed in the 14th century, compiled by Carile, Rendita, 117–74, passim. 128 See Jacoby, “Encounter,” 894–95. 129 See Jacoby, “Cretan Cheese,” 51–54; Gallina, “Finanza,” 34–40; idem, Societa`, 133. However, in some contracts the parties agreed that payment would be made according to the market price at the time of delivery.

David Jacoby [ 219 ] In addition to sale credit, there were also loans and leases based on other types of contract, which stimulated market and export-oriented investments of money and labor in land cultivation and the raising of animals.130 In Crete the extension of cash crops is illustrated by contracts stipulating the expansion of vineyards, especially the planting of high-quality Malvasia and Athiri vinestocks from the fourteenth century on, in response to a growing external demand for high-grade Cretan wines. Wine became a major item in Cretan exports.131 Irrigation was already practiced in the Byzantine period. It is attested in 1118 for cotton and vegetable cultivation in Crete and by a letter of Pope Innocent III written in 1209, thus shortly after the conquest, referring to gardens and orchards in the bishoprics of Athens and Negroponte.132 Cistern and well water was occasionally used for small pieces of land, the latter case attested by a Cretan contract of 1280,133 yet irrigation was mostly based on the diversion of streams. Irrigation appears to have been substantially extended after the Latin conquest. It was promoted from around the turn of the thirteenth century on the mainland, in Crete and in some Aegean islands, especially by great landholders and investors, who more easily than peasants could muster the large resources needed for the construction and maintenance of expensive watering systems.134 Cotton growing expanded in Crete, and by 1307 the island was exporting its fiber to Venice. By the second half of the fourteenth century, cotton was also being grown in the countryside of Coron and Modon, as well as in Negroponte, Santorini, and Corfu, obviously in response to the increasing demand of the Venetian cotton industry. In Frankish Morea, cotton was cultivated in the area south of Corinth and presumably also in other areas, since it was partly exported from Chiarenza by Ragusan and Anconitan ships.135 Irrigation may have occasionally been used for luxury crops fetching a high price on the market, such as cherries, marasca or sour cherries, pomegranates, peaches, and pears.136 It was indispensable for the growing of citrus, known in

130

See above, p. 200. See Topping, “Viticulture,” 509–20; Gallina, Societa`, 41–44, 57, 133. See also above, note 23. On external demand and destinations, see Jacoby, “Creta,” 80, 85–86, 88, 90–94, 96, 100–102. 132 See Harvey, Economic Expansion, 127–33; Gallina, Societa`, 19–20; Innocent III, Epist., XI.256, PL 215:1560: “flumina unde rigantur horti.” 133 Marcello, no. 251. 134 See Jacoby, “Migration,” 124–25; Gallina, Societa`, 19–21, 53–54; Gaspares, H gh, 105–10. In 1352 the authorities in Crete intervened in a dispute between Venetian landholders over the use of water for irrigation: see Thiriet, Romanie, 310–11. The construction of an adequate irrigation system was among the heavy investments envisaged by Marco de Zanono, who in 1428 had obtained a monopoly for the growing of sugar cane in Crete: see D. Jacoby, “La production du sucre en Cre`te ve´nitienne: L’e´chec d’une entreprise e´conomique,” in RODWNIAÚ Timh` sto` n M. I. Manou´ saka, ed. C. Maltezou, T. Detorakes, and C. Charalampakes (Rethymno, 1994), 1:172, 175–77, repr. in Jacoby, Trade, no. . 135 For Venetian territories, see Jacoby, “Creta,” 79; above, note 22, on cotton and flax grown under irrigation at Lombaro before 1307; Thiriet, Romanie, 321–22; C. N. Sathas, Documents ine´dits relatifs a` l’histoire de la Gre`ce au Moyen Age (Athens-Paris, 1880–90), 2:119, no. 336; 126, no. 347; 154, no. 385 (respectively in 1404, 1405, 1406). For the Morea, see bambaso in Longnon and Topping, Documents, 176–78, 188–92, and M. F. Mazzaoui, The Italian Cotton Industry in the Later Middle Ages, 1100–1600 (Cambridge, 1981), 43. 136 See Gallina, Societa`, 43–44; Jacoby, “Migration,” 124–25. According to a Cretan contract of 1352, the lessees of a fruit garden undertook to plant at least ten fruit trees each year, citrus included, the species being determined by the lessor: A. Lombardo, ed., Zaccaria de Fredo, notaio in Candia (1352–1357), Fonti per la storia di Venezia, Sez. III, Archivi notarili (Venice, 1968), no. 82. 131

[ 220 ] Changing Economic Patterns in Latin Romania Constantinople in the second half of the twelfth century and around 1300 in Crete.137 Citron, lemon, and bitter orange are mentioned in the Pwrikolo´ go", a Byzantine satirical work that has been dated to the late thirteenth or the early fourteenth century.138 Citrus growing appears to have been introduced in Latin Romania around that time, yet remained rather limited for a long period.139 Significantly, in his commercial manual, completed between 1330 and 1340, Francesco Balducci Pegolotti includes cetrine and cederni, citron and cedrate fruit, among the spezierie, a term broadly applied in the Middle Ages to spices and other expensive luxury products traded in small amounts.140 A Greek account book from Rhodes dated to the last two decades of the fourteenth century mentions two shipments including a total of 600 lemons, 40 oranges, and 6 thickskinned citrons.141 Venice was importing lemons from Alexandria in 1396 and 1404, a further proof that yields in Latin Romania were still small.142 By 1450, however, citrons and oranges were being exported from Coron and Modon in larger amounts.143 As elsewhere, there were two basic patterns of maritime trade and transportation in Latin Romania. One of them was centered on bilateral exchanges between specific areas within that region or between them and other regions. The other was inserted within the broader trans-Mediterranean framework. Despite the collapse of the empire, subsequent warfare, and the establishment of new political entities, the maritime trade of western Romania displayed basic continuity with respect to the Byzantine period in its ranges and orientations. Nevertheless, new geopolitical conditions no less than economic factors both required and enabled several major adjustments. The thirteenth and fourteenth centuries witnessed changes in the relative importance of destinations, as well as a progressive growth in volume and an acceleration of exchanges, which called for a partial restructuring of commercial and shipping networks. Given the limitations of space, only the most important aspect of this evolution will be examined here. The losses inflicted upon the infrastructure of Constantinople by widespread fires and warfare in 1203 and 1204 were compounded by a massive exodus of population from all 137 Twelfth-century evidence in Ptochoprodromos, 4.328, ed. and trans. H. Eideneier (Cologne, 1991), 157: to` diakitri´on, translated “Zitruskonfekt” (208), yet in fact the reference is to citron; the origin of the fruit is not stated. Alice-Mary Talbot has kindly drawn my attention to an epigram by Theodore Balsamon, Eij" ne´ rant´ louÚ K. Horna, ed., “Die Epigramme des Theodoros Balsamon,” WSt 25 zan th'" monh'" tou' Arguropw j ´ rwnÚ see R. Janin, La (1903): 193–94, ep. 31; the correct name of the Constantinopolitan monastery is Argu j ge´ographie eccle´siastique de l’Empire byzantin: Premie`re partie. Le sie`ge de Constantinople et le patriarcat oecume´nique, vol. 3, Les e´glises et les monaste`res, 2d. ed. (Paris, 1969), 51. For Crete, see below, note 139. 138 In W. Wagner, ed., Carmina graeca medii aevi (Leipzig, 1874), 199; see M. Bartusis, “The Fruit Book: A Translation of the Porikologos. Translated from Byzantine Greek with an Introduction,” Modern Greek Studies Yearbook 4 (1988): 205–12, esp. 206 for the dating and 208 for the relevant references. 139 See Jacoby, “Migration,” 124–25. 140 Pegolotti, Mercatura, 294. 141 P. Schreiner, ed., Texte zur spa¨tbyzantinischen Finanz- und Wirtschaftsgeschichte in Handschriften der Biblioteca Vaticana, ST 344 (Vatican City, 1991), 70 and 73, lines 46–47 (text and trans.), 66–67 (dating), 78 (commentary). 142 Noiret, Documents, 80, 150: freight charge for each casket of lemons. This was not lemon juice, as would seem at first glance. In maritime trade, caskets also served as containers for a variety of solid goods: see H. Zug Tucci, “Un aspetto trascurato del commercio medievale del vino,” in Studi in memoria di Federigo Melis (Naples, 1978), 3:335–37. 143 Giorgio di Lorenzo Chiarini, El libro di mercantantie et usanze de’ paesi, ed. F. Borlandi (Turin, 1936), 55.

David Jacoby [ 221 ] walks of life, including the Byzantine imperial court, the social elite, and craftsmen. As a result, Constantinople ceased to be the major consumption and industrial center of Romania it had been. The political and territorial fragmentation of Romania further undercut the city’s economic centrality. After 1204 it was merely the capital of a reduced territorial entity, the size of which was continuously shrinking from the 1220s. This factor further contributed to a substantial reduction in the flow of cash and goods to Constantinople, whether in the form of fiscal revenue or in the framework of self-supply and trade. The Latin imperial court in Constantinople was chronically impoverished, the Latin nobility suffered from economic stress, and the rather meager Latin settlement in the city did not offset the consequences of large-scale depopulation. The economic contraction resulting from these events was only partly overcome in the last two decades of Latin rule.144 In these circumstances, surpluses exported from western Romania to Constantinople before 1204 had to be redirected toward other destinations. The need for a reorientation of trade networks occurred precisely in a period witnessing a rise in Western demand for specific commodities produced in Latin Romania, namely, foodstuffs, especially grain, cheese, wine, and salt, as well as industrial raw materials such as silk and colorants. This demand was linked to demographic expansion, a rise in living standards and purchasing power, as well as a growth in industrial production and an increase in the volume of goods available for exchange. These processes, already well under way in the twelfth century, gained momentum in the following period and had a decisive impact on the economy of Latin Romania. Soon after 1204 they contributed to a major shift in the orientation of this region’s economy. Instead of being mainly geared toward Constantinople and the internal Byzantine market, as before 1204, it became rapidly inserted within the patterns of the Western supply system. This shift was decisively enhanced by Venetian presence and economic activity in the region, discussed below. By 1261 it had already become irreversible, despite the renewed expansion of Constantinople’s economy after the Byzantine reconquest of 1261. This is not to say that Latin Romania failed to take advantage of the subsequent intensification of trade in Constantinople and the Black Sea.145 The West, however, especially Italy, remained henceforth its main trade partner. There was yet another powerful factor in Latin Romania itself that contributed to the shift in orientation of its economy, namely, the firm correlation existing between political factors, settlement, and economic activity in this region. This correlation, several instances of which we have already encountered, was particularly strong in Venetian

144 L. B. Robbert, “Rialto Businessmen and Constantinople, 1204–61,” DOP 49 (1995): 43–58, claims that Latin Constantinople experienced a continuous decline in economic activity, substantially enhanced since the 1230s. Yet see a different interpretation of the evidence by D. Jacoby, “Venetian Settlers in Latin Constantinople (1204–1261): Rich or Poor?” in Plousi´oi kai ptwcoi´ sthn koinwni´a th" ellhnolatinikh¯ " Anaj tolh¯ " (⫽ Ricchi e poveri nella societa` dell’Oriente grecolatino), ed. C. Maltezou, Biblioteca dell’Istituto ellenico di Studi bizantini e postbizantini di Venezia 19 (Venice, 1998), 181–204, partly based on unpublished documents. 145 Venetian refugees from the imperial capital, mentioned below, presumably contributed their share in this respect. On some aspects of this trade, see Jacoby, “Creta,” 80–87.

[ 222 ] Changing Economic Patterns in Latin Romania territories. Venetian rule over Crete, Coron, Modon, and Venice’s quarter in Negroponte reinforced existing links or created new ones with Venice. The Venetian nuclei in Latin Romania were strengthened after the Byzantine reconquest of Constantinople in 1261, which triggered an exodus of some three thousand Latins, most of whom must have been Venetian citizens or subjects.146 An unknown number among them, induced by earlier trade relations, the presence of relatives, Venetian rule, or the combination of these factors, settled in the city of Negroponte or in Crete.147 The partial reorientation of Latin Romania’s maritime trade toward the West resulting from a spontaneous shift was significantly enhanced by Venice’s policies. Protectionist measures as well as state intervention in and supervision over economic processes favored Venetian citizens in trade and shipping, yet also entailed for them some serious limitations. These citizens enjoyed preferential custom rates, and Venetian carriers benefited from a virtual monopoly on maritime transportation, since returning merchants were compelled to ship their goods to Venice exclusively on board their ships. The same rule applied to specific commodities that were to travel exclusively on board state galleys, in service since the early fourteenth century, unless the authorities issued other instructions.148 In addition, the Commune strictly regulated the rhythm of navigation between Venice and other ports, especially with respect to returning ships, so as to prevent an overflow of merchandise and a slump in prices on the Venetian market.149 Venice’s restrictive and discriminatory policy toward foreigners in the field of seaborne trade did not deter fairly large numbers of them from settling in the Venetian colonies. While in Coron from 1289 to 1293 the notary Pasquale Longo recorded the names of settlers originating in areas of Italy extending from the Veneto and Lombardy in the north to Barletta in the south.150 Similarly, there were numerous foreigners in Candia from the early stages of Latin settlement.151 In business transactions with Venice, however, foreigners largely depended upon privileged Venetian citizens, with whom they sought joint ventures, at times in an attempt to defraud the Commune’s treasury.152 Not surprisingly, these foreigners were particularly eager to obtain Venetian citizenship, granted only to a small number of them, or at least Venetian nationality.153 State intervention also affected other aspects of trade and shipping. As early as 1211, 146

For this figure, see Geanakoplos, Michael Palaeologus, 113–14. Ibid., 379–80, no. 2; Tsirpanles, Kata´ stico, 182–84, nos. 99–100. No other territories of Latin Romania are known to have served as havens for these refugees, yet this is surely due to the paucity of evidence regarding them. 148 On state galleys, see D. Sto¨ckly, Le syste`me de l’incanto des gale´es du marche´ a` Venise ( fin XIIIe–milieu XVe sie`cle) (Leiden, 1995). 149 See D. Jacoby, “La Venezia d’oltremare nel secondo Duecento,” in G. Cracco and G. Ortalli, eds., Storia di Venezia, vol. 2, L’eta` del Comune (Rome, 1995), 290. 150 See Borsari, Studi, 113–14. 151 See Jacoby, “Colonisation,” 307. 152 See Jacoby, “Venezia d’oltremare,” 291. 153 On the distinction between the two, see above, note 73. In his letters the Pisan Pignol Zucchello, a merchant who after residing in Crete settled in Venice, repeatedly requested intervention on his behalf to obtain Venetian citizenship, which he eventually was granted; his correspondents also deal with that subject: see R. Morozzo della Rocca, ed., Lettere di mercanti a Pignol Zucchello (1336–1350), Fonti per la storia di Venezia, Sez. IV, Archivi privati (Venice, 1957), 23, 47–49, 54, 75, 97, 102, 113–15. 147

David Jacoby [ 223 ] when the first military settlers were about to leave for Crete, Doge Pietro Ziani stipulated that exports of wheat from the island would be subject to state control, as noted earlier.154 Later in the thirteenth century the Cretan authorities imposed on the Latin military settlers and Greek archontes quotas of wheat, which they undertook to buy at state regulated prices determined each year according to the anticipated yield. Although preventing bargaining by local producers, landlords, and wholesale merchants, this disposition was advantageous in various ways to those who delivered the wheat, since it ensured a convenient marketing, a rapid payment in cash, partly in advance of delivery, and a guaranteed price floor even in case of overflow due to good harvests. Prices appear to have been slightly on the rise in the second half of the thirteenth and again in the second half of the fourteenth century. There was also a free grain trade, though exports were controlled by the Commune and restricted to Venice, Venetian territories overseas, and other destinations when authorized.155 Candia shared the export of wheat with Canea, a port serving as outlet for the large production of its own hinterland. Shipping from Crete to the ports of Latin Romania was often carried out by the Greeks of the region. Transportation to Venice was handled by both Venetian citizens and subjects, some residing in Crete and others in Venice.156 In addition to its regulation of the wheat trade, the Venetian government established in 1279 a salt monopoly in Crete, and two years later imposed a new overall salt policy requiring all ships to carry salt on their return voyage to Venice and sell it to the Commune.157 The documentation regarding Latin Romania, which is largely Venetian, has created the wrong impression that Venice dominated the economy of that region beginning in the early thirteenth century. In fact, the strengthening of Venice’s position was slow to come and its supremacy achieved only by the mid-fourteenth century. Several factors explain this rather lengthy process. In non-Venetian territories the absence of a strong, centralized government and of heavy-handed state intervention attracted settlers widely differing in origin and afforded more latitude for their activity, as well as more variety in the orientation of their operations. To be sure, Venetians also traded and resided in Thebes, Chiarenza, and Patras.158 By early 1389 the Venetian Marco Morosini was settled in Nauplia, before Venice took hold of that city, and such was the case with Albano Contarini and other Venetians in Argos.159 Yet in addition to Venetians, traveling 154

See above, p. 207. See Gallina, “Finanza,” 63–64, 127–32; D. Tsougarakes, “ JH sitikh` politikh` th'" Beneti´a" sth` n Krh´ th to` n 13o–14o aijw´ na,” Mesaiwnika` kai` ne´ a eJllhnika´ 3 (1990): 333–85, a full review of the policy. 156 See C. Gaspares, “ JH nautiliakh` ki´nhsh ajpo` Krh´ th pro` " th` n Pelopo´ nnhso kata` to` n 14o aijw´ na,” Ta` JIstorika´ 9 (1988): 287–318, passim, esp. 309–10, table 4. Two ships sailing in 1310 from Venice to load ´ sei", 62–63, nos. 19–20. The measures used in Candia and Canea were wheat in Crete: Theotokes, Apofa j identical, as revealed by their comparison with the one used in Rhodes: Pegolotti, Mercatura, 104 and 113. In the first case the ratio mentioned is 55:100, in the second 870:1560. 157 See Jacoby, “Creta,” 89, and for 1281, J.-C. Hocquet, Le sel et la fortune de Venise, vol. 2, Voiliers et commerce en Me´diterrane´e (Lille, 1979), 199–208, 249–55. 158 On 27 February 1275 Domenico Spadaro, a resident of Thebes, promises to maintain the local Venetian church of San Marco: Venice, Archivio di Stato, S. Nicolo di Lido, b. 2 perg. For Chiarenza, see Jacoby, “Venezia d’oltremare,” 272–73, and E. Gerland, Neue Quellen zur Geschichte des lateinischen Erzbistums Patras (Leipzig, 1903), 30–66, 89–107, also relevant for Patras. 159 Chrysostomides, Monumenta, 105, lines 163–66. 155

[ 224 ] Changing Economic Patterns in Latin Romania and resident merchants from Modena, Parma, Cremona, Piacenza, and other cities of central Italy known for their commercial and banking activities are attested in Negroponte and Thebes from the 1220s.160 Others from Milan and Ancona were active in Negroponte around 1270,161 while Italians, French, and German immigrants, as well as an Englishman resided in Patras in the fourteenth century.162 The close political connections between the kingdom of Sicily and Frankish Morea beginning in the late 1260s resulted in a marked increase in exchanges between the two regions for about an entire century, a process surprisingly not reflected by commercial manuals. Grain was the main commodity shipped to the Peloponnesos from south Italian ports, which also served as transit stations for silk and kermes from the principality on their way to textile manufacturing centers.163 The intensification of trade between the two regions was greatly enhanced by the operation and settlement in the Peloponnesos and neighboring areas of merchants, bankers, and officials hailing from central and southern Italy or involved in the operation of its economy. Prominent among them were Sienese and Florentine citizens, some of them operating with fairly large amounts of capital. They were originally attracted by the prospects of trade and credit operations in connection with the transfer of funds to or on behalf of the papal treasury, which they were already practicing in the West and in the Crusader Levant. Some of them acted on their own, like the Sienese merchants settled in Chiarenza who obtained fiefs between ca. 1260 and ca. 1325, and the Sienese banker based in Negroponte, who in 1310 supplied the duke of Athens, Walter V of Brienne, with the funds needed for the hiring of the Catalan Company. Others served as resident partners of the companies or as their resident agents in the branches they established beginning in the 1260s in Latin Romania.164 Among these companies the Bardi, Peruzzi, and especially the Acciaiuoli of Florence were the most important ones until they collapsed in the 1340s.165 Sienese and Florentine merchants, whether acting on their own or on behalf of com160 See Jacoby, “Migration,” 107; P. Racine in P. Castignoli and M. A. Romanini, eds., Storia di Piacenza, vol. 2, Dal vescovo conte alla signoria (996–1313) (Piacenza, 1984), 200. 161 TTh, 3:204, 209–10. 162 Gerland, Quellen, 89. 163 On Chiarenza and its trade with southern Italy, see Bon, More´e, 320–25; Jacoby, “Migration,” 105–8, 112, and 120. On several occasions King Charles I ordered royal wheat and barley to be sold in Chiarenza in order to finance his military operations in the principality, for instance in 1273: R. Filangieri et al., eds., I registri della cancelleria angioina (Naples, 1950–), 10:42–43, reg. 40, no. 146. The relevant evidence on trade appearing in this series of documents, of which more than thirty volumes have been edited thus far, has yet to be exploited. On trade of Apulia and Naples with Chiarenza and Negroponte in 1274, see R. Cessi, “La tregua fra Venezia e Genova nella seconda meta` del sec. XIII,” Archivio veneto-tridentino 4 (1923): 35–38, and for the dating, 16–18; trade between Apulia and Nauplion in 1272: TTh, 3:274–76, with correct dating in G. Morgan, “The Venetian Claims Commission of 1278,” BZ 69 (1976): 429, no. 60; Venetians in regular trade between Chiarenza and Apulia: R. Cessi, ed., Deliberazioni del Maggior Consiglio di Venezia (Bologna, 1931–50), 2:135 and 3:25–26 (1282 and 1283); Chrysostomides, Monumenta, 33, in 1381. On south Italian ceramics found in the Morea, see below, p. 232. 164 See Jacoby, “Migration,” 98–99, 107–18. On the Levant: D. Jacoby, “Migration, Trade and Banking in Crusader Acre,” in Balka´ nia kai Anatolikh´ Meso´ geio", 12o"–17o" aiwne" (⫽ The Balkans and the Eastern Mediterranean, 12th–17th Centuries), The National Hellenic Research Foundation, Institute for Byzantine Research, Byzantium Today 2, ed. L. Mavromatis (Athens, 1998), 114–19. 165 Ibid., 114–18.

David Jacoby [ 225 ] panies, significantly contributed to economic growth in the non-Venetian territories of Latin Romania. Credit operations and transfers of money were partly achieved by investment in maritime trade. Sienese merchants conveyed wax from Romania to the fairs of Champagne, as attested in 1265 by Andrea Tolomei, member of an important company based in Siena, in a letter sent from Troyes.166 An unidentified Sienese company apparently exported silk from Chiarenza in the 1270s, and kermes from Romania reached Siena, as attested by a Sienese custom list compiled between 1273 and 1313.167 Continental Greece, especially the Peloponnesos, as well as some islands of the Aegean and the Ionian Sea, were major exporters of this expensive dyestuff used in textile manufacture.168 The Florentine mercantile banking house of the Alberti shipped kermes from Corinth and Monemvasia, as well as from the islands of Cerigo and Cephalonia.169 By the late thirteenth century, some Italian mercantile and banking companies had extended their operations to Corinth, a thriving and affluent economic center until the Catalan attack of 1312.170 Their large-scale financing of warfare and of conspicuous consumption brought large infusions of liquid capital. These activities, combined with the introduction of advanced business methods into Latin Greece from the second half of the thirteenth century, such as deposit accounts with payment on demand, transfer banking, double-entry bookkeeping and management, contributed to an intensification of trade and an acceleration of the monetary flow.171 The banking practices introduced into the Frankish territories also spread to Venetian Coron and Modon, the economy of which was tightly linked to theirs.172 I have already noted the integration of Italians within the knightly class of Frankish Morea beginning in the second half of the thirteenth century, among them merchants and bankers. In the following century we find Italian intendants administering large feudal estates, several of which belonged to absentee Italian landlords. Italian lords and intendants were familiar with sophisticated business techniques and had a clear impact on the exploitation of Moreot estates. They introduced structural changes in their management, a more rational organization of space and use of resources, whether land, water, beasts of labor

166

C. Paoli and E. Piccolomini, eds., Lettere volgari del secolo XIII scritte da Senesi (Bologna, 1871), 57. See Jacoby, “Migration,” 112–13; M. A. Ceppari and P. Turrini, eds., “Documenti: Il commercio delle stoffe; l’abbigliamento e le provvisioni di lusso; arredi sacri e profani,” in “Drappi, velluti, taffeta` et altre cose.” Antichi tessuti a Siena e nel suo territorio, ed. M. Ciatti (Siena, 1994), 245. 168 See Jacoby, “Silk Production,” 45–47, 61. 169 A. Sapori, I libri degli Alberti del Giudice (Milan, 1952), 71, 101, and 229. 170 See Jacoby, “Migration,” 103–4, 114. The assessment of destruction inflicted by the Catalans has recently been somewhat tempered by the suggestion that major destruction may have been caused by an earthquake that occurred ca. 1300: see C. K. Williams II, E. Barnes, and L. M. Snyder, “Frankish Corinth: 1996,” Hesp 66 (1997): 41–42. 171 See Jacoby, “Migration,” 113, 121–27. Accounts are mentioned by Marino Sanudo Torsello in Hopf, Chroniques, 101–2: “Nel suo tempo fu` nel principato tanta cortesia e amorevolezza, che non solamente li cavallieri ma` anche li mercadanti andavano su` e giuso senza denari . . . e con il semplice loro scritto di mano se li dava denari.” Since Sanudo refers to the reign of Prince William II, who ruled from 1248 to 1278, the introduction of deposit accounts into Greece should be placed in the latter’s reign, thus earlier than in Venice, on which see R. C. Mueller, The Venetian Money Market: Banks, Panics and the Public Debt, 1200–1500 (Baltimore, 1997), 9–18, esp. 15–16. 172 See below, pp. 227–28, 231. 167

[ 226 ] Changing Economic Patterns in Latin Romania or the workforce, diversified crops, and achieved a rise in agricultural and pastoral productivity, a growth in output, and improvements in the marketing of products.173 In the thirteenth century, Genoese interests and trade were also important factors in the portion of Latin Romania not subject to Venetian rule, far more so than commonly assumed.174 While importing silk fabrics from the eastern Mediterranean, Genoa also acted as the main supplier of raw materials to the expanding Lucchese silk industry from the beginning of its operation around the mid-twelfth century.175 Not surprisingly, therefore, silk textiles, raw silk, and dyestuffs appear to have been the main commodities that Genoese merchants sought in Latin Romania, where their activity is documented as early as 1210. In that year two partners, one of them a Lucchese, bought in Genoa a certain amount of grana de Romania, kermes clearly originating in Latin Romania.176 The silk called seta de Romania, documented in Genoa in 1269, also came from that region.177 Sixteen contracts drafted in Genoa between 1274 and 1345 refer to trade with Chiarenza, some of 1287 explicitly mentioning the export of woolens, the sale of which was to finance purchases there.178 Between 1330 and 1340 Pegolotti lists samite, by then a medium-grade silk textile, among the commodities in which merchants on business in Chiarenza reinvest proceeds from the sale of the goods they import.179 Chiarenza was also a major exporter of silk cocoons, silk, and kermes collected from its own hinterland and neighboring areas.180 Significantly, according to a Pisan trade manual of 1278, the units of Lucca were the standard used for the weighing of raw silk in Frankish Morea, which suggests that the Genoese, in view of their role as suppliers of Lucca, were then also the main exporters of silk from the principality.181 Negroponte too shipped cocoons, silk, and silk fabrics on a fairly large scale to Italy.182 A Genoese consul is attested in 1236 in the city of Negroponte, which implies Genoese trade and presumably also a resident community there. Genoese merchants are again documented in that city from 1245 to 1251.183 By 1240 there was a well-established resident Genoese community headed by

173

See Jacoby, “Migration,” 121–27. E.g., by Balard, “Ge´nois en Romanie,” 467–89. 175 See D. Jacoby, “Genoa, Silk Trade and Silk Manufacture in the Mediterranean Region (ca. 1100–1300),” in A. R. Calderoni Masetti et al., eds., Tessuti, oreficerie, miniature in Liguria, XIII–XV secolo, Istituto internazionale di Studi liguri, Atti dei Convegni, III (Bordighera, 1999), 16–29, 38. 176 Lanfranco, no. 915. See also above, p. 225. 177 Mentioned by E. Basso, “Le relazioni fra Genova e gli stati latini di Grecia nei secoli XIII–XIV,” in Studi balcanici: Pubblicati in occasione del VI Congresso internazionale dell’Association internationale d’Etudes Sud-Est Europe´ennes, Sofia, 1989, ed. F. Guida and L. Valmarin (Rome, 1989), 23, yet this was not silk textile as assumed by the author. Although Asia Minor was included in Romania, silk from that region had other names. 178 See M. Balard, La Romanie ge´noise (XIIe–de´but du XVe sie`cle), vol. 1 (Rome, 1978), 163–64 and n. 211. A Pisan document drafted in Chiarenza in 1317 refers to three Genoese merchants who had apparently been on business there: ed. C. Otten-Froux in M. Balard, A. E. Laiou, and C. Otten-Froux, Les Italiens a` Byzance (Paris, 1987), 175–77, no. 8. 179 Pegolotti, Mercatura, 117. See also below, p. 228. 180 See Jacoby, “Silk Production,” 46–48, 55 n. 52, 60–61. 181 R. Lopez and G. Airaldi, eds., “Il piu` antico manuale italiano di pratica della mercatura,” in Miscellanea di studi storici, vol. 2 (Genoa, 1983), 127, fol. 360, lines 15–16. See Jacoby, “Migration,” 120–21. 182 Jacoby, “Silk Production,” 61, and see also below, p. 228. 183 See Balard, “Ge´nois en Romanie,” 480; Basso, “Relazioni,” 24. A sailing contract of March 1254 mentions Negroponte among the ports at which the merchants may unload their goods: ed. E. H. Byrne, Genoese Shipping in the Twelfth and Thirteenth Century (Cambridge, Mass., 1930), 125–28, no. 37. 174

David Jacoby [ 227 ] its consul in Thebes, the major center of silk manufacture in Latin Romania. Some time earlier Genoese entrepreneurs had begun to finance production in a number of this city’s silk workshops, ordered textiles from others, and exported local fabrics. It follows that the infusion of cash into the economy of Latin Romania and the function of credit were not restricted to the rural sector or to trade, and also affected industrial production. However, by the late thirteenth century Genoese trade in Latin Romania appears to have been on the decline as a result of several developments. The continuous growth of highgrade silk manufacture in Lucca, as well as large-scale Genoese imports of eastern textiles and of high-grade silk from the countries around the Black Sea and the Caspian Sea offered advantageous alternatives to the products of Latin Romania.184 In this region the Genoese also faced what appears to have been aggressive Venetian competition, backed by the Commune. Pisans also traded in Latin Romania. They are attested in Frankish Morea from the 1270s, yet appear to have been active there earlier. In 1273 King Charles I of Sicily asked the Moreot prince William II to dispense justice to a Pisan citizen having a claim against one Scottus, apparently an Italian banker. The attention paid by the Pisan trade manual of 1278 to the silk standard of Lucca used in the Morea implies that Pisans too were involved in the export of this commodity to Lucca and other Italian manufacturers, though presumably on a much smaller scale than the Genoese. Pisans are mentioned in Chiarenza in 1303, 1307, and 1313, some of them being apparently settled in that city. The Pisan trade manual compares the measures for grain used in Negroponte and Pisa, which hints at Pisan trade in the island. No later evidence has surfaced until now, yet the continuation of Pisan trade in Byzantium in the Palaiologan period suggests that some activity must also have taken place in the ports of call of Latin Romania located along the maritime routes linking Pisa with Constantinople.185 The Venetians were familiar with the territories of Latin Romania ruled by feudal lords, in which they had traded before the Fourth Crusade. As early as 1209 Venice obtained full tax exemptions for its merchants, confirmed subsequently, from the lord of Frankish Morea and those of Euboea.186 Coron, Modon, and the Venetian quarter of Negroponte depended heavily on the flow of products from their respective hinterlands, ruled by these lords. This explains why Venetian Coron and Frankish Chiarenza used the same light pound for the weighing of silk and kermes.187 Around 1290 silk and kermes were arriving in Coron from both Venetian and Frankish territory, and this must have been customary for some time already.188 It is not clear to what extent tax exemptions promoted Venetian penetration inland. Among the products handled, a few deserve particular attention. Since none of the Venetian colonies of Latin Romania pro184 See Jacoby, “Migration,” 118–20, and above, note 175. It is likely that Venetian merchants too acted as entrepreneurs in Thebes: see below, note 190. 185 For Latin Romania: Filangieri et al., I registri della cancelleria angioina 10:93, no. 373, and Jacoby, “Migration,” 112; above, note 181; R. Predelli, ed., I libri commemoriali della repubblica di Venezia: Regesti (1293–1787) (Venice, 1876–1914), 1:26, 80–81, lib. 1, nos. 108, 339, 344; above, note 178; Lopez and Airaldi, “Manuale,” 127, fol. 360, line 10. For Byzantium: Otten-Froux in Balard, Laiou, and Otten-Froux, Italiens (as in note 178), 159 and n. 36, and documents in ibid., 168–91, nos. 3, 9, 11, 12, 16. 186 TTh, 2:91, 94, 97, 176, 181, and 3:55. 187 See Jacoby, “Silk Production,” 55 n. 52. 188 For this period and later, see Jacoby, “Silk Production,” 41, 43–47, 55–56, 60–61. See also above, p. 216.

[ 228 ] Changing Economic Patterns in Latin Romania duced quality silk textiles, Venetian merchants, like others, purchased them in the Peloponnesos, at Thebes, and at Negroponte.189 Far more important in the long run was their quest for raw silk and dyestuffs. The development of the Venetian silk industry, apparently initiated after the Fourth Crusade, seems to have contributed significantly to the growth of Venetian activity and presence in Frankish Morea and Negroponte.190 Yet for much of the thirteenth century the Venetian share in the handling of these commodities must have been fairly limited compared with that of the Genoese. The intensification of long-distance Venetian shipping between Venice and Constantinople, especially from the 1270s, enhanced Venetian trade in the areas situated along the navigation routes. The convoys of state galleys, in service since the early fourteenth century, stopped at Chiarenza, or else some ships were sent there to pick up Venetian merchants and the goods they had bought in Frankish Morea.191 Such was the case in 1338, in connection with the fair of St. Demetrius. The attendance of Venetians at that fair was by no means exceptional. It illustrates a pattern of penetration inland on quite an impressive scale. The six merchants from prominent families in Venice invested in that trade venture a total of 16,005 hyperpers in woolens and cash.192 The Venetian silk trade in Patras around 1351 illustrates further expansion.193 By the mid-fourteenth century the Venetians had consolidated their hold on trade and navigation in the portion of Latin Romania not ruled by the Commune. Genoese sources documenting trade with Chiarenza apparently cease after 1345. Genoa’s failure to gain a foothold in the western Aegean, the Genoese occupation of Chios in 1346, and the Venetian-Genoese war of 1350–55 resulted in a shift of Genoese interest to the eastern Aegean.194 These factors apparently also put a virtual end to Genoese involvement in the economy of continental Greece and neighboring islands. The late eleventh century witnessed the establishment of a triangular transMediterranean trade network linking Italy with Egypt and Byzantium, as well as these states one to the other. The integration of western Romania within this network was enhanced after the First Crusade by the establishment of the Crusader states in the Levant and the intensification of maritime trade in the eastern Mediterranean. It was further promoted after the Fourth Crusade by the growing impact of Venice on the economy of Latin Romania. Ships engaging in free navigation as well as regular convoys of state galleys sailed from Venice via Modon, Coron, and Negroponte on their way to Constantinople and the Black Sea, and via Modon and Candia on the way to Cyprus, Lesser Armenia, Beirut, and Alexandria.195 Thirteenth-century Byzantine, Muslim, and Crusader coins, as well as pharmaceutical containers from Egypt found at Corinth illustrate 189 D. Jacoby, “The Production of Silk Textiles in Latin Greece,” in Technology in Latin-Occupied Greece, ed. C. Maltezou and H. Kalligas (Athens, 1999), in press. 190 See D. Jacoby, “Tra Bisanzio, il Levante e Venezia: Dalla materia prima ai drappi nel medioevo,” in Dal baco al drappo: La seta in Italia tra Medioevo e Seicento, ed. R. Mueller (Venice, 1999), 265–304. 191 See Sto¨ckly, Gale´es, 101–8, esp. 103, 105. 192 See above, pp. 215–16. 193 Gerland, Quellen, 33 n. 194 On the first two aspects, see M. Balard, “The Genoese in the Aegean (1204–1566),” Mediterranean Historical Review 4 (1989): 158–62, repr. in Arbel, Hamilton, and Jacoby, Latins (as in note 15). 195 See Sto¨ckly, Gale´es, 96–152, and Jacoby, “Creta,” 94–102.

David Jacoby [ 229 ] the pivotal location of Latin Romania within the triangular trade network.196 The seaborne trade of Latin Romania became subordinated to the requirements, routes, and seasonal rhythm of long-distance maritime trade, increasingly dominated by Venetian merchants and carriers who took advantage of Venice’s naval and diplomatic protection and of the infrastructure offered by its colonies and commercial outposts. Short-distance trade and shipping, carried out with small vessels often practicing cabotage, were partly redirected in order to convey goods collected in Latin Romania and neighboring regions to the main ports inserted within the trans-Mediterranean traffic. These same ships also took charge of the diffusion of imported goods. While the functions of the major transit ports have drawn some attention, their contribution to the economy of Latin Romania itself has been largely overlooked and underestimated. These ports offered infrastructures and services assisting transiting merchants, ships, and goods.197 Storage and transshipment, the supply of provisions, ship maintenance and repairs performed by qualified craftsmen, as well as money changing and banking provided substantial infusions of cash. The sailing of ships engaged in surveillance and the protection of convoys, naval warfare, piracy, and the recruitment of sailors, archers and crossbowmen, the latter especially in Crete, had similar effects.198 To these we may add the building and enlargement of arsenals, as in Coron, Modon, Negroponte, and Candia, as well as repeated improvements in harbors that, although not always successful, ensured a flow of public money collected as taxes back into the local economy, instead of being siphoned off to Venice.199 There is yet another function of transit ports that should be underlined. In addition to handling passing goods, resident merchants also served as middlemen in complex trade ventures between several regions. The function of intermediaries was particularly important in Crete.200 All these activities generated profits, which were reinvested in Latin Romania’s own economy, whether in the rural sector, in trade, or in transportation. Greeks are clearly underrepresented in the extant, overwhelmingly Western documentation bearing on Latin Romania. The bias is less acute with respect to Venetian Crete beginning in the fourteenth century, in view of the large number of notarial deeds referring to Greeks. It is nevertheless impossible to arrive at a fair evaluation, let alone a quantitative assessment of the Greek share in the region’s economy, dominated by the 196 The coins are recorded in numerous annual reports of the excavations, published in the 1990s in Hesp. For the containers, see C. K. Williams II and O. H. Zervos, “Frankish Corinth: 1994,” Hesp 64 (1995): 16–22. 197 See F. Thiriet, “Candie, grande place marchande dans la premie`re moitie´ du XVe sie`cle,” Krhtika` Cronika´ 15–16 (1961–62): 343–47, repr. in idem, Etudes, no. . 198 On Crete in the Venetian naval defense system, see Thiriet, Romanie, 243–51; D. Jacoby, “Les gens de mer dans la marine de guerre ve´nitienne de la mer Ege´e aux XIVe et XVe sie`cles,” in Le genti del Mare Mediterraneo (XVII Colloquio internazionale di storia marittima, Napoli, 1980), ed. R. Ragosta (Naples, 1981), 1:172–74, repr. in idem, Studies, no. ; Jacoby, “Creta,” 103–5. 199 On work in harbors, see Hodgetts, “Colonies,” 146–51; R. Gertwagen, “The Venetian Port of Candia, Crete (1299–1363): Construction and Maintenance,” Mediterranean Historical Review 3 (1988): 141–58; eadem, “L’isola di Creta e i suoi porti (dalla fine del XII alla fine del XV secolo),” in Ortalli, Venezia e Creta (as above, note 68), 350–74; F. Thiriet, “Re´thimo et son district au quinzie`me sie`cle,” in Pepragme´ na tou' G⬘ Dieqnou'" Krhtologikou' Sunedri´ou, vol. 2 (Athens, 1974), 305–6, repr. in idem, Etudes, no. . 200 See Jacoby, “Creta,” 80–81, 83–84, 92–103, 105–6.

[ 230 ] Changing Economic Patterns in Latin Romania Latins. Fourteenth-century sources point to the growing integration of Greeks within the latter’s patterns and networks. They appear as middlemen and wholesalers in local trade, invest in exports, participate in seaborne trade, sail on business to foreign countries, and enter into partnerships with Latins.201 It would seem that a decree issued in Crete in 1313 prohibited the exit of Greeks from the island or their sailing to other regions. Their participation in maritime trade would thus have been severely limited. In fact, the purpose of that decree was to prevent the flight of villeins belonging to the Commune, regardless of whether they were indigenous or of foreign origin.202 It did not deal with villeins attached to landlords.203 More extensive measures were introduced in 1349 to prevent both slaves and villeins belonging to individuals from leaving Crete and to ensure their arrest, should they reach other Venetian territories, as well as their return to the island.204 It follows that there were no legal restrictions preventing free Greeks from engaging in maritime trade or in shipping, once the proper administrative formalities before departure had been completed.205 Indeed, there is abundant evidence showing that Greek merchants, shipowners, skippers, and sailors from Crete participated in exchanges between their island and other areas. Greeks residing in other Venetian territories must have been equally active in this respect, yet the surviving evidence regarding them is fairly restricted. The bulk of Greek maritime activity took place within the Aegean,206 yet it also extended further. In 1357 a Cretan nobleman, Marco Salamone, shipped more than 15 tons of cheese from Sitia to Cyprus on board the griparia of the Greek shipowner Costa Vlisma, who took along more than 5 tons of his own cheese.207 In 1361 the papas Dimitrius Siropulo from Candia sailed with his own griparia from Famagusta to Rhodes.208 In the first half of the fifteenth century, Greek merchants and ships from Crete fairly regularly reached Alexandria with cargoes of cheese, wine, and other commodities and returned with spices among other goods. In the same period they also exported wine to Constantinople and reached Venice. Moreover, they expanded their activity beyond bilateral exchanges between Crete and the Byzantine capital, sailing from this city to Caffa in the Crimea, Cyprus, and Beirut, and via Crete also to Messina. In addition, Greek merchants and ships from Latin 201 See A. E. Laiou, “Quelques observations sur l’e´conomie et la socie´te´ de Cre`te ve´nitienne (ca. 1270–ca. 1305),” in Bisanzio e l’Italia: Raccolta di studi in memoria di Agostino Pertusi (Milan, 1982), 177–98, esp. 193 ff, repr. in eadem, Gender, no. ; eadem, “Venetians and Byzantines: Investigation of Forms of Contact in the Fourteenth Century,” Qhsauri´smata 22 (1992): 33–35. 202 Bandi, no. 8: “nullus villanus comunis tam terrigena quam forensis.” The villein was allowed to reside in Candia or in the countryside, “sicut sibi placuerit,” without fear of being considered the villein of a military settler, “et non timeat capi pro villano militum.” Since state villeins enjoyed freedom of movement within Crete, it was more difficult to prevent their escape from the island. 203 On the distinction between the two categories of villeins and on the legal limitations imposed upon them, see above, note 25. 204 P. Ratti Vidulich, ed., Duca di Candia: Quaternus consiliorum (1340–1350), Fonti per la storia di Venezia, Sez. I, Archivi pubblici (Venice, 1976), no. 233, esp. p. 131. 205 Two sailing permits referring to this decree, issued by the Cretan authorities in 1356 and 1368, have been published by Gaspares, Nautiliakh` ki´nhsh, 289–90 nn. 9 and 8, respectively. 206 Ibid., 287–318, on 14th-century Greek shipping from Candia and Canea; an earlier example of Greeks from Modon in TTh, 3:236–37. 207 On this transaction and numerous others of Marco Salomone, see Gallina, Societa`, 99–100, 119. 208 A. Lombardo, ed., Nicola de Boateriis, notaio in Famagosta e Venezia (1355–1365), Fonti per la storia di Venezia, Sez. III, Archivi notarili (Venice, 1973), no. 125.

David Jacoby [ 231 ] Romania reached Dubrovnik, and traders also traveled to Barcelona on their way to Saragossa in Spain, though it is unclear on which ships. In short, by the fourteenth century the Greeks were firmly inserted within the geographic pattern of Latin Romania’s seaborne trade, with certain limitations.209 The activity of these Greeks calls for a few remarks. It was generally based on relatively small amounts of capital and conducted with the help of small and medium-sized ships having a limited carrying capacity, manned by Greek or mixed crews.210 Greeks only seldom participated directly in trans-Mediterranean traffic and, in any event, were totally excluded from commercial patterns sponsored by the Venetian government and directly subject to its control. The transportation of precious goods on board state galleys and the leasing of these ships were the exclusive preserve of Venetian citizens. It is therefore obvious that, despite their numbers, in terms of capital turnover and profits the Greeks of Latin Romania had a fairly limited share in medium and especially in long-distance maritime trade. Nevertheless, some of them based in the major transit ports greatly benefited from the general increase in economic and maritime activity and managed to accumulate considerable wealth in related activities, such as banking. C ¸ an Cremolisi, a Greek resident of Coron, provided loans totaling 35,000 gold ducats to the lord of Corinth, Nerio Acciaiuoli, the reimbursement of which he sought for several years after the latter’s death in 1394.211 In the first half of the fifteenth century, high officials of Byzantine Morea were depositing valuables and cash in Coron and Modon in private banks operated by Greeks, who used the same sophisticated commercial techniques as their Latin counterparts.212 In the same period a Greek family, the Filomati from Crete, settled in Venice and adopted a business strategy resting on the dispersal of its members, as commonly practiced by Venetian mercantile families, positioning some in Crete and one in Constantinople. Incidentally, some prominent Byzantine families appear to have acted in the same way both with respect to the Byzantine provinces and the cities of Latin Romania, namely, Venetian Coron and Modon.213 So far we have noted two important economic functions fulfilled by Latin Romania 209 On this pattern, see Jacoby, “Venezia d’oltremare,” 272–73; idem, “Cretan Cheese,” 57–60; idem, “Creta,” 83–84; S. Borsari, “Ricchi e poveri nelle comunita` ebraiche di Candia e Negroponte,” in Maltezou, Plousi´oi (as above, note 144), 213–14, 216–18, on Cretan Jews providing maritime loans for trade with Cyprus, Alexandria, the Aegean islands, and Venice, partly on board Greek ships; B. Krekic´, Dubrovnik (Raguse) et le Levant au moyen aˆge (Paris, 1961), 99–100, 103, 125–50; Gallina, Societa`, 123–27. 210 See Gaspares, Nautiliakh` ki´nhsh, 293–305 (tables), and for a mixed crew with the sailors’ names, ibid., 289–90 n. 9; mixed crew also on a small Venetian ship around 1270: TTh, 3:274–75. Further examples of small vessels appear above. 211 See J. Chrysostomides, “Merchant versus Nobles: A Sensational Court Case in the Peloponnese (1391– 1404),” in Praktika` tou' D⬘ Dieqnou'" Sunedri´ou Peloponnhsiakw'n Spoudw'n (1992–93), 2:116–31, and for the documents, see now Chrysostomides, Monumenta, 626, General index, s.v. “Cremolisi, Court case.” 212 See K.-P. Matschke, “Geldgescha¨fte, Handel und Gewerbe in spa¨tbyzantinischen Rechenbu¨chern und in der spa¨tbyzantinischen Wirklichkeit: Ein Beitrag zu den Produktions- und Austauschverha¨ltnissen im byzantinischen Feudalismus,” Jahrbuch fu¨r Geschichte des Feudalismus 3 (1979): 187–88; idem, “Griechische Kaufleute am ¨ bergang von der byzantinischen Epoche zur Tu¨rkenzeit,” in Lauer and Schreiner, Kultur Griechenlands (as in U note 7), 78–79. 213 Ibid., 77–78; Matschke, “Geldgescha¨fte,” 195–96; on the Filomati, see now D. Jacoby, “I Greci ed altre comunita` fra Venezia ed oltremare,” in I Greci a Venezia, nel V centenario della fondazione della comunita` greca, ed. M. F. Tiepolo (Venice, 2000), in press; for the Latins, see D. Jacoby, “La dimensione demografica e sociale,” in Cracco and Ortalli, Storia (as above, note 149), 2:703.

[ 232 ] Changing Economic Patterns in Latin Romania within the Mediterranean trade system: as a source of foodstuffs, raw materials, and semifinished or finished products, and as a supplier of services to trade and shipping operations. In addition, the region was also a market for Western goods. Though concentrated in specific locations, the Latins remained a small minority in the midst of the Greek population. In particular within their upper ranks, this may have sharpened their awareness of the evolving attitudes, lifestyle, and consumption patterns of their peers in the West. Their approach in this respect stimulated the demand for Western manufactured goods, especially high-grade products serving as status symbols. This demand, also enhanced by a rise in the standard of living and a refinement in taste, was not confined to members of the knightly class in the feudalized territories of Latin Romania.214 It also extended to the social elite in Venetian territories. In Crete, members of the Venetian elite granted their daughters dowries “according to the custom of the noblewomen of Venice.”215 Furthermore, conspicuous consumption was widespread among Latin commoners, Greeks and Jews alike. The display of luxury among prosperous Cretans prompted the Venetian authorities to publish in 1339 sumptuary laws regarding dress and jewelry, though with little effect.216 The rising volume of medium and high-grade consumption is evidenced by various imports, which partly enabled Western merchants to finance their purchases in Latin Romania. Fine ceramics manufactured in southern Italy, the Veneto, and Pisa, as well as Italian glassware found in recent excavations, reached Corinth from the 1260s.217 They reflect the range of commercial exchanges of this city with Italy and the activity of south Italian, Venetian, and Florentine merchants, already encountered above. We have already noted in passing the woolens brought by Genoese and Venetian merchants to Chiarenza.218 Among the luxury items imported to Crete in the 1370s, we find fine Flemish woolens and Italian silk textiles, most of which were presumably Venetian products.219 In 1444 the eighteen-year-old Quirina, daughter of Alexios Kallergis, ordered from Venice a gold-interwoven silk garment costing between 90 and 100 ducats, as well as expensive pieces of velvet.220 These growing imports were not only related to the impact of Western consumption patterns and fashions on local demand. They were also connected with, and even enhanced by, the state of textile manufacture in Latin Romania itself. To be sure, the 214 See D. Jacoby, “Knightly Values and Class Consciousness in the Crusader States of the Eastern Mediterranean,” Mediterranean Historical Review 1 (1986): 158–86, repr. in idem, Studies, no. . 215 See S. McKee, “Households in Fourteenth-Century Venetian Crete,” Speculum 70 (1995): 40–41. 216 J. Jegerlehner, “Beitra¨ge zur Verwaltungsgeschichte Kandias im XIV. Jahrhunderts,” BZ 13 (1904): 464– 65, para. 14–22. On sums above 500 hyperpers appearing in contracts regarding Latin, Greek, mixed LatinGreek, and Jewish marriages, see McKee, “Households,” 40–41, 46, 50–51; M. Gallina, “Diversi livelli di richezza e di penuria negli atti matrimoniali rogati a Candia nel corso del secolo XIV,” in Maltezou, Plousi´oi (as above, note 144), 268, 272, 280–91. 217 See esp. Williams and Zervos, “Frankish Corinth: 1994,” 16–24. See also A. Oikonomou-Laniado, “La ce´ramique protomajolique d’Argos,” in La ceramica nel mondo bizantino tra XI e XV secolo e suoi rapporti con l’Italia, ed. S. Gelichi (Siena, 1997), 307–16. 218 See above, pp. 226, 228. 219 See S. Borsari, “Il mercato dei tessuti a Candia (1373–1375),” Archivio veneto, 5th ser., 143 (1994): 5–30. 220 See C. Maltezou, Benetikh` mo´ da sth` n Krh´ th (Ta fore´ mata mi´a" Kallergopou´ la"), in Byzantium: Tribute to Andreas N. Stratos (Athens, 1986), 1:139–47, esp. document on pp. 145–46.

David Jacoby [ 233 ] Theban silk industry was stimulated in the first half of the thirteenth century by Genoese and possibly also other Latin merchant-entrepreneurs, as noted above, yet it is not clear how long this support lasted. While silk textiles of Latin Romania continued to be exported to the West in the fifteenth century, they faced mounting competition from the high-grade products of the expanding Venetian and Luccan silk industries, which availed themselves of advanced technologies. The growing import of Italian silks into Romania reversed an age-old trend: instead of being exclusively the supplier of the West, Romania had also become its customer. Both foreign and local merchants increasingly viewed Latin Romania as a source of industrial raw materials for Western industries, rather than of finished products. The absence of new investments and technology transfers prevented the silk industries of Latin Romania from expanding and upgrading their production in a significant way, and Western imports further undermined their ability to compete.221 The painted glassware manufactured in Venice for export to Romania from the late thirteenth century must have had a similar effect on some centers of glass production in Latin Romania.222 This survey has dealt with large portions of Latin Romania ruled by the Latins for more than two centuries after the Fourth Crusade. The evidence bearing on this region reveals that all the sectors of its economy underwent important structural changes in the period extending roughly to the mid-fifteenth century. The main factors contributing to this evolution were the constant interplay between micro- and macroeconomic factors, as well as between private initiative and political powers; geopolitical developments both within and outside Latin Romania; and, finally, the broader economic systems within which the provinces of the empire conquered by the Latins were integrated. After the conquest the economy of Latin Romania swiftly geared itself to Western demand, yet also took advantage of conjuncture to develop its bilateral exchanges with other regions.223 Credit was a major factor stimulating a growth in export-oriented products, as well as in short- and medium-range trade and transportation.224 In this respect, the economy of Venetian Crete appears to have been particularly dynamic, its landholders, peasants, and merchants being more responsive to market incentives than in the feudalized areas of Latin Romania. In addition to the factors just mentioned, the intensification of local, regional, and trans-Mediterranean trade and shipping and the supply of services in their framework generated substantial infusions of cash into Latin Romania, which from major ports trickled through the various sectors of its economy. As a result, the whole region experienced an ever stronger economic interaction between the countryside, the cities and maritime trade, as well as an acceleration of monetary circulation. The Hebrew University, Jerusalem 221

See Jacoby, “The Production of Silk Textiles.” On this production, see A. E. Laiou, “Venice as a Center of Trade and Artistic Production in the Thirteenth Century,” in H. Belting, ed., Il Medio Oriente e l’Occidente nell’arte del XIII secolo, Atti del XXIV Congresso internazionale di storia dell’arte, Bologna, 1975 (Bologna, 1982), 2:14–15, 18–19. 223 More evidence in this respect in Jacoby, “Creta,” 80–106. 224 On credit in seaborne trade, see Gallina, “Finanza,” 13–21; idem, Societa`, 111–27. 222

The Crusades and the Development of Islamic Art Oleg Grabar Among the numerous catchy but arrogant as well as intellectually dubious aphorisms attributed to great men is Napoleon’s statement “l’intendance suivra.” The idea is, I guess, that, once a brave and well-led army has moved forward conquering territories and defeating enemies, the paraphernalia of practical institutions and needs required to make an army function and to lead it to other successes or, alternately, to keep conquered territories under control, this practical and necessary context of a significant event follows automatically. Historians of art and of anything else have followed this Napoleonic adage in assuming that major events affect culture and the arts. It is, so the assumption goes, legitimate to argue that the French or Russian revolutions, Alexander the Great’s conquests in western Asia, the Mongol invasions, the appearance of Islam, the spread of Buddhism, and other such momentous episodes in the history of mankind had an impact on the arts or modified existing ways of doing or seeing things in some significant manner. Such impacts or modifications can be the culmination or expression of internal, culture-bound, seeds which would be shaken into revolutionary novelty because of an event, as with constructivism in twentieth century Russia, the evolution of David and the formation of Ingres in late eighteenth- and early nineteenth-century France, or the merging of Hellenistic with Roman sculpture. Or else these novelties can be attributed to the sudden appearance of the new and foreign element, like the art of sculpture in India, apparently revolutionized by Hellenistic models, floral ornament introduced into Chinese art by the spread of Buddhism, or French painting and architecture transformed by the invasion of Italy in the last years of the fifteenth century. The mode of operation of these changes varies. There are destructions of the old and its replacement with something new through the will and decision of whoever wields power, as often happened with changes in religious spaces when temples became churches and churches or fire-temples became mosques. At other times, artists and artisans as well as techniques of all sorts are brought in and become the instruments of the transformation, as when Persian painters went to India in the sixteenth century, or when the Umayyads in the seventh and eighth centuries, like the Normans of the twelfth, This paper is more or less the one delivered as a lecture at the occasion of the Dumbarton Oaks symposium on the Crusades. Only the most elementary bibliographical references have been added and a few rhetorical changes made.

[ 236 ] The Development of Islamic Art acquired the financial means to hire whomever they wanted for their ambitious programs of construction and decoration. The assumption is that the revolutionary or unusual event, whether or not it involves an invasion, leads either to some triggering within the affected culture which brings out new and perhaps unexpected features hidden in the existing cauldron of memories and competencies or else compels new patterns of cultural life through the importation of foreign technicians, models, taste, or behavior. A priori, the phenomenon of the Crusades can legitimately be considered as such an epic or at least momentous event. New men and women with a Western European and Latin culture established themselves in a land that had existed for more than three hundred years in an Islamic culture blissfully ignorant, for the most part, of the Latin Christianity about to appear on its territory. That new and unexpected event was kept alive and active for nearly two hundred years, until the late thirteenth century, and survived far longer in cultural memory. There were large-scale, if temporary, movements of people, including artisans and construction workers brought from afar to replicate as many aspects as possible of the homelands of the Crusaders. We have an enormous program of construction, as Jerusalem, Bethlehem, Hebron, the coastal cities of Syria, Palestine, and Lebanon and several interior areas (I am thinking of the Crac des Chevaliers to the north, now in Syria, and Kerak or Shaubak to the south in Jordan) are all covered with monuments of a new imported architecture. Manuscripts are illustrated in the Latin Kingdom of Jerusalem and, although I am not aware of many remaining objects made specifically for the Latin inhabitants of Palestine, the emblems of their presence, are still visible in their coins and seals, and a number of reliquaries were probably done in Palestine as well.1 All of this was happening at the height of the Romanesque artistic explosion and of the sophisticated Byzantine pictorial wealth under the Komnenian dynasty. It is equally important to recall that the twelfth century, and especially its second half, witnessed an extraordinary set of quantitative and qualitative changes in Islamic art. Hundreds of new buildings, for the most part madrasahs and other social functions of pious architecture but also caravanserais and bazaars, were sponsored in every city from Central Asia to the Mediterranean coast in order to meet a set of newly developed ideological and economic purposes. It is in the twelfth century that the muqarnas, that ubiquitous and unique form of Islamic architecture about which I will have more to say shortly, spread all the way to Morocco from its origins in Iraq and Central Asia, with all sorts of intermediary stops like the Cappella Palatina in Palermo. Thanks to a patronage expanded from the courts to the mercantile elites of the cities, new techniques in the decorative arts made possible representations of people, animals, and at times whole narratives, on relatively common and inexpensive media like ceramics. Metalwork was modified by the expansion and refinement of the technique of silver inlaying, which allowed for clear and very detailed representations and ornament. At the end of the century the first known dated manuscript with illustrations inaugurates the short-lived but brilliant school of Arab painting in the thirteenth century. These novelties also 1

The bibliography is enormous on the art of the Crusades. The most recent thorough survey is J. Folda, The Art of the Crusades in the Holy Land, 1098–1187 (Cambridge, 1995).

Oleg Grabar [ 237 ] affected Christian art within or at the periphery of the Muslim world, as is clear by the changes that occurred within Armenian, Syriac, and Egyptian Coptic traditions.2 Conventional academic wisdom attributes these changes in Islamic and Islamicate art to an expanded patronage made more sophisticated through international trade, through the industrialization of the manufacture of paper with all sorts of important ramifications for all the arts, and through new techniques, especially in ceramics and metalwork, developed first in northeastern Iran and in the area of Baghdad.3 The main question is, then, a relatively simple one: Is there a connection between the Latin (and perhaps more generally Christian) forceful reappearance in Syria and Palestine and the major, often revolutionary, modifications brought into Islamic art at about the same time? And there are subsidiary questions: Was there an impact of Islamic forms on the arts made for the newly arrived Christians and their co-religionists back home? Or are there examples of the opposite, Western features in Islamic objects? Did the Crusaders affect the long-range development of Islamic art in the Levant? The only one of these questions that had been the subject of discussion and of published studies is that of a possible or actual impact of Islamic art on Western art, and I shall not deal with it at all except for one small observation later on. I shall be brief on the matter of direct or indirect traces of Crusader art in contemporary or later Islamic art, because there is, to my knowledge, no systematic survey of existing evidence; I cannot, therefore, propose even a reasonable scheme for the organization of the material. What follows are a few random examples. In Cairo the complex of Muhammad al-Nasir (ca. 1303) (Fig. 1) has a Gothic portal brought by boat from Acre and used as a trophy; it is also possible to argue that certain features of the slightly earlier complex of Qala’un, for instance the articulation of the exterior facade and of the windows, owes something to Western architecture, although it may simply be that it uses forms associated with the holy city of Jerusalem and not with the art of another culture.4 In Anatolia the mosque and hospital at Divrigi dated around 1229 contain unusual features on their facades which are difficult to connect to local Anatolian, Christian or Muslim, traditions; this is especially true of their elaborate gates with most unusual splayings (Fig. 2) and with a prominent stone sculpture in high relief with vegetal rather than figurative motifs (although some figuration is actually hidden in the vegetation) belonging to an interesting subgroup of late Romanesque art found from southwestern France 2 The easiest access to this material is in R. Ettinghausen and O. Grabar, The Art and Architecture of Islam, 650–1250 (London, 1987), in particular chaps. 7 and 8; a new and considerably revised version of the book should appear in late 2000. For Christian material from the Near East, see J. Leroy, Les manuscrits syriaques a` peintures (Paris, 1964); J. Leroy, Les manuscrits coptes et coptes-arabes illustre´s (Paris, 1974); H. C. Evans, “Cilician Manuscript Illumination,” in Treasures in Heaven: Armenian Illuminated Manuscripts (New York, 1994), 66–83. 3 Some of the interpretations were formulated a long time ago by O. Grabar, “Les arts mineurs de l’Orient musulman,” CahCM 11 (1968): 181–90. Additional and alternate views appear in J. Allan, “Silver: The Key to Bronze in Early Islamic Iran,” Kunst des Orients 11 (1976–77): 5–21, and in Jonathan Bloom’s forthcoming work. 4 For these buildings, the basic publication is by K. A. C. Creswell, Muslim Architecture of Egypt, vol. 2 (Oxford, 1959), 190 ff. Summary in S. Blair and J. Bloom, The Art and Architecture of Islam, 1250–1800 (New Haven, Conn., 1994), 72–77.

[ 238 ] The Development of Islamic Art to east of Moscow. Whether in Anatolia or in Russia, the motifs of this art and, even more so, the very fact of its existence seem to reflect Western and most particularly Latin models. The assumption of such models may actually explain other innovative features of thirteenth-century Anatolian architecture, although none are as original as those of Divrigi. One should add the alternate possibility of an impact coming from Armenia and the Caucasus.5 Examples of Crusader parts reintegrated in Muslim monuments of the Mamluk and even later periods abound in Jerusalem, Hebron, Bethlehem, Beirut, and Tripoli. Although I am not well acquainted with military architecture, I take it as a valid proposition that large Crusaders’ fortifications had an impact on the citadels of the Muslim world in Syria, Jordan, and Anatolia and merged with an older and somewhat different tradition of the urban citadel which had already begun in eastern Iran in the tenth century, but for entirely different reasons.6 It is more difficult to detect a visual or formal impact of the Crusades on arts other than architecture. What did, however, happen is the use of forms to communicate the political and ideological ambitions issued from the Crusades. Such is the case of Nur alDin’s inscriptions, which charge madrasahs and mosques from Mosul in northern Mesopotamia to Damascus in Syria with a new militancy, as has been so well demonstrated by Yasser Tabbaa.7 And, in a particularly spectacular way, such was the meaning given to the minbar, now destroyed, made in Aleppo for a Jerusalem that had not yet been reconquered. In all these instances, forms and functions owe little, if anything, to the Crusades, but the meaning to be given to these forms is very much tied to the existence of the Latin kingdom. These examples from the Levant lead to the conclusion that the art of the Crusaders did leave traces in the Muslim world, but these traces are, relatively speaking, minimal (except perhaps in military architecture) and almost any one of them is a unique case which can be explained through special circumstances. Only in Palestine is it possible to argue for a sort of symbiosis of imported Western and local forms and techniques. This symbiosis, as it appears for instance in the eastern wall of the Aqsa mosque or in the fountain at the bab al-silsilah also in Jerusalem, was a natural meshing together of building traditions, and, if it strikes us today as awkward, it probably was not so at the time. The handsome screen of wrought iron built by the Crusaders around the rock in the Dome of the Rock remained there until the fifties of this century. One can understand the historical (actually more antiquarian than historical) and national reasons for its removal, but the interesting point is that it made visual sense inside the building—it enclosed and protected its holy spot—even if it was not the building’s original message to highlight the rock in this fashion. A possibly more curious impact of that screen occurs in Cairo. There the mausoleum of Qala’un was based on the Dome of the Rock and provided 5

Among other places, R. Hillenbrand, Islamic Architecture (Edinburgh, 1994), 96–97. H. Kennedy, Crusader Castles (Cambridge, 1994). D. Pringle, The Churches of the Crusader Kingdom of Jerusalem (Cambridge, 1993). 7 Y. Tabbaa, “Monuments with a Message,” in The Meeting of Two Worlds: Cultural Exchange between East and West during the Period of the Crusades, ed. V. P. Goss and C. Bornstein (Kalamazoo, Mich., 1986), 223–40, and for the specific case of Aleppo, idem, Constructions of Power and Piety in Medieval Aleppo (University Park, Pa., 1997). 6

1 Cairo, mosque madrasah of Sultan al-Nasir, portal from Ascalon, ca. 1303 (courtesy of Nasser Rabbat)

2 Divrigi, mosque, northwest portal, thirteenth century (courtesy of Ülkü Bates)

3 Louvre, ewer or aquamanile in the shape of a peacock, inv. no. MR 1519 (drawing after Adrien de Longpérier)

4 Urtuqid plate, f irst half of the twelfth century, Innsbruck

5 Mantle of Roger II, dated 1133–34, Vienna

Oleg Grabar [ 239 ] with a fancy wooden screen around the tomb of the sultan, because it was there in the model as the early Mamluks knew it, not as it had been at the beginning. Some scholars have even argued for a Palestinian quality to the Romanesque sculpture in the Holy Sepulcher and elsewhere in the Christian sanctuaries of the Holy Land.8 These arguments have not been accepted by all scholars, but, even if they are valid in part or as a whole, they would still be mere examples of what I would call the micro effect of the Crusades on the arts: an intrusion within the archaeological texture of Palestine and of the Syro-Lebanese coast and occasionally monuments elsewhere in which, for known or obscure reasons, something Western pops up. This micro effect recalls the French women of one of Zoe Oldenbourg’s novels dealing with the Crusades who were remembering the lush landscape of Picardy while working as indentured servants in the harsh and dry fields of central Syria. They were touching minor episodes within a grand history. This could be the end of the paper with the simple conclusion that the Crusades hardly mattered in the artistic life of the Muslim world. Matters are far more complicated when one turns to the reflection of Islamic forms in Western art. They are complicated because the range of these reflections is much wider and the specificity of the Crusades in their existence more difficult to determine. There is the simple exoticism of imported forms and techniques, as with so many textiles, bronzes, or ivories reused for relics or for the ornamentation of churches and of ecclesiastical vestments, with the construction of a very Syrian Islamic mausoleum for the Norman prince Bohemond, or with the random imitation of the Arabic script found all over medieval art and especially in textiles. There may well have been subtler impacts, as the memory of the Holy Land may well have affected the architecture of late Romanesque cloisters. Altogether, at this mini level, the presence of Islamic forms was greater in the West than of Western forms in the Islamic world, that presence is only partly to be related to the Crusades, and it was but a minor bit player in the very active changes of the twelfth and thirteenth centuries.9 On the levels of exchanges of forms or of affecting artistic creativity directly and immediately, there is not, I believe, much more to say except to continue listing examples. But there still looms the broader issue that too many changes occurred in the twelfth century, and especially all around the Mediterranean, not to feel that the most unusual event of that century should not somehow be connected with these changes. In other words, there may well have been a macro impact of the Crusades. Perhaps some other term than “impact” is the appropriate one for a very different type of relationship than that of immediate connection. Let me investigate the matter around two series of documents. One, the thirteenth-century Islamic metalwork with Christian scenes, is a tight 8

N. Kenaan, “Local Christian Art in Twelfth Century Jerusalem,” IEJ 23 (1973): 167–75, 221–29. There is no overall survey of Islamic themes in Western art. See P. Soucek, “Artistic Exchange,” in C. Bornstein and P. Soucek, The Meeting of Two Worlds: The Crusades and the Mediterranean Context (Ann Arbor, Mich., 1981), 15–16; A. Shalem, Islam Christianized (Frankfurt, 1996); G. Sievernich and H. Budde, eds., Europa und der Orient, 800–1900 (Berlin, 1989); and the very thoughtful observations of R. Ettinghausen, “The Impact of Muslim Decorative Arts and Painting on the Arts of Europe,” in The Legacy of Islam, ed. J. Schacht and C. E. Bosworth (Oxford, 1974), 292–320, repr. in his Collected Papers (Berlin, 1984), 1074 ff. All these studies are provided with good bibliographies. 9

[ 240 ] The Development of Islamic Art and small group of objects from a very limited and specific area, northern Mesopotamia and Syria. The other one roams all over the place and consists of works of art, of motifs, and even of problems that do not fit into any common category of understanding but illustrate something quite important about the twelfth century. There are eighteen remaining inlaid bronzes with Christian scenes.10 Some have often been illustrated, others are hardly known, very few have received the detailed attention they merit, in part because each one contains details in its alleged or demonstrated Christian features that are unusual or do not make sense. Three broader characteristics of these objects are more important for our purposes than a search for an explanation of individual details. One is that they are for the most part in a technique—bronze or brass inlaid with silver, occasionally with copper—and for purposes—lighting, serving, writing, washing, pouring water or other liquids—that are common in metalwork made for feudal lords and for the notables of Muslim cities since the middle of the twelfth century. This new metalwork appeared first in northeastern Iran and culminated in the great thirteenth-century series of works associated with the city of Mosul in northern Mesopotamia, but found all over Egypt, Yemen, and the Levant. All objects with Christian scenes are in the same technique, and the shapes and functions are also quite common, except for an extraordinary canteen in the Freer Gallery of Art in Washington whose actual purpose and function are still very much of a mystery.11 One can still wonder whether there is any connection between the possible uses of these objects and the Christian motifs of their decoration, but, so far, none has been discovered, and it is more reasonable to conclude that these objects belonged to a common body of things and designs for the upper classes of society in Muslim lands during the first half of the thirteenth century regardless of ethnic or religious affiliation. It is curious that one of them, a tray now in St. Petersburg, was found near Kashgar, today Kashi, at the frontier of the Islamic world and of China. In short, Christian themes simply became one of the possible options in the standard imagery available within the Muslim world. The second characteristic of these objects is that, even though several among them are provided with inscriptions, none of the inscriptions implies that the object was made for a unique purpose or a specific individual, least of all that it was made for a Christian. Most of the inscriptions are standard series of good wishes without, insofar as I have been able to figure out, anything separating them from other metal objects of the same class. There are three references to the Ayyubid amir al-Malik al-Salih, who ruled, at various times, in northern Mesopotamia, Egypt, and Damascus between 1232 and 1249 and who was involved in complicated alliances with or against the Crusaders. But even these references are of a general nature, not personalized, and the more frequent lists of

10 The standard book on the subject is E. Baer, Ayyubid Metalwork with Christian Images (Leiden, 1989); see also R. Katzenstein and G. D. Lowry, “Christian Themes in Thirteenth-Century Islamic Metalwork,” Muqarnas 1 (1983): 53–68. 11 An interesting and original explanation of this canteen has been proposed by Nuha Khoury, “Narratives of the Holy Land,” Orientations 29.5 (1998): 63–69.

Oleg Grabar [ 241 ] good wishes have an anonymous quality suggesting objects made on speculation, for sale at some court or to a wealthy buyer possibly from the nonmilitary class. These objects reflect the needs or opportunities of a market, and it is the thoughtful but cautious requirements of a market that may explain the absence of scenes like the Crucifixion or the Ascension, which are uniquely peculiar to Christianity and which could be seen as offensive or at least inappropriate by Muslims, as well as the prominence in nearly all these objects of rows of standing personages, these priestly and monastic figures that were admired by Muslims since the time of the Prophet. These objects breathe a quiet pietism that is not so much ecumenical as areligious. It may well, as has been suggested, correspond to the spirit of the area, at least at the upper feudal level, after the 1229 treaty between Frederick II and Malik al-Kamil, a time when conflicts were feudal and territorial rather than religious and national. The sources of the Christian images are probably for the most part local Syriac ones, possibly Byzantine, very little Western, although one Western iconographic detail has apparently been identified. And the third important character of this group of motifs is that they disappear after the middle of the thirteenth century, whereas the technique of inlaying remains, as do the functions of most of these objects. It is, of course, true that representations of other kinds also diminish in the latter part of that century, and Christian images could have been the victim of a general revival of aniconism all over the Levant, in spite of major exceptions like the Baptiste`re de St. Louis in the Louvre.12 But, even if this is so, it is remarkable that Christian themes entered, for a short while, within the mainstream of Muslim private art. Iconographically, this phenomenon cannot easily be connected with the Crusades since, at least within the boundaries of research known to me, no representation on the bronzes has been connected with the art of the Crusades. Nor are there any stylistic parallels between the images on objects from Muslim workshops and the little we know of comparable Latin or Byzantine art. But it may be possible to argue that the very special political mood established by the negotiations that followed the recapture of Jerusalem by Saladin in 1187 permitted, at least for a while and within a certain class of Muslims or of dignitaries within the Muslim world, an acceptance of identifiably Christian but nonthreatening motifs as a sort of worldliness acceptable on secular objects. This possible fact of contemporary taste, rather than the motifs themselves, would have been the result of the presence of the Crusaders in the Levant. What remains to be worked out is the probably considerable role played in this development by the Christians of the Near East, who, as is well known, ended up as the main victims of the whole adventure but served as important agents of change in the twelfth and thirteenth centuries. However they are to be interpreted, the Islamic bronzes with Christian topics form a neat and coherent set. Incoherence is what characterizes my second group of examples. It consists of several peculiar objects, of an architectural invention, and of a chronological 12 This fascinating object has received much recent attention long after its initial publication by D. S. Rice, The Baptiste`re de St. Louis (Paris, 1953). See D. Behrens-Abouseif, “The Baptiste`re de Saint Louis: A Reinterpretation,” Islamic Art 3 (1989): 3–13.

[ 242 ] The Development of Islamic Art problem. Individually these items have nothing to do with the Crusades, but as a group they may perhaps best be explained as willed or accidental effects triggered by the Crusades. The first of these items is a container in the shape of a peacock now in the Louvre (Fig. 3).13 Art historians have tended to date the object and the whole group of zoomorphic containers with which it forms a set to the twelfth century and to attribute it to Spain or to Sicily, with one, to my mind unsuccessful, attempt to give this peacock an Iranian origin. The curiosity of the object lies in part in the manner in which it was meant to function, but this will not concern me in these remarks. The other reason for its notoriety is that it has two inscriptions certainly engraved at the same time in a space provided for them. One is in Arabic reads: ‘amal Abd al-Malik al-Nasrani, “made by Abd al-Malik the Christian.” There has been some controversy about the correctness of the translation, but the alternate (“made by the servant of the Christian king”) poses both grammatical and historical difficulties. The other inscription is in Latin and has been read as opus Solomonis erat, a curious sentence, which should be translated as “it was the work of Solomon,” suggesting either a maker by that name or, in a metaphoric way, a “beautiful work,” so to speak worthy of Solomon. A Spanish attribution for the bird seems reasonable because it is only in Spain that this particular Latin formula occurs on a couple of other objects. The peacock was thus seen as belonging to some sort of exchange between Muslims and Christians for which the following scenario can be proposed: someone, a merchant perhaps, orders from a Christian metalworker by the name of Abd al-Malik a fancy container of a known type (we have at least two other examples) and decides to broadcast the value of the object by calling it Solomonic in beauty or perhaps to imply, by the use of erat instead of the more normal est or the more accurate fuit, that this object imitates or actually is something belonging to Solomon and brought from the Holy Land, at the very least copying something that would have belonged to Solomon. The market for the object would have been Latin. But value is provided by the artisan’s name written in Arabic, yet identified as a Christian. It could have been an example of the sort of mercantile and vanity-driven artistic contact made possible by the Crusades. The trouble with this traditional interpretation is that it may have been based on a misreading of the Latin text. In an article that came out almost twenty years ago but which did not attract much attention among historians of art, the palaeographer RobertHenri Bautier argued that, after opus Solomonis, we should read era T and then a small x.14 T, as it turns out, was in central Spain and until the twelfth century, the symbol for the Latin M to mean 1000, and the era involved in this system began in 38 .. Thus we have an object dated to 962 or 972, depending on how one interprets the doodad after T. And historians of art throw their arms up in despair, for something they would have sworn to be twelfth century turns out to be two centuries earlier and dated to boot. The

13

The object has often appeared in catalogues, e.g., L’Islam dans les collections nationales (Paris, 1977). The first publication was by A. de Longpe´rier, “Vase arabo-sicilien,” RA, n.s., 6 (1865): 356–67. A paper of mine entitled “About a Bronze Bird” is to appear in a volume in honor of Ilene Forsyth. 14 R.-H. Bautier, “Provenance du paon aquamanile,” BullSocAntFr (May 1978): 92–101.

Oleg Grabar [ 243 ] whole neat and elaborate scenario I proposed a few minutes ago falls apart in its most interesting part; certainly the Crusades could not have had anything to do with this object, unless it is legitimate to talk of a pre-Crusade Crusade culture in Christian Spain. But is Bautier’s reading right? I do not have a solution at this stage, and I certainly have difficulties putting this bird in the tenth century. Should one imagine for the late tenth century in Spain some prefiguring of a type of mercantile and symbolic activity more common later on? Should it belong on the Spanish frontier or something comparable to Akhtamar on the Anatolian one and from which the Crusades were absent indeed, but which exhibit multicultural features to be found later more frequently? The second object I will deal with is the celebrated Innsbruck plate (Fig. 4).15 It was made for a minor Urtuqid ruler of eastern Anatolia between 1114 and 1142, during the most successful decades of the crusading enterprise. The plate is unique for being enameled on both sides with themes that find parallels in Byzantine art from Constantinople, in Georgian art, and even in Limoges. I can easily imagine a successful prince in the upper Euphrates valley being persuaded by a Georgian goldsmith, traveling, as many artisans did at that time, from one court to the other, that, given the proper financial reward, he could have on one copper bowl all the themes that appear in Byzantine imperial art, that he could thereby impress Frankish knights with the inventiveness of his patronage and his subjects with Arabic and Persian statements, the latter being quite ungrammatical and for practical purposes illegible. There is something slightly vulgar and nouveau riche in this display of colorful wealth, but it may well illustrate the taste of many other feudal lords and barons than a relatively minor Urtuqid prince. The third example of an odd object is even better known. It is the mantle of Roger II (Fig. 5) now in Vienna, to which I shall also return in a slightly different way in my conclusion. All that matters at this stage is to point out two obvious and well-known contrasts in it. One is that it contains an inscription that gives its date (1133–34) and place of manufacture (Palermo) as well as many good wishes but no indication of owner, maker, or use. This inscription is in Arabic, but the shape and probable function are Latin. The carefully woven decoration of the mantle is without direct parallel anywhere but probably reflects a very concrete astrological or astronomical configuration on one side and the royal court on the other.16 The Crusades may not have been directly involved in the manufacture of this mantle, but they are so much part of the wealth and ambitions of Roger II that it is perhaps not unreasonable to see them as part of the climate that made the robe possible. My last two examples are of a different type. One is the peculiar adventure of the muqarnas, that quintessentially Islamic form developed probably in Iraq in the tenth century and appearing in northeastern Iran and Egypt in the tenth and eleventh centuries. What is interesting about it for our purposes is its spread westward: new mosques in

15 Innsbruck, Institut fu¨r Kunstgeschichte, Die Artuqiden-Schale (Innsbruck, 1995), with all appropriate references. 16 A large study of the mantle will be found in the contribution I made to the fifteenth Levi della Vida Award symposium “The Experience of Islamic Art,” hopefully to be published at UCLA. The object itself is often illustrated.

[ 244 ] The Development of Islamic Art Tinmal and Fez, both in Morocco, among many other places, use muqarnas ceilings in the thirties of the twelfth century, and the largest and most brilliant example of the technique is the ceiling of the Cappella Palatina completed by 1142.17 The form itself has been given various symbolic and possibly religious meanings within the Muslim tradition, and a cosmic meaning is implied by Philagatas in a celebrated sermon pronounced in Palermo, although there is some doubt whether such meanings should be attributed to all uses of the form or to its inception. What really matters is that it is after the formation of the Latin kingdom in Palestine and after the Mediterranean had become the lifeline of the kingdom’s existence that the new form spread to become the common ceiling ornament of religious as well as secular buildings in Muslim as well as Christian lands. And, finally, there is a peculiar and difficult problem connected with the chronological rhythm of figural representations in Islamic art. As is by now well known, such representations were never entirely given up, but they diminished in number and in variety during Abbasid cultural preeminence in the ninth and tenth centuries. This was especially true for objects that had some public visibility. Thus Spanish courtly and private ivories could have masses of images in the tenth century, although fewer in the eleventh, and the representational vocabulary on ceramics was on the whole quite limited; there were on them more rabbits and birds than people. All this changed in the middle of the twelfth century, most strikingly in the ceramics and metalwork from Iran. Yet the earliest signs of change occurred in the eleventh century and in the primarily luster ware of Fatimid Egypt, where representations of considerable thematic and stylistic variety make their appearance. We do not know very well how to date these ceramics and how to establish stylistic and chronological sequences within them. But it is on the ceilings of the Cappella Palatina that many of these Fatimid themes appear before the middle of the twelfth century and before the true explosion of images elsewhere in the Muslim world. Is it entirely an accident that a major change in Islamic art, which had begun to take place in the Mediterranean area, spread elsewhere just as the Crusader state and culture were establishing themselves in the Levant? These examples—and there are others—do not demonstrate nor even suggest an impact of the Crusades, neither on concrete levels of forms and techniques nor even in ideological or sociopolitical programs. What they do show is the existence of a concentration of creative energy in the twelfth century, wherever one looks. The Crusades are part of that energy, and they established an enabling focus of power in an artistically underdeveloped area, for such was the case with Syria and Palestine in the eleventh century. That power led to major changes in that area and, even more remarkably, in adjacent and surrounding areas: a revitalized Syrian interior, an Anatolia bursting with building activities for all sorts of religious and secular purposes, a northern Mesopotamia

17 Much has been written on the muqarnas over the years, but there is as yet no definitive and coherent explanation. See Y. Tabbaa, “The Muqarnas Dome,” Muqarnas 3 (1985): 61–74, and “Muqarnas,” Dictionary of Art, ed. J. Turner, vol. 22 (London, 1997), 321–25; also his Constructions of Power and Piety, esp. 144. For Palermo, arguments and interpretations are found in W. Tronzo, The Cultures of His Kingdom (Princeton, 1997).

Oleg Grabar [ 245 ] with new cities and centers growing along reestablished trade routes, and even a transformation of the coast and hinterland of North Africa. That focus of power led to the expansion and development of motifs and ideas like the muqarnas and like representations of almost anything to satisfy the needs of a newly and intensely recharged Muslim society and ethos. The results were sometimes bizarre, as with the Innsbruck plate or the Louvre bird, but the point is that of a wealth of accomplishments often based on seeds planted earlier in a more haphazard fashion in the East or in Egypt and in Andalus, perhaps, as had been suggested by some, in the pan-Mediterranean culture of the eleventh century.18 The seeds might have taken much longer to spread all over the Muslim world and to grow in quality and quantity, had it not been for the new impulse created by the Crusades, an impulse of energy, not of forms. In a sense and quite paradoxically, one could argue that the most spectacular Islamic art of the time of the Crusades is the Norman art of Sicily primarily under Roger II, but also under William II, when, in a spirit of unwitting ecumenism, muqarnas, figures, inscriptions, and astronomy all combine to create a stellar series of highly original works of art. They do not form an eclectic combination of forms from different sources but a genuine entity, illustrating not a clash of civilizations, to use an abominable, recently coined, expression, but a truly operating manipulation and enjoyment of commonly accepted forms. Of course, like the Crusades themselves, this Norman creativity left no posterity, but, because it is a work of great art and artifice charged with obscure memories, a robe with a camel subdued by a lion, framed by an Arabic inscription praising sensual pleasures, became the formal coronation robe of the holy Roman emperors north of the Alps, and no one ever worried about the original meaning of the motifs on it. The century of the Crusades was thus far more interesting and far more creative than the Crusades themselves. In the Norman art of Sicily it almost succeeded in managing something that was hardly imagined by either the Crusaders or their opponents: a formal setting accessible visually and intellectually to all the actors of the century in the Mediterranean. On a much more limited scale, something comparable may have been present with the bronzes with Christian scenes in Syria and northern Mesopotamia and possibly in the architecture of thirteenth-century Anatolia. On the whole, however, after 1250 or thereabouts, the creation of man became Muslim or Christian, Saracenic or Infidel, ours or alien. The latter became the others who could be hated and despised or, at best, exotic. The modern age had begun. Institute for Advanced Study

18

P. Soucek in The Meeting of Two Worlds and several articles by Hans Belting.

The Impact of Frankish Architecture on Thirteenth-Century Byzantine Architecture Charalambos Bouras In the eleventh and twelfth centuries, Byzantine ecclesiastical architecture in both the capital and the provinces was mature and self-contained, meeting to the full the church building’s liturgical needs and dogmatic symbolism, as well as realizing possibilities in the development of other art forms such as sculpture and monumental painting. Even though many important churches were built at this time, this mature and self-contained architecture helped to create a conservative climate and at the same time gave rise to certain reservations about the production of new types and the conception of the church’s interior space. This trend began at the beginning of the eleventh century and became especially prevalent toward the end of the Komnenian period. In contrast, architecture in the West during the same period was progressing in leaps and bounds in matters of type, architectural form, decoration, and, chiefly, construction methods. Throughout Western Europe, much larger churches were being built, while a series of technological innovations beginning in France realized the transformation from Romanesque to Gothic to create formal systems that sooner or later predominated throughout Latin Christendom. And at the time when the two cultures of Byzantium and Latin Europe came into immediate contact, particularly after the Fourth Crusade, Gothic architecture in the West was creating its most beautiful and important works. The thirteenth century witnessed the zenith of its dynamism and majesty, especially in France during the reign of Louis IX (1226–70). How did the Byzantines regard this hugely impressive cultural development, the agents of which were the invading Crusaders themselves? How did they respond to this challenge? The written sources are silent when it comes to architecture. Answers to these questions, therefore, can only come from the monuments themselves, in particular those of the thirteenth century when the Crusaders dominated much of Byzantine territory. This essay discusses two distinct groups of localized ecclesiastical monuments: on the one hand, the Gothic-style churches built by the Latins in Byzantine lands and intended for the Latin rite and, on the other, the Greek churches intended for the Orthodox rite and built during the same period. The latter will be examined to find traces of Frankish influence. The arrangement of the sanctuary area and more particularly its division from the congregation (a tall templon screen in Orthodox structures and a low balustrade in Catholic ones) permit us to make a clear distinction between these two groups. In Constantinople and its immediate sphere of influence, Thrace and Macedonia, archi-

[ 248 ] The Impact of Frankish Architecture on Byzantine Architecture tectural activity before the recapture of the capital in 1261 was almost nil. In contrast, the Laskarid empire in Asia Minor saw many noteworthy buildings being erected where Byzantine forms and methods continued to be used unabated.1 In the capital itself, subjected to poverty and insecurity, the Latin emperors did nothing more than renovate two or three chapels of already existing churches.2 The absence, therefore, of Frankish influence in the architecture of the Empire of Nicaea and of Constantinople after the restoration of Byzantine rule during the last quarter of the thirteenth century is self-evident. The situation was quite different in Greece and the islands, where the Latins not only attempted but also succeeded in establishing themselves on a permanent basis, organizing the administration and economy on Western models and establishing monasteries by calling upon the monastic orders of the Latin church.3 In the Peloponnesos in particular, where the principality of the Morea had been established, we find not only the most important monuments of Gothic architecture, but also the most characteristic examples of Western influence on embellished Byzantine church buildings. Thanks to studies made by R. Traquair, C. Enlart, A. Bon, A. Boetticher, A. Orlandos, N. Moutsopoulos, B. Kitsikis, and others,4 we know a fair amount about the Gothic monuments of the Morea erected by Latin monastic orders during the period under examination. The most important can be summarily listed as follows: Hagia Sophia5 and Hagios Iakovos6 at Andravida, the churches of the Panagia7 and Hagios Niko¨ B 28 (1979): 261–96. H. Buchwald, “Laskarid Architecture,” JO On the chapel of St. Francis in Kalenderhane Camii, see C. Striker and D. Kuban, “Work at Kalenderhane Camii in Istanbul, Second Preliminary Report,” DOP 22 (1968): 185–93. The stained glass windows of the Pantokrator church and the Chora monastery of Constantinople, once considered as works of the 12th century (A. H. S. Megaw, “Notes on Recent Work of the Byzantine Institute in Istanbul,” DOP 17 [1963]: 333–71), are now considered to belong to Frankish renovations ( J. Lefond, “De´couverte de vitraux historie´s du Moyen Age a` Constantinople,” CahArch 18 [1968]: 231–38). See also C. Mango, Byzantine Architecture (New York, 1976), 243, 245. For Thessaloniki, see A. Xyngopoulos, “OiJ Fra´ gkoi sth´ n Qessaloni´kh,” Makedoniko´ n JHmerolo´ gion (1965): 37–40. 3 Already in 1210 the Cistercians of Hautecombe, an abbey near Lac Bourget in Savoie, were officially invited to the Peloponnesos by the Latin bishop of Patras, according to a letter of Pope Innocent III (anno XIII, ep. 168, Nov. 1210, PL 216:341–42). 4 R. Traquair, “Frankish Architecture in Greece,” RIBA Journal 31.2 (1923): 33–50, and 31.3 (1923): 73– 86. C. Enlart, “Quelques monuments d’architecture gothique en Gre`ce,” RArtChr 8 (1897): 309–14; A. Bon, La More´e franque: Recherches historiques, topographiques et arche´ologiques sur la principaute´ d’Achaı¨e (Paris, 1969); A. Boetticher, Die frankischen Bauten in Morea, Beilage zur Allgemeine Zeitung (Munich, 1885); A. C. Orlandos, “ JH fragkikh´ ejkklhsi´a th'" Stumfali´a",” in Me´langes offerts a` Octave et Melpo Merlier (Athens, 1955), 1–18. N. Moutsopoulos, “Fragkike´ " ejkklhsi´e" sth´ n JElla´ da,” Tecnika´ Cronika´ 37 (1960): 13–33; idem, “ JH Panagi´a kai´ oJ ”Agio" Niko´ lao" th'" “Isoba",” Tecnika´ Cronika´ 33 (1956): 95–101; idem, “Le monaste`re franc de Notre-Dame d’Isova,” BCH 80 (1956): 80 ff. B. Kitsiki Panagopoulos, Cistercian and Mendicant Monas´ ou konta´ teries in Medieval Greece (Chicago-London, 1979); C. Bouras, “ jEpanexe´ tash tou' legome´ nou Agiole J sth´ n Meqw´ nh,” in Fi´lia “Eph, Volume in Honor of G. Mylonas, vol. 3 (Athens, 1989), 302–22. 5 Bon, La More´e franque, 319–20, 537–53; Traquair, “Frankish Architecture,” 17–20; Moutsopoulos, Fragkike´ " ejkklhsi´e", 19–22; Kitsiki Panagopoulos, Monasteries, 65–77; C. D. Shepherd, “Excavations at the Cathedral of Hagia Sophia, Andravida, Greece,” Gesta 25 (1986): 139–44; N. K. Cooper, “The Frankish Church of Hagia Sophia at Andravida, Greece,” in The Archaeology of Medieval Greece, ed. P. Lock and G. D. R. Sanders (Oxford, 1996), 29–45. 6 Bon, La More´e franque, 319 n. 4; Kitsiki Panagopoulos, Monasteries, 66. 7 Bon, La More´e franque, 352, 354, 537–44; Traquair, “Frankish Architecture,” 2–7; Moutsopoulos, Fragkike´ " ejkklhsi´e", 15–17; idem, “Le monaste`re franc de Notre-Dame,” 80 ff; Kitsiki Panagopoulos, Monasteries, 42–52. 1 2

Charalambos Bouras [ 249 ] laos8 at Isova, the katholikon of the Zarakas monastery at Stymphalia,9 the church of the Virgin De Verge near Methone,10 perhaps that of Hagios Ioannis ho Theologos at Methone,11 and the chapel of Hagios Nikolaos at Aipeia.12 The Crusaders were also responsible for the erection of a number of strong fortresses, perhaps a part of the palace at Mistra, and certain other new settlements.13 The monuments of the second group, the thirteenth-century Byzantine churches of the Peloponnesos, are many in number, usually smaller, and often included in monastic complexes. Ten of them display evident Frankish influence. 1. The katholikon of the Blachernae monastery in Elis,14 a relatively large three-aisled basilica whose construction began during the twelfth century but was not finished until the early Frankish period (Figs. 1, 2). Its western part, the upper story of the tripartite narthex in particular, has pointed arch windows, slender engaged columns running along the corners of the structure, rain spouts, and other Gothic formal elements. The western arcade was evidently rebuilt during the eighteenth century. Recent studies15 indicate that the half barrel vaults over the side aisles of the church do not belong to the twelfth- or thirteenth-century phase but are of a much later date. The church is well preserved. 2. Hagios Georgios in the cemetery at Androusa.16 This single-aisled cross barrelvaulted church preserves Gothic style doorways with pointed arches and jamb molds, cornices, and string courses, but is built with typical mid-Byzantine masonry. 3. The Dormition of the Theotokos outside Anilio17 (formerly Glatsa) near Olympia. This was a small three-aisled, timber-roofed Byzantine basilica whose Gothic elements

8 Bon, La More´e franque, 544–47; Traquair, “Frankish Architecture,” 7–10; Moutsopoulos, Fragkike´ " ejkklhsi´e", 17–19; Kitsiki Panagopoulos, Monasteries, 52–56. 9 Bon, La More´e franque, 553–59; Orlandos, “ JH fragkikh´ ejkklhsi´a”; Kitsiki Panagopoulos, Monasteries, 27–42; Moutsopoulos, Fragkike´ " ejkklhsi´e", 20–24. 10 Bouras, “ jEpanexe´ tash.” 11 Ibid., 318 nn. 94–98; A. Blouet, Expedition scientifique de More´e, vol. 1. (Paris, 1831), 12, pl. 14a,b. 12 G. Dimitrokallis, “Agnwstoi buzantinoi´ naoi´ iJera'" Mhtropo´ lew" Messhni´a" (Athens, 1990), 215–32, esp. 231. 13 Bon, La More´e franque, 601 ff: castles of Glarenza; 602 ff: Chlemoutsi; 608 ff: Karytaina; 629 ff: Kalavryta, Acova, Hagios Basilios, Androussa, Mistras, Geraki, and others. A. C. Orlandos, “Ta´ pala´ tia kai´ ta´ spi´tia tou' Mustra',” Arc.Buz.Mnhm. j JEll. 3 (1937): 13–21. The so-called Building A was possibly built by the Crusaders before the delivery of the Mistra castle to the Greeks. J. Alchermes, “Medieval Towns of Santomeri and the Countryside of Frankish Morea,” in Architectural Studies in Memory of R. Krautheimer, ed. C. Striker (Mainz, 1996), 13–16; J. M. Downs, “The Medieval Settlement at the Hexamilion Fortress at Isthmia,” BSCAbstr 22 (1996): 42. 14 A. C. Orlandos, “AiJ Blace´ rnai th'" jHlei´a",” Arc. j jEf. (1923): 5–34; Bon, La More´e franque, 561 ff; Traquair, Frankish Architecture, 20–24; G. Millet, L’e´cole grecque dans l’architecture byzantine (Paris, 1916), 7, 20, 31, 33–35, 53, 125; A. H. S. Megaw, “The Chronology of Some Middle Byzantine Churches,” BSA 32 (1931– 32): 113, 114, 116, 117, 119, 124, 129; A. Bon, “Monuments de l’art byzantin et de l’art occidental dans le Pe´loponne`se au XIIIe sie`cle,” in Caristh´ rion eij" A. K. jOrla´ ndon, 4 vols. (Athens, 1966), 3:86 ff; Mango, Byzantine Architecture, 254. 15 F. Drossogianni, “Buzantina´ kai´ mesaiwnika´ mnhmei'a Peloponnh´ sou,” Arc.Delt. j 25.2 (1970): 206–7; J. P. Michaud, “Chronique des fouilles en 1971,” BCH 96 (1972): 673. 16 ´ sh",” in Caristh´ rion (as in note 14), 2:270–85; Bon, La C. Bouras, “ JO ”Agio" Gew´ rgio" th'" Androu j More´e franque, 582–84, pl. 97; Kitsiki Panagopoulos, Monasteries, 151. 17 C. Bouras, “ JH fragkobuzantinh´ ejkklhsi´a th'" Qeoto´ kou sto´ Anh j ´ lio (te´ w" Gkla´ tsa) th'" jHlei´a",” Delt.Crist. Arc. j JEt. 12 (1984): 239–64.

[ 250 ] The Impact of Frankish Architecture on Byzantine Architecture (Fig. 3) are confined to the doorways and the double-colonnette icon frames inside the church. Today the church is a ruin. 4. The church of the Rachiotissa at Phlious,18 partly ruined and disfigured today, also included Gothic engaged columns and column capitals on its doorways (Fig. 4). Typologically, it belongs to a very common type of mid-Byzantine domed inscribed cross church. 5. The katholikon of the Palaiomonastero of the Phaneromene19 near Corinth has exactly the same peculiar doorway elements as the Rachiotissa and belongs to the same type of church (Fig. 5). 6. The single-aisled cross barrel-vaulted church of Hagios Georgios at Aipeia20 in Messenia is dominated by a double-pointed arch window with Gothic style jamb molds. 7. The Dormition of the Theotokos at Merbaka in the Argolid.21 This is one of the best known and most beautiful churches of Greece, consisting of an inscribed cross with a dome in an excellent state of preservation. In general, it copies the neighboring church of Hagia Moni in Nauplion (1149) but contains a wealth of Gothic details, such as engaged columns in the trilobed sanctuary window (Fig. A), jamb molds on the arches of the porches that once existed in front of the entrances, column capitals with crockets in the dome (Fig. 6), and other elements. A characteristic feature of the church is the absence of white marble and the use of carved poros stone for all the sculptured architectural elements. The date of the church is much disputed,22 but recent excavations23 and research into the many glazed bowls that decorate the facades24 indicate that it belongs

C. Bouras, “Fliou'", Panagi´a hJ Raciw´ tissa,” Delt.Crist. Arc. j JEt. 16 (1991–92): 39–46; D. Pallas, “ Ana j ´ glufo" sth´ lh tou' Buzantinou' Mousei´ou,” Arc. j jEf. (1953–54), 3:296 n. 3. 19 A. C. Orlandos, Buzantinoi´ naoi´ th'" Anatolikh' j " Korinqi´a", Arc.Buz.Mnhm. j JEll. 1 (1935): 88–90, figs. 36–38; Bon, “Monuments de l’art byzantin,” 89, pl. . 20 I. Staboltzis, “Parathrh´ sei" ejpi´ triw'n cristianikw'n naw'n th'" Messhni´a",” in Praktika´ tou' A⬘ Dieqnou'" Sunedri´ou Peloponnhsiakw'n Spoudw'n, vol. 2 (Athens, 1976–78), 270–81. 21 A. Struck, “Vier byzantinische Kirchen der Argolis,” AM 34 (1909): 201–10, pl. 10; A. Bon, Le Pe´loponne`se byzantin (Paris, 1951), 92–93, 145–46, 149–51; A. H. S. Megaw, “Chronology,” 95, 101, 108, 111, 114, 117–18, 124–25, 127; G. Hadji-Minaglou, L’e´glise de la Dormition de la Vierge a` Merbaka (Hagia Triada) (Paris, 1992). See also the book review of the latter by U. Peschlow in BZ 89 (1996): 470–71. 22 A. Struck suggested a date of about 1140 for the Merbaka church. A. H. S. Megaw (“Chronology,” 129) later dated the church to the last quarter of the 12th century. Recently, G. Hadji-Minaglou suggested a date be´ tween 1130 and 1135. G. Velenis ( JErmhnei´a tou' ejxwterikou' diako´ smou sth´ n Buzantinh´ Arcitektonikh j [Thessaloniki, 1984], 182 n. 2 and 268 n. 1) dated the church of Merbaka before the katholikon of Hagia Moni of Areia, securely dated by an inscription to 1149. A date before 1204 was also accepted by D. Pallas (“ Ana j ´ glufo" sth´ lh,” 296–99; idem, “Eujrw´ ph kai´ Buza´ ntio,” in Byzantium and Europe: First International Byzantine Conference [Athens, 1987], 24–30) and others. 23 By the local Ephoreia of Byzantine Antiquities ( J. Varalis and A. Oikonomou). The results of the excava´ shmeiw´ mata,” in Mnh´ mh D. A. Zakuqhnou', ed. tion are not yet published. See also B. Konti, “ Argolika j G. Moschonas (Athens, 1994), 249–68. 24 Of the 53 originally attached glazed bowls, only 22 can now be studied. A. H. S. Megaw proved that the bowls were set into the masonry of the church while it was built and that most of them were protomajolica bowls (“Glazed Bowls in Byzantine Churches,” Delt.Crist. Arc. j JEt. 4 [1964–65]: 147–48, 153–58). He did not clearly express his opinion on the problem. The date of these bowls in the late 13th century is proved by G. Nikolakopoulos, jEntoicisme´ na kerameika´ III, Ta´ kerameika´ th'" Panagi´a" tou' Me´ rmpaka (Athens, 1979), 37; C. Tsouris, JO keramoplastiko´ " dia´ kosmo" tw'n uJsterobuzantinw'n mnhmei´wn th'" Boreioduti18

Fig. A Merbaka, Argolid, church of the Dormition, exterior of the bema apse

Fig. B Daphni, Attica, monastery, west facade of the exonarthex after the repair by Cistercian monks (based on a drawing of the actual state of the facade in 1955 by E. Stikas)

Fig. C Sykamino, Attica, church of the Panagia Eleousa, south door

1 Elis, Peloponnesos, Blachernae monastery, katholikon, southwest angle of the exonarthex

2 Elis, Peloponnesos, Blachernae monastery, katholikon, northwest angle of the exonarthex

3 Anilio (formerly Glatsa), Peloponnesos, church of the Dormition, shrine on the south side of the bema

4 Phlious, Peloponnesos, church of the Virgin Rachiotissa, capital of door jamb

5 Near Corinth, Palaiomonastero of the Phaneromene, katholikon, capital of door jamb

6 Merbaka, Argolid, church of the Dormition, detail, upper part of the dome

7 Geraki, church of Hagios Georgios, shrine on the north side of the ikonostasis

8 Geraki, church of Hagia Paraskevi, detail of an arcosolium

9 Galatsi, near Athens, Omorphe Ekklesia, ribbed cross vault of the side chapel

10 Avlonari, Euboea, the cross barrel-vaulted church of Hagios Demetrios

11 Arta, church of the Panagia Paregoritissa, interior

12 Ammochostos, Cyprus, Hagios Georgios of the Greeks, interior looking east

13 Geraki, church of Zoodochos Pege, entrance

14 Sykamino, Attica, church of Hagioi Saranta, templon colonnette

Charalambos Bouras [ 251 ] to the late thirteenth century. There is, however, no evidence to associate the church with the learned Latin bishop of Corinth William of Moerbeke (1278–84),25 to whom other, no longer extant buildings may be attributed.26 8. The bell tower of the church of Zoodochos Pege in Karytaina27 has zigzag string courses and engaged corner columns of a typical Gothic type. 9. Hagios Georgios in the castle at Geraki28 preserves an icon frame (Fig. 7) containing sculptural elements executed in the high-relief Western style of the period. 10. Hagia Paraskevi in the castle at Geraki29 is covered with a pointed barrel-vaulted roof and has Gothic formal elements at the entrance to the sanctuary (Fig. 8). In the area held by the former Duchy of Athens, similar monuments of both groups can be found. The Cistercian monks who took over the monastery of Daphni remodeled the open stoa attached to the facade of the Byzantine katholikon30 (Fig. B) and may have built the square enclosed cloister to its south.31 The de la Roche family transformed the Propylaea of the Athens Acropolis into a fortified palace with the addition of many purely Gothic formal elements32 and constructed other buildings that have now disappeared leaving only spolia behind.33 The third phase of the church of Hagios Ioannis Magoutis in Athens34 (demolished, but known to us today thanks to the drawings of A. Couchaud) probably belongs to the same group of purely Western architectural monuments. The structure had three aisles and a square sanctuary covered with a groin vault. kh'" JElla´ do" (Kavala, 1988), 102, 113–14; and G. Sanders, “Peloponnesian Churches,” in Recherches sur la ce´ramique byzantine, ed. V. Deroche and J.-M. Spieser (Paris, 1989), 189–94. 25 The connection of the Dormition church with the archbishop was first suggested by A. Struck (“Vier byzantinische Kirchen,” 234–35). A. Bon accepted the idea and considered the church as a building of the 13th century (“Monuments de l’art byzantin,” 93). For the archbishop of Corinth William of Moerbeke (1278–84), see G. Verbeke, DHGE 22 (1988), 963–66. 26 According to G. Sanders, the Latin cathedral of Corinth of which nothing remains. We may note a few architectural members (spolia), of purely Gothic style, which were recently found in Corinth and belonged to an unidentified building. C. Williams, E. Barnes, and L. M. Snyder, “Frankish Corinth: 1996,” Hesp 66 (1997): 32, pl. 9. 27 N. Moutsopoulos, JH ajrcitektonikh´ tw'n ejkklhsiw'n kai´ tw'n monasthri´wn th'" Gortuni´a" (Athens, 1956), 56–59, fig. 32; Bon, La More´e franque, 588–89, pl. 73b. 28 A. J. B. Wace, “Laconia. Frankish Sculptures at Parori and Geraki,” BSA 11 (1904–5): 144, fig. 4; R. Traquair, “Laconia. The Mediaeval Fortresses,” BSA 12 (1905–6): 265–66, pl. ; A. Van de Put, “Note on the Armorial Insignia in the Church of St. George, Geraki,” BSA 13 (1906–7): 282–83; N. Moutsopoulos and G. Dimitrokallis, JH eJllhnikh´ hJmise´ lhno" (Athens, 1988), 30–31, figs. 21–23. 29 Wace, “Laconia,” 142; Traquair, “Laconia,” 267. 30 G. Millet, Le monaste`re de Daphni (Paris, 1899), 25–42, pl. .2; E. Stikas, “Stere´ wsi" kai´ ajpokata´ stasi" tou' ejxwna´ rqhko" tou' kaqolikou' th'" monh'" Dafni´ou,” Delt.Crist. Arc. j JEt. 3 (1962–63): 1–3, pls. 1, ´a th'" 4; Moutsopoulos, Fragkike´ " ejkklhsi´e", 30–31, figs. 41–42; G. Lampakis, Cristianikh´ Arcaiologi j Monh'" Dafni´ou (Athens, 1889), 96. 31 As a type, the present square court and surrounding porticoes can be considered a Frankish concept, but as architectural forms and constructions they belong to the Turkish period. See also A. C. Orlandos, “New´ tera euJrh´ mata eij" th´ n monh´ n Dafni´ou,” Arc.Buz.Mnhm. j JEll. 8 (1955–56): 67–71, figs. 3–4. 32 ´ polh" kata´ to´ n Mesai´wna (Athens, 1997). ¨kh'" Akro T. Tanoulas, Ta´ Propu´ laia th'" Aqhnai j j 33 ´ polh,” lecture at the Gennadeion LiT. Tanoulas, “Fragkika´ stoicei'a kai´ tecnologi´a sth´ n Akro j brary, Athens, 8 Feb. 1997. 34 A. Xyngopoulos, “Buzantina´ kai´ Tourkika´ mnhmei'a tw'n Aqhnw' j n,” EuJreth´ rion tw'n Mesaiwnikw'n Mnhmei´wn th'" JElla´ do" 2 (1929): 85–87, figs. 1–3.

[ 252 ] The Impact of Frankish Architecture on Byzantine Architecture The Hypapante church was once considered of the same group,35 but it would seem to be of a much earlier date.36 Byzantine churches with Western architectural elements in the same geographical area include the following. 1. The chapel of the Omorphe Ekklesia in Galatsi near Athens,37 a single-aisled structure covered by three purely Gothic groin vaults (Fig. 9). These have pronounced jamb moldings, but the arches of their facades are semicircular. The exterior formal elements are genuinely mid-Byzantine. 2. The church of the Panagia Eleousa, Sykamino in Attica.38 A single-aisled domed church, much altered today, which preserves a Gothic pointed doorway on its south side (Fig. C). 3. The church of the Hagioi Apostoloi, Oropos, also in Attica,39 a single-aisled basilica with a tripartite sanctuary. Two arcosolia preserve arches of a purely Gothic style with pronounced jamb molds. 4. The church of Hagios Georgios, Oropos,40 a three-aisled timber-roofed basilica with a large pointed arch in the colonnades. The church is now a ruin. 5. An unknown church in Athens from which the white marble Gothic-style sculptured arches carved in the round, now exhibited at the Byzantine Museum, originated.41 Their Greek inscriptions may indicate that these were intended for a church of the Orthodox rite. In Chalkis the Venetians made a third(?) and large-scale intervention in the basilica of Hagia Paraskevi,42 completely renovating the sanctuary on the Western model. At the same time, a host of individual towers and fortresses were built in mainland Greece and in Euboea.43 Of the many Orthodox churches of the thirteenth century on the island,

35

Ibid., 63, fig. 48; C. Enlart, “Quelques monuments d’architecture gothique en Gre`ce,” RArtChr (1897): 311 ff; Moutsopoulos, Fragkike´ " ejkklhsi´e", 30, fig. 40. The last remains of the ruined church (a three-aisled basilica) were removed during the late 1950s, when the whole area was incorporated into the Athenian Agora region and excavated by the American School of Classical Studies. 36 A. Xyngopoulos, “Fragkobuzantina´ glupta´ ejn Aqh j ´ nai",” Arc. j jEf. (1931): 69 nn. 1–2. 37 A. C. Orlandos, JH “Omorfh jEkklhsi´a (Athens, 1921); C. Bouras, Buzantina´ stauroqo´ lia me´ neurw´ sei" (Athens, 1965), 62, 64, 68–69, fig. 17. 38 A. C. Orlandos, “Buzantina´ mnhmei'a jWrwpou' kai´ Sukami´nou,” Delt.Crist. Arc. j JEt., ser. 1, 4 (1927): 42–44, fig. 4b, 16, 17; I. N. Koumanoudis, “Peri´ tino" ijdiomo´ rfou tumpa´ nou ojktapleu´ rou trou´ llou tou' 12ou aijw'no",” Tecnika´ Cronika´ 224 (1963): 1–15. 39 Orlandos, “Buzantina´ mnhmei'a,” 31, fig. 4a, 15 no 1. 40 Ibid., 31–34, figs. 5–8. The date suggested by Orlandos was correctly disputed by M. Chatzidakis, “Buzantine´ " toicografi´e" sto´ n jWrwpo´ ,” Delt.Crist. Arc. j JEt. 1 (1959): 87–107; see esp. 107 (early decades of the 13th century). 41 Xyngopoulos, “Fragkobuzantina´ glupta´ ,” 69–102; G. Sotiriou, Guide du Muse´e byzantin d’Athe`nes (Athens, 1932), 48–49; A. Liveri, Die byzantinischen Steinreliefs der 13. und 14. Jahrhunderts im griechischen Raum (Athens, 1996), 177–84, figs. 70–75; J. Maksimovic´, “La sculpture byzantine du XIIIe sie`cle,” in L’art byzantin du XIIIe sie`cle (Symposium de Sopoc´ani), ed. V. J. Djuric´ (Belgrade, 1967), 32–33, figs. 19–20. 42 Traquair, “Frankish Architecture,” 10–16, figs. 13–21; J. Koder, Negroponte, Vero¨ff. TIB 1 (Vienna, 1973), 92 ff. 43 A. Bon, “Forteresses me´die´vales de la Gre`ce centrale,” BCH 61 (1937): 136–208; P. Lock, “The Frankish Towers of Central Greece,” BSA 81 (1986): 101–24; idem, “The Mediaeval Towers of Greece,” Mediterranean

Charalambos Bouras [ 253 ] only Hagios Demetrios at Chania of Avlonari shows signs of Western influence.44 This is a three-aisled basilica with pointed arches both in the longitudinal rows of piers and in the windows of the south facade (Fig. 10). In the despotate of Epiros, which was completely Greek, we have monuments only of the second group, namely, Byzantine churches with Frankish elements. The first two of these are, indeed, foundations of the Komnene-Doukas dynasty itself and thus are of special importance. 1. The Panagia Paregoritissa (1294–96)45 represents a very ambitious design as regards both size and mosaic and fresco decoration. Typologically, it remains unique: a square below and a domed cross-in-square at the level of the vaults. Its vivid originality can be discerned not only in the support system used for the dome—pronouncedly dynamic in the interior—but also in the ornamental trefoil arches and representational reliefs of a purely Gothic style46 (Fig. 11). 2. The Pantanassa at Philippias,47 erected by the despot Michael II, now a ruin, has been excavated over the last twenty years. It was a large domed inscribed-cross church with two side chapels and pairs of colonnettes at the entrances that once formed Gothicstyle porches. It has been argued that both the porches and the column capitals therein are of a later date.48 3. Hagia Theodora in Arta49 had a colonnaded outer pi-shaped stoa covered with Gothic-style pointed ribbed groin vaults. Only the south part of the stoa exists today.50 4. On the narthex of the Porta-Panagia, yet another pointed Gothic arch was added

Historical Review 4 (1989): 129–45. I. Papadimitriou, “Fragkika´ ka´ stra kai´ ojcurw´ mata ejn Eujboi´a,” BNJ 7 (1928–29): 462–64; Koder, Negroponte, 95–99, 105 ff, figs. 27–48. 44 Koder, Negroponte, 137, 163, figs. 63–64; C. Bouras, “A Chance Classical Revival in Byzantine Greece,” in Byzantine East—Latin West: Art Historical Studies in Honor of Kurt Weitzmann, ed. C. Moss and K. Kiefer (Princeton, 1995), 585–90; C. Farantos, “Buzantine´ " kai´ metabuzantine´ " ejkklhsi´e" sti´" perioce´ " tw'n ´ ri . . . th'" N. Eujboi´a",” Arcei' cwriw'n Alibe j j on Eujjboi¨kw'n Meletw'n 23 (1980): 368–70. 45 A. C. Orlandos, JH Parhgorh´ tissa th'" “Arth" (Athens, 1963); L. Theis, Die Architektur der Kirche der Panagia Paregoritissa in Arta/Epirus (Amsterdam, 1990); eadem, “Die Architektur der Kirche der Panagia Paregoritissa,” Praktika´ Dieqnou'" Sumposi´ou gia´ to´ Despota'to th'" jHpei´rou, ed. E. Chrysos (Arta, 1992), 475–93. 46 Orlandos, JH Parhgorh´ tissa, 66–93; L. Safran, “Exploring Artistic Links between Epiros and Apulia in the Thirteenth Century: The Problem of Sculpture and Wall Painting,” Praktika´ Sumposi´ou gia´ to´ Despota'to (as in note 45), 456 ff. 47 ´ Pantana´ ssh" Filippia´ do",” Arcaiologika ´ Ana P. Vokotopoulos, “ Anaskafh j j j ´ lekta ejx Aqhnw' j n5 (1972): 87–97, and 10 (1977): 149–64, figs. 1–3, 17. 48 This purely Gothic arrangement concerns two entrances of the south and one of the north from the Pshaped ambulatory to the main church. P. Vokotopoulos’ suggestion that such ambulatories are in general later than 1250 (Delt.Crist. Arc. j JEt. 1 [1980–81], 372) rather than the archaeological evidence led G. Velenis to the opinion that the ambulatory of the Pantanassa (and consequently the Gothic entrances) are later additions. See G. Velenis, “Thirteenth-Century Architecture in the Despotate of Epirus,” in Studenic´a et l’art byzantin autour de l’anne´e 1200, ed. V. Korac´ (Belgrade, 1988), 281 n. 17. 49 A. C. Orlandos, “ JH Agi J ´a Qeodw´ ra th'" “Arth",” Arc.Buz.Mnhm. j JEll. 2 (1936): 88–104, esp. 103–4, fig. 3; Bouras, Buzantina´ stauroqo´ lia, 63, fig. 15H, pl. 20. 50 The ambulatory was demolished long before the study of the monument by A. C. Orlandos. The form of the Gothic ribs can be seen in an old photograph in the collection of the Ecole Pratique des Hautes Etudes, Paris.

[ 254 ] The Impact of Frankish Architecture on Byzantine Architecture over the entrance at a later date, if the redating of the narthex to the twelfth century is valid.51 It should also be noted that both the sculptural ornamentation52 and the ceramic decorative plaques53 in Artan monuments of the thirteenth century are evidently products of Western influence. The latest research, however, has concluded that this is less due to Crusader activity than to the relations between the despotate and Apulia.54 The churches built by the Venetians in the towns of Crete were as a rule large threeaisled, timber-roofed basilicas:55 Hagios Markos in Herakleion, the church of the Panagia in Canea, those of St. Francis in Herakleion, Hagios Ioannis Prodromos in Canea, the Savior, Hagios Petros Martyr, and the Panagia in Herakleion, and finally St. Francis and Hagios Nikolaos in Canea. These churches were built without particular artistic pretensions and with Gothic elements confined to the interior colonnades and the openings on the facades. Correspondingly, the Orthodox churches in Crete during the thirteenth century are humble structures, usually single-aisled vaulted buildings, small basilicas, or cross barrel-vaulted churches of the simplest variety. Pronounced Western influence appears at times only on the facades: occasional pointed arch openings or decorative sculptural relief, usually much simplified in a popular idiom.56 Unfortunately, the architecture of these monuments has been insufficiently studied, and their dating is still problematic. It would appear evident that many do not belong to the thirteenth century but to the following three centuries. Three small monuments dated by their wall paintings to the thirteenth century (Hagios Georgios in Kouneni, Hagios Ioannis in Gerakari, and Hagios Georgios of Sklavopoula in Selinos) are single-aisled vaulted chambers built in a popular style without evident Western influence.57 Noteworthy exceptions, however, can be found in the church of Hagios Ioannis in Episkopi at Mylopotamos58 with its multilobed arching in the conch of the sanctuary, and in the Timios Stavros church at Monochori, where a Gothic ribbed groin vault covers the entire chapel.59 51 A. C. Orlandos, “ JH Po´ rta Panagia´ th'" Qessali´a",” Arc.Buz.Mnhm. j JEll. 1 (1935): 5–40, esp. 23– 24, fig. 11. On the redating of the narthex, see C. Bouras, “Twelfth- and Thirteenth-Century Variations of the Single Domed Octagon Plan,” Delt.Crist. Arc. j JEt. 9 (1977–79): 27. The redating is recently disputed by S. Mamaloukos, who attempted new measurements and observations of the narthex of the church. 52 Liveri, Die byzantinischen Steinreliefs; Safran, “Exploring Artistic Links.” 53 A. C. Orlandos, “ JO ”Agio" Basi´leio" th'" “Arth",” Arc.Buz.Mnhm. j JEll. 2 (1936): 122–26, figs. 7–8; Tsouris, JO keramoplastiko´ " dia´ kosmo", 76–95, pls. 66–71. 54 Safran, “Exploring Artistic Links.” 55 G. Gerola, Monumenti veneti nell’isola di Creta, vol. 2 (Venice, 1908), 17 ff, Le chiese latine. Pages corresponding to the above nine monuments: 17, 102–5, 112–17, 117–20, 119–21, 122–27, 217–19, 130–34, 135–40. See also S. Curuni and L. Donati, Creta veneziana (Venice, 1988), 111, 210, 242, 252; KitsikiPanagopoulos, Monasteries, 64–127 passim. 56 M. Borbudakis, K. Gallas, and K. Wessel, Byzantinisches Kreta (Munich, 1983), 57–60. For other examples of Cretan monuments that adopted debased Gothic forms, see Curuni and Donati, Creta, 35, 254–55, 322, 328, 333, 335, 343, 384, 390, 395, 398, 400, 412, 414, 423. 57 Borbudakis, Gallas, and Wessel, Byzantinisches Kreta, 200, fig. 149; 282, fig. 57; 239 and 211, fig. 58; and 162–63, respectively. 58 Gerola, Monumenti, 79–83; Curuni and Donati, Creta, 322; Borbudakis, Gallas, and Wessel, Byzantinisches Kreta, 300–302. 59 Borbudakis, Gallas, and Wessel, Byzantinisches Kreta, 333, fig. 295.

Charalambos Bouras [ 255 ] Cyprus is an exception. In the hands of the Crusaders from 1191, it witnessed an outstanding flourishing of Gothic architecture with a series of large and impressive monuments, unique in the East,60 which clearly originate in northern France in their first and major phase (1209–80).61 During the thirteenth century, the architectural activity of the island’s Orthodox inhabitants was very limited. Later, perhaps beginning with the church of Hagios Georgios of the Greeks in Ammochostos62 (Fig. 12)—an Orthodox church built in a purely Gothic style around 1370—Gothic formal elements began to be used by the indigenous architectural tradition to such a degree that it becomes difficult in certain instances to distinguish one group of churches from another. Despite this, C. Enlart maintains that “in Cypriot medieval architecture we do not find the lively and fruitful fusion of elements that true syncretism could produce.”63 The pitched timber roofs that cover Greek churches in Cyprus—whether vaulted or not—have been studied, but their origins have been variously attributed.64 Finally, we have the monuments on the island of Rhodes. Both those of the Knights of St. John,65 as well as those of the Orthodox inhabitants who accepted the use of piecemeal Gothic elements66 (Hagios Demetrios67 and the Panagia church68 in Lindos, the church of the Dormition at Salakos,69 and the Hurmali Medresse70 in the town of Rhodes), are much later, all dating to after the island’s capture in 1309. It is now necessary to analyze the relatively small number of Byzantine monuments influenced by Gothic architecture at its height, discussed above, if we are to understand this phenomenon. Traces of Gothic influence on the typology and function of Byzantine churches remain unconfirmed or simply hypothetical. Greek monuments continued to use the church types known from the earlier period, whether basilicas or domed, almost always with narthex and tripartite sanctuary area. Four elements of thirteenth-century churches have been suggested as betraying Western typological influences. (1) A new type of cross barrel-vaulted church—both of a single and three-aisled vari60 On the Gothic monuments of Cyprus, the book by Camille Enlart, published in 1899, is still invaluable. An English translation under the title Gothic Art and Renaissance in Cyprus was published in London in 1987. 61 Mainly during the reign of the Louis IX. Enlart, Gothic Art, 53 n. 9. 62 Ibid., 255; A. W. Carr, “Byzantines and Italians on Cyprus: Images from Art,” DOP 49 (1995): 340 n. 6, figs. 2, 16–18. 63 Enlart, Gothic Art, preface, 2. 64 ´ " to´ mo" 50ethri´do" periodikou' Apo A. Papageorgiou, OiJ xulo´ stegoi naoi´ th'" Ku´ prou. Anamnhstiko j j ´stolo" Barna´ ba" (Nicosia, 1975), 3–198, esp. 196–98; C. Feraios, Xulo´ stego" fragkobuzantinh´ ajrcitektonikh´ th'" Ku´ prou (Nicosia, n.d.), esp. 233. 65 Also invaluable is the old publication by A. Gabriel, La cite´ de Rhodes (Paris, 1921–23). The architectural activity of the Hospitallers of Rhodes is principally limited to secular buildings and fortifications. 66 A. C. Orlandos, “Buzantina´ kai´ metabuzantina´ mnhmei'a th'" Ro´ dou,” Arc.Buz.Mnhm. j JEll. 6 (1948): 107–8. 67 Ibid., 65, fig. 51. 68 Ibid., 98, fig. 86. 69 Ibid., 100–104, figs. 90–91; A. Alpago-Novello and G. Dimitrokallis, JH buzantinh´ te´ cnh sth´ n JElla´ da (Athens, 1995), 175–76. 70 Orlandos, “Buzantina´ ,” 99–106, figs. 89, 93.

[ 256 ] The Impact of Frankish Architecture on Byzantine Architecture ety—appears in Greece during the thirteenth century.71 The transverse barrel-vault with a pitched roof, which generally takes the place of the dome,72 has been interpreted by some scholars73 not only as a technical simplification of the standard mid-Byzantine domed cross-in-square church, but also as having been influenced by the transept typical of the Romanesque or Gothic cathedral. Indeed, certain large buildings dating to the immediately preceding period and mostly in Apulia74 and Sicily contain the characteristic raised single transverse aisle, and in Gothic cathedrals also the transverse aisle is the formal element that dominates the long sides of the building. Support for this view involves three factors: (a) that the cross barrel-vault type appears and survives in the Peloponnesos, Euboea, Arta, and Crete during the thirteenth century, while it is completely absent from Asia Minor as well as from Constantinople and its immediate sphere of influence; (b) that many of the churches that contain formal elements of Western influence do indeed belong to the cross barrel-vault type;75 and (c) that at least one important monument of the same type76 is clearly attested as having been erected by a Western master craftsman. But the problem has not been definitively solved. It has long been observed that the single-aisled variation of the type already existed in tenth-century Greece in the arrangement of the nartheces of domed cross-in-square churches77 and also that the attempt to do away with interior columns in inscribed-cross churches had led to transitional arrangements that prefigured the cross barrel-vault type even before the Frankish period.78 71 On the type of cross barrel-vaulted churches generally, see A. C. Orlandos, “OiJ staurepi´stegoi naoi´ th'" JElla´ do",” Arc.Buz.Mnhm. j JEll. 1 (1935): 41–52; H. M. Ku¨pper, Der Bautypus der griechischen Drachtranseptkirche (Amsterdam, 1990); M. Doris, Pro´ tash gia´ th´ n tupologi´a tw'n staurepiste´ gwn naw'n (Athens, 1991). 72 The origin of the type according to Orlandos’ theory is connected with the form of troullokamara. However, this uncommon architectural form is usually the product of later alterations and repairs. Orlandos, “OiJ staurepi´stegoi naoi´,” 50–52; idem, “Eine unbeachtete Kuppelform,” BZ 30 (1930): 577–82. 73 ´ sh",” 285 n. 36; H. Hallensleben, review of R. Krautheimer, Bouras, “ JO ”Agio" Gew´ rgio" th'" Androu j Early Christian and Byzantine Architecture, in BZ 66 (1973): 129; Ku¨pper, Der Bautypus, 90–91, 147, 151–55. On the same problem of the origin of the type, see also Borbudakis, Gallas, and Wessel, Byzantinisches Kreta, 62–66, 476; Koder, Negroponte, 160–61. 74 With typical examples St. Nicholas of Bari (F. Schettini, La basilica di San Nicola di Bari [Bari, 1967]) and the Trani cathedral. 75 E.g., the churches of St. George of Androusa, St. George of Aipeia, and St. Demetrius of Avlonari on Euboea. 76 The church of the Savior, near Galaxeidi, a foundation of the despot of Epiros Michael II, of the mid13th century, was built by Nicolo Carouli, who was “peri´fhmo" tzinie´ rh" . . . pou´ ejsta´ qhke sth´ dou´ leyh th'" Fraggi´a"”; see P. Vokotopoulos, “Parathrh´ sei" sto´ n nao´ tou' Swth'ro" konta´ sto´ Galaxei´di,” Delt.Crist. Arc. j JEt. 17 (1993–94): 199–210, and Euthymios hieromonachos, Croniko´ tou' Galaxeidi´ou (Athens, 1985), 28. The name of the arch builder is Italian, but the architectural forms of the church are purely Byzantine. 77 G. Dimitrokallis, “ JH katagwgh´ tw'n staurepiste´ gwn naw'n,” in Caristh´ rion (as in note 14), 2:187– 211; Ku¨pper, Der Bautypus. Three of the oldest examples of nartheces of this type are those of the Prophet Elias of Staropazaro in Athens, of the Panaxiotissa in Gavrolimni, and of the Metamorphosis at Koropi, in Attica. 78 A. H. S. Megaw correlated a very rare type of 12th-century church with transverse barrel vault surmounted by a small dome (considered a simplified version of the inscribed-cross domed type) with the very common one-aisled cross barrel-vaulted church. A. H. S. Megaw, “Byzantine Architecture in Mani,” BSA 33 (1932–33): 160–61; C. Bouras, “Sthri´xei" suneptugme´ nwn trou´ llwn se´ monokli´tou" naou´ ",” Eujfro´ sunon. Afie j ´ rwma sto´ n Mano´ lh Catzhda´ kh, vol. 2 (Athens, 1992), 413 nn. 31–33.

Charalambos Bouras [ 257 ] The difficulty faced in the construction of the support system for a dome is clearly the most important factor in the development of the new type of cross barrel-vault churches in an age when donors, as we shall see, requested smaller and less expensive churches. (2) The tendency toward an elongated ground plan. This was asserted by A. Bon79 and Cyril Mango80 but is not in fact confirmed by the material evidence. On the contrary, certain Frankish monuments of the Peloponnesos (Hagios Nikolaos at Isova and the chapel of Hagios Nikolaos of Aipeia) have almost square ground plans, even though they belong to the basilica type. (3) Bon81 and Mango82 also attribute the erection of bell towers to Western influence. There are indications, however, that even by the twelfth century these had purely Byzantine origins in churches in the northern regions immediately under Constantinople’s influence, such as Kursˇumlija,83 Bjelo-Polje,84 and the Omorphoekklisia (Galista) near ´ urcˇic´, basing himself Kastoria.85 Other examples are to be found in Greece.86 Slobodan C on certain later examples, argues for a Byzantine origin of the bell tower in these regions.87 (4) According to G. Millet, the timber-roofed, three-aisled basilica begins to be used once again in Greece in the thirteenth century, thanks to Frankish influence.88 Certain early mid-Byzantine monuments, however, such as the basilicas at Zourtsa,89 of Episkopi at Mastron,90 at Metzaina,91 and of the Blachernae at Arta92 indicate, to the contrary, that this type continued to be used in Greece from Early Christian times. In terms of architectural form and decoration, however, Western influence is evident. But in virtually all the Byzantine ecclesiastical monuments noted above, these elements appear as isolated formal elements and are not widespread throughout the building. We have, in other words, one or two pointed openings (doorways or windows), slender columns occasionally engaged to the corners, pointed arcosolia, colonnettes in doorways fashioning portals (Fig. 13), and, in the interior, icon frames with Gothic column capitals. It may be said that these isolated Gothic details (Fig. 14) were meant to provide a Bon, La More´e franque, 587. Mango, Byzantine Architecture, 254. 81 Bon, La More´e franque, 588. 82 Mango, Byzantine Architecture, 254. 83 B. Vulovic´, “Die Heilige Nikola Kirche bei Kursumlija,” Zbornik Arhitektonskog Fakulteta 3 (1956–57): 3–22. 84 ´ anak-Medic´ and D. Boskovic´, L’architecture de l’e´poque de Nemanja, vol. 3 (Belgrade, 1986), 47–72. M. C 85 E. Stikas, “Une e´glise des Pale´ologues aux environs de Castoria,” BZ 51 (1958): 102 n. 6. The 12thcentury frescoes on the upper floor of the belfry prove that it was built long before the actual church; ibid., 105, 108. 86 C. Bouras, “ jEpanexe´ tash tou' kaqolikou' th'" Zwodo´ cou Phgh'", Derbenosa´ lesi,” Delt.Crist. Arc.j JEt. 17 (1993–94): 31 nn. 25–27. 87 ´ urcˇic´, “Byzantine Legacy in Ecclesiastic Architecture of the Balkans after 1453,” in The Byzantine LegS. C acy in Eastern Europe, ed. L. Clucas (New York, 1988), 69 ff. 88 Millet, L’e´cole grecque, 21 ff. 89 C. Bouras, “Zourtsa, une basilique byzantine au Peloponne`se,” CahArch 21 (1971): 137–49. 90 P. Vokotopoulos, JH ejkklhsiastikh´ ajrcitektonikh´ eij" th´ n Dutikh´ n Sterea´ n JElla´ da kai´ th´ n “Hpeiron, 2d ed. (Athens, 1992), 11–20. 91 Ibid., 35–41. 92 As it was originally in the first building phase; ibid., 25–28. 79 80

[ 258 ] The Impact of Frankish Architecture on Byzantine Architecture certain variety to the church, which otherwise generally preserved the traditional middle Byzantine form. Indeed, the random and eclectic addition of Frankish architectural forms represents nothing more than the intensification of a tendency toward variety. Byzantine texts often refer to variety as a positive factor in structural aesthetics. The inclusion of dissonant formal elements in architecture betrayed the spirit of picturesque irregularity that generally characterized all of Byzantine architecture. On occasion, we can also discern the transformation of Gothic formal elements. For example, in the windows of the church of the Blachernae in Elis, the dentil courses of purely Byzantine technique are accompanied by a slender Gothic molding and a projecting hood carved with a large, well-cut dog-tooth. At Hagios Georgios in Aipeia, the double Gothic window has a marble Byzantine colonnette and a large ceramic bowl in place of the standard quatrefoil or trefoil Gothic rosette. Much debate has taken place over whether Western architectural forms appeared in Greece prior to the Crusaders’ invasion of 1205.93 This discussion was based, however, on the mistaken dating of the church of the Dormition at Merbaka in the Argolid, mentioned earlier. The whole question is thus no longer of relevance. Furthermore, those Western elements known in Byzantine churches belong to the mature Gothic and not to the Romanesque style of the twelfth century. Two questions should be broached here. (1) Slightly pointed arches of a simple nature without moldings and accompanying Gothic characteristics have been identified in Byzantine churches securely dated before the appearance of the Crusaders. Examples can be found in the south doorway of the katholikon of Hagia Moni near Nauplion (1149)94 and in the chapel of the Virgin and refectory of the monastery of St. John the Theologian on Patmos.95 This type of arch is easily scribed with two center points during construction and can be found in the East in both Byzantine96 and Islamic architecture.97 (2) The fine workmanship used in the cutting of stone masonry blocks that characterizes Romanesque and especially Gothic architecture is also encountered in many twelfth-century Byzantine churches in Greece.98 Here, also, associations were made between Frankish and Byzantine techniques. It is very likely, however, that what we have here is a parallel phenomenon of technological progress, unrelated at least till 1205. The most significant achievements of Gothic architecture, however, do not concern conventional matters of form. They involve the vertical extension of the interior space, accentuated dynamism of exterior facades and interior spaces, and the resolution into 93 Pallas, “ Ana j ´ glufo" sth´ lh,” 296 ff; idem, “Eujrw´ ph kai´ Buza´ ntio,” 25 ff; Bouras, Buzantina´ stauroqo´ lia, 65–73; Bon, “Monuments de l’art byzantin,” 93 ff. 94 The entrance is now walled in. There is no indication of the entrance and the pointed arch on the drawings of A. Struck in AM 34 (1909), pl. . 95 A. C. Orlandos, JH ajrcitektonikh´ kai´ aiJ buzantinai´ toicografi´ai th'" monh'" Qeolo´ gou Pa´ tmou (Athens, 1970), 78 ff, fig. 60. 96 Ibid., 80–81 nn. 1–8. 97 K. A. C. Creswell, Early Muslim Architecture (Harmondsworth, 1958), 55, 85–86, 101–4, 116, 131, 143, 157, 195. 98 ´ sei"’ kai´ ajrcitektonikh´ tou' Bouras, Buzantina´ stauroqo´ lia, 65–73; idem, “Buzantine´ " ‘ Anagennh j 11ou kai´ 12ou aijw'no",” Delt.Crist. Arc. j JEt. 5 (1969): 268–71.

Charalambos Bouras [ 259 ] colonnettes of the main supports to correspond to the ribbing that dynamically forms the vaulting. To all these factors, essential to the Gothic style, the Byzantine response was totally negative. The only instance where one could hypothesize a certain analogous addition can be found in the church of the Paregoritissa in Arta, which, as we saw, contains Gothic formal elements.99 Indeed, this church of the despotate of Epiros possesses a very lively interior dynamism and represents a unicum of Byzantine architecture due to the almost acrobatic support system used for the dome and the complete priority of the vertical axis. If, however, we look closely at this system, we see that it has nothing in common with Gothic techniques. The marble columns, which via arches support the very tall dome, stand on corbels arranged by height on two levels.100 These columns are, moreover, Early Christian spolia; they have nothing in common with the complex Gothicstyle vaulting supports. The small pointed trefoil arches are clearly decorative and have no relation to the ribbed arches of Gothic architectural works.101 Even if we accept that the addition of this intense dynamism is related to contemporary European architectural trends, the techniques and methods used here are improvisational, original, and foreign to the Western tradition. Finally, in matters of architectural technology, where the Latins of the thirteenth century were innovators, we once again find that the Byzantines made very little use of these advances. Latin ribbed cross-vaults in their simplest form were known from the eleventh and twelfth centuries102 but never realized their full potential. And when during the Frankish occupation the pointed ribbed cross-vault was introduced, it was usually of restricted size and was never widely disseminated.103 Another kind of vault, the half tunnel-vault, appears in Byzantine churches in Corinthia and the Mani, and was also considered to have its origins in Latin models.104 Its extremely limited dissemination, however, indicates that it constituted an improvised solution to particular roofing problems encountered in the construction of inscribed cross-domed churches.105 It is not difficult to interpret the very limited and eclectic influence of Western architecture on Byzantine architecture in the thirteenth century. The international Gothic 99 Note the instructive drawings of Orlandos in JH Parhgorh´ tissa th'" “Arth", figs. 51–55, 57–58. See also the comments of R. Krautheimer, Early Christian and Byzantine Architecture (Harmondsworth, 1986), 417–18. 100 Orlandos, JH Parhgorh´ tissa, 60–64. 101 Ibid., figs. 64, 76, 78. 102 Bouras, Buzantina´ stauroqo´ lia, passim. 103 Ibid., appendix, 62 ff. 104 Bon, “Monuments de l’art byzantin,” 88–90, fig. b, with reference to previous known examples, 89 n. 7. Additional cases of half-tunnel vaults appear in St. Ioannis at Platsa in the Mani (A. Christophidou, 17th Annual Symposium of the Christian Archaeological Society [Athens, 1997], 86) and in a number of chapels of the Ottoman period (I. Kakouris, “To´ monasth´ ri tou' Taxia´ rch Micah´ l th'" Moura'" Kunouri´a",” Peloponnhsiaka´ 16 [1985–86]: 320 ff). 105 St. Mamaloukos, “Parathrh´ sei" sth´ n diamo´ rfwsh tw'n gwniakw'n diamerisma´ twn tw'n dikioni´wn naw'n th'" JElla´ do",” Delt.Crist. Arc. j JEt. 14 (1987–88): 200, 203, fig. 8.

[ 260 ] The Impact of Frankish Architecture on Byzantine Architecture style did not predominate in Byzantine lands, despite its conscious introduction by the Crusaders in the form of a few large buildings. Something similar occurs in southern Italy and the area of Rome, where there was a strong and self-contained local tradition accompanied by opposition to foreign trends and styles.106 While the purely Gothic buildings of the Morea may have been large and impressive, they were too few in number to serve as models, and, apart from those built in areas held by the Venetians, their life-span as functioning churches was very short, perhaps not more than a century from the date of their completion.107 A more significant factor, however, in the Byzantines’ unwillingness to accept Western architectural models involves the founder-donors, those who had the initiative to fund and erect individual churches. We can approach the subject only through donor inscriptions, which, even though found in very few of the monuments discussed here, provide us with a sufficiently satisfactory picture of the erection and decoration of many Orthodox churches, especially in the thirteenth century. The recent publication by S. Kalopissi-Verti greatly facilitates the study of the question.108 The donor inscriptions for the building and decoration of more than seventy churches in the examined area of Greece represent all social strata: from despots and officials of the Epirote state and members of the local Greek aristocracy to ordinary people, bishops, priests, and simple monks.109 The disproportionately large number of wall-painting groups, but also of small churches in the Greek countryside during the Frankish period,110 indicates that under a regime of religious tolerance—but also of relative prosperity—donors could include even Greeks of limited financial resources. The subject of whether Greek archons were incorporated into the Western feudal system in the principality of the Morea has been meticulously studied, but only insofar as the legal ramifications of the question are concerned.111 The gap that clearly existed between the two religious dogmas was never bridged, either by the archons or the populace at large. Consequently, differences existed in all religious and cultural matters. Analogous observations have been made for Cyprus.112 Manolis Chatzidakis’ observation is revealing: Byzantine painting in the thirteenth and fourteenth centuries is also characterized by the absence of Western influence.113 Contemporary texts, such as the works of 106

R. Krautheimer, Rome: Profile of a City (Princeton, 1980), 211–12. From the Cistercian regulation (statute) of the Daphni monastery we are informed that in 1276 all the Latin monasteries in the former Byzantine territory were occupied and destroyed by the Greeks, except that of Daphni; F. Lenormant, “Le monaste`re de Daphni pre`s d’Athe`nes sous la domination des princes croise´s,” RA 24.2 (1872): 238. 108 S. Kalopissi-Verti, Dedicatory Inscriptions and Donor Portraits in Thirteenth-Century Churches of Greece (Vienna, 1992). 109 Ibid., 28–41. 110 M. Chatzidakis, “ JH mnhmeiakh´ zwgrafikh´ sth´ n JElla´ da, Posotike´ " proseggi´sei",” Praktika´ th'" ´a" Aqhnw' Akadhmi j j n 56 (1981): 375 ff; idem, “Les aspects de la peinture monumentale byzantine en Gre`ce,” Glas 338 de l’Acade´mie Serbe des Sciences et des Arts, Classe des sciences historiques, 3 (1983): 1–10. 111 D. Jacoby, “Les archontes grecs et la fe´odalite´ en More´e franque,” TM 2 (1967): 421–81; J. Ferluga, “L’aristocratie byzantine en More´e au temps de la conqueˆte latine,” ByzF 4 (1972): 76–87; D. Jacoby, “Un re´gime de coseigneurie greco-latine en More´e, Les ‘casaux de parc¸on,’” Me´lRome 75 (1963): 111–25. 112 Enlart, Gothic Art, 2. 113 At least in Greece; see Chatzidakis, Posotike´ " proseggi´sei", 389 n. 15. 107

Charalambos Bouras [ 261 ] Michael and Niketas Choniates, along with the Chronicle of the Morea,114 give us an idea of the attitude of distrust and contempt in which the Orthodox Greeks held the Latins.115 A second factor involved in the interpretation of the entire phenomenon concerns the builders, technicians, and craftsmen in general. There is no doubt that along with monks from Italy and France, skilled craftsmen were also called in to erect those Gothic monuments in which the Latin rite would be celebrated. Stylistic details in the churches at Zarakas, Isova, Andravida, and Chalkis are proof of this. For the Cypriot monuments, the employment of Western craftsmen is also verified by written documents. Whether they themselves also played a part in the careful application of Gothic forms to Orthodox buildings while working thereupon remains a hypothesis that is proved only in extremely rare cases, such as in the sculpture of the Paregoritissa in Arta and the marble arches in the Byzantine Museum in Athens. Foreign craftsmen seem to have played a decisive role only in other places where local architecture had not previously been developed; such is the case in the churches of the so-called Rasˇka School in Serbia during the twelfth century. Here, however, the situation was completely different. It seems, nevertheless, that local Greek craftsmen did at times work in Crusader building projects. The funerary plaque (1286) of Agnes, wife of William II, in the church at Andravida was undoubtedly made by Byzantine marble workers,116 and fourteenthcentury inscriptions attest that Greek craftsmen worked on the town walls of Rhodes and on the buildings of the Gattilusi in Ainos and Samothrake.117 Their specific contribution, however, is not possible to gauge, in either formal or technological terms. We arrive then at the conclusion that the influence of Frankish on Byzantine architecture in the thirteenth century was insubstantial and is evinced only in certain limited and isolated formal elements in buildings that preserve the general style of the mid-Byzantine period. These are found on mainland Greece and in certain islands; in the capital, Asia Minor, Thrace, and Macedonia they are completely absent. The limited nature of this influence is due on the one hand to the fact that there preexisted in Byzantine territory a lively, self-contained, local architecture, and on the other to the great cultural and religious divide between the invading Crusaders and the locals. Majestic Gothic architecture was not well known and found few admirers among the Greeks. The suspension of the influence of the great cultural center of Constantinople for almost sixty years brought about a general decline in Byzantine architecture, but imme114 Niketas Choniates, Historia, ed. I. Bekker (Bonn, 1835), 847 ff; P. Kalonaros, ed., Croniko´ n tou' More´ w" (Athens, 1940), 35, lines 760–71. 115 H. Ahrweiler, L’ide´ologie politique de l’Empire byzantin (Paris, 1975), 103–14. Kalopissi-Verti, Dedicatory Inscriptions, 46; A. Alieva, Frankish Morea: Socio-Cultural Interaction between the Franks and the Local Population (Athens, 1991); B. Ferjancˇic´, “Rapports entre les grecs et les latins apre`s 1204,” ZRVI 10 (1967): 171. 116 A. Bon, “Dalle fune´raire d’une princesse de More´e (XIIIe sie`cle),” MonPiot 49 (1957): 129–39; idem, “Pierres inscrites ou armorie´es de la More´e franque,” Delt.Crist. Arc. j JEt. 4 (1964–65): 96, pl. 29; idem, La More´e franque, 590–91, pl. 21a,b. 117 Gabriel, La cite´ de Rhodes, 98, inscription no. 57 of the year 1457; F. W. Hasluck, “Monuments of the Gatelusi,” BSA 15 (1908–9): 254, 256; A. Konze, Reise auf den Inseln des Thrakischen Meeres (Hannover, 1860), 55, pl. .7–8; C. Asdracha and C. Bakirtzis, “Inscriptions byzantines de Thrace,” Arc.Delt. j 35.1 (1980): 271 ff.

[ 262 ] The Impact of Frankish Architecture on Byzantine Architecture diately after the recapture of the city in 1261 a new flourishing began under the Palaiologan dynasty.118 This architecture is characterized by a trend toward a neatness of line and an increase in the decoration of exterior building surfaces. In the Morea, this new acme manifests itself in the erection of the beautiful churches of Mistra. During the course of the fifteenth century, Gothic formal elements attempting to create a sense of variety appear in the church of the Pantanassa, the refectory of the monastery of the Peribleptos, and in the new wing of the palace built by the Palaiologoi.119 These are to be interpreted as indications of an eclectic taste that was popular for a short period prior to the end of Byzantium and of Byzantine architecture in general. National Technical University of Athens

Krautheimer, Early Christian and Byzantine Architecture, 415 ff; Mango, Byzantine Architecture, 252 ff. On the three monuments respectively, see G. Millet, Monuments byzantins de Mistra (Paris, 1910), pl. 30.5, 35–37; D. Mouriki, “Palaeologian Mistra and the West,” in Buza´ ntio kai´ Eujrw´ ph, ed. A. Markopoulos (Athens, 1987), 209 ff, esp. figs. 2–5; Orlandos, Ta´ pala´ tia kai´ ta´ spi´tia tou' Mustra', 32 ff; St. Sinos, “Organisation und Form des Palastes von Mystras,” Architectura 17 (1987): 105–28; Mango, Byzantine Architecture, 290–91. 118 119

Art and Identity in the Medieval Morea Sharon E. J. Gerstel In 1204 the Frankish knight Geoffroy de Villehardouin arrived unexpectedly at the southwestern corner of the Peloponnesos. As Villehardouin’s uncle later noted in his chronicle: “It happened by chance that the wind carried his ship to the port of Methone [Modon], where it was so badly damaged that he was obliged to spend the winter in those parts.” 1 Within a decade Villehardouin had conquered the Morea. In 1210 Pope Innocent III referred to him as the prince of Achaia.2 When, in that year, Villehardouin’s wife and son arrived from Champagne, he installed them in his residence in Kalamata and thereby confirmed his intentions to settle permanently in the region. Geoffroy’s second son, Guillaume II Villehardouin (1246–78), ruled over most of the Peloponnesos. But in 1261, as a result of his imprisonment following the battle of Pelagonia, Guillaume was forced to cede to Byzantium the southeastern Morea, a triangle of land demarcated by the castles of Maı¨na, Monemvasia, and Mistra (see Map).3 The barony of Geraki was turned over to the Byzantine Empire shortly thereafter. Although the Latin principality of the Morea survived the death of Guillaume II, the weak leadership of Italian and Angevin landlords oversaw its decrease in size and power. By 1430 Latin rule in the region had come to an end. The effect of the Frankish conquest on the Morea has been a fruitful topic for historians and philologists.4 Art historians and archaeologists have studied Latin influences on Orthodox ecclesiastical architecture, but the impact of Frankish rule on the region’s monumental painting has proven difficult to assess.5 Some evidence of the Latin presence has been found in the details of narrative scenes, from the occasional embossing of haloes 1

Joinville and Villehardouin, Chronicles of the Crusades, trans. M. R. B. Shaw (New York, 1963), 113. Innocentii III opera omnia: PL 216:221. 3 D. A. Zakythinos, Le despotat grec de More´e, vol. 1 (Paris, 1932), 15–25; A. Bon, La More´e franque: Recherches historiques, topographiques et arche´ologiques sur la principaute´ d’Achaie (1205–1430) (Paris, 1969), 125–29. 4 Among the many studies, see D. Jacoby, “The Encounter of Two Societies: Western Conquerors and Byzantines in the Peloponnesus after the Fourth Crusade,” AHR 78 (1973): 873–906; P. Topping, “Co-Existence of Greeks and Latins in Frankish Morea and Venetian Crete,” 15th CEB (Athens, 1976), 3–23, repr. in P. Topping, Studies in Latin Greece, .. 1205–1715 (London, 1977), no. ; J. Horowitz, “Quand les Champenois parlaient le grec: La More´e franque au XIIIe sie`cle, un bouillon de culture,” in Cross Cultural Convergences in the Crusader Period: Essays Presented to Aryeh Grabois on His Sixty-fifth Birthday, ed. M. Goodich, S. Menache, and S. Schein (New York, 1995), 111–50. 5 For Western influences on Byzantine architecture, see the study by C. Bouras in this volume, with collected bibliography. 2

[ 264 ] Art and Identity in the Medieval Morea to unusual representations of soldiers at the Arrest and Crucifixion of Christ.6 Such details, for the most part, have left an impression of only minimal Frankish influence on the decoration of Orthodox churches.7 In this study, I examine certain distinctive elements of monumental decoration in the Morea in order to reassess Latin influence.8 I begin with works sponsored by the Frankish rulers of the region and then examine Orthodox churches decorated during and immediately following the Latin occupation of the southern Peloponnesos.

Monumental Decoration of the Frankish Morea Our books have informed us that the pre-eminence in chivalry and learning once belonged to Greece. Then chivalry passed to Rome, together with that highest learning which now has come to France. God grant that it may be cherished here, and that it may be made so welcome here that the honor which has taken refuge with us may never depart from France: God had awarded it as another’s share, but of Greeks and Romans no more is heard, their fame is passed, and their glowing ash is dead. Chre´tien de Troyes, Clige´s9 What remains of Crusader-sponsored painting is central to any discussion of Western impact on monumental decoration in the Morea. Two painted programs are known 6 For embossed haloes in the Morea, see N. Drandakes, S. Kalopissi, and M. Panayotidi, ““Ereuna sth` Messhniakh` Ma´ nh,” Prakt. Arc. j JEt. (1980): 208–9. Comments on the “Latinizination” of military costume in narrative scenes are too widespread to cite individually. 7 In a study of late Byzantine painting at Mistra, D. Mouriki suggests three reasons for the absence of Latin influence on Orthodox monumental painting: “a) Conservative Byzantium would have felt that iconographical and stylistic changes might entail alterations in dogma and doctrinal error; b) Byzantium was rightfully snobbish about its achievements in painting, a province in which it felt itself unsurpassed; and c) the Byzantine understanding of painting remained conservative less out of fear of the danger of heresy and more out of a conviction that a static and hieratic approach was the only one proper to religious painting.” D. Mouriki, “Palaeologan Mistra and the West,” in Byzantium and Europe: First International Byzantine Conference, Delphi, 20–24 July, 1985 (Athens, 1987), 239. See also A. Grabar, “L’asyme´trie des relations de Byzance et de l’Occident dans le domaine des arts au moyen aˆge,” in Byzanz und der Westen: Studien zur Kunst des europa¨ischen Mittelalters, ed. I. Hutter (Vienna, 1984), 9–24. 8 This study has benefited from discussions on site with J. C. Anderson, A. Bakourou, E. Jeffreys, and J. Papageorgiou. I thank, in particular, Aimilia Bakourou, director of the Fifth Ephoreia of Byzantine Antiquities, whose comments have helped shape this study, and the anonymous readers for their invaluable suggestions. I also thank C. Jolivet-Le´vy for her careful reading of a draft of the paper. This article is based solely on published monuments, although the substance of the argument includes unpublished material. It is not my intention to discuss Orthodox painting in the northern Morea and Attica in this study since the pattern of interaction between Greeks and Latins in that region differed substantially. For painting in the northern Peloponnesos and the region around Athens, and comments on the painted results of cultural exchange, see S. Kalopissi-Verti, Die Kirche der Hagia Triada bei Kranidi in der Argolis (1244) (Munich, 1975), and N. Coumbaraki-Panselinou, “”Agio" Pe´ tro" Kalubi´wn Koubara' Attikh' j ",” Delt.Crist. Arc. j JEt. 14 (1987–88): 173–87. 9 Chre´tien de Troyes, Arthurian Romances, trans. W. W. Comfort (New York, 1975), 91. Chre´tien, who is associated with the court of Marie, countess of Champagne, in Troyes, wrote Clige´s around the year 1176. The holdings of the Villehardouin family were to the northeast of Troyes and to the west of Brienne. For the family’s holdings in France, see J. Longnon, Recherche sur la vie de Geoffroy de Villehardouin (Paris, 1939), 6–13.

Sharon E. J. Gerstel [ 265 ] from the northern Peloponnesos, though neither decorates the interior of a church. The first is a reception hall in the archbishop’s hospitium in Patras, which was described in a traveler’s account written by Niccolo` de Martoni in 1395. According to Niccolo`, the walls of the prelate’s residence were decorated with scenes from the destruction of Troy.10 The fall of Troy was a popular subject in courtly art and literature produced in the medieval West. A full-length account, the Roman de Troie, was composed in Old French by the Benedictine monk Benoıˆt de Ste.-Maure in 1160–70.11 The epic was widely illustrated in manuscripts and other media.12 The Trojan tale held enormous appeal for the Latin Crusaders as not only a work of romance and chivalry but also an expression of Crusader ideology.13 According to the Western view, the medieval Greeks were to be held responsible for the defeat of Troy. In 1204 the French knight Pierre of Bracheux justified the conquest of Byzantine territory by asserting that escaped Trojans had settled in France and were the direct ancestors of the Crusaders.14 The selection of the Trojan tale for the decoration of the archbishop’s hospitium demonstrates the story’s popularity in Frankish Morea, the lands once held by Agamemnon. The second example of Latin-sponsored painting in the northern Morea survives in the antechamber of a gatehouse at Akronauplion.15 The gatehouse, which functioned until 1463, was accessible to Latin and Greek residents of the city. Three identifiable coats of arms painted over the antechamber’s west portal suggest a date for the program. Wulf Schaefer, who published a preliminary study of the gatehouse decoration, associated the coats of arms with Hugues de Brienne, count of Lecce and bailie of the duchy of Athens from 1291 to 1294, Isabelle of Villehardouin, princess of Achaia, and Florent of Hainault, Isabelle’s husband and the Latin ruler of the principality of the Morea from 1289 to 1297.16 Schaefer’s description of the painted insignia matches the coat of arms of Isabelle of Villehardouin and her husband carved on a capital in the cathedral of St.

10 “Habet unam salam longam paxus XXV, in cujus sale parietibus est picta in circuytu tota ystoria destructionis civitatis Troye.” L. Legrand, “Relation du pe`lerinage a` Je´rusalem de Nicolas de Martoni, notaire italien (1394–1395),” ROL 3 (1895): 661. 11 Le Roman de Troie, ed. L. Constans, 6 vols. (Paris, 1904–12). For a more recent analysis of Benoıˆt de Ste.Maure’s text, see M. Jung, La le´gende de Troie en France au moyen aˆge: Analyse des versions franc¸aises et bibliographie raisonne´e des manuscrits, Romanica Helvetica 114 (Basel-Tu¨bingen, 1996). 12 H. Buchthal, Historia Troiana: Studies in the History of Medieval Secular Illustration, Studies of the Warburg Institute 32 (London, 1971). 13 D. Jacoby, “Knightly Values and Class Consciousness in the Crusader States of the Eastern Mediterranean,” Mediterranean Historical Review 1 (1986): 170–73; idem, “La litte´rature franc¸aise dans les e´tats latins des croisades,” in Essor et fortune de la Chanson de geste dans l’Europe et l’Orient latin, vol. 2 (Modena, 1984), 635–36; E. M. Jeffreys, “Place of Composition as a Factor in the Edition of Early Demotic Texts,” in Origini della letteratura neogreca, ed. N. Panayotakis (Venice, 1993), 319. 14 Jacoby, “Knightly Values,” 171. 15 W. Schaefer, “Neue Untersuchungen u¨ber die Baugeschichte Nauplias im Mittelalter,” JDAI 76 (1961): 156–214; D. Pallas, “Eujrw´ ph kai´ Buza´ ntio,” in Byzantium and Europe (as in note 7), 9–61; M. Hirschbichler, “The Crusader Paintings in the Gatehouse of Akronauplia, Greece” (M.A. thesis, University of Maryland, 1998). A comprehensive study of the gatehouse is being prepared by Aimilia Bakourou and Monika Hirschbichler. 16 Schaefer, “Neue Untersuchungen,” 198–209. Schaefer suggests that the gatehouse decoration celebrated the treaty of Glarentza, ratified in 1290.

[ 266 ] Art and Identity in the Medieval Morea Sophia in Andravida.17 The late thirteenth-century date of the gatehouse decoration and its Latin patronage are thus secure. The subjects selected for the gatehouse fully reflect its Western patronage.18 On the upper, arched surface of the west portal, the artist painted a medallion containing the Agnus Dei carrying the Crusader banner, a red cross on a white ground.19 Two fulllength holy men, St. Anthony the Great and St. James, decorate the south wall, and both are identified with Latin inscriptions (Fig. 1). St. Anthony, dressed as a monk, carries an open scroll in his left hand. With his short, divided beard and covered head, he was presumably recognizable to both Latin and Greek viewers. The image of St. James, however, must have perplexed the Orthodox inhabitants of the city. His undistinguished facial features could belong to a number of Byzantine male saints, but his costume is unique. The painter depicted James according to Western fashion, wearing a brimmed hat decorated with a cockle shell, the badge awarded pilgrims upon successful completion of the arduous journey to Santiago de Compostela, the saint’s shrine in distant Spain.20 In his right hand James carries a long staff, the characteristic walking stick of the medieval pilgrim, and in his left, the pouch or scrip often associated with the saint and visitors to his shrine.21 Similar pouches are carried by a Western couple represented on a column in the church of the Nativity in Bethlehem, a program that has been associated with Crusader patronage (Fig. 2).22 The two pilgrims kneel at the feet of St. James; their black purses are decorated with a white shell to show that they had traveled to Spain before undertaking the journey to the Holy Land. Although the Life of Meletios the Younger mentions the saint’s journey from Rome to “the Gauls of James,” Compostela appears to have been a rare destination for Byzantine pilgrims.23 The selection and depiction of St. James in the Nauplion gatehouse reflects Latin patronage, and especially the patronage of the local rulers of the Morea. The saint’s importance to the region’s Crusader lords was demonstrated by the use of the church of St. James in Andravida as the burial site for the Villehardouin family.24 Moreover, the Assizes of Romania, the region’s 17 A. Bon, “Pierres inscrites ou armorie´es de la More´e franque,” Delt.Crist. Arc. j JEt. 4 (1964): 90. For Isabelle’s coat of arms, see also her private seal in G. Schlumberger, F. Chalandon, and A. Blanchet, Sigillographie de l’Orient latin (Paris, 1943), 185, pl. , 5. 18 Several figures in the gatehouse, including St. Christopher carrying the Christ child, have not yet been published. 19 Schaefer, “Neue Untersuchungen,” fig. 9. 20 For the cockle shell, see C. Hohler, “The Badge of St. James,” in The Scallop: Studies of a Shell and Its Influences on Humankind, ed. I. H. Cox (London, 1957), 49–70. For examples of shells as badges, see Santiago de Compostela: 1000 jaar Europese Bedevaart (Ghent, 1985), 291–97. 21 On the representation of St. James in the West, see E. Maˆle, Religious Art in France: The Twelfth Century (Princeton, 1978), 293–99. 22 G. Ku¨hnel, Wall Painting in the Latin Kingdom of Jerusalem (Berlin, 1988), 40–43, pl. ; J. Folda, The Art of the Crusaders in the Holy Land, 1098–1187 (New York, 1995), 456, pl. 10.17c, d. 23 See D. Abrahamse, “Byzantine Views of the West in the Early Crusade Period: The Evidence of Hagiography,” in The Meeting of Two Worlds: Cultural Exchange between East and West during the Period of the Crusades, ed. V. Gross and C. Verza´r Bornstein (Kalamazoo, Mich., 1986), 193. 24 According to the Chronicle of the Morea, “He [Guillaume de Villehardouin] commanded and ordered that after he had died, and not before a full year had passed, his bones alone were to be placed in a coffin in the church of St. Jacob [James] of Morea in Andravida, in this church which he had built and which he had given to the Temple, in the tomb he had built and in which lay his father; his brother to lie to the right of him, he to

Patras

Lakedaemona (23) Trypi (12)

Andravida Nauplion

Mystra Mystra

Kalamata

Hagioi Anargyroi (13)

Geraki Passava Monemvasia Maina

Geraki (7, 17, 21, 25)

Vrontamas (10)

Kastania (18)

Platsa (33) Langada (27)

St. Andreas (32, 34)

Krokrees (9)

24 1415 22, 28, 31 3Tsopakas (6) Polemitas (26) Kepoula (2) 20 1, 511 Ano Boularioi (4)

Pakia (16)

19

Kato

Kastania (8)

Pharaklos (30)

Lachi (29)

The Byzantine Morea

Churches with portraits of equestrian saints, 13th–15th century. The numbers correspond to those in the list of churches in the appendix of published sources. See pages 281–85.

1 Nauplion gatehouse, Sts. Anthony and James (photo: after W. Schaefer, “Neue Untersuchungen über die Baugeschichte Nauplias im Mittelalter,” JDAI 76 [1961]: f ig. 12a)

2 Bethlehem, church of the Nativity, male pilgrim (photo: after J. Folda, The Art of the Crusaders in the Holy Land, 1098–1187 [New York, 1995], pl. 10.17c)

3 Nauplion gatehouse, St. George (photo: after Schaefer, “Neue Untersuchungen,” f ig. 10a)

4 Mount Sinai, monastery of St. Catherine, St. Sergios (reproduced courtesy of the Michigan-Princeton-Alexandria Expedition to Mount Sinai)

5 Mount Sinai, monastery of St. Catherine, Sts. Sergios and Bacchos (reproduced courtesy of the Michigan-Princeton-Alexandria Expedition to Mount Sinai)

6 Seal of Florent de Hainault (photo: after G. Schlumberger, F. Chalandon, and A. Blanchet, Sigillographie de l’Orient latin [Paris, 1943], pl. IX.2)

7 Kastoria, Hagioi Anargyroi, Sts. George and Demetrios

8 Ano Boularioi, Mani, Hagios Strategos, St. George

9 Geraki, St. John Chrysostom, St. George

10 Mani, St. Nicholas Polemitas, Sts. George and Kyriake (photo: after N. Drandakes, Buzantin¢w toixograf¤ew t∞w M°sa Mãnhw [Athens, 1995], 146)

11 Mani, St. Nicholas Polemitas, St. George, detail

12 Vrontamas, Palaiomonastero, frieze of equestrian saints

13 Vrontamas, Palaiomonastero, St. George

14 Lakedaemona, monastery of the Forty Martyrs, St. Demetrios (photo: after N. Drandakes, “TÚ Paliomonãsthro t«n ÑAg¤vn Sarãnta stØ Lakeda¤mona ka‹ tÚ éskhtariÒ tou,” Delt.Xrist.ÉArx.ÑEt. 16 [1992]: f ig. 19)

15 Geraki, St. John Chrysostom, Women at the Tomb, detail of soldiers

16 Geraki, church of St. George, St. George

Sharon E. J. Gerstel [ 267 ] law code, makes specific mention of James. According to the code, pilgrimage to Rome, Jerusalem, and Santiago de Compostela constituted three reasons for which a lord could not deny his vassal permission to leave the principality.25 The representation of St. James in the Nauplion gatehouse served to remind Western viewers of their devotion to this saint and the obligations of pilgrimage. The saint’s representation on the walls of the gatehouse, the passageway of the wayfarer, was singularly appropriate. The north wall of the gatehouse is decorated with a monumental image of an equestrian St. George, a figure whom the Crusaders considered patron and protector (Fig. 3).26 The saint’s fame derived from the widespread belief that he was one of three warrior saints who had rescued the Crusaders at the battle of Antioch on 28 June 1098. There were more local reasons, however, to favor the representation of St. George in the gatehouse. George’s military skill was not lost on the Crusaders of the Peloponnesos. According to the Chronicle of the Morea, the saint took the part of the Franks at the battle of Prinitza in 1263: “Some of those who took part in that battle saw and testified that they saw a knight mounted on a white charger, carrying a naked sword and always leading the way wherever the Franks were. And they saw and affirmed that it was St. George and that he guided the Franks and gave them courage to fight.”27 The passage echoes the description of the earlier battle of Antioch included in the Gesta Francorum, where witnesses identify the saints as “a host of men on white horses.”28 The location of George’s tomb at Lydda in the Holy Land has also been cited as a reason for the frequency of his portrayal in the Crusader East.29 The Frankish territories also boasted a close connection with the saint. The head of St. George was the principal relic of Livadia, a town located between Thebes and Delphi. When Gautier I, the son of Hugues de Brienne (whose coat of arms appears in the gatehouse), prepared for the battle of Kephissos in March 1311, he made out a will that left 100 hyperpera to the church of St. George of Livadia.30 be on the left, and his father in between. He instructed and endowed four chaplains, whom all the Romans call hiereis, to continue without cease unto eons of eons to chant and celebrate masses everlastingly for their souls; he ordered as a commandment and excommunicable offense, and it was put into writing, that never should they have interference from any man of the world.” Crusaders as Conquerors: The Chronicle of the Morea, trans. H. Lurier (New York, 1964), 290. To my knowledge, no architectural remains of this church have been found. For early reports concerning the location of the church, see J. A. Buchon, La Gre`ce continentale et la More´e (Paris, ` ejn Andrabi ´di,” Ne´ o" JEll. 17 (1923): 101–3; Bon, La More´e franque, 1843), 510; S. Lambros, “ Anaskafai j j 319–20. For the cover slab from the grave of Agnes, the third wife of Guillaume II de Villehardouin, originally located in this church, see A. Bon, “Dalle fune´raire d’une princesse de More´e (XIIIe sie`cle),” MonPiot 49 (1957): 129–39. 25 P. Topping, Feudal Institutions as Revealed in the Assizes of Romania: The Law Code of Frankish Greece (London, 1949), 63. I thank D. Jacoby for this reference. 26 For representations of the equestrian St. George within a Crusader context in the West, see P. Deschamps, “Combats de cavalerie et e´pisodes des Croisades dans les peintures murales du XIIe et du XIIIe sie`cle,” OCP 13 (1947): 454–74. 27 Crusaders as Conquerors, 211; J. Longnon, Livre de la conqueste de la prince´e de l’Amore´e (Paris, 1911), 338. For the location of Prinitza, see Bon, La More´e franque, 354–56. 28 Gesta Francorum et aliorum Hierosolimitanorum, trans. and ed. R. Hill (London, 1962), 69. 29 R. Cormack and S. Mihalarias, “A Crusader Painting of St. George: ‘Maniera greca’ or ‘lingua franca’?” Burlington Magazine 126 (1984): 133. On the shrine at Lydda, see Gesta Francorum, 87; M. Benvenisti, The Crusaders in the Holy Land ( Jerusalem, 1970), 169–70. 30 K. M. Setton, “Saint George’s Head,” Speculum 48 (1973): 4. Gautier’s will also stipulated donations to several other churches and monastic orders.

[ 268 ] Art and Identity in the Medieval Morea It is impossible to determine whether this donation and the saint’s representation in the gatehouse satisfied a pious impulse or arose from links between Hugues’s family and the saint. Nonetheless, the equestrian portrait of St. George, like that of St. James, reflects well-established Latin devotional practices. The portrait of St. George in the Nauplion gatehouse shares a number of traits with representations of military saints from the Crusader East. The white badge with a red cross on his shield is reminiscent of the banners and saddle decoration on two representations of St. Sergios from Mount Sinai that have been dated to the late thirteenth century (Figs. 4, 5).31 The coat of arms that marks the cantle of his saddle further demonstrates his Latin affiliation; moreover, his right leg is thrust forward in the position favored by Latin horsemen in the twelfth century.32 The personal seal of Florent of Hainault, whose coat of arms decorates the gatehouse, illustrates the characteristic pose of the Latin knight; Florent is represented on a galloping horse with his right leg thrust forward in its stirrup, a triangular shield in his left hand, and a drawn sword in his right (Fig. 6).33 Many scholars have commented on the popularity of equestrian saints in lands held by the Crusaders.34 This mounted warrior, inscribed in Latin and marked with emblems of his Western sponsor, might thus be grouped with other images of the period that have been linked to Crusader patronage. Yet the style of the representation indicates that the painter who created the Nauplion St. George was not foreign born, but a local artist who was trained in the region.35 Although the image shares certain characteristics with the Sinai panels, the portrait-type, pose of the body, and rendering of the military costume all find closer parallels in contemporaneous decoration of Orthodox churches in the Morea (see Figs. 8, 10). This image of St. George stands at the center of the problem of lasting Crusader influence on the Byzantine art of the Morea, for it raises—far more sharply than the story of Troy or representations of St. James of Compostela—the question of how images expressed cultural identity.

31 Demetrios Pallas, recognizing stylistic affinities to known Crusader works, compared the St. George at Nauplion to the Sinai icon of St. Sergios. See Pallas, “Eujrw´ ph kai´ Buza´ ntio,” 56–60. The icon, which has been identified as a Crusader work on the basis of both style and subject matter, has been associated with painters working in Cyprus, Sinai, and Syria. See G. and M. Soteriou, Eijko´ ne" th'" Monh'" Sina', vol. 1 (Athens, 1956), fig. 187; vol. 2 (Athens, 1958), 171; K. Weitzmann, “Icon Painting in the Crusader Kingdom,” DOP 20 (1966): 71–72, repr. in idem, Studies in the Arts at Sinai (Princeton, 1982), 345–46; D. Mouriki, “ThirteenthCentury Icon Painting in Cyprus,” The Griffon (Gennadius Library, American School of Classical Studies, Athens), n.s., 1–2 (Athens, 1985–86): 66–71, 76–77; L. Hunt, “A Woman’s Prayer to St. Sergios in Latin Syria: Interpreting a Thirteenth-Century Icon at Mount Sinai,” BMGS 15 (1991): 96–145. Closer to home, “Crusader banners” are carried by the equestrian saints Theodore Teron and George in the church of the Virgin in Attica. For this composition, most likely dated to the 14th century, see N. Coumbaraki-Panselinou, Saint-Pierre de Kalyvia-Kouvara et la chapelle de la Vierge de Me´renta (Thessaloniki, 1976), pl. 78. 32 B. S. Bachrach, “Caballus et Caballarius in Medieval Warfare,” in The Study of Chivalry: Resources and Approaches, ed. H. Chickering and T. Seiler (Kalamazoo, Mich., 1988), 194–96. 33 Schlumberger et al., Sigillographie, 185, pl. , 2. 34 See, e.g., A. Cutler, “Misapprehensions and Misgivings: Byzantine Art and the West in the Twelfth and Thirteenth Centuries,” Mediaevalia 7 (1984): 51. 35 A discomfort with the formation of Latin characters may be cited as further proof that the painter was Greek; the letters are uneven in size and shape and are disproportionately large for the figures.

Sharon E. J. Gerstel [ 269 ]

Monumental Decoration in the Byzantine Morea The day on which he [Manuel Komnenos] set out on his dangerous march, a certain man, an interpreter, and a Roman by race, whose surname was Mavropoulos, came to him and related that he dreamed he entered a church named after [Saint] Kyros, and as he was making a propitiatory offering he heard a voice coming from the icon of the Mother of God saying, “The emperor is now in the utmost danger,” and “Who will go forth in my name to assist him?” The voice of one unseen answered, “Let [Saint] George go.” “He is sluggish,” came the reply. “Let [Saint] Theodore set forth,” then suggested the voice, but he was also rejected, and finally came the painful response that no one could avert the impending evil. So much for these matters. Niketas Choniates36 From the Middle Byzantine period, warrior saints such as George and Theodore formed a vital component of the Orthodox church program. In the late twelfth-century Hagioi Anargyroi in Kastoria, for example, military saints stand along the walls of the naos at the level of the faithful (Fig. 7). Placed in this location, they served as popular devotional figures and as personal and communal guardians. The tradition of representing military saints standing in a frontal pose continued through the Late Byzantine period in areas with close artistic ties to Constantinople, for example, in Macedonia, which was recaptured from the Latins at an early date.37 The same is true for the Late Byzantine ecclesiastical decoration at Mistra, which demonstrates a number of stylistic and iconographic links to painting in the distant capital.38 The monumental portrayal of military saints on foot and in frontal pose mirrors their representation on ivory and steatite icons and on imperial coins and seals, particularly ones of the Komnenian period.39 Only rarely do seals present an image of a military saint on horseback; the overwhelming majority stand in frontal pose, with spear and shield in hand.40 The Byzantine manner of rendering 36

O City of Byzantium: Annals of Niketas Choniates, trans. H. Magoulias (Detroit, 1984), 107–8. With one exception, surviving Constantinopolitan churches of the late Byzantine period maintain frontal portraits of military saints. Only the late 13th-century program of St. Euphemia contains a votive image of George and Demetrios on horseback. See R. Naumann and H. Belting, Die Euphemia-Kirche am Hippodrom zu Istanbul und ihre Fresken (Berlin, 1966), 188, pl. 38a, b. Among the numerous churches of Byzantine Macedonia, St. Demetrios in Thessalonica contains a representation of the equestrian saints George and Demetrios on the east wall of the south aisle. The painting, adjacent to the entrance of St. Euthymios, must date to the 13th or the early 14th century. G. and M. Soteriou, JH basilikh` tou' Agi J ´ou Dhmhtri´ou Qessaloni´kh" (Athens, 1952), 1:219; 2:93a. An early image of an equestrian St. George is also found in the north aisle of Hagioi Anargyroi in Kastoria. This image, however, is part of a narrative scene. 38 Personal observation. 39 For military imagery and inscriptions on ivory icons, see N. Oikonomides, “The Concept of ‘Holy War’ and Two Tenth-Century Byzantine Ivories,” in Peace and War in Byzantium: Essays in Honor of George T. Dennis, S.J., ed. T. Miller and J. Nesbitt (Washington, D.C., 1995), 62–86. For steatite icons, see I. KalavrezouMaxeiner, Byzantine Icons in Steatite (Vienna, 1985), 63–65. Kalavrezou dates two icons of military saints on horseback to the 13th century. The remaining saints in her catalogue are represented in a frontal, standing position. 40 Only two seals in the Zacos and Veglery catalogue present military saints on horseback. The first, an unidentified saint, dates to the late 7th/early 8th century; the second, on a seal belonging to Sebastokrator Alexios Komnenos, represents St. George on the obverse and is dated to the late 12th century (G. Zacos and A. Veg37

[ 270 ] Art and Identity in the Medieval Morea military figures on seals forms a stark contrast with Western sigillographic preferences. From 1100 to 1250, equestrian seals representing mounted warriors were the overwhelming choice of feudal lords and simple knights (Fig. 6).41 The seals are the visual manifestations of a knightly ideal that was intimately associated with equestrian activity. Like those in regions closer to Constantinople, village churches of the southern Morea that were painted before the Fourth Crusade present portraits of military saints exclusively in a frontal, standing position following the established Byzantine format.42 But in the thirteenth and fourteenth centuries, military saints in village churches of the Morea mount their horses and begin to gallop across the interior walls—a phenomenon that has been noted in border regions of the empire at an earlier period and, contemporaneously, in areas under Crusader control such as Cyprus and the Holy Land.43 Although church decoration under Crusader rule in the Morea follows patterns established in the Middle Byzantine period, the change in the format and the enormously increased scale of the equestrian saint give the image disproportionate importance vis-a`-vis the remaining program. Superimposed layers of fresco decoration in Hagios Strategos at Ano Boularioi in the Mani help pinpoint the moment at which military saints mounted their horses in the Mani. In the northwest compartment of the naos, the figure of an equestrian St. George, painted on a layer dated 1274/75 by an inscription, covers a late twelfthcentury portrait of a standing military saint, most likely also St. George (Fig. 8).44 Supporting evidence for the time of the transition is found in St. Mamas in Karavas, dated 1232, which contains an image of an equestrian saint (Theodore?).45 The church was painted in the period of Frankish rule over the Mani, which ended with the transfer of the nearby castle of Maı¨na to Byzantium in 1262.46 The evidence is unequivocal: in Orthodox churches of the region along the borders of the Frankish Morea, the shift from standing to equestrian military saints took place during the period of Latin overlordship. Nikolaos Drandakes has estimated that more than fifty churches in the Mani were

lery, Byzantine Lead Seals, vol. 1:3 [Basel, 1972], no. 274). Two other seals with representations of the equestrian saints George and Demetrios were recently auctioned from the Zacos collection. See Spink Auction 127. Byzantine Seals from the Collection of George Zacos, Part I (London, Wednesday 7 Oct. 1998), nos. 79, 88. 41 B. Bedos Rezak, “Medieval Seals and the Structure of Chivalric Society,” in The Study of Chivalry (as in note 32 above), 313–72; eadem, “The Social Implications of the Art of Chivalry: The Sigillographic Evidence,” in The Medieval Court in Europe, ed. E. Haymes (Munich, 1986), 142–75. 42 See, e.g., the church of the Episcopi near Stavri, St. Peter near Gardenitsa, and Hagios Strategos in Ano Boularioi, all in the Mani (Drandakes, Buzantine` " toicografi´e", 151–212, 259–306, 392–458), and the Evangelistria church in Geraki (N. K. Moutsopoulos and G. Demetrokalles, Gera´ ki. OiJ ejkklhsi´e" tou' oijkismou' [Thessaloniki, 1981], 85–136). 43 Among other studies of churches in these regions, see J. Leroy, “De´couvertes de peintures chre´tiennes en Syrie,” AArchSyr 25 (1975): 95–113; J. Folda, “Crusader Frescoes at Crac des Chevaliers and Marqab Castle,” DOP 36 (1982): 194–95; D. Mouriki, “The Wall Paintings of the Church of the Panagia at Moutoullas, Cyprus,” in Byzanz und der Westen (as above, note 7), 192–95; Hunt, “A Woman’s Prayer to St. Sergios,” 104–10; E. Dodd, “The Monastery of Mar Musa al-Habashi, near Nebek, Syria,” Arte medievale 6 (1992): 84–91, 112–18. 44 Drandakes, Buzantine` " toicografi´e", 404. 45 C. Konstantinide, “ JO ”Agio" Ma´ ma" Karaba' Me´ sa Ma´ nh",” Lak.Sp. 10 (1990): 140–65. The equestrian saint is not illustrated. 46 For this castle, see Bon, La More´e franque, 502–7.

Sharon E. J. Gerstel [ 271 ] decorated in the second half of the thirteenth century following the return of the region to imperial control.47 The village churches of the Mani and other areas of the Byzantinecontrolled Morea form the basis of the following discussion. In village churches of the Byzantine Morea, the equestrian saint, whose horse easily occupies the space of three standing figures within the painted program, often dominates the lower register of the interior decoration. The newly invested importance of the equestrian saint is seen in the church of St. John Chrysostom in Geraki. In this small, single-aisled church of roughly 1300, the portrait of St. George on horseback is placed directly opposite the entrance to the building (Fig. 9).48 Although he is mounted to do battle with the serpent coiled below the hooves of his horse, the upper part of his torso and his face are turned to confront the viewer directly. The first impression of the church interior is thus formed by an encounter with the votive portrait of the holy rider. In other churches, repeated equestrian figures dominate the lower registers of the decorative program. St. Theodore (Trisakia) at Tsopaka, dated on the basis of style to the late thirteenth century, has a long nave articulated by four recessed arches paired on the north and south walls.49 At least two of the arches were decorated with equestrian saints who sit astride horses facing the sanctuary.50 In other churches, such as St. Eustratios in Pharaklos, equestrian saints are paired in recessed arches on opposite sides of the nave.51 From the mid-thirteenth through the fourteenth century, more than thirty churches in the southern Morea were decorated with one, and often more, military figures on horseback (see the Appendix). Two explanations can be offered for the introduction and proliferation of equestrian saints in village churches of the Byzantine Morea. First, the equestrian saint appeared in the decorative program as an indigenous response to regional danger. Second, the representation of military saints at this moment in Peloponnesian history demonstrates the appreciation of Frankish chivalric customs and reveals a certain degree of cultural emulation and symbiosis. I begin with the first of the two explanations. Although the painted evidence indicates that equestrian saints appear in monumental decoration in the southern Morea only from the thirteenth century, mounted warriors have a long history in Byzantine art.52 In their protective capacity, equestrian saints are occasionally painted on the exterior of Orthodox churches, as is the case at St. George, Kurbinovo (1191).53 Before the Fourth Crusade, mounted saints had already entered the interior decoration of churches located on the borders of the empire. Painted churches 47 This estimate includes both new church programs and additional painted layers. N. Drandakes, “Parathrh´ sei" sti´" toicografi´e" tou' 13ou aijw´ na pou´ sw´ zontai sth´ Ma´ nh,” 17th CEB, Major Papers (Washington, D.C., 1986), 684. 48 Moutsopoulos and Demetrokalles, Gera´ ki, 7, 40. 49 Drandakes, Buzantine` " toicografi´e", 49, 51, fig. 18. 50 Theodore is located on the south wall in the second recessed arch from the east. A second equestrian saint is located on the north wall in the first recessed arch from the west. 51 N. Drandakes et al., ““Ereuna sth` n jEpi´dauro Limhra´ ,” Prakt. Arc. j JEt. (1982): 436, pl. 244a. 52 For early Byzantine magical amulets decorated with mounted warriors, see C. Bonner, Studies in Magical Amulets Chiefly Graeco-Egyptian (London, 1950), 208–28, 302–8. The discussion of these early amulets is beyond the scope of this paper. Their use as apotropaia may foreshadow the protective function of later equestrian saints. 53 L. Hadermann-Misguich, Kurbinovo: Les fresques de Saint-Georges et la peinture byzantine du XIIe sie`cle (Brussels, 1975), 275–82, pls. 146–50.

[ 272 ] Art and Identity in the Medieval Morea in Cappadocia, the Byzantine border par excellence, feature a number of equestrian saints, as do churches in Georgia and Egypt.54 The sudden appearance of these saints in the southern Morea may reflect the new border status of this region in the thirteenth century.55 Frankish incursions between 1248 and 1262 threatened the Orthodox population; a foreboding inscription in St. John the Baptist in Megale Kastania may allude to military activity in the Mani in the mid-thirteenth century.56 The construction of churches named for sainted warriors and the inclusion of equestrian saints in churches close to the Latin-held Morea demonstrate a desire to protect rural communities through the invocation of powerful military guardians. Approximately twenty churches in the southern Morea contain Greek inscriptions that securely provide their date and dedication during the thirteenth and fourteenth centuries. Roughly half of these churches were dedicated to military saints or protective figures like George, Demetrios, Theodore, and the Archangel Michael.57 As with other categories of holy men and women, military saints were invoked by individuals for personal or familial protection.58 Evidence for this practice is found in church decoration. A supplicatory inscription adjacent to the equestrian St. Theodore Stratelates in the early fifteenth-century church of Aı¨-Sideros near Pyrgos in the Mani names the servant of God, Theodore.59 The unique pairing of St. Kyriake and the equestrian St. George in the fourteenth-century church

54 For images of military saints in Cappadocian churches, see M. Restle, Byzantine Wall Painting in Asia Minor (Recklinghausen, 1967), 2:28, 29, 30, 32, 246–47; 3:436, 510, 516. On the representation of military saints in the region, see also A. W. Epstein, “Rock-Cut Chapels in Go¨reme Valley, Cappadocia: The Yılanlı Group and the Column Churches,” CahArch 24 (1975): 115–26; G. P. Schiemenz, “Felskapellen im Go¨reme-Tal, Kappadokien: Die Yılanlı-Gruppe und Saklı kilise,” IstMitt 30 (1980): 291–319; C. Jolivet-Le´vy, “Hagiographie cappadocienne: A propos de quelques images nouvelles de Saint Hie´ron et de Saint Eustathe,” in Eujfro´ sunon. Afie j ´ rwma sto` n Mano´ lh Catzhda´ kh (Athens, 1991), 205–18. For icons representing equestrian saints that have been attributed to painters from Coptic Egypt and Georgia, see K. Weitzmann, The Monastery of Saint Catherine at Mount Sinai: The Icons, vol. 1 (Princeton, 1976), 71–73, 78–79; pls. , ; V. N. Lazarev, “Novyj pamjatnik stankovoj zˇivopisi XII v. i obraz Georgija-voina v vizantijskom i drevnerusskom iskusstve,” VizVrem 5 (1952): 205–6. 55 The border between the Byzantine and Frankish Morea was still fluctuating in the early 14th century. The southwest chamber of the Hodegetria church in Mistra is decorated with paintings of four chrysobulls that enumerate the privileges and holdings of the monastery. One of these texts, dated September 1332, provides the monastery with two towns (Zourtza and Mountra) that had been occupied by Latins. See G. Millet, “Inscriptions byzantines de Mistra,” BCH 23 (1899): 115–18; Zakythinos, Despotat grec, 1:82. 56 See P. Drosogianne, “Sco´ lia sti`" toicografi´e" th'" ejkklhsi´a" tou' Agi J ´ou jIwa´ nnou tou' Prodro´ mou sth` Mega´ lh Kasta´ nia th'" Ma´ nh"” (Ph.D. diss., University of Athens, 1982), 302–14, pl. ; Kalopissi-Verti, Dedicatory Inscriptions, 65–66. 57 These churches include St. George, Karinia, Mesa Mani (1285), St. George, Oitylon (1331/32), St. George, Longanikos (1374/75), St. Demetrios, Krokees (1286), Sts. Theodoroi, Kaphiona, Mesa Mani (1263– 71), Sts. Theodoroi, Mistra (before 1296), St. Michael, Polemitas, Mesa Mani (1278), St. Michael, Charouda, Mani (1371/72), Taxiarches, Goritsa Laina (mid-13th century). For the inscriptions, see Kalopissi-Verti, Dedicatory Inscriptions, 66–67, 71–75, 80–81, 106, 107; A. Philippidis-Braat, “Inscriptions du Pe´loponne`se,” TM 9 (1985): 314–17, 318–19, 328–30, 338–40. A number of churches dedicated to military saints could be added to this list solely on the basis of iconography. 58 S. E. J. Gerstel, “Painted Sources for Female Piety in Medieval Byzantium,” DOP 52 (1998): 89–111. 59 ⫹De´ hsi(") tou' dou´ lou tou' Q(eo)u' Qeodw´ rou tou' . . . (e“)t(ou") . . . . . (1423) (N. Drandakes, S. Kalopissi, and M. Panayotidi, ““Ereuna sth` Ma´ nh,” Prakt. Arc. j JEt. [1979]: 176). Supplicatory inscriptions are also found adjacent to the military saint in the church of St. George in Kambinari (undated) and in St. George, Geraki (Philippides-Braat, “Inscriptions,” 345–46).

Sharon E. J. Gerstel [ 273 ] of St. Nicholas Polemitas may also have satisfied the devotional requirements of the church’s patron and his family (Fig. 10).60 These examples demonstrate the personal relationship between the medieval supplicant and the protective saint, whose equestrian portrait newly augmented the church program from the thirteenth century. As votive images, the equestrian saints are placed within recessed arches in a number of churches (Figs. 8, 9). The majority, holding spears in their right hands and shields over their left shoulders, are engaged in conflict with a large serpent or dragon, a symbol of evil referred to in legends that recount the saints’ holy miracles.61 The representation of military saints on horseback within churches of the southern Morea thus conforms to a well-established tradition of securing the borders through the assistance of the holy cavalry. The fear generated by the Latin conquest of the region may have thus spurred the introduction of equestrian saints into thirteenth-century decorative programs.

Art and Cultural Identity The alternative argument for the popularity of equestrian saints rests on a reciprocal interchange between cultures. The Capetian court served as a model for the construction of a feudal society on Greek soil, and the knights of the Morea showed their allegiance to this court through the retention of chivalric behavior in their new land.62 With ancestral roots in the vicinity of courtly Troyes, Guillaume II Villehardouin tried his own hand at verse composition.63 Such courtly pretensions were admired by visitors to the region. In the early fourteenth century, following a century of settlement in the Peloponnesos, the Catalan chronicler Ramon Muntaner commented: “it is said that the most noble chivalry of the world is that of the Morea, and they speak as beautiful French as in Paris.”64 The Orthodox inhabitants of the Morea had ample opportunity to observe the Latins. The Greek version of the Chronicle of the Morea indicates that the Franks had a firsthand knowledge of southern Peloponnesian topography. During the winter of 1248–49, for example, when Guillaume de Villehardouin lodged in Lakedaemonia, “he went riding with his retinue and strolled among the villages in the neighborhood of Monemvasia, and to Helos and Passava and to the lands in that direction; with joy he went around and passed his time.”65 Yearly trade fairs, such as the one held at Vervaina in the north of the Morea, also provided ample opportunities for the Orthodox population of the region to come into contact with Latins, whose language and mannerisms reflected their country Drandakes, Buzantine` " toicografi´e", 147. H. Delehaye, Les le´gendes grecques des saints militaires (Paris, 1909); J. B. Aufhauser, Das Drachenwunder des heiligen Georgs in der griechischen und lateinischen u¨berlieferung (Leipzig, 1911). 62 Jacoby, “The Encounter of Two Societies,” 883–84. 63 J. Longnon, “Le prince de More´e chansonnier,” Romania 65 (1939): 95–100. 64 The Chronicle of Muntaner, trans. Lady Goodenough, vol. 2 (London, 1921), 627. 65 Crusaders as Conquerors, 158. Helos, near the mouth of the Eurotas River, is located on the Gulf of Lakonia. Passava, at the northeastern corner of the Mani, may have been in the hands of the Greeks by 1263. See Bon, La More´e franque, 508–9; M. Breuillot, “To` ka´ stro tou' Passaba' sto` n More´ a,” Lak.Sp. 11 (1992): 299–338. 60 61

[ 274 ] Art and Identity in the Medieval Morea of origin.66 The Orthodox population in the region seems to have been receptive to the Latin chivalric ethos, which emphasized courtoisie and romance. Neither romance nor certain aspects of chivalry were new to Byzantium, though they were not articulated as behavioral guidelines that defined a specific segment of society. In the twelfth century, Byzantium had seen the development of the Akritan epicromance, whose two-blooded protagonist patrolled the borders of the empire on horseback. The impact of the Akritan legend has been noted artistically in ceramic wares produced in Corinth, in the northern Morea, in the late twelfth and early thirteenth century, and in other locations.67 Fragments of pottery found at Mistra, in the Byzantine Morea, display incised images of soldiers and warriors, evidence of the spread of military imagery that derived from myth and everyday experience.68 Nonreligious imagery also entered church decoration in the region. The icon screen of St. Demetrios, Mistra, is decorated with the carved figure of a centaur, and a similar figure is roughly sketched in brown paint on the east side of the masonry templon screen in St. John the Baptist, Megachora.69 These representations reveal a sustained or renewed interest in vernacular literature and legend. While Orthodox residents of the Morea were wary of Latin military exploits, they may have reacted favorably to certain aspects of Frankish culture that resonated within their own background. An impact of Western culture in the Morea has been documented through linguistic evidence. Studies by J. Schmitt and H. and R. Kahane have demonstrated the frequency of Gallicisms in the Chronicle of the Morea.70 The evidence of cultural exchange, not surprisingly, is found in loanwords that concern feudal regulations and chivalric custom. In the Morea, medieval Greek added such Gallo-Romance terms as kourtesi´a, nta´ ma, and li´zio". The impact of a burgeoning cultural symbiosis can also be traced in the development of the Byzantine romance. Byzantine vernacular literature contemporary with many of the churches that contain portraits of equestrian saints features knights, castles, chivalry, and amor—concepts at home in the Morea. In the late fourteenthcentury Byzantine romance Livistros and Rodamni, for example, the protagonist is described as “a noble Latin from abroad, a capable and good-looking young man. . . . He was mounted on a horse, in his hand was a hawk and behind him a dog followed on a leash.”71 Elizabeth Jeffreys has suggested that several romances of this period reveal an awareness of Frankish customs and contain regional clues that hint at the Morea as the 66

W. Miller, The Latins in the Levant (London, 1908), 189. C. Morgan, “Several Vases from a Byzantine Dump at Corinth,” AJA 39 (1935): 76–78; C. Morgan, Corinth: The Byzantine Pottery, Corinth 11 (Cambridge, Mass., 1942), no. 1685, pl. ; A. Frantz, “Digenis Akritas: A Byzantine Epic and Its Illustrations,” Byzantion 15 (1940–41): 87–91; J. Notopoulos, “Akritan Iconography on Byzantine Pottery,” Hesp 33 (1964): 108–33; E. Ioannidake-Dostoglou, “Parasta´ sei" polemistw´ n kai kunhgw´ n sta buzantina´ aggei´a,” Arc.Delt. j 36 (1981): 127–38. 68 These sherds, in the museum housed at St. Demetrios, Mistra, are unpublished. 69 G. Millet, Monuments byzantins de Mistra (Paris, 1910), pls. 44.2, 45.3; Drandakes et al., ““Ereuna sth` n jEpi´dauro Limhra´ ,” 419. The sculpted centaur has been dated to the 12th century. 70 J. Schmitt, The Chronicle of the Morea (Groningen, 1967), 599–622; H. and R. Kahane, “The Western Impact on Byzantium: The Linguistic Evidence,” DOP 36 (1982): 127–54. 71 Three Medieval Greek Romances, trans. G. Betts (New York, 1995), 95. 67

Sharon E. J. Gerstel [ 275 ] place of their initial composition. These works include the War of Troy, the Greek rendering of Ste.-Maure’s text, in which the hero Ajax is called Ai“a" th'" Ma´ nh", a topographical reference not found in the French edition.72 It is only natural to assume that, like the linguistic, the artistic vocabulary could be expanded through contact with Latins in the region. An isolated cave church near Geraki contains an unusual image that betrays such expansion. The Old Monastery at Vrontamas presents a painted frieze of six military riders (Fig. 12).73 The frieze, located within the cave on the masonry facade of the church (which also serves as the east wall of the exonarthex), has been dated by N. Drandakes to 1201. The chronology of the frescoes, however, is problematical, and I believe that the paintings of the narthex interior and the facade should be dated later in the thirteenth century.74 The six saints on the frieze, which include Menas, Eustathios, George, and Niketas, are divided into two groups that face each other in mock tournament.75 The equestrian confrontation recalls the Western chivalric practice, known to have had its Byzantine admirers. Niketas Choniates describes how the Byzantine emperor Manuel Komnenos jousted with Latin knights in Antioch in 1159: “He carried his lance upright and wore a mantle fastened elegantly over his right shoulder which left the arm free on the side of the brooch. He was borne by a war-horse with a magnificent mane and trappings of gold which raised its neck and reared up on its hind legs as though eager to run a race, rivaling its rider in splendor.”76 The description of the emperor matches the representations of the Byzantine saints mounted on horses that prance above the arched opening to the church. The saints’ mantles flow behind their right shoulders, and their spears are held diagonally over their chests or extended in anticipation of the encounter. The representation of a frieze of mounted warriors, like the Nauplion St. George, seems to cross borders. Although jousting was popular among certain Komnenian emperors, it is doubtful that their activities in distant lands directly influenced the tournament scene

72 E. M. Jeffreys, “Place of Composition,” 320. See the discussion on sources in M. Papathomopoulos and E. Jeffreys, ÔO Po´ lemo" th'" Trwa´ do" (Athens, 1996), xli–lxvii. 73 N. Drandakes, “To Paliomona´ sthro tou Brontama´ ,” Arc.Delt. j 43 (1988): 184–85, pls. 66, 97. 74 Although Drandakes dates the narthex decoration to 1201, the date provided by an inscription on its west wall, I believe that the paintings should be dated later on the basis of their style. It may be possible that the inscription, painted on an added wall that supports the vaulted ceiling of the narthex, copies an earlier text painted on a primary wall associated with the first phase of the church’s construction and decoration. Such repainted inscriptions are known in the region. See, for example, inscriptions in the churches of Sts. Theodoroi, Kaphiona and St. Zachariah, Lagia, both in the Mani (N. Drandakes, “Les peintures murales des SaintsThe´odores a` Kaphiona [Magne du Pe´loponne`se],” CahArch 32 [1984]: 163–65; N. Drandakes et al., ““Ereuna sth` n Ma´ nh,” Prakt. Arc. j JEt. [1978]: 142 n. 2). 75 Drandakes, “To Paliomona´ sthro,” 184–85. Adjacent to St. Eustathios, Drandakes observed a small deer, a reference to the vision of the saint. For a second depiction of St. Eustathios in the Morea (St. John the Baptist, Megachora, Epidauros Limera, dated 1282), see Drandakes et al., ““Ereuna sth` n jEpi´dauro Limhra´ ,” 420. The vision of Eustathios is also represented in St. Kyriake in Keratea in Attica and the church of St. Thekla in Hagia Thekla, Euboia. For collected references to representations of this saint (primarily in Cappadocia), see A. Coumoussi, “Une repre´sentation rare de la vision de saint Eustache dans une e´glise grecque du XIIIe sie`cle,” CahArch 33 (1985): 51–60. 76 O City of Byzantium, 62.

[ 276 ] Art and Identity in the Medieval Morea at the isolated Vrontamas cave. Closer to the Morea, Frankish knights engaged in jousting.77 The French version of the Chronicle of the Morea describes the region around Lakedaemona (Sparta) as “une bonne marche pour gens et pour chevaux.”78 The Vrontamas frieze is unusual. In general, churches contain only one or two representations of equestrian saints, and they are represented in the lower register of the painted program.79 We may see the influence of the Crusaders in their pose, their military equipment, and the use of heraldic devices on shields. Let us begin with the pose. In paintings of the Byzantine Morea, the majority of equestrian saints confront serpents, a portrait-type that continues the long-standing representation of holy riders in art of the Christian East. In the Morea, however, the sainted horsemen generally extend their legs forward in their stirrups, according to the Latin fashion (Figs. 9, 10, 13, 15). A number of the equestrian saints are rendered without serpents, in a pose that has been termed “parade format.” This pose is common on personal seals in the West and in icon and monumental painting in the Crusader East (Figs. 4, 5).80 According to this format, the saint holds a lance-flag aloft or rests his spear on his shoulder with the tip upright; his horse marches with its foreleg raised, as if in military procession. The titular saints in two thirteenth-century churches in the Mani, St. Niketas in Karavas and St. George in Kastania, carry lance-flags, though decorated with a pattern of diamonds forming a cross and rays in place of the red cross on white ground.81 Painters in the Byzantine Morea commonly represented equestrian saints with their spears held aloft. In the thirteenthcentury narthex decoration of Vrontamas, portraits of Niketas and George, with spears uplifted, occupy the east wall (Fig. 13). The saints, who confront each other, were painted by two artists who employed different styles for the faces, although the horses and military costumes may have been painted by a single hand. Niketas is rendered in Byzantine fashion; his face is modeled, and skin tones are layered on an ochre base. The style of the second saint, George, is flatter and resembles the linear treatment of the equestrian St. Sergios on the two Crusader icons at Mount Sinai dated to the late thirteenth century (Figs. 4, 5).82 The faces possess a limner quality; they are flattened and highlighted by thick lines. The eyes of the saints are articulated in a similar fashion; eyebrows arch from the bridge of the nose to the extended crease of the outer corner of the lid. The hollow of the neck on all the figures is emphasized by a shaded “V.” Like the Sinai figures, the Vrontamas George rides on a white horse; his spear is held across 77 A contest that was held on the Isthmus of Corinth in 1305 was attended by more than one thousand knights. 78 Longnon, Livre de la conqueste, sec. 386. 79 Three equestrian saints are represented on the north wall of St. Demetrios in Krokees, dated 1286. See N. Drandakes, “ Apo j ´ ti´" toicografi´e" tou' Agi J ´ou Dhmhtri´ou Krokew'n (1286),” Delt.Crist. Arc. j JEt. 12 (1984): 203–7. 80 For examples of the lance-flag or gonfanon on Western seals, see Bedos-Rezak, “Medieval Seals,” 332. According to the author, such flags implied military leadership in the feudal army. For the cross banner in monumental painting in the East, see Hunt, “A Woman’s Prayer to St. Sergios,” fig. 5; Dodd, “The Monastery,” fig. 28. 81 N. Gkioles, “ JO nao` " tou' Ag. J Nikh´ ta sto` n Karaba' Me´ sa Ma´ nh",” Lak.Sp. 7 (1983): 172; N. Drandakes, S. Kalopissi, and M. Panayotidi, ““Ereuna sth` Messhniakh` Ma´ nh,” Prakt. Arc. j JEt. (1980): 198. 82 Drandakes, “To Paliomona´ sthro,” 176–77, pl. 86b. Drandakes compares the representation of St. George to the Sinai icon of Sergios and Bacchos. The portrait of St. Niketas is not illustrated in his article.

Sharon E. J. Gerstel [ 277 ] his chest in his right hand; a scabbard is attached to his saddle in place of a quiver; his weight is supported by triangular stirrups; and prick-spurs encircle his ankle. The saint wears the short mail hauberk and the long surcoat of the cavalryman.83 These common elements suggest that the Vrontamas painter was aware of certain features that have been identified with military imagery in the Crusader East. Once the parade format was adopted in the Morea, it was repeated in a number of smaller churches, some of which may have been decorated by a common workshop. Close to Vrontamas, in St. Nicholas near Geraki (1280–1300), a portrait of St. Demetrios decorates the recessed arch on the north wall.84 Although the image is damaged, the main lines of the representation are clear: Demetrios carries his spear in his right hand, and his roan horse rears on its hind legs in anticipation of imminent confrontation. Another example of a military saint in parade format is found in the nearby monastery of the Forty Martyrs near Sparta (Fig. 14).85 The wall paintings of the cave church are dated by one inscription to 1304/5; a second inscription provides the name of the painter (historiographos) as Constantine Manasses. In this church, St. Demetrios sits astride a white horse with his spear held across his chest. The saddle has a raised cantle and pommel, features found on the Sinai portraits of Sergios and the images of St. George at Nauplion and St. John Chrysostom, Geraki. In Demetrios’ portrait, the saint wears a lamellar hauberk and a long surcoat over ornately patterned leggings. The painter has added an unusual element to the composition—a scarf tied around the horse’s head and knotted below its neck.86 The depiction of equestrian saints in parade format is a departure from the serpent-spearing riders of earlier periods. The courtly display of intricate weaponry and equestrian prowess falls closer to representations that have been termed Crusader. Details of military costume and equipment also suggest that Byzantine artists in the Morea looked to Crusader fashion. When discussing the two Sinai icons, K. Weitzmann drew attention to the quiver “of decidedly Oriental form and ornament.”87 He associated the quiver with the Near East and posited a Persian model. In St. Nicholas, Polemitas, St. George wears an object that is similar in shape and occupies the same position as the Sinai quivers (Fig. 10).88 On closer inspection, the “quiver” is revealed to be a scabbard, although its shape is unusually blunt for the long sword of the equestrian saint (Fig. 11). 83 For the military costume, see D. C. Nicolle, Arms and Armour of the Crusading Era, 1050–1350 (New York, 1988); A. Hoffmeyer, “Military Equipment in the Byzantine Manuscript of Scylitzes in Biblioteca Nacional in Madrid,” Gladius 5 (1966): 8–160. 84 A. Gkiaoure (Bakourou), “ JO nao´ " tou' Agi J ´ou Nikola´ ou konta´ sto´ Gera´ ki,” Arc.Delt. j 32 (1977): 105–6, pl. 39. Bakourou relates the portrait to the Vrontamas saints. 85 N. Drandakes, “To´ Paliomona´ sthro tw'n Agi J ´wn Sara´ nta sth´ Lakedai´mona kai´ to´ ajskhtario´ tou,” Delt.Crist. Arc. j JEt. 16 (1992): 125, fig. 19. In his analysis of the portrait, Drandakes observes that the work is reminiscent of a portable icon. Yet another saint in parade format is found in the church of St. Niketas in Karavas in the Mani, dated 1270–90. See Gkioles, “ JO nao` " tou' Ag. J Nikh´ ta,” 172–73, pl. 3a. 86 For a similar detail, a red scarf on a white horse, see the church of the Virgin at Rustika, Crete, dated 1381–82 (M. Bissinger, Kreta: Byzantinische Wandmalerei [Munich, 1995], fig. 150). In this church, it is George’s horse who wears the ornamental scarf. In the church of St. Paraskeve, in St. Andreas (Epidauros Limera), dated to the second quarter of the 15th century, Demetrios’ horse has a scarf around the neck. See N. Drandakes, S. Kalopissi, and M. Panayotidi, ““Ereuna sth` n jEpi´dauro Limhra´ ,” Prakt. Arc. j JEt. (1983): pl. 170b. 87 Weitzmann, “Icon Painting,” 71. 88 Another church containing an equestrian saint with a quiver will be published by A. Bakourou.

[ 278 ] Art and Identity in the Medieval Morea George does carry a case that contains both bow and arrows, but it is an added detail on the left side of the saddle. It appears that the artist, in copying a work similar to the Sinai icons, must have misunderstood the shape and function of the quiver. Like the rendering of saints in parade format, the inclusion of a quiver suggests the infiltration of artistic details from beyond the Morea. The last detail that remains for discussion is the use of heraldic symbols to express the political and religious allegiance of the saint. Such markings must have become increasingly necessary as Latins and Greeks pressed claims on the same corps of saints. At Nauplion, the portrait of St. George was painted by a local, Orthodox artist, who identified the saint as a Latin by painting a specific badge on his shield, a red cross on a white ground (Fig. 3). A number of Orthodox saints in the Morea carry shields marked by insignia, perhaps in imitation of the Latin custom. In the West, the knights’ seals regularly incorporated such heraldic devices to establish identity.89 As Florent of Hainault’s shield demonstrates, the Latin rulers of the Morea followed this practice (Fig. 6). The typical badge of the Orthodox saint in this region is formed of a crescent moon and star. The motif appears in several churches in Geraki and in two churches in nearby Epidauros Limera.90 The earliest church in the Morea to include a saint holding a shield marked by the crescent and star may be St. John Chrysostom, which has been dated on the basis of style to ca. 1300, some forty years following the return of Geraki to Byzantium (Fig. 9). The north wall of the church presents a unique juxtaposition of two shields in superimposed scenes that are opposite the entrance. In the vault above the portrait of St. George, the scene of the Women at the Tomb forms part of the narrative cycle.91 In the lower right corner of the scene is a group of sleeping soldiers (Fig. 15). The shield of one of the soldiers is decorated with a small castle, a well-known symbol in the region. The reverse sides of contemporaneous Frankish coins minted in the Morea feature the castle tournois, a schematic rendering of a Latin castle consisting of a central turret flanked by two smaller towers.92 Through the use of this motif, the Roman soldiers are associated with Latins; the coat of arms that marks the shield is identified with the Frankish denier tournois, the local replacement of the debased Byzantine currency. The shield of the equestrian St. George is also marked with a badge, though comprised of a crescent and star. Two sets of scholars have attempted to trace the origins of this motif on the shields of Orthodox military saints, and a convincing argument has been made for its association 89

Bedos Rezak, “Medieval Seals,” 340–48. In Geraki: St. John Chrysostom (ca. 1300); St. George (14th century); St. Athanasios (14th century); Taxiarchs (15th century?). The later churches of Geraki are being redated by J. Papageorgiou. For images of saints in Geraki whose shields are marked by the crescent and star, see N. Moutsopoulos and G. Demetrokalles, Le croissant grec (Athens, 1988), figs. 5–8, 24. For the chapel of the Taxiarchs, traditionally dated to the 13th century, see R. Traquair, “Laconia: The Fortresses,” BSA 12 (1905–6): 266; M. Panayotidi, “Les e´glises de Ge´raki et de Monemvasie,” CorsiRav (1975): 347; M. Soteriou, “ JH prw´ imo" palaiolo´ geio" ajnage´ nnhsi" eij" ta` " cw´ ra" kai` ta` " nh´ sou" th'" JElla´ do" kata` to` n 13on aijw´ na,” Delt.Crist. Arc. j JEt. 4 (1964–65): 264–65. In Epidauros Limera: St. Eustratios, Pharaklos (ca. 1400); St. George Babylas, Lachi (mid-14th century). See Drandakes et al., ““Ereuna sth` n jEpi´dauro Limhra´ ,” 436, 447. 91 Moutsopoulos and Demetrokalles, Gera´ ki, 12. Based on the symbols decorating their helmets and their facial features, Moutsopoulos has identified the warriors as Mongols. See N. K. Moutsopoulos, “Siniko´ ideo´ gramma se toicografi´a tou Geraki´ou,” Byzantiaka 18 (1998): 15–31. 92 D. M. Metcalf, Coinage of the Crusades and the Latin East (London, 1983), 70–77, pls. 30–33. 90

Sharon E. J. Gerstel [ 279 ] with coins minted in Constantinople, especially an anonymous bronze follis of Alexios I Komnenos (1081–1118).93 This coin, decorated with a cross on crescent, has been found in large numbers in excavations at Corinth and Athens.94 The use of the motif for the painted shield of military saints in churches of the Morea may evoke the memory of the last powerful Byzantine dynasty to reign before Villehardouin came ashore at Methone. The Geraki painter juxtaposed motifs borrowed from Byzantine and Frankish coinage in order to identify Orthodox and Latin warriors. In a period in which numismatic imagery carried powerful associations, it is significant that coin motifs were used as signifiers of political allegiance. In the church of St. George in the kastro of Geraki, painted after St. John Chrysostom, the painter also combined the mark of the Latins with that of the Greeks, but here on the shield of a single figure, the standing St. George (Fig. 16). At the top of the shield is the crescent and star, identical to that found in the church of St. John Chrysostom. At the bottom of the shield is the checkered coat of arms of the Frankish lords of Geraki, a sign that was carved above the entrance to the church and also decorates the apex of the stone proskynetarion in the north aisle of the church. What did it mean for the artist to combine these two symbols on a single shield? Andreas and Judith Stylianou suggest that the combination reflects the “forced and uneasy coexistence of the new conquerors and the conquered.”95 But the composition was painted decades after Geraki had been ceded to Byzantium. The combination of Orthodox and Latin emblems on the shield of St. George may simply acknowledge the absorption of the church’s Latin identity into the newly painted Orthodox program. The incorporation of coats of arms and other insignia into Orthodox painting reveals an awareness of heraldic devices, and may signal their acceptance and adoption by Byzantine artists and patrons in emulation of their Latin neighbors. The insignia established an identity for the equestrian saints and marked political boundaries. Related in style to the representations that have been discussed in the Morea, the equestrian St. George in the Nauplion gatehouse was painted for a Latin patron. The hybrid image demonstrates that both Latins and Greeks could employ the same artistic vocabulary to assert their ethnicity and to proclaim their religious and cultural heritage. The equestrian saint, whether as a single portrait or as part of a frieze of galloping horsemen, seems to have been an image that was capable of crossing cultural barriers. Only a mark on the shield or a costume detail might differentiate St. George of the Latins from St. George of the Orthodox. Monumental painting in the Morea, as in all lands held by the Crusaders, engages in several polemical dialogues to which we should be attentive. In the practical representa93 A. and J. Stylianou, “A Cross inside a Crescent on the Shield of St. George, Wall-Painting in the Church of Panagia Phorbiotissa Asinou, Cyprus,” Kupr.Sp. 26 (1982): 133–40; Moutsopoulos and Demetrokalles, Le croissant grec. 94 The coin is typed as Class J of the anonymous folles of the period. See A. R. Bellinger and P. Grierson, Catalogue of the Byzantine Coins in the Dumbarton Oaks Collection, vol. 3.2 (Washington, D.C., 1973), pl. / J.3. For the Athenian Agora, see M. Thompson, The Athenian Agora, vol. 2 (Princeton, 1954), 111, 114–15. For Corinth, most recently, see C. K. Williams, II and O. Zervos, “Frankish Corinth: 1995,” Hesp 65 (1996): 48. In the 1995 season alone, three Class J folles of Alexios I were recovered (Corinth inv. nos. 95–12, 14, 41). For the typology of this coin, see A. R. Bellinger, The Anonymous Byzantine Bronze Coinage (New York, 1928). 95 Stylianou, “A Cross inside a Crescent,” 138.

[ 280 ] Art and Identity in the Medieval Morea tion of the holiest figures of Orthodoxy and narrative scenes painted as guides to moral behavior and ecclesiastical rite, the paintings depend on the ideological underpinnings of the Byzantine Empire. The representation of equestrian saints was linked to a long tradition of depicting mounted warriors for apotropaic and protective purposes. The recognition of the saints linked the faithful to a heritage of church and empire. But a new manner of representation gave coherence to a sense of national identity that had been challenged by the Latin occupation and may even betray a grudging admiration for Latin overlords that is often absent in historical sources. The depiction of equestrian saints affirmed boundaries based on political and religious grounds, but also signaled that the Byzantines could look beyond the boundaries. The rhetorical appeal of these paintings to the medieval viewer was thus complex: they served to unite the faithful and, paradoxically, to link them to the artistic and cultural production of their foreign overlords. Through a common mode of representation, Latins and Greeks, whether patrons or painters, were able to exchange artistic styles and compositional formulae. Whether these styles and formulae were borrowed directly from Western models, such as seals and coins that circulated in the region, or whether they were filtered through Crusader art of the East, such as icons, remains an open question. Artistic symbiosis, best viewed in the image of the equestrian saint, places the Morea in the midst of a number of Mediterranean locations where indigenous populations were confronted by Crusader overlords and where hybrid art forms arose from the interaction of two, and perhaps more, cultures. University of Maryland

Appendix Equestrian Saints in Byzantine Churches of the Morea according to Published Sources See the map in the plate section for locations. 1

St. Mamas, Karavas Kounou, Mani, 1232 North wall: St. Theodore? (N. Drandakes, N. Gkioles, and C. Konstantinide, ““Ereuna sth` Lakwnikh` Ma´ nh,” Prakt. jArc. E J t. [1981]: 255; C. Konstantinide, “ O J ”Agio" Ma´ ma" sto` n Karaba' Kou´ nou Me´ sa Ma´ nh",” Lak.Sp. 10 [1990]: 153)

2

Sts. Kosmas, Damianos, and Theodota, Kepoula, Mani, 1265 North wall: St. Niketas (N. Drandakes, Buzantine` " toicografi´e" th'" Me´ sa Ma´ nh" [Athens, 1995], 336, fig. 35)

3

Sts. Theodoroi, Kaphiona, Mani, 1263/64 North wall: St. Theodore Stratelates (N. Drandakes, “Les peintures murales des Saints-The´odores a` Kaphiona [Magne du Pe´loponne`se],” CahArch 32 [1984]: 170–71; Drandakes, Buzantine` " toicografi´e", 92, figs. 21, 25)

4

Hagios Strategos, Ano Boularioi, Mani, 1274/75 (second layer) Northwest compartment: St. George Narthex, north wall: St. George (Drandakes, Buzantine` " toicografi´e", 458, 466, figs. 19, 78)

5

St. Niketas, Karavas Kounou, Mani, 1270–90 North wall: St. Niketas North wall: St. Demetrios (Drandakes, Gkioles, and Konstantinide, ““Ereuna sth` Lakwnikh` Ma´ nh,” 258; N. Gkioles, “ O J nao` " tou' JAg. Nikh´ ta sto` n Karaba' Me´ sa Ma´ nh",” Lak.Sp. 7 [1983]: 172–73, pl. 3a)

[ 282 ] Art and Identity in the Medieval Morea 6

St. Theodore, Tsopaka, Mani, 1275–1300 South wall: St. Theodore North wall: Unidentified saint (George?) (N. Drandakes, “ O J ”Agio" Qeo´ dwro" sto` n Tso´ paka th'" Ma´ nh",” Peloponnhsiaka´ 16 [1985–86]: 252–53, fig. 15; Drandakes, Buzantine` " toicografi´e", 49–51, fig. 18)

7

St. Nicholas near Geraki, Lakonia, 1280–1300 North wall: St. Demetrios (A. Gkiaoure [Bakourou], “ O J nao´ " tou' JAgi´ou Nikola´ ou konta` sto` Gera´ ki,” jArc.Delt. 32 [1977]: 105–6, pl. 39)

8

St. John the Baptist, Kato Kastania, Epidauros Limera, 1282 West wall: St. Eustathios (N. Drandakes et al., ““Ereuna sth` n E j pi´dauro Limhra` ,” Prakt. jArc. E J t. [1982]: 420)

9

St. Demetrios Krokees, 1286 North wall (east to west): Sts. Demetrios, George, Theodore (N. Drandakes, “ jApo´ ti´" toicografi´e" tou' JAgi´ou Dhmhtri´ou Krokew'n (1286),” Delt.Crist. jArc. E J t. 12 [1984]: 203–7)

10

Palaiomonastero, Vrontamas, Lakonia, thirteenth century Narthex exterior: Frieze of six equestrian saints (left to right: Sts. Theodore [?], Theodore [?], Niketas, George, Eustathios Plakidas, Menas) Narthex, east wall: St. George, St. Niketas (N. Drandakes, “To Paliomona´ sthro tou Brontama´ ,” jArc.Delt. 43 [1988]: 176–77, 184–85, pls. 66, 86b, 97)

11

St. George, Keria, Mani, thirteenth century St. George (N. Drandakes, ““Ereunai eij" th` n Ma´ nhn,” Prakt. jArc. E J t. [1974]: 128)

12

Hermitage of St. Nikon, Trypi, Lakonia, second half of thirteenth century South wall: St. Demetrios (K. Diamanti, “To` ajskhtario` tou' JAgi´ou Ni´kwna sth` n Tru´ ph,” Lak.Sp. 9 [1988]: 351–52, fig. 3)

13

St. John the Baptist, Zoupena (Hagioi Anargyroi), Lakonia, late thirteenth century St. George (?) on white horse (N. Drandakes, “ O J sphlaiw´ dh" nao` " tou' JAi-Gianna´ kh sth` Zou´ pena,” Delt.Crist. jArc. E J t. 13 [1985–86]: 89, fig. 21)

Sharon E. J. Gerstel [ 283 ] 14

Holy Asomatoi, North Church, Phlomochori, Mani, late thirteenth century North wall: Two confronted equestrian saints (N. Drandakes, E. Dore, S. Kalopissi, and M. Panayotidi, ““Ereuna sth` Ma´ nh,” Prakt. jArc. E J t. [1978]: 145)

15

St. Nicholas in Skaltsotianika, Mani, late thirteenth century North wall: St. George “tropaiªoºfo´ ro"” South wall: St. Niketas (Drandakes, Dore, Kalopissi, and Panayotidi, ““Ereuna sth` Ma´ nh,” 149)

16

St. Nicholas, Pakia, Lakonia, late thirteenth century North wall: St. George West wall: Two equestrian saints (S. Koukiares, “Du´ o buzantinoi` naoi` sta` Pa´ kia Lakwni´a",” Lak.Sp. 10 [1990]: 181, 194)

17

St. Nicholas, Geraki, Lakonia, late thirteenth century North wall: St. Demetrios, St. George (N. K. Moutsopoulos and G. Demetrokalles, Gera´ ki. OiJ ejkklhsi´e" tou' oijkismou' [Thessaloniki, 1981], 56)

18

St. George, Kastania, Mani, late thirteenth century North wall: St. George (N. Drandakes, S. Kalopissi, and M. Panayotidi, ““Ereuna sth` Messhniakh` Ma´ nh,” Prakt. jArc. E J t. [1980]: 198)

19

St. Marina, Agios Phokas (Voutama), Epidauros Limera, ca. 1300 North wall: St. George (Drandakes et al., ““Ereuna sth` n E j pi´dauro Limhra` ,” 405)

20

St. Nicholas near Kitta, Mani, ca. 1300 South wall: Unidentified saint (N. Drandakes, S. Kalopissi, and M. Panayotidi, ““Ereuna sth` Ma´ nh,” Prakt. jArc. E J t. [1979]: 186)

21

St. John Chrysostom, Geraki, Lakonia, ca. 1300 North wall: St. George