Post-classical history


The March to Constantinople

The polity of western Christendom comprised regions rather than kingdoms. Consequently, recruitment, politics, structure and command of the First Crusade were dominated by provincial lords, not kings. Writers on and of the expedition to Jerusalem took pains to identify different regional identities. Sigebert of Gembloux specified recruits from Provence, Aquitaine, Brittany, Scotland, England, Normandy, Francia (i.e. roughly, in this context, the area from the Loire to the Meuse), Lotharingia (i.e. greater Lorraine), Burgundy, Germania, Lombardy and Apulia. From his Lotharingian perspective, Albert of Aachen listed Franks, Lotharingians, Alemans, Bavarians, Flemings, ‘all the people of the Teutons’, Swabians, Normans, Burgundians and Bretons. From the south, Raymond of Aguilers distinguished between Franks, northern French, and Provençals, southern French, amongst whom he further separated those from Provence itself, Burgundy (probably the county east of the Saône/Rhône corridor, not the duchy), the Auvergne, Gascony and ‘Gothia’ (i.e. what might now be called Languedoc). Fulcher of Chartres described his companions as western Franks; Albert of Aachen mentioned East Franks. Raymond commented that the Muslims called them all Franks, clearly well informed of the Arabic catch-all for western European Christians, ‘al-ifranj’. The anonymous, possibly Normano-Italian author of the Gesta Francorum, who often used general terms such as ‘Christiani’, carefully differentiated those from Italy who joined Peter the Hermit at Constantinople as ‘Lombardi’, from the Po region, and ‘Longobardi’, his neighbours from the centre and south of the peninsula. The Gesta retains the older name ‘Gauls’ for the geographic France. The fiercely xenophobic Guibert of Nogent insisted on a fabricated nationalism, arguing that Urban II had specifically summoned the ‘Franks’, not the Germans, to protect Christendom from the Turks, a distortion of the events that rapidly gained favour with other ‘French’ writers such as Robert of Rheims and Baldric of Bourgeuil: thus were invented the Gesta Dei per Francos, the Deeds of God through the Franks, the title of Guibert’s admiring account, a national gloss that concealed the nature and structure of the expedition itself. So keen was Guibert on the Frankish monopoly on the Gesta Dei that he insisted that Bohemund – an Italian Norman – through his family’s origins and later marriage ‘might very well be considered a Frank’. Judged by their own letters, the members of the expedition called themselves ‘Christiani’, their clerics as ‘Latini’, in contrast to the local ‘Graeci’.1

Given the fame and aura of sanctity that surrounded the First Crusade, the Francophile gloss on the racial and regional diversity of the expedition played a part in the elevation and consolidation of a new sense of national identity apparent in twelfth- and thirteenth-century France, one exploited vigorously by the Capetian kings, not least in their own crusading ventures in 1147, 1190 and 1248. This nascent consciousness of unity encouraged by historians of the First Crusade such as Guibert of Nogent or Robert of Rheims was to contrast strongly with the older traditions of particularism maintained in Germany and Italy, whose actual experience of crusading differed little but lacked any specifically national dividend. The image of the French as dominating the crusades was not entirