The Byzantine background to the First Crusade, by Paul Magdalino

The Byzantine Background to the First Crusade

By Paul Magadalino

Published by the Canadian Institute of Balkan Studies (1996)

Just over nine hundred years ago, Pope Urban II closed a provincial church council at Clermont Ferrand with a rousing call to arms that launched the First Crusade. Every day from November 1995 to August 1999 will have marked the nine-hundredth anniversary of one of the extraordinary series of events that led huge companies of men from France and Germany to Constantinople, then on to Nicaea, Edessa, Antioch and Jerusalem, suffering great losses from dehydration, starvation, disease and ambush, yet managing to capture three formidably defended cities and to defeat three massive counter-attacks by the local Muslim powers who seemed to have all the advantages. However we look at the First Crusade - whether we look at the causes of a mass movement that defies rational explanation; whether we look at the gripping story of incredible success against seemingly impossible odds; or whether we focus on the consequences of an enterprise that was surely the most decisive moment in western civilization's long rise towards global hegemony - however we look at it, whatever we make of it, and whether we like it or not, the crusade is important to all of us who deal with the Middle Ages, the Mediterranean and the Near East, and the arrival of this symbolic anniversary invites us to reflect on its relevance for us.

Fortunately for those of us who are not historians of the crusades, there is no lack of up-to-date literature to help us concentrate our minds. Indeed, the last thirty-five years have seen a remarkable flowering of crusade scholarship in North America, in Israel, in Germany, and above all in Britain, where a generation of productive and dedicated scholars, mainly under the leadership of Jonathan Riley-Smith, is rewriting the whole history of the crusading movement, in practice and in theory, from beginning. to end. They have already published significant contributions to the re-interpretation of the First Crusade - I mention Riley-Smith's own book on the subject, published in 1986,1 John France's military history of the crusade,2 and Marcus Bull's study of lay piety in south western France in the eleventh century 3 - and we can no doubt look forward to the appearance of more volumes occasioned by the ninth centenary celebrations. The day is hopefully not far off when we will know all that can be known, or at least will have heard all that can usefully be said, about what was in Pope Urban's mind when he preached at Clermont, and what was in the minds of the laymen who answered, or did not answer, his call to arms. What is certain is that the labours of so many energetic milites Christi make it unnecessary, and indeed impertinent, for a Byzantinist to pronounce on such matters. We have come a long way since the 1950s, when a Byzantinist, Sir Steven Runciman, could write an authoritative, best-selling history of the crusades.4

And yet, in moving on from Runciman, we have lost as well as gained. We have lost not only the combination of scholarly erudition and narrative ease which still make Runciman an unbeatable first introduction to the subject; we have also lost the sense of Byzantium as something integral to the crusading movement. ~ Whereas Runciman stressed the role of the Byzantine emperor Alexios I (1081-1118) in initiating the crusade, directing the crusaders and helping them on their way, the two recent monographs by Jonathan Riley-Smith and John France have cast him as a much more marginal player.5 Riley-Smith suggests that relations with Byzantium were not uppermost in Urban II's calculations, and France, after dismissing the evidence of the main Byzantine source, the biography of Alexios I by his daughter Anna Comnena, argues that Alexios did little to help the crusade and was indifferent to its fate once he had used it to recapture Nicaea. Neither of these judgements is highly momentous in itself, but taken together and in the context of the general drift of modern crusade scholarship, they can be regarded as symptomatic of a tendency to write Byzantium out of the script, or, to choose a metaphor appropriate to our small academic world of conferences and symposia, not to invite Byzantium to the party. Byzantium is barely mentioned in the Oxford History of the Crusades, edited by Jonathan Riley-Smith and published in 1995. Byzantium thus comes to be seen as an unwilling, passive and even obstructive channel that the crusade had to pass through in order to get from Western Europe to the Holy Land: the place that happened to be on the way, or in the way. To use another metaphor, Byzantium was the temporary conductor of an electrical charge which the empire played no part in creating and had no interest in retaining. The implications of the fact that it received the current because its own battery needed recharging are not even considered.

Why has Byzantium thus been written out of the crusading script? Partly, perhaps, as a result of the increasing specialisation of historical scholarship, but more fundamentally, I think, as the result of a tendency on the part of both Byzantinists and crusading historians to identify with the mind-set of their respective sources, which prevents them from joining forces or exploring common ground. I think it is fair to say that most Byzantinists instinctively share the perception of Anna Comnena that the crusade came upon the empire as a nasty surprise.6 Some of us Ralph-Johannes Lilie,7 Jonathan Shepard 8 and myself,9 have suggested, in different ways, that the crusade and its consequences had a positive place in the Byzantine imperial agenda, but it is not yet clear whether our message has got through. Certainly, another scholar, Athena Kolia-Dermitzaki, has encountered powerful opposition to her thesis that the Byzantines had their own variety of holy war.10 As for the crusading historians, what distinguishes the school of Jonathan Riley-Smith and his cohorts is their unfailingly upbeat assessment of the whole crusading phenomenon, even at its most perverted. It is sufficient to note that Riley-Smith gave an inaugural lecture entitled "Crusading as an act of love",11 and, in a review of Lilie's book Byzantium and the Crusader States, complained of "that reverential frame of mind which is typical of so many byzantinists when they deal with the Greeks".12 This is not the remark of a completely neutral observer. One is reminded of the historian of the Second Crusade, Odo of Deuil, and his disapproval of the excessive reverence which the Greeks showed to their emperor and their empire.13

The chasm between Byzantine history and crusading history is unfortunate, because it is not as if they have nothing more to say to, or learn from, each other. This is very clear from a recent work which, encouragingly, goes some way towards closing the gap: a study of the defence of the Latin East in the twelfth century, which gives due credit to the part that Byzantium played in the survival of the crusader states.14 The investigation of the First Crusade can also benefit from a more integrated approach. There seems to be general agreement that Urban II would not have preached the crusade if Alexios I had not asked for help against the Turks. It is also fairly evident that when the second, and main, wave of the crusade left Constantinople for Asia Minor in the spring of 1097, it was an army under Byzantine imperial command. The nature of the imperial appeal, the pope's attitude on receiving it, and the exact tenor of the negotiations between Alexios and the crusading leaders are crucial to any understanding of the event, and since they are still open to debate, they deserve to figure on the agenda of any centenary discussions. Crusading historians have long been concerned to distinguish the self-conception with which the crusaders set off from that which they acquired on the long march to Jerusalem, and from that which was projected on them retrospectively by chroniclers. If history is the study of origins as well as the study of outcomes, it is about initial expectations as well as the events which supersede them. And if expectations are the stuff of history as much as end results, it is important to consider relations between crusaders and Byzantines before they started to go wrong. This means, in particular, that it is important to look both critically and imaginatively at the evidence of the main Byzantine source, the Alexiad of Anna Comnena. This evidence is suspect in more ways than one. The author makes no secret of her bias in favour of her father and against the Latins. She was only a young girl at the time of the crusade, and she did not write about it until at least forty years later. Her perception of her father's reign may also have been coloured by the frustration of her own ambition to succeed him as empress, or as the wife of his chosen successor. She wrote with the knowledge that the crusade had gone badly for Alexios, and that his son, John II, and grandson, Manuel I, were still having to cope with the consequences, in the form of the crusader states, including, most seriously, an independent Norman principality of Antioch. If, as is quite likely, she wrote her account of the First Crusade in or after 1146, she would have known that the unfinished business of the First Crusade had resulted in the calling of a Second Crusade which posed an equal or greater threat to the empire's security.

Anna Comnena's hindsight has been noted but its full significance has yet to be investigated.15 It may mean, as John France has argued, that she exaggerated Alexios' fatherly concern for the crusading army.16 However, the implications are broader than this. They include the distinct possibility that she, or her source, deliberately suppressed the embarrassing fact that the crusade did not come as a total surprise to Alexios, but was a response to a Byzantine appeal for help.17 There is also the possibility that Anna misrepresented the substance of her father's negotiations with his old enemy the Norman Bohemond, when the latter joined the other crusaders at Constantinople. Bohemond was the crusading leader who spoiled the crusade for Alexios by keeping Antioch for himself, in violation of his sworn agreement to hand over all former imperial territory. We may wonder, therefore, whether Alexios was as cautious of Bohemond's good behaviour as Anna says he was, and whether he did, in fact, politely decline Bohemond's request to be appointed domestic of the East - that is, imperial commander of the crusade. Is it not just as likely that Alexios offered Bohemond the position, and that Bohemond refused it because it would hamper his freedom of action?

But this is speculation, and speculation will not get us very far unless it involves a thorough analysis of Anna's retrospection, which would take us away from the First Crusade. It seems to me more profitable, in re-assessing the Byzantine contribution to the crusade, to take a fresh look at the situation and the sources prior to the crusade. If I may return to my electrical metaphor, a major preoccupation of western medievalists has long been to explain the impulse that set tens of thousands of westerners on the road to Jerusalem. Even if Byzantium did no more than conduct the charge from west to east, its readiness to do so cannot be taken for granted. Why did the Byzantine Empire happen to be a suitable and efficient conductor of crusading energy? In fact, did it merely serve as a conductor of an external force, or did its own structural evolution in the late eleventh century play an active part in attracting, directing and concentrating the expansionist energies of the West?

There is an obvious answer to this question, and it is summed up in i one word: Mantzikert. The defeat of the emperor Romanos IV Diogenes by the Turks at this fateful battle in 1071 led to the collapse of the empire's eastern frontier and the complete disarray of its armed forces.18 The great Christian power of the Near East thus became transformed into a power vacuum which was bound to attract the expansionist energies of its warlike neighbours, and to bring the increasingly militant Christianity, or Christian militarism, of the Latin West into direct confrontation with a newly aggressive Islam. However, this observation offers a framework for an explanation rather than an explanation capable of satisfying any but the most superficial curiosity. Just as "expansionism" is a blanket term which neither explains nor defines the complex of forces taking shape in the West, so "collapse" and "crisis" do not precisely represent what was happening in Byzantium. The important point is that the empire was not a helpless and passive spectator to the confrontation between Latin Christendom and Islam; the confrontation occurred because of, not in spite of, Byzantium. It is the purpose of this lecture to suggest that Mantzikert and its aftermath accelerated developments that were already taking shape, and that even had the empire not undergone a military crisis in the 1070s and 1080s, it would still have brqught the Latin West into contact as well as conflict with the Muslim East, perhaps even creating the conditions for some sort of Latin colonisation. The information on which this suggestion is based is not unknown to scholarship, but it has not been viewed as an ensemble in the perspective from which it is presented here.

The first crusaders were not totally unfamiliar with Byzantium and the eastern Mediterranean. Two of their leading figures, Peter the Hermit, the leader of the so-called People's Crusade, and Count Robert of Flanders, had made their pilgrimage to the Holy Land via Constantinople, and Count Robert on his return had despatched five hundred warriors to serve Alexios I.19 However, the group with the longest and closest experience of the empire were undoubtedly the Normans. Bohemond, the leader of the Normans from southern Italy, had invaded the empire in 1082-3 in the service of his father Robert Guiscard; some of Bohemond's companions on the crusade may have accompanied him on that campaign. They came from an area which had recently been within the Byzantine orbit, and where there were still sizeable Greek communities, among whom Byzantine religious and administrative traditions were still very much alive.20 Moreover, many Normans in Bohemond's army, as well as many in the army led by Count Robert of Normandy, would have encountered Byzantium, at least at second hand, through the experiences of those Normans who had served and still served as mercenaries in the imperial armed forces.21 Three aspects of this service are of interest to us in the light of the crusade.22 Firstly, the "Franks" constituted a substantial and self-contained element in the Byzantine army, numbering some 2000 in 1057 23 and rising to at least 3000 by 1071.24 Secondly, by the time of the crusade they had served the empire for almost fifty years, long enough for some of them to feel thoroughly at home,' and for much of this period, from c.1050 to 1076, one important Frankish unit was settled in a particular area of Asia Minor, the theme of the Armeniakoi. This unit included the three Frankish commanders who figure most prominently in the sources: Herve, surnamed Phrangopoulos by the Byzantines, Robert Crispin, and Roussel of Bailleul. In 1057 Herve had an estate in the province, which to judge from his social position, and his social aspirations, must have been quite considerable.26 Crispin and Roussel both held castles in the Armeniakon in the early 1070s.27 All three leaders therefore exercised some kind of lordship in the area. The nature and origins of this lordship are far from obvious, and require lengthier discussion than can be devoted to them here.28 However, this much is clear: although all three became rebels, their rebellion was not a cause but a consequence of their tenures in the locality. Of the three, only Roussel of Bailleul was motivated by political ambition rather than by a sense of grievance, and his ambition may well have been fuelled by his evident popularity among the local people. In the troubled aftermath of Mantzikert, his lordship offered them a degree of protection from the Turks, and therefore value for taxation, which they were not getting from the government of Michael VII in Constantinople. And although the government put an end to Roussel's lordship in the Armeniakon in 1075-6, it did so only after his refusal to accept the important court title of kouropalates, and it managed to do so only through the wiles of the young Alexios Komnenos, who bribed the Turks to betray the Norman into his hands.29

If Roussel had held out longer, the government might have been induced to recognise his independence, in the way that it recognised the de facto independence of another Norman warlord, Robert Guiscard, who had recently conquered the Byzantine provinces of southern Italy. In 1074, Guiscard finally agreed to a marriage alliance with the emperor Michael VII, which took effect two years later, when Guiscard's daughter Olympias-Helen was despatched to Constantinople to await her eventual marriage to Michael's infant son Constantine.30

In seeking this humiliating union with a parvenu barbarian adventurer, Michael VII was no doubt partly concerned to legitimise the status of southern Italy.31 However, the explanation offered by the earliest Byzantine source for the event is that the emperor wanted Guiscard to help him fight the Turks,32 and this explanation gains credibility from evidence that the overture had originally come from Romanos IV Diogenes,33 whose top priority had been the war on the eastern front. This brings us to our third observation concerning the "Franks" in imperial service before the First Crusade. Although the first Norman units in the regular armed forces seem to have been brought in by Constantine IX Monomachos to deal with an internal crisis in 1047,34 later recruits from Normandy were solicited specifically to fight the Turks.35 Certainly, the Frankish mercenaries are most often mentioned in connection with the war in the east. We hear of them fighting in Edessa, Amida, Erzerum, Mantzikert, and in the region of Antioch under the Armenian warlord Philaretos Brachamios.36

Thus for decades before the crusades, and most conspicuously in the 1070s, the Normans in Byzantine service anticipated many of the experiences which the crusaders were to have in the East. They got to know Greeks, Turks and Armenians as friends and enemies, they visited many of the places in Cilicia and Syria where crusaders were to fight and settle in 1097-8, they formed lordships on imperial territory, and they showed that this could even be done with the emperor's blessing. Their recruitment opened up the East as a land of opportunity for their compatriots, and, perhaps most importantly, they established a reputation for the Franks, both in their homeland and in Byzantium, as being the only warriors who were capable of taking on the Turks and winning. We shall return to the question of how they were perceived in Byzantium. Here the point I wish to make is that the Byzantine recruitment of the Normans created habits of mind and patterns of movement which made crusading thinkable and practicable in the late eleventh century.

The success of the crusading movement depended, primarily, on the fighting qualities of its warriors. However, it could not have been sustained without regular sea communications between Europe and Syria. After the establishment of the crusader states, these communications were provided by the three great trading cities of northern Italy: Genoa, Pisa, and Venice.37 Of these three cities, only Venice was associated with the Byzantine Empire at the time of the First Crusade, and it was the last to join the crusading movement. However, Venetian merchants frequented Antioch before the First Crusade.38 During the crusader siege of Antioch in 1097-8, the Byzantine navy itself, in conjunction with English and Genoese flotillas and a few Venetian and Pisan ships, was able to keep the besieging army supplied with basic necessities.39 And when Bohemond became lord of the captured city in 1098, he rewarded the Genoese for their help by granting them a trading quarter adjacent to the "street of the Amalfitans" (rugam Melfitanorum).40 This reminds us that before Antioch fell to the Turks in 1085, there already existed another network of maritime communications linking Syria with Europe and operated by an Italian trading city, the city of Amalfi near Naples.41 Amalfi, like Venice, was only loosely and nominally subject to the Byzantine emperor, and at the end of the tenth century its merchants were, if anything, less favourably treated by the imperial government than the Venetians. However, this changed in the course of the eleventh century, as the empire became increasingly preoccupied with first the expansion and then the defence of its territorial position in southern Italy. In the power politics of the region Amalfi was a valuable ally against the Lombards and the Normans, as well as a valuable channel of communication with Rome and the influential Benedictine abbey of Monte Cassino.42 Relations between Constantinople and Amalfi intensified towards the middle of the century, when Duke John II spent twelve years in exile at the Byzantine court before returning to power in 1052, no doubt with Byzantine help. It is surely no coincidence that the bulk of the evidence for Amalfitan activity in the east dates from after this point. We hear of Amalfitan religious houses in Constantinople, Mount Athos, Antioch and Jerusalem; we hear of Amalfitan traders in Dyrrachion, Constantinople and Antioch; most importantly, we hear of aristocratic Amalfitan entrepreneurs with business interests ranging from Italy to Palestine. Indeed, the sources give the impression that the Amalfitan presence in the East c.1060-1080 was closely tied to the business empire of one family, the descendants of count Mauro. The tantalisingly brief but striking details of their activities have often received comment, most famously by Yves Renouard in his book on Italian businessmen in the Middle Ages.43 We shall have more to say about one member of this family, Pantaleone, who flourished in the 1070s. For now it is sufficient to point out that they commanded an organisation which was able to finance the establishment of hostels in Antioch and Jerusalem, and to arrange the manufacture and shipment of bronze doors produced by Syrian craftsmen in a workshop of the Byzantine imperial palace for churches throughout central and southern Italy.44

In addition to fighting men and adequate supply lines, crusading required spiritual authorisation by the church, which in western Europe meant the papacy. As every crusading historian knows, Urban II was not the first pope to call for an expedition to help the eastern Christians against the Turks. In 1074, Pope Gregory VII wrote three letters announcing his intention to lead such an expedition in person and calling on the faithful to join him. Although Gregory did not describe this initiative as a response to a Byzantine VII in 1073, expressing the desire for restoration of the "ancient concord" between the churches appeal, it was undoubtedly connected with the embassy which the pope had received from Michael of Rome and .Constantinople.45

Thus, in the 1070s, Byzantium attracted the convergence of three developments which had not occurred before and would all recur in the crusade: the recruitment of Frankish knights to fight the Turks, the formation of a network of maritime communications linking Italy with Syria and Palestine, and papal sponsorship of a holy war of liberation. But does this mean that the imperial government of Michael VII actively anticipated the crusade? It is not hard to find objections. First of all, it may be observed that the three developments in question were entirely separate: the Papacy, Amalfi and the Normans were not yet natural allies, and each became involved with Byzantium for reasons which had nothing to do with the others.46 Only Gregory VII had a religious motive, but he did not invite either Amalfi or the Normans to join in his expedition, and he neither sponsored nor blessed the marriage alliance between Michael VII and Robert Guiscard. His main supporter in southern Italy was Gisulf of Salerno, a determined opponent of both the Normans and Amalfi, with his own designs on the Byzantine throne.47 Besides, at the very time when the marriage alliance was concluded, in 1076, Gregory had to abandon his plans for an eastern expedition in order to devote his energies to the struggle with the German emperor Henry IV which occupied him until his death.

In response to this objection, it may reasonably be argued that Gregory VII did not mean to exclude the Normans from his vision of a united Latin Christendom marching to the relief of its eastern brethren. Furthermore, it should be noted that from 1073, when Amalfi submitted to Robert Guiscard, the foreign interests of the two were effectively identified. The marriage alliance between Guiscard and Michael VII should therefore have made it easier for Amalfitan merchants to operate in Byzantine ports, and indeed they may well have been instrumental in causing the alliance to happen.

A more fundamental objection is that all three developments can be seen as alien to Byzantium, which, according to the accepted view, had no time for foreigners, and cultivated them only for short-term political gains.48 The Byzantine ruling elite would have detested the Normans as barbarians, despised the Amalfitans as traders, and distrusted the Reform Papacy as a threat to the unity and the traditions of the universal church. If these interests were beginning to make Byzantium the theatre of new international movements, they did so, surely, under their own momentum, and not as a result of conscious policy on the part of the imperial government. Rather, their very involvement in the empire's affairs would seem to be symptomatic of the failure of imperial policy in the period 1050-1080. For was it not the emperors of this period, with their reliance on the so-called civil party, who squandered the legacy of the great Basil 11, by diverting resources to the capital away from the provinces and the frontiers, and by allowing a corrupt, extortionate bureaucracy to flourish at the expense of military performance and military morale?49 And is it coincidence that the crusading plans of Gregory VII and the marriage alliance with Robert Guiscard both occurred after Mantzikert, during the reign of the amiable but ineffectual Michael VII Doukas (1071-1078), who, thanks largely to his negative portrayal by the historian Michael Attaleiates, became synonymous with the worst excesses and failures of the civilian regime?

These considerations are all superficially valid, but they are not the whole story.. Based on the record of sources written after the fall of Michael VII, they overlook the positive aspects of his regime, and they fail to do justice to the potential that existed, in the 1070s, for Byzantium to go a different route, not only in foreign policy, but also in official ideology, from the way it went under the dynasty of Alexios Komnenos. At the same time, however, and in qualification of this last point, the view that the trends of the 1070s had no future overlooks the extent to which they were resumed by Alexios.

Michael VII may have been personally incompetent, but he was served by able and energetic ministers, the metropolitan of Neocaesarea, the metropolitan of Side, the sebastos Constantine, and, last but not least, the notorious eunuch Nikephoros, known opprobriously by the diminutive form "Nikephoritzes". As Paul Lemerle recognised, if we divest the facts of Michael's' administration of the negative interpretation put on them by Attaleiates and other historians, we can see that his government did have a serious programme for the military and financial revitalisation of the state.50 I suggest that Michael VII’s foreign policy, which his government inherited from his father, Constantine X, and ultimately from Constantine IX Monomachos, was equally serious and responsible. Specifically, it would be my contention that the recourse to Norman manpower, Amalfitan sea-power and alliance with the papacy was in the best tradition of Byzantine imperialism: a tradition which embraced the Latin West, just as it embraced the Syriac and Armenian Orient, as an integral part of the Christian Roman oikoumene.

Of course, the contraction of the empire in the seventh and eighth centuries had made it rather less ecumenical, and rather more exclusively "Greek Orthodox" in dimensions, culture and outlooks.51 However, Byzantium had never formally renounced its claims to its lost provinces, and from the ninth to the eleventh centuries it was in a position to reassert those claims, by expanding on all fronts to become, once again, the dominant power in southern Italy and northern Syria. By the time of Mantzikert, these areas where non Greek Christians either formed a significant minority, as did the Latin Catholics in southern Italy, or were in the majority, as was the case with the Armenians and Syrians in Cilicia and Syria, had been ruled from Constantinople for a century - and more in the case of southern Italy.52 We tend to assume, from our reading of narrative sources which dwell, retrospectively, on the iniquities of the imperial administration, that these extremities of empire were peripheral to the interests of the ruling elite in Constantinople, which alienated their inhabitants by a combination of fiscal greed, military fecklessness and religious intolerance. But there is another side to the coin, which appears on more careful consideration of the evidence. Southern Italy and northern Syria were desirable assets. Quite apart from their strategic value, they were rich and populous, with well-developed economies.53 Though far from Constantinople, they offered proximity to two other ancient centres of the Roman world, Rome and Jerusalem, on which Byzantine emperors kept their eye. Furthermore, the capital of Byzantine Syria was Antioch, another great city of antiquity, an apostolic see and one-time imperial residence.54 Modern histories of Byzantium consistently underestimate the significance of the reintegration of this city after three and a half centuries in which Constantinople had had no urban rival within the empire.55 By all accounts, Antioch was an extremely flourishing place.56 The list of Byzantine governors of the province from 969 to 1085 certainly suggests that this was a key political appointments.57

The effort to bind the imperial periphery to the imperial centre, to contain frontier regions as distant and diverse as southern Italy and northern Syria within the framework of a highly centralised state, was ultimately unsuccessful, but while it lasted it created patterns of interaction that did not end with the failure of the system, and were perhaps at their most developed when the system was on the point of collapse in the 1070s, by which time they had acquired a momentum of their own. The system created strains, but it also created contact, movement and relationships. It imposed the centre on the periphery, but it also caused the periphery to impinge on the centre, and brought opposite extremities of empire into contact with each other, in all sorts of direct and indirect ways. The large-scale migration of Armenians and Syrians into the reconquered eastern frontier areas, and from these areas to Constantinople along with numbers of Jews;58 the coming of the Normans to southern Italy and their subsequent engagement on the eastern front; the activities of the Amalfitans in Constantinople and Antioch; the contributions made by the emperors Romanos III (1028-34), Michael IV (1034-42) and Constantine IX (1042-55) to the rebuilding of the church of the Holy Sepulchre in Jerusalem;59 the generosity of Michael VII to the cathedral church of Anagni 60 and the abbey of Monte Cassino 61: these were all products or by-products of the Byzantine imperial presence in southern Italy and northern Syria.

So, in a sense, was the schism of 1054. This event, more than any, was symbolic of the rupture in the Christian oikoumene, and symptomatic of the growing refusal of the Latin West to regard the church and empire of Constantinople as having more than a regional significance.62 Yet the issues and personalities that brought the conflict to a head reflected the multi-cultural, supra-regional character of the Byzantine state. The issue that provoked Pope Leo IX to reprove and his legates to excommunicate the Patriarch Michael Keroularios, was the condemnation by certain influential Greek clerics of the use of unleavened bread (azyma/Azymes) in the Eucharist, and the reported closure of Latin churches in Constantinople that followed this practice. The practice was common to both Latins and Armenians, and the Greek criticism of its usage in Latin churches was probably an attempt to achieve consistency in a polemic directed initially, and principally, at the empire's far more numerous Armenian community.63 The issue that underlay all others was the papal claim to ecclesiastical supremacy, which Leo IX and his associates were asserting as the basis of their programme of church reform in the West. It is likely that their application of this principle to the eastern patriarchates was prompted by the letter in which Peter, the newly elected Patriarch of Antioch, sent to Rome in 1052 announcing his enthronement, in accordance with ancient but lapsed custom.64 Peter was completely loyal to the traditions of the Constantinopolitan church, in which he had been formed.65 He certainly could not have intended that his letter should elicit, as it did, an ecclesiological critique of his alma mater and a papal offer of support for the eastern patriarchs against the intrusions of their overbearing Constantinopolitan colleague.66 Yet from the viewpoint of an eastern patriarch like Peter, the distant Pope was a respected colleague rather than the dangerous rival he appeared to be in Constantinople. Peter was no doubt encouraged in this fraternal view by his contacts with the Latins who took advantage of the pax byzantina to visit Antioch on business and on their way to the holy places of Palestine. He was impressed by the piety of these pilgrims,67 and it was to one of them that he entrusted his letter to the Pope, sending it to "the most illustrious magistros and doux of Italy, Argyrus", for forwarding to Leo IX.68 Peter either did not know at this point, or did not care, that Argyrus, a Latin noble from Bari with a family history of rebellion against imperial rule, was an implacable enemy of Michael Keroularios and a staunch adherent of Latin ritual.69 Keroularios was convinced that. Argyrus had inflamed the controversy by misrepresenting him to the Pope,70 and it is not unlikely that Argyrus and Leo IX were planning a political solution to the problems of southern Italy which would involve the return of the local bishoprics to the jurisdiction of Rome.71 Yet Argyrus was in many ways a Byzantine figure: his upbringing in Constantinople, his intimacy with the emperor Constantine IX Monomachos, his elevated court titles of patrikios and magistros, his titled relatives,72 and his function as military governor (doux) of Byzantine Italy, all made him an integral member of the imperial establishment. He and Keroularios represented alternative versions of Byzantine imperial interests, and the emperor, if anything, preferred his version.73

The periphery, then, was not so peripheral, and the centre was not completely self-centred. We cannot get inside the minds of imperial policy-makers and advisers in the 1070s, but we can build up a cultural profile of the circles surrounding Michael VII and his ministers from the writings and careers of individuals who moved in those circles. They comprised three men of non-Greek origin from southern Italy and northern Syria. One was Pantaleone, the noble entrepreneur from Amalfi. That he enjoyed favour at court is evident not only from his family house in Constantinople and his court title of anthypatos, but also from the fact that he was able to commission bronze doors for Italian churches from the workshops of the imperial palace. As Cyril Mango has suggested, he may actually have been responsible for operating the workshops, given that the craftsmen were Syrian and he had trading interests in Antioch.74 Pantaleone may or may not have been identical with the Amalfitan noble who took advantage of the confusion at the overthrow of Michael VII in 1078 to steal a piece of the True Cross from the imperial palace; if not, there was at least one other Amalfitan in the emperor's entourage.75

Another southern Italian close to Michael VII was the philosopher John Italos. The main facts of his biography are well known from Anna Comnena and from the records of his trial for impiety under Alexios I.76 His father was a soldier from the Italian mainland who served in Sicily, and the young John was apparently destined for a military career, until the family was forced to flee to Apulia, from where John emigrated to Constantinople, possibly in the entourage of Argyrus. Here he studied under various teachers, including Michael Psellos, and came to enjoy the patronage of Michael VII. Because of his Italian background, the emperor sent him on a diplomatic mission to southern Italy; this, however, went wrong, and fearing the worst, Italos fled to Rome. Anna Conmena says that he betrayed imperial interests, but his delinquency cannot have been too serious, since he subsequently returned to Constantinople to take up an important teaching position, succeeding Michael Psellos as "consul of the philosophers". Although his teaching provoked an official enquiry from the patriarch Kosmas, he was not as yet dismissed or silenced. Indeed, there is fairly good evidence, which has been entirely overlooked by Byzantinists, that he was appointed to ecclesiastical office. A diatribe written in 1116 against one of Italos' pupils, Eustratios of Nicaea, says that Eustratios was taught by "John, the former chartophylax of Antioch the Great".77 It is not clear when the appointment took place; nevertheless, the discovery of the connection with Antioch adds a new dimension to Italos' personality, particularly in the light of the trading connection maintained by his fellow Italian Pantaleone of Amalfi.

According to Psellos, Italos on one occasion hit back at his critics by composing a speech in which he lamented the fact that the wisdom of the Greeks had migrated to the eastern barbarians, "to the Assyrians and Medes and Egyptians".78 This was not just a deft rhetorical rejoinder to the accusation of being a western barbarian; it was a commonplace among Arab intellectuals of the tenth and eleventh centuries,79 which Italos would have encountered in his contacts with easterners who came to Constantinople to study with Psellos and seek patronage at the imperial court.80 We know two of these men, the Egyptian astrologer Eleutherios Zebelenos,81 and the philosopher Symeon Seth, who wrote two treatises on astronomy and one on diet, all dedicated to Michael VII.82 Later, at the request of Alexios I, he translated the Indian fable Kalila wa Dimna from Arabic into Greek as Stephanites and Ichnelates,83 and, according to Anna Comnena, he accurately predicted by astrology the date of the death of Robert Guiscard.84 Although he was very learned in Greek philosophy,85 his knowledge of Arabic and his surname would seem to indicate a Semitic origin. The heading of his treatise on diet calls him "Antiochene", which suggests that he may have been a native Syriac speaker.86

From these three cases we get an impression of a court milieu which was receptive to men of wealth and talent from southern Italy and northern Syria, and promoted contacts between them. But how were such outsiders viewed by the native Constantinopolitans who ran the bureaucracies of church and state and the great monasteries of the Byzantine world? The terms and the context in which Anna Comnena refers to John Italos and Symeon Seth suggest that they were likely to be regarded as undesirable aliens, along with the barbarian "Franks". The exclusive sense of Hellenic cultural identity and Orthodox religious identity which Anna expresses in her Alexiad 87 had deep, if separate, roots that are clearly visible a century earlier, in the cultural and metropolitan snobbery of Michael Psellos,88 and in the increasingly confrontational attitudes towards the Armenians and the Latins that were surfacing in some clerical and monastic circles - the same circles that disapproved of John Italos, and that Alexios I wanted to impress by bringing John Italos to trial.89

However, such attitudes were not necessarily representative of the clergy as a whole. The church was notoriously slow to condemn or silence Italos. Kosmas, the patriarch to whom the philosopher's case was first referred, declined to investigate it. It may or may not be relevant that Kosmas came from Jerusalem,90 where he would have come into contact with many different types of Christians, including the Amalfitans with their hostel of St John the Almsgiver. But the fact that he and other members of the ruling elite either came from or had direct experience of the frontier areas must have helped to broaden their horizons.

A broad, imperial outlook that transcends cultural divisions certainly informs the writings of two senior army officers, Kekaumenos (writing c.1077-8) and Gregory Pakourianos (writing in 1084). In his brief survivor's guide to the imperial system, Kekaumenos, though an advocate of keeping foreigners in their place and no admirer of Robert Guiscard, can see things both from the viewpoint of a Constantinopolitan bureaucrat and from that of a semi-independent Serbian or Armenian lord; his own ancestors had operated on both sides of the eastern frontier.91 In the foundation charter for his monastery in Thrace, Gregory Pakourianos, a Georgian of Armenian culture, though evincing a strong dislike of Greeks, whom he excludes from his monastery, is unfailingly loyal to the empire and takes as his monastic model a Greek community in Constantinople that was combined or associated with a Latin Benedictine house. He began his career as a landed magnate in the Caucasus and Asia Minor, and ended it with a domain in Thrace and Macedonia.92

Of course, both Kekaumenos and Pakourianos were of foreign descent, like many military men, and military men were, arguably, by temperament and training more flexible and pragmatic than sedentary civilians with their heads full of doctrinaire ideas.93 But even civilians could be flexible and pragmatic, especially if they had experience of provincial administration. A case in point may lx the notorious Nikephoritzes, who had twice served as governor of Antioch, and once as governor of central and southern Greece, before becoming Michael VII’s chief minister.94 According to Attaleiates, Nikephoritzes' worst crime was to establish a government grain monopoly at Raidestos, the main maritime outlet for the wheatfields of Thrace, by compelling producers to sell and exporters to buy at a purpose-built, state controlled exchange called the phoundax, whose effects were as injurious as its name was outlandish.95 Of the many problems posed by this institution, I would like to raise just one: why did it have an outlandish name? Why was an institution which had good antecedents in imperial tradition not called by an obvious and traditional name, such as παυδoχειυ or απoθήκη, or even μιτάτου - the name, of Latin derivation, used of the entrepots in Constantinople where foreign merchants sold their wares? Why, instead, was it called by an Arabic corruption of the name παυδοχειυ? The answer to my mind can only be that the phoundax of Raidestos was directly modelled on the Arab funduq, which was, above all, a facility for foreign merchants, and that the model was adopted because the merchants who acted as the privileged middlemen at the corn exchange were primarily oriented towards the Islamic world. Of the two most likely sets of candidates, the Venetians and the Amalfitans, the clear favourites must be the Amalfitans. Not only were they more in evidence in Constantinople at this time, and more vital to the empire's Italian interests, but Nikephoritzes would have encountered them during his governorships at Antioch, where their trading quarter was no doubt called funduq or fondaco.

This has to be conjecture, because Attaleiates tantalisingly does not identify the corn dealers at the phoundax. But then there is much that Attaleiates does not tell; indeed, his silences are as telling as his statements, and together they make him the key witness for the outlook of Michael VII’s administration. People who know Attaleiates may think it odd that I should be calling him to testify in defence of Michael VII, since he is so clearly the chief witness for the prosecution.96 However, Attaleiates wrote to gratify, and justify, Nikephoros Botaneiates; the ruler who deposed and succeeded Michael. He was an expert lawyer, which meant that he was even better trained than most Byzantine historians in the rhetoric of arguing both sides of a case. He may genuinely have liked Botaneiates better than Michael VII - it was easier to like an indiscriminately generous regime than a rigorous fiscal regime. But Attaleiates had actually done quite well under Michael VII. This much is clear, mainly from his foundation charter for his monastery in Constantinople and his poor-house at Raidestos,97 but also from the fact that he composed a legal textbook at Michael's request,98 and from his history, where he reveals that he was on sufficiently good speaking terms with Nikephoritzes to be able to warn the minister that the unpopularity of the phoundax at Raidestos was helping the cause of the rebel Nikephoros Bryennios.99 Attaleiates was hurt by the phoundax, but he kept his estates, his titles, and his high judicial office, which was no small achievement in the unstable political conditions of those years. In short, he can be regarded as a fairly typical, conformist civil functionary of the time, more educated than most, but not an outstanding intellectual like his great contemporary Psellos; conventionally but not excessively pious. This is what makes his testimony especially valuable.

In general terms, Attaleiates makes the appropriate noises about heretics and barbarians. He quotes, with apparent agreement, the opinion that the Turkish inroads into the eastern frontier provinces were a manifestation of divine "wrath against the heretics, the Armenians who inhabit Iberia and Mesopotamia, as far as Lykandos and Melitene and the adjacent lands, and those who profess the Jewish heresy of Nestorius and that of the Akephaloi; for those territories are full of that kind of evil doctrine.100 In another passage, he writes that "the pangs of death tightened about us (Psalm 18.5), as the Gothic and most foul nations prevailed throughout the whole of the east and the west".101 It should be noted, however, that both these remarks are made in the context of the argument that the Byzantines are their own worst enemies, and have brought divine wrath upon themselves by their flagrant disregard for justice and their propensity to fight each other to the death in civil war - in contrast to the feebleness of their war effort against outside invaders. The kakodoxy of the heretics and the foulness of the barbarians are therefore evoked in order to express the gravity of the corruption at the heart of the Doukas regime.

In other contexts, and in more specific instances, Attaleiates is quite willing to express solidarity with, or admiration of, non-Greek Christians. At the beginning of his history, he blames the governor of southern Italy, Michael Dokeianos, for having alienated the empire's confederates to the north, the Albans and the Latins, "who share equal citizenship with us, as well as our religion".102 He thus, presumably, regarded the Amalfitans as fellow Romans, which has important implications for our interpretation of their complete absence from his history. Given the favour in which they were held by the Doukas emperors, Constantine X and Michael VII, the fact that he does not hold this favour against either emperor means either that he considered the Amalfitans unimportant or that he did not want to criticise them. I personally think that the Amalfitans were important enough for his silence to be deliberate, and I would therefore interpret his silence as an indication, not only that the Amalfitans remained in favour under Nikephoros III Botaneiates, but also that Attaleiates knew and liked Pantaleone and his compatriots. I would further connect his silence with his curious lack of attention to the actual merchants who operated the phoundax. And following the same line of interpretation, I would interpret his failure to mention John Italos as an indication of sympathy for the controversial philosopher.

As for Symeon Seth, he may be implicated in a disapproving reference to Michael VIPs consultation of astrologers, but it is perhaps significant that Attaleiates does not mention names, and I wonder whether it is pure coincidence that Seth had a house at Raidestos, where Attaleiates had substantial interests and connections, and sold some books to Attaleiates' monastic foundation.103 There is no doubt that he would have admired Seth's cultural formation. He writes admiringly of another man, Peter Libellisios, governor of Antioch in 1070, whose background and learning sound very similar: "a man who was Assyrian by race, a native of Antioch the Great, who was consummately trained both in Roman wisdom and learning, and in that of the Saracens, through his talented and lofty nature".104 It is interesting to learn, incidentally, that Seth was not an isolated case.

Perhaps it is not surprising that a cultivated Byzantine would have felt an affinity for civilised Christians from neighbouring Mediterranean cultures. The real surprise is to find Attaleiates entirely positive on the subject of the Normans, who were barbaric by any Mediterranean standards. He is, in effect, one of the earliest exponents of the "Norman myth".105 In his account of the reign of Constantine IX Monomachos, he singles out two anonymous Latin soldiers, probably Normans, for special commendation. The Latin commander entrusted with defending the garrisons of Thrace against the Pechenegs, possibly to be identified with the Norman Herve, is called "a man particularly brave in combat, and second to none in thinking of the right course of action".106 At the first Turkish siege of Mantzikert, a formidable enemy siege engine was destroyed by the brave action of a certain Latin, whom God, "who always takes care of the Christian race, inspired with strength and purpose beyond imagination".107

Elsewhere, Attaleiates describes the Normans in the Byzantine army as "bloodthirsty and warlike men", but in the context that may be a compliment.108 In connection with the rebellion of Robert Crispin against Romanos IV Diogenes, he writes that "the Frankish race is faithless by nature".109 Again, however, the exact context is important: Attaleiates is explaining why, even after their reconciliation, the emperor failed to treat the Norman leader with honour, and listened to suggestions that Crispin's submission was not genuine. The episode, as told by Attaleiates, in fact reflects quite well on Crispin and quite badly on Romanos and the Byzantines. Crispin is portrayed as a brave and fearless fighter against the Turks who might have defied imperial authority and robbed imperial tax-collectors but had deep scruples about shedding Christian blood, in stark contrast to the Byzanfnes, who set upon him in overwhelming force on Easter Day, only to be ignominiously routed. The clear implication is that it was both hypocritical and small-minded of Romanos to doubt the sincerity of this particular Frank on the grounds that all Franks are fickle by nature. Attaleiates was on the whole an admirer of Romanos, but he had to explain why God allowed this brave and well-intentioned emperor to be defeated at Mantzikert, and then to lose the civil war against Michael VII. He clearly saw divine justice in the fact that the turning-point in the civil war came when Crispin joined the forces of Michael VII.110

Attaleiates certainly has no hesitation in showing sympathy for the main Norman character in his narrative, Roussel of Bailleul, who rebelled against Michael VII. He does not excuse, but neither does he condemn, Roussel's insurrection, and he repeatedly praises Roussel's martial prowess. A propos of the second phase of the insurrection, when Roussel withdrew from western Asia Minor to his old base in the Armeniakon, Attaleiates clearly implies that the emperor should have let him stay there, because he was doing such a good job of driving the Turks out of the region. Instead, Michael VII, at the instigation of Nikephoritzes, "preferred that the Turks should take over and direct the entire Roman state than that this Latin should be confined to one place and repel their incursions".111 This passage, which goes on to describe how Roussel was taken captive by Alexios Komnenos, comes immediately after the passage lamenting how the Byzantines have made themselves a prey to "the Gothic and most foul nations". It is then followed by the section describing the iniquitous phoundax at Raidestos, which Attaleiates explicitly introduces in order to show what Nikephoritzes was doing while Asia Minor was being overrun by the Turks. The sequence clearly shows that Attaleiates thought of the Latin Roussel not as one of the empire's Gothic and most foul enemies, but rather as its greatest asset, which Michael VII and Nikephoritzes were incapable of appreciating. The point is further developed when the narrative, after denouncing the phoundax, returns to the subject of Roussel and relates what happened when Alexios delivered his captive to Constantinople. Michael VII had Roussel tortured and imprisoned, whereas what he should have done, says Attaleiates, was to try him for treason, sentence him to death, and then pardon him, thereby winning Roussel's undying gratitude and "preserving such a great soldier and general for the Roman Empire".112

Attaleiates may have found it politic to eulogise Roussel because the latter favoured Nikephoros Botaneiates;113 however, his comments on Roussel are consistent with his comments on other Normans, and together they add up to the statement that the Normans were a superior race of beings, clever as well as brave, who would have saved the Empire if only its rulers had handled them properly. He more or less says that they were well worth any price they demanded for their services. Most importantly, he does not consider them to be aliens. He notes their Orthodox Christian piety. He refers to them as Phrangoi and Latinoi, words which did not yet have the negative connotations they were to acquire in the twelfth century. The Franks, according to an official tradition articulated in the tenth century by Constantine Porphyrogenitus, enjoyed a special relationship with the Roman Empire.114 The Latins, by Attaleiates' own definition, enjoyed equal citizenship with the Byzantines. This contrasts strikingly with later histories of the same events, which not only insist on the Normans' bad points,115 but call Roussel a barbarian 116 - a term which Attaleiates never applies to the Normans. Moreover, the negative image of the Normans in later historiography owes much to the inclusion of Robert Guiscard, who is conspicuously missing from the pages of Attaleiates. Here, I suggest, we have another significant silence. If Attaleiates had disapproved of the marriage alliance between Michael VII and Guiscard, he would surely have cited it as an example of Michael's misrule, especially since Botaneiates put an end to it. The inference must be that Attaleiates regretted the destruction of the alliance but in loyalty to Botaneiates could not say so.

In conclusion, Michael Attaleiates, writing in 1080, presents a much more inclusive view of the Latin West, and, to a lesser extent, of the Christian East, than is to be found in writers of the Comnenian period. This inclusiveness may have something to do with his own temperament and personality, but mostly it has to do with the fact that he was writing fifteen years before the First Crusade. We know that others shared his admiration of Roussel. Attaleiates lived at the end of a long era of imperial expansion, and his outlook reflects the expansive side of the Byzantine imperial mentality - an expansiveness which had accumulated not only under the great soldier emperors, but also, more recently, under the civilian successors of Basil II.

What nearly all these emperors, from Romanos III to Michael VII, had in common was their interest in law and justice. Beyond its practical application, this interest had an ideological dimension, which can be seen Constantine IX's foundation of a school for the academic study of Roman law (1047).117 As P. Speck and M. Angold have noted, this foundation was unlikely to have been concerned solely with improving the legal system; it was also an assertion of Roman imperial renewal.118 Constantine IX envisaged that the students would learn Latin; this was no doubt optimistic, but it is surely significant that Psellos refers to law as the "Italian art", the "Italian Muse", and the "Italian wisdom".119 In the sixth century, the study of law had created a deep sense of Romanness and Latinity in bureaucrats like Tribonian and John the Lydian, both natives of Greek-speaking Asia Minor.120 Something very similar, I suggest, was at work in the eleventh century, and is the key to understanding Michael Attaleiates, a native of southern Asia Minor, like Tribonian, and possibly a graduate of the law school of Constantine IX.121 Attaleiates, who refers to his legal education as his "sacred initiation into the laws ",122 wrote a legal manual which shows a clear appreciation of the historical evolution of Roman law,123 and in his history draws his classical exempla from ancient Rome, not ancient Greece.124 It was this Romanitas of Attaleiates which disposed him to consider the Latin West, including the Normans,125 as a natural and necessary component of the empire's defence against its real enemies, the Turks. And in expressing his Romanitas he was very much a man of his age 126 - an age which peaked under Michael VII, and began to end in 1081.

By the time Alexios Komnenos seized power in 1081, the imperial foreign policy of the past decade was in ruins. Robert Guiscard was invading the Balkans with the declared intention of avenging the deposed Michael VII. He now had the blessing of Pope Gregory VII, who needed his help against the German emperor Henry IV and obligingly excommunicated both Nikephoros Botaneiates and Alexios Komnenos. Because Guiscard controlled Amalfi, the Amalfitans in Constantinople were inevitably identified with his camp. A grand imperial counter-offensive against the Turks based on Norman manpower, Amalfitan sea-power and the spiritual backing of the papacy was now out of the question. Instead, Alexios had to make his peace with the Turks, seek alliance with the German emperor, and offer improved trading privileges to Amalfi's main rival, Venice, in order to counter the Norman threat. At home, Alexios was obliged - and no doubt quite pleased - to advertise himself as a champion of hardline Orthodoxy in order to legitimise his violent takeover and his controversial dynastic regime, which amounted to a kind of privatisation of the state.127 John Italos was a conspicuous casualty of both the internal and the external realignment.

To twelfth-century Byzantines it was clear that the war against Guiscard and Bohemond was the beginning of a recurring and unresolved conflict between the empire and the barbarian Franks. Yet until the crusade - or rather, until the crusade went wrong - Alexios and his advisers were not to know that the conflict was going to recur or to resist resolution by traditional means; they were not to know that the Norman war invalidated the assumptions and attitudes with which they had all grown up in the 1060s and 1070s. Internally, Alexios perpetuated the administrative ethos and methods of the Doukas regime, especially after 1090, when the civil administration was taken over by the son of one of Michael VIPs leading ministers.128 Externally, in the course of his reign, Alexios renewed the empire's ties with several old friends, including Monte Cassino, Amalfi, and, of course, the Papacy.129 His rapprochement with the new pope, Urban II, which began in 1089 and culminated in the preaching of the crusade, marked a clear and conscious reversion to the policies of the 1070s. The pope again sought to restore the "ancient concord" between the churches of Rome and Constantinople, and the emperor again sought military aid from the West against the Pechenegs and the Turks. a Alexios' experience of the Normans evidently did not deter him from wanting to recruit large numbers of Frankish mercenaries.130 That experience, we may recall, had begun in 1075-6 with the capture of Roussel of Bailleul. Alexios could not have got the better of this formidable adversary without developing a healthy respect, not only for Roussel's military abilities, but also for the admiration and loyalty which the Norman inspired among many Byzantines. It is hard to believe that Alexios did not share the conviction, expressed by his own relative Theodore Dokeianos as well as by Michael Attaleiates, that the empire needed to use, rather than break, a man of this calibre.131

It is also hard to believe that Alexios did not feel the same way about his later Norman adversaries, Robert Guiscard and Bohemond. For Anna Comnena, who knew the sequel, the relationship between her father and these men could only be one of confrontation, but for Alexios himself there was always the hope of winning them over, as he won over many of Guiscard's and Bohemond's men, by gifts and by his own growing reputation for sound and successful military leadership.132 By the time of the First Crusade, the empire's finances had improved dramatically and he had proved his worth as a leader. He could therefore confidently expect to attract the crusading leaders into dependence on him, and he could look back on the empire's previous problems with the Normans as the result of weakness and mismanagement on the part of his predecessors. As Jonathan Shepard has observed, his treatment of the crusading leaders suggests a care to avoid the mistakes which had led Herve to rebel against Michael VI, Crispin to rebel against Romanos IV, and Roussel to rebel against Michael VII. Not only was he generous with his money, as even the hostile western sources admit, but he sought to bind the western lords to him by personal ties which corresponded to their own ideas of political loyalty: ties of adoptive kinship, and ties of liege homage.133 This treatment had the desired effect on at least one crusader, Stephen of Blois.134

So what went wrong? The obvious explanation is that Alexios failed to understand what the crusade was all about;135 that he looked on the crusaders as mercenaries, comprehending neither their religious zeal nor their fixation on Jerusalem, because his vision was limited to the restoration of his empire in Asia Minor. But this is where we must be particularly careful of assuming that Anna Comnena is representative of her father's attitudes in the 1090s. Another Byzantine historian, writing in the thirteenth century, but drawing on material for Alexios' reign that is now lost, states that Alexios himself called for the liberation of Jerusalem because he knew that this would motivate the western knights.136 Since Peter Charanis drew attention to this source almost fifty years ago, it has not received the attention it deserves.137 There are certainly indications that after the crusader conquest of Jerusalem Alexios included it in his plans of imperial restoration.138 And even if Jerusalem was an afterthought in these plans, Antioch was not. A city which had been an important centre of the eleventh-century empire; a city of which Alexios' own brother had been governor,139 and which had remained nominally subject to Constantinople until 1085, could not be low on his list of priorities. The very fact that Bohemond, the most byzantinised of the crusaders, coveted Antioch and allowed it to distract him from completing his pilgrimage to Jerusalem, is indicative of the city's value in Byzantine eyes. I see no reason to doubt that the crusader siege of Antioch was undertaken on his advice and was of vital interest to him, and contrary to the scepticism of John France, I consider it likely that he did intend to join the crusaders there, before he was discouraged by Stephen of Blois and the other deserters.140 There can certainly be no doubt that he felt deeply disappointed and betrayed by Bohemond's appropriation of the city in violation of his oath. We can appreciate the depth of this disappointment if we take seriously the indications of two western sources that Alexios had treated Bohemond with extraordinary trust and generosity. The Gesta Francorum claims that he offered Bohemond land to the extent of fifteen days' (north and?) south of Antioch, and eight days march east of it.141 Raymond of Aguilers says that the imperial commander, Tatikios, handed over to Bohemond the three Cilician cities of Tarsus, Mamistra and Adana before departing from the siege of Antioch.142

From our vantage point it seems inconceivable that Alexios could possibly have maintained cooperation with an expedition in which Bohemond was a leader. In this essay I have been arguing that the pertinent question is rather: why did Alexios and Bohemond fail to reach a working agreement? Bohemond was a disinherited son in search of a role. Why did he not find a niche in imperial service - a niche shaped by the trial and error of the empire's previous experience with Herve, Crispin and Roussel of Bailleul? The explanation may lie in the fate, real or legendary, of these figures. Herve was said to have been drowned by order of Constantine X;143 Crispin was said to have been poisoned "through the envy of the Greeks";144 and Roussel would have ended his days in prison if Michael VII had not - called on him in desperation.145 Serving the Byzantine emperor was a risky business, as his own subjects knew well.146 The rewards of the imperial system were worth the risks only as long as there was no alternative. But the crusade did offer an alternative: an opportunity to gain fame and fortune in the service of God, as mediated by the Pope, who, though he offered no better job security than the Byzantine emperor, did offer the infinitely greater reward of the remission of sins. No offer that Alexios could make could possibly equal this. The crusade created a context in which Bohemond could carve a lordship out of imperial territory and not only escape dependence on the emperor, but also, by defying the emperor, win a heroic reputation such as he could never have gained as imperial commander-in-chief.

This more than anything confirmed the worst suspicions of those Byzantines who had always maintained that "the race of the Franks is faithless by nature". From now on, this negative stereotype gained at the expense of the image of the Franks as the cavalry charging to the rescue. The positive image did not disappear entirely, and can be seen in Anna Comnena's grudging admiration for Guiscard and Bohemond.147 But this was an admiration for the "otherness" of the outsider, the admiration of the civilised Hellene for the untamed, unprincipled barbarian, rather than the admiration of a Roman imperialist for another Roman of a different, but compatible, culture. The crusade had made this ecumenical perception difficult to sustain. Yet without this perception there would have been no crusade. And ecumenism and Romanitas were far from spent: they were to enjoy at least one further flowering in the imperial foreign policy, and the Kaiserkritik, of the twelfth century.148

End Notes

1. J. Riley-Smith, The First Crusade and the Idea of Crusading (London, 1986).

2. J. France, Victory in the East. A Military History of the First Crusade (Cambridge, 1994).

3. M. G. Bull, Knightly Piety and the Lay Response to the First Crusade: The Limousin and Gascony, c.970-c.1130 (Oxford, 1993).

4. S. Runciman, A History of the Crusades, 3 vols. (Cambridge, 1951-4).

5. Above, nn. l and 2.

6. Anna Comnena, Alexiad, X.5, 4ff: ed. B. Leib (Paris, 1937-45), II, 206ff; tr. E.R.A. Sewter (Harmondsworth, 1969), 308ff. See also the German translation, based on a forthcoming new edition of the text, by D.R. Reinsch (Cologne, 1996), 334ff.

7. R.-J. Lilie, Byzanz und die Kreuzfahrerstaaten (Munich, 1981), tr. J.C. Morris and J.E. Ridings (Oxford, 1994).

8. J. Shepard, "Aspects of Byzantine Attitudes and Policy towards the West in the Tenth and Eleventh Centuries," in J. Howard-Johnston (ed.), Byzantium and the West, c.850-c.1200 (Amsterdam, 1988) [= Byzantinische Forschungen, 13 (1988)], 102-15; idem, "When Greek meets Greek: Alexius Comnenus and Bohemond in 1097-8," Byzantine and Modern Greek Studies 12 (1988), 185-277; "'Father' or 'scorpion'? Style and substance in Alexios's diplomacy," in M. Mullett and D. Smythe (eds.), Alexios I Komnenos, I: Papers (Belfast, 1996), 69-132.

9. P. Magdalino, The Empire of Manuel I Komnenos, 1143-1180 (Cambridge, 1993), chapter 1.

10. A. Laiou, "On Just war in Byzantium," in S. Reinert, J. Langdon, J. Allen (eds.), Studies in Honor of Speros Vryonis Jnr. (New Rochelle, N.Y., 1993), I, 153-74; N. Oikonomides, "The Concept of 'Holy War' and Two Tenth-Century Byzantine Ivories," in T.S. Miller and J. Nesbitt (eds.), Peace and War in Byzantium. Essays in Honor of George T. Dennis S.J. (Washington, D.C., 1995), 62-86.

11. The published text is in History 65 (1980), 177-92.

12. English Historical Review, 99 (1984), 148.

13. Ed. and tr. V. G. Berry, De profectione Ludovici VII in orientem (New York, 1948), 56-7.

14. J. Phillips, Defenders of the Holy Land. Relations between the Latin East and the West, 1119-1187 (Oxford, 1996).

15. A new starting point has been set by the provocative article of J. Howard-Johnston, "Anna Komnene and the Alexiad," in Mullett and Smythe (eds.), Alexios I Komnenos, 260-302. Some of the implications for Anna's representation of the crusade have been discussed, from different viewpoints, by R.-J. Lilie, "Der Erste Kreuzzug in der Darstellung Anna Komnenes," in Varia II (Bonn, 1987), 49-148, and R.D. Thomas, "Anna Comnena's account of the First Crusade. History and politics in the reigns of the emperors Alexius I and Manuel I Comnenus," Byzantine and Modern Greek Studies 15 (1991), 269-312; see also next note.

16. J. France, "Anna Comnena, the Alexiad and the First Crusade," Reading Medieval Studies 10 (1983), 20-32; idem, Victory in the East, 111-12, 117ff, 144-5, 382. France seems unaware of the extent to which Byzantinists (see following note) share his reservations about Anna's evidence.

17. Shepard, "Aspects," loc. cit.; Magdalino, Empire, 30-34.

18. M. Angold, The Byzantine Empire, 1025-1204: A Political History (London, 1984), 21-5.

19. Peter the Hermit: Alexiad, X.5. 5, 10 (ed. Leib, II, 207, 209-10); cf. E.O. Blake and C. Morris, "A Hermit goes to War: Peter and the Origins of the First Crusade," in W.J. Shiels (ed.), Monks, Hermits and the Ascetic Tradition, Studies in Church History 22 (Oxford, 1985), 79-107. Robert of Flanders: Alexiad, VII.6.1, 7.4, VM.3.5 (ed. Leib, II, 105, 109-10, 135); cf. F.L. Ganshof, "Robert le Frison et Alexis Comnene," Byzantion 31 (1961), 57-74.

20. For Bohemond's own familiarity with Byzantine ways, see the cultural profile that emerges from recent studies of his artistic patronage and linguistic knowledge: A.W. Epstein, "The Date and Significance of the Cathedral of Canosa in Apulia, South Italy," Dumbarton Oaks Papers 37 (1983), 79-80; Shepard, "When Greek meets Greek," passim and 251ff.

21. For an example of the kind of contacts that existed, see the letter of St Anselm attempting to dissuade a novice of Bec from joining his brother in the service of the emperor of Constantinople: ed. F.S. Schmitt, Sancti Anselmi Opera Omnia (Edinburgh, 1946), III, 253-4.

22. For what follows I am greatly indebted to the excellent article of J. Shepard, "The Uses of the Franks in Eleventh-Century Byzantium," Anglo-Norman Studies 15 (1993), 275-305.

23. Shepard (ibid., 289) estimates a minimum of 1500, based on references to three Frankish tagmata in the Byzantine army, two stationed at Koloneia in Asia Minor and one in Constantinople. However, this does not seem to take into account the Franks, numbering at least 500, who were stationed in the Armeniakon: John Skylitzes, Synopsis Historiarum, ed. I. Thum (Berlin-New York, 1973), 485.

24. Michael Attaleiates, Historia, ed. I. Bekker (Bonn, 1853), 188 (3000), 189 (2700). The reference is to the Franks in the army of the rebel commander Roussel of Bailleul when he was encamped opposite Constantinople in 1074. Many other Franks must have remained in the eastern frontier regions, notably the 800 in the service of the Armenian Philaretos: Matthew of Edessa, 11.60: tr. A. Dostourian, Armenia and the Crusades, 10th to 12th Centuries. The Chronicle of Matthew of Edessa (Lanham, MD-London, 1993), 138. Cf. Shepard, "Uses of the Franks," 301, 303-4.

25. E.g. the Constantine Humbertopoulos who helped Alexios Komnenos seize power in 1081 and later conspired against him: Alexiad, II.4.7; IV.4.3; VI. 14.4; VIII.5.5, 7.1; X.2.6: ed. Leib, I, 74, 152; II, 83, 141, 146-7, 193.

26. Skylitzes, ed. Thum, 485: Herve had the title of vestes, and rebelled against Michael VI when the latter denied him the title of magistros, which he did however receive from a later emperor: Shepard, "Uses of the Franks," 296-7.

27. Attaleiates, ed. Bekker, 122-5, 198-9; Shepard, "Uses of the Franks," 287-8, 298-301.

28. I have made some further observations and suggestions in "The Byzantine Army and the Land: from Stratiotikon Ktema to Military Pronoia," in N. Oikonomides (ed.), Byzantium at War.

29. Attaleiates, ed. Bekker,187; Nikephoros Bryennios, Historiarum Libri IV, ed. and tr. P. Gautier (Brussels, 1975), 182-94; Alexiad, 1.1-3 (ed. Leib, 1, 10-16); Shepard, "Uses of the Franks," 300-302.

30. See H. Bibicou, "Une page d'histoire diplomatique de Byzance au XP siecle: Robert Guiscard et la pension des dignitaires," Byzantion 29-30 (1959-60), 43-75; V. von Falkenhausen, "Olympias, eine normannische Prinzessin in Konstantinopel," Bisanzio e l’Italia. Raccolta di studi in memoria di Agostino Pertusi (Milan, 1982), 56-72; W. B.. McQueen, "Relations between the Normans and Byzantium 1071-1112," Byzantion 56 (1986), 428-37; R. Macrides, "Dynastic marriages and political kinship," in J. Shepard and S. Franklin (eds.), Byzantine Diplomacy (Aldershot, 1992), 270-1, 2734, 277-8.

31. This is implied by Amatus of Montecassino, Histoire de li Normant, VII.26.

32. Skylitzes Continuatus, ed. E. Tsolakis, (Thessaloniki, 1968), 167-70 esp. 170.

33. Michael VII’s first letter to Guiscard, drafted by Michael Psellos, contains an allusion to this effect: ed. K. Sathas, (Venice, 1876), 387.

34. I .e. the revolt of Leo Tomikes: John Mauropous, ed. P. de Lagarde and J. Bollig, Iohannis Euehaitorum metropolitae quae in Codiee vaticano Graeco 676 supersunt, Abhandlungen der historisch-philologischen Classe der koniglichen Gesellschaft der Wissenschaften zu Gottingen, 28 (Gottingen, 1882), 192; cf. Shepard, "Uses of the Franks," 285-6. The Byzantine armed forces already included some Normans who had served under the rebel commander George Maniakes, and had passed into imperial service after his defeat in 1043: Skylitzes, ed. Thum, 484; Skylitzes Continuatus, ed. Tsolakis, 167. In 1051 Constantine IX also attempted to recruit Normans through his agent in southern Italy, the Italian Argyrus: William of Apulia, Gesta Roberti Wiscardi, 1154-60: ed. M. Mathieu, La geste de Robert Guiscard (Palermo, 1961), 134-5.

35. Ibid.; William of Poitier, Gesta Gulielmi, 1.59, 11.32: ed. R. Foreville, Histoire de Guillaume le Conquerant (Paris, 1952), 144-5, 228.

36. Attaleiates, ed. Bekker, 46-7; Matthew of Edessa, tr. Dostourian, 99-101, 107-8, 138.

37. See in general M. Favreau-Lilie, Die Italiener im Heiligen Land von; ersten Kreuzzug bis zum Tode Heinrichs von Champagne (Amsterdam, 1989).

38. R. Morozzo della Rocca and A. Lombardo, Documents del commercio veneziano nei secoli XI-XIII, 2 vols. (Turin, 1940), I, no. 24, p. 27 (July 1095): "dedi in taxegio de Antiochia in nave in qua nauclerus fuit Valentinus Christianus"; Skylitzes Continuatus, ed. Tsolakis, 165-6. The correspondence between Peter, Patriarch of Antioch and Dominic, Patriarch of Grado, in 1052-3, was no doubt facilitated, if not occasioned, by the Venetian merchants in Antioch: ed. C. Will, Acta et scrspta de controversies ecclesiae Graecae et Latinae (Leipzig-Marburg, 1861), nos. XVI-XVII, pp. 205-28.

39. France, Victory in the East, 209-20: "This great maritime endeavour, led and supported by the Byzantines, was one of the key factors which enabled the crusader army to survive the bitter nine-month siege of Antioch and to triumph over their enemies" (p. 220).

40. Codice diplomatsco della Repubblica di Genova, ed. Imperiale di sant'Angelo, I (Rome, 1936), nos. 7, 12, pp. I1-12, 17-18. The expression actually occurs in the second of these documents, which is Tancred's confirmation (22 November 1101) of Bohemond's original grant.

41. The bibliography on Amalfi and its overseas trade is vast, especially in proportion to the small amount of evidence. The most useful recent surveys are: M. Balard, "Amalfi et Byzance (Xe-XIIe siecles)," Travaux et Memoires 6 (1976), 85-95; U. Schwarz, Amalfi im fruhen Mittelalter (9.-11. Jahrhundert) Untersuchungen zurAmalfitaner Uberlieferung (Tiibingen, 1978); A.O. Citarella, "Merchants, Markets and Merchandise in Southern Italy in the High Middle Ages," Mercati a mercand nell'alto medioevo: Parea euroasiatica e 1 area mediterranea. Settimane di studio del Centro Italiano di studi sull'alto medioevo, 40 (Spoleto, 1993), 239-82; V. von Falkenhausen, "La chiesa amalfitana nei sui rapporti con 1'impero bizantino (X-XI secolo)," Rivista di studi bizantini a neoellenici, n.s. 30 (1993), 81-115. For the Amalfitan presence in Constantinople, see most recently P. Magdalino, Constantinople medievale. etudes sur L evolution des structures urbaines (Paris, 1996), 75-6, 85-8 and Appendix II.

42. On the latter see H.M. Willard, Abbot Desiderius and the Ties between Montecassino and Amalfi in the Eleventh Century, Miscellanea Cassinense 37 (Montecassino, 1973); H. Bloch, Monte Cassino in the Middle Ages, 3 vols. (Cambridge, MA, 1986).

43. Renouard, Les hommes d’affaires italiens du Moyen Age, 2nd ed. (Paris, 1968), 34-41.

44. See M.E. Frazer, "Church Doors and the Gates of Paradise: Byzantine Bronze Doors in Italy," Dumbarton Oaks Papers 27 (1973), 147-62; C. Mango, "Storia dell'arte," in La civilta bizantina dal IX all XI secolo, University degli Studi di Bari, Centro di Studi Bizantini, Corsi di studi II, 1977 (Bari, 1978), 249-51; Bloch, Monte Cassino, I, 139-66.

45. Gregorii VII Registrum 1.18, 46, 49, ed. E. Caspar, Monumenta Germaniae historica, epistulae selectae II, 1 (Berlin, 1920), 29-30, 69-71, 75-6. Cf. A. Tuilier, "Michel VII et le Pape Gregoire VII: Byzance et la r6forme gregorienne," Actes du XVe Congres International des etudes Byzantines, N (Athens, 1980), 350-64; H.E.J. Cowdrey, "Pope Gregory VIPs 'Crusading' Plans of 1074," in B.Z. Kedar, H.E.Mayer, R.C. Smart (eds.), Outremer. Studies in the History of the Crusading Kingdom of Jerusalem presented to Joshua Prawer (Jerusalem, 1982), 27-40 [repr. in idem, Popes, Monks and Crusaders]; idem, "The Gregorian Papacy, Byzantium and the First Crusade," in Howard Johnston (ed.), Byzantium and the West [=Byzantinische Forschungen, 13], 145-64; M. Kaplan, "La place du schisme de 1054 daps les relations entre Byzance, Rome et 1'Italie," Byzantinoslavica 54 (1993), 34-6.

46. For a recent historical overview, with reference to earlier bibliography, see K.B. Wolf, Making History. The Normans and Their Historians in Eleventh-Century Italy (Philadelphia, 1995), 10-45.

47. The main evidence for Gisulf s eastern ambitions comes from the laudatory poems of Alfanus of Salerno: see Bloch, Monte Cassino, 1, 93f.

48. This view informs even the most sophisticated recent treatments of Byzantine foreign policy, the articles of Jonathan Shepard cited above.

49. For this classic view of the eleventh century, see, notably, G. Ostrogorsky, History of the Byzantine State, tr. J. Hussey, 2nd ed. (Oxford, 1968), 320-50; Sp. Vryonis, "Byzantium: The Social Basis of Decline in the Eleventh Century," Greek, Roman and Byzantine Studies 2 (1959), 159-75. Neither the rehabilitation of the eleventh century by P. Lemerle, Cinq Etudes sur le XIe siecle byzantin (Paris, 1977), 251-312, nor the recent textbook on the period (Angold, The Byzantine Empire, 1025-1204;) challenged the thesis of the civilian-military dichotomy, which was reinforced by the influential studies of A.P. Kazhdan, Sotsialnyj sostav gospodsvujushchego klassa Yizantii XI-XII vv. (Moscow, 1974), and 1. Sevcenko, "Constantinople Viewed from the Eastern Provinces in the Middle Byzantine Period," Harvard Ukrainian Studies 3-4 (1979-80), 712-47 [repr. in idem, Ideology, Letters and Culture in the Byzantine World (London, 1982)]. The challenge has come, most recently, from the important work of J.-C. Cheynet, Pouvoir et contestations a Byzance (963-1210) (Paris, 1990); see also the same author's reassessment of the empire's military strength in the period: "La politique militaire byzantine de Basile II a Alexis Comnene," Zbornik Radova Vzzantolofkog Instituta 29-30 (1991), 61-73; "Les effectifs de 1'armee byzantine aux X°-X11` s.," Cahiers de civilisation medievale 38 (1995), 319-35. However, Cheynet's conclusions have yet to find their way into any general work of synthesis.

50. Lemerle, Cinq etudes, 300-302.

51. This development is the defining theme of the recent book by M. Whittow, The Making of Orthodox Byzantium, 600-1025 (London, 1996). For religious and ideological changes in the same period, see above all G. Dagron, in Histoire du Christianisme, IV (Paris, 1993), chapters 1-5.

52. There is no monograph on the second Byzantine occupation of Cilicia and Syria, though see G. Dagron, "Minorites ethniques et religieuses dans 1'Orient byzantin: 1'immigration syrienne," Travaux et Memoires 6 (1976), 177-216 [repr. in idem, La romanite chretienne en Orient (London, 1984)]. For Italy see V. von Falkenhausen, La dominazione bizantina nelltafa meridionale dal IX al XI secolo (Bari, 1978).

53. Cheynet, "Politique militaire," 65.

54. See G. Downey, A History of Antioch in Syria from Seleucus to the Arab Conquest (Princeton, 1961).

55. Exceptions are Dagron, "Imigration syrienne," 205, and Cheynet, Pouvoir et contestations, 402-3.

56. See, e.g., Leo the Deacon, Historia, ed. C.B. Hase (Bonn, 1828), 73, on the city at the time of the Byzantine conquest in 969, and Ibn Butlan, who visited Antioch in 1051: tr. G. Le Strange, Palestine under the Moslems. A Description of Syria and the Holy Land from A.D. 650 to 1500 (London, 1890; repr. Beirut, 1965), 370-75.

57. See V. Laurent, "La chronologie des gouverneurs d'Antioche sous la seconde domination byzantine," Melanges de l'Universite Saint-Joseph 38 (1962), 219-54.

58. Dagron, "Minorites ethniques," 179-98; D. Jacoby, "The Jews of Constantinople and their demographic hinterland," in C. Mango and G. Dagron (eds.), Constantinople and its Hinterland (Aldershot, 1995), 2234.

59. Skylitzes, ed. Thum, 387-8; William of Tyre, Chronicon, 1.6: ed. R.B.C. Huygens, I, Corpus Christianorum, Continuatio medievalis 63A (Turnholt, 1986), 113.

60.See Bruno of Segni in Acta Sanctorum, August, vol. 1, 236-7.

61. The benefactions of Michael and other Byzantine emperors to the abbey are discussed in full by Bloch, Monte Cassino, I, 40-71. "

62. See above all E. Petrucci, "Rapporti di Leone IX con Costantinopoli," Studi Medievali, 3rd series, 14 (1973), 733-831. Other useful recent discussions are: Angold, Byzantine Empire, 27ff; Bloch, Monte Cassino, I, 32-40; Dagron in Histoire du Christianisme, IV, 338-47; von Falkenhausen, "La Chiesa amalfitana," 104ff; Kaplan, "La place du schisme"; idem, "Le 'schisme' de 1054. Quelques e1ements de chronologie," Byzantinoslavica 56 (1995), 147-57.

63. Petrucci, "Rapporti," 769-71.

64. The letter is attested by the Pope's reply, and by Peter's mention of it in his letter to the Patriarch of Grado: ed. Will, Acta, 168-71, 227-8.

65. Ibid., 191.

66. Ibid., 168-71.

67. Ibid., 202, 217-18.

68. Ibid., 227-8.

69. Petrucci, "Rapporti," 771-96.

70. Will, Acta, 160, 167, 175, 189, 203.

71. Petrucci, "Rapporti," 791ff.

72. An imperial letter quoted in the synodal report of July 1054 refers to a vestarches who was Argyrus' son-in-law, and his son who had the title of vestes: Will, Acta, 167.

73. The Latin polemics against Keroularios refer to the emperor in glowing terms. In addition to the documents of Cardinal Humbert, published by Will, Acta, see the treatise of the Amalfitan Pantaleone, ed. A. Michel, Amalfi und Jerusalem im griechischen Kirchenstreit (1054-1090), Orientalia Christiana Analecta 121 (Rome, 1939), 52. Constantine Monomachos also receives favourable mention in the preface and epilogue of the Neapolitan manuscript version of the Latin translation of Barlaam and Ioasaph, which was done by a Latin, possibly in Argyrus' entourage, resident in Constantinople in 1048-9: P. Peeters, "La premiere traduction latine de 'Barlaam et Joasaph' et son original grec," Analecta Bollandiana 49 (1931), 276-312, esp. 277-8; R.L. Wolff, "Barlaam and Ioasaph," Harvard Theological Review 32 (1939), 131-9, esp. 137; cf. von Falkenhausen, "La Chiesa amalfitana," 100-101.

74. See above, n. 44.

75. Leo of Ostia, Chronicle of Montecassino, III.55: ed. H. Hoffmann, Die Chronik von Montecassino (Hanover, 1986), 436.

76. Alexiad, V.8-9 (ed. Leib, II, 32-40); J. Gouillard, "Le proces officiel de Jean 1'Italien: les actes et leurs sous-entendus," Travaux et Memoires 9 (1985), 133-69. Cf. the introductory chapter of P.P. Ioannou, Christliche Metaphysik in Byzanz: L Die Illuminationslehre des Michael Psellos and Johannes Italos (Ettal, 1956); and L. Clucas, The Trial of John Italos and the Crisis of Intellectual Values in Byzantium in the Eleventh Century (Munich, 1981).

77. Ed. Th. Zisis, (Thessaloniki, 1976), 65.

78. Ed. A.R. Littlewood, Michaelis Pselli oratorio minora (Leipzig, 1985), 70-71; cf. N.G. Wilson, Scholars of Byzantium (London, 1985), 155-6.

79. See Mas'udi, Les prairies d'or, tr. Barbier de Meynard and Pavet de Courteille, rev. Ch. Pellat, lI (Paris, 1965), 278; cf. A. Shboul, "Byzantium and the Arabs: the Image of the Byzantines as Mirrored in Arabic Literature," in E. and M. Jeffreys and A. Moffatt (eds) Byzantine Papers, Byzantina Australiensia I (Canberra, 1981), 57-8.

80. On the diverse geographical origins of Psellos' students, see his letter, ed. Sathas, V, 508; Wilson, Scholars, 164.

81. Fragments of his work are published in the Catalogus codicum astrologorum Graecorum, II, 132ff; V, 227-8. From the second of these passages it emerges that he was born in 1040 and was writing in 1079-80. He can thus be identified with the Eleutherios at the court of Alexios I, mentioned in Alexiad, VI.7.5: ed. Leib, II, 59.

82. Ed. A. Delatte, Anecdota Athenensia, II (Liege-Paris, 1939), 1-127; ed. B. Langkavel, Simenois Sethi syntagma de alimentorum facultatibus (Leipzig, 1868); cf. H. Hunger, Die hochsprachlicheprofane Literatur der Byzantiner (Munich, 1978), II, 241, 275, 308-9.

83. Ed. L.-O. Sjoberg, Stephanites und Ichnelates (Uppsala, 1962); cf. H.G. Beck, Geschichte der byzantinischen Volksliteratur (Munich, 1971), 41-5.

84. Alexiad, VI.7.1: ed. Leib, II, 57.

85. See the list of authors cited in his astronomical treatises, ed. Delatte, 127, and his citation of Proclus in his treatise on diet, ed. Langkavel, 118; cf. G. Faraggiana di Sarzana, "Una testimonianza bizantina finora ignorata sulla Filosofza Galdaica di Proclo," Prometheus 16 (1990), 279-83.

86. Ed. Langkavel, 18. Two translators of the Alexiad (Leib, II, 58; Sewter, 194) have taken Seth to be the same as the "Alexandrian" from Egypt among the astrologers at the court of Alexios I whom Anna Comnena goes on to describe (VI.7.4), and this identification is adopted by B. Skoulatos, Les personnages byzantins de l’Alexiade (Louvain, 1980), 279-80. Other translators have recognised that it does less violence to the Greek to identify the Alexandrian as a separate person: E. Dawes, The Alexiad of the Princess Anna Comnena (London, 1928), 149; Reinsch, Alexias, 207-9 and n. 75. But if the two are distinct, it is a curious coincidence (a) that Symeon Seth had visited Alexandria, where he observed a solar eclipse in 1058 (Delatte, Anecdota Athenensia, II, 53), and (b) that Alexios I sent the Alexandrian from Constantinople to live in pensioned seclusion at Raidestos in Thrace, where the "house of Seth" passed into the possession of the Pantokrator monastery: ed. P. Gautier, "Le Typikon du Christ Sauveur Pantokrator," Revue des etudes Byzantines 32 (1974), 114-15. That the Seth in question was Symeon Seth seems likely in view of the fact that the latter sold some books to the monastery of Christ Panoiktirmon, which had a poor-house at Raidestos: ed. P. Gautier, "La Diataxis de Michel Attaliate," Revue des Etudes Byzantines 39 (1981), 126-7.

87. See D.R. Reinsch, "Auslander and Byzantiner im werk der Anna Komnene," Rechtshistorisches Journal 8 (1989), 257-74, and A.E. Laiou, "The Foreigner and the Stranger in 12th-century Byzantium: Means of Propitiation and Acculturation," in M.Th. Fogen (ed.), Fremde der Gesellschaft (Frankfurt, 1991), 71-97.

88. See, e.g., Chronographia, ed. E. Renauld, II (Paris, 1928), 25-6, and Orationes forenses et acta, ed. G.T. Dennis (Leipzig, 1994), 97 - a dismissive comment on astrologers and soothsayers who owe their reputation not to their technical competence but to their ethnic origin; cf. G. Dagron, "Former et fonctions de pluralisme linguistique a Byzance (IX-XIIe siecle)," Travaux et Memoires 12 (1994), 237. Cf. in general P. Magdalino, "Byzantine Snobbery," The Byzantine Aristocracy, IX to XIII Centuries, ed. M. Angold, British Archaeological Reports, International Series 221 (Oxford, 1984), 58-78 [repr. in idem, Tradition and Transformation in Medieval Byzantium (Aldershot, 1991)]; idem, "Hellenism and Nationalism in Byzantium," in Tradition and Transformation, no. XIV.

89. Dagron, "Minoritbs ethniques,"passim, esp. 213-4; Angold, Church and Society, 27-35, 50-54.

90. Skylitzes Continuatus, ed. Tsolakis, 176.

91. Ed. B. Wassiliewsky and V. Jemstedt (St Petersburg, 1896; repr. Amsterdam, 1965); ed. G. Litavrin, Sovety i rasskazy Kekavmena (Moscow, 1972). Cf. P. Magdalino, "Honour among Romaioi: the framework of social values in the world of Digenes Akrites and Kekaumenos," Byzantine and Modern Greek Studies 13 (1989), 183-218 [repr. in idem, Tradition and Transformation]; C. Galatariotou, "Open space/closed space: the perceived worlds of Kekaumenos and Digenes Akrites," in Mullett and Smythe (eds.), Alexios I Komnenos, 303-28.

92. Ed. P. Gautier, "Le typikon du sebaste Grbgoire Pakourianos," Revue des Etudes Byzantines 42 (1984), 5-145; cf. Lemerle, Cinq etudes, 115-91; Dagron, "Formes et fonctions du pluralisme linguistique," 233-4. On the monastery of Panagiou in Constantinople, which I have argued is identical with the monastery of St Mary of the Latins, see Magdalino, Constantinople medievale, appendix 2.

93. Sevcenko, "Constantinople Viewed from the Eastern Provinces".

94. Attaleiates, ed. Bekker, 180-83; Bryennios, ed. Gautier, 143, 203, 211-13.

95. Attaleiates, 201-4; cf. P. Magdalino, "The grain supply of Constantinople, ninth-twelfth centuries," in Mango and Dagron (eds.), Constantinople and its Hinterland, 39-44.

96. See A.P. Kazhdan and S. Franklin, Studies on Byzantine Literature of the Eleventh and Twelfth Centuries (Cambridge, 1984), 23-86.

97. Ed. P. Gautier, "La Diataxis de Michel Attaliate," Revue des Etudes Byzantines 39 (1981), 5-143, esp. 100-108; cf. Lemerle, Cinq etudes, 67-112.

98. Ed. P. and 1. Zepos, Jus Graecoromanum, VII (Athens, 1931), 411-97.

99. Attaleiates, ed. Bekker, 244-6.

100. Ibid., 96-7.

101. Ibid., 198.

102. Ibid., 9.

103. Ibid., 257; see above, n. 86.

104. Ibid., 110-11.

105. The expression is borrowed from R.H.C. Davis, The Normans and their Myth (London, 1976). See further Wolf, Making History.

106. Attaleiates, ed. Bekker, 35.

107. Ibid., 46-7.

108. Ibid., 107. The context is that of blaming the commander to whom Romanos N entrusted the Frankish troops for not having taken more decisive action against the enemy.

109. Ibid., 125.

110. Ibid., 171.

111. Ibid., 199.

112. Ibid., 206-7.

113. Ibid., 252-3.

114. De Administrando Imperio, ed. and tr. G. Moravcsik and R.J.H. Jenkins (Washington, D.C., 1967), 70-71.

115. Anna Comnena's comments on the Normans are well known. See also Skylitzes Continuatus, ed. Tsolakis, 135, on Crispin's revolt against Romanos IV; the author not only amplifies Attaleiates' short remark about the faithlessness of the Franks (ed. Bekker, 125), but suppresses Attaleiates' information that this was the reasoning of the emperor's misguided advisers.

116. Skylitzes Continuatus, ed. Tsolakis, 158; Bryennios, ed. Gautier, 189.

117. Lemerle, Cinq etudes, 207-12.

118. Angold, Byzantine Empire, 43; P. Speck, "Konstantinopel - ein Modell fur Bologna?," in P. Speck et al., Varia, III (Bonn, 1991), 307-48. See also W. Wanda-Conus, "L'ecole de droit et 1'enseignement du droit i Byzance au XP siecle: Xiphilin et Psellos," Travaux et Memoires 7 (1979), 1-103, at 103: "L'6cole de droit de Constantinople au XP siecle, c'est le terrain de la derni6re bataille livree - et perdue pour la sauvegarde des traditions du droit romain dans sa forme ancienne, et non pas une 'pdpiniere' de jeunes juristes preparant de brillantes carrieres de cadres de 1'Etat".

119. Ed. E. Kurtz and F. Drexl, Michaelis Pselli scripta minora, (Milan, 1936-41), I, 146-7; II, 143, 187; ed. G.T. Dennis, Michaelis Psellis orationes panegyricae (Stuttgart-Leipzig, 1994), 151; G. Weiss, Ostromische Beamte im Spiegel der Schriften des Michael Psellos (Munich, 1973), 279. Cf. P. Magdalino, "Die Jurisprudenz als Komponente der byzantinischen Gelehrtenkultur des 12. Jahrhunderts," in L. Burgmann, M.Th. Fogen, A. Schminck (eds.), Cupido Legum (Frankfurt, 1985), 176.

120. See M. Maas, John Lydus and the Roman Past (London, 1992), chapter 2; T. Honore, Tribonian (London, 1978), chapter 2.

121. Wolska-Conus, "L'ecole de droit," 97-101, considers Attaleiates an opponent of the academic legal tradition of the time, as represented by Michael Psellos and John Xiphilinos, the first nomophylax and teacher at the law-school. However, he refers to the foundation of the school in positive terms (ed. Bekker, 21). It seems to me that his sense of professional mystique, and admiration for ancient Rome, which he shared with Psellos, reflect the more intellectual approach to jurisprudence that was introduced by Constantine IX. If born as early as 1020 (cf. Gautier, "Diataxis," 12), he would have been past school age at the foundation of the school in 1047, though it is conceivable that he attended it as a mature student.

122. Ed. Gautier, "Diataxis," 31.

123. Ed. Zepos, Jus Graecoromanum, VII, 415-17; cf. Wolska-Conus, "L'ecole de droit," 97-101.

124. Ed. Bekker, 193-5, 217-20.

125. Note that the Normans who visited the imperial court were not all uneducated soldiers; they included Odo Stigand, an accomplished doctor who spoke several languages and spent three years at the court of Constantine X, where he was given the title of protospatharios: La chronique de Sainte-Barbe-en Auge, ed. R.-N. Sauvage (Caen, 1907), 57-8.

126. His enthusiasm for things Roman was shared by Michael Psellos, and by one John Xiphilinos, nephew of the homonymous patriarch, who produced a synopsis of the Roman History of Dio Cassius for the emperor Michael VII: ed. U.P. Boissevain, III (Berlin, 1891), 479-709, esp. 526; cf. M. Angold, "The Byzantine State on the Eve of the Battle of Manzikert," Byzantinische Forschungen 16 (1991), 24-7.

127. For a selection of recent views on this subject, see Mullett and Smythe, Alexios I Komnenos, passim, especially chapters 1 (Mullett), 6 (Karlin-Hayter), 7 (Magdalino), 18 (Angold).

128. I.e., Michael sebastos, son of Constantine: see P. Magdalino, "Justice and Finance in the Byzantine State, Ninth to Twelfth Century," in A.E. Laiou and D. Simon (eds.), Law and Society in Byzantium, Ninth-Twelfth Centuries (Washington, D.C., 1994), 93-115, esp. 108ff.

129. See in general, A. Becker, Papst Urban II. (1088-1099), Teil 2: Der Papst, die griechische Christenheit und der Kreuzzug, Schriften der Monumenta Germaniae Historica 19, II (Stuttgart, 1988).

130. Shepard, "Uses of the Franks," 302-5.

131. Bryennios, ed. Gautier, 195; Attaleiates, ed. Bekker, 207.

132. Alexiad, VI. 1.4, 8.1: ed. Leib, II, 43, 60; William of Apulia, V.387-90, ed. and tr. Mathieu, 256-7; Orderic Vitalis, Historia Ecclesiastica, VII.8: ed. and tr. M. Chibnall, N (Oxford, 1973), 38-9.

133. Shepard, "Uses of the Franks," 304-5; idem, "'Father' or 'Scorpion'," 105ff.

134. See his enthusiastic letter to his wife Adela, the daughter of William the Conqueror, informing her, inter alia, that her father's munificence was "almost nil" compared with the emperor's: ed. H. Hagenmeyer, Die Kreuzzugsbriefe aus den Jahren 1088-1100 (Innsbruck, 1901), 138-40.

135. Shepard, "Uses of the Franks," 305: "the Crusaders were for the most part pursuing goals which lay, in every sense, beyond him."

136. Anonymous, ed. Sathas, VII, 184-5.

137. P. Charanis, "Byzantium, the West and the Origin of the First Crusade," Byzantion 19 (1949), 27-9 [repr. in idem, Social, Economic and Political Life in the Byzantine Empire (London, 1973)].

138. Magdalino, Empire, 34.

139. Bryennios, ed. Gautier, 201ff.

140. France, Victory in the East, 302ff.

141. §6, ed. R. Hill (London, 1962; repr. Oxford, 1972, 1979), 12: "XV dies eundi terre in extensione ab Antiochia retro daret et VIII in latitudine"; cf. Shepard, "When Greek meets Greek," 219-27. If the grant was meant to exclude Antioch, it would have corresponded to the land which Bohemond's descendants were to receive by the terms of the treaty of Devol in 1108: Alexiad, XIII.12.24: ed. Leib, III, p.136.

142. Ed. J.H. and L.L. Hill, Le "Liber" de Raymond d Aguilers (Paris, 1969), 55, tr. eidem (Philadelphia, 1968), 37; cf. Shepard, "When Greek meets Greek," 195, 270-71.

143. Matthew of Edessa, tr. Dostourian, 101.

144. Patrologia Latina, vol. 150, col. 737.

145. Attaleiates, ed. Bekker, 252-3.

146. Kekaumenos, ed. Wassiliewsky-Jernstedt, 3ff, 76 (ed. Litavrin, 122ff, 298).

147. Alexiad, VI.7.6-7, XIII.10.4: ed. Leib, 11, 59-60; 111, 122-3.

148. P. Magdalino, "Aspects of Twelfth-Century Byzantine Kaiserkritik," Speculum 58 (1983), 326-46, esp. 342-5; idem, Empire, chapter 1 and pp. 360, 404-5; R. Macrides and P. Magdalino, "The Fourth Kingdom and the Rhetoric of Hellenism," in P. Magdalino (ed.), The Perception of the Past in Twelfth-Century Europe (London, 1992), 126-31.

This work was originally published by the Canadian Institutute of Balkan Studies as a pamphlet (Toronto, 1996). We thank Paul Magdalino for his permission to republish this item.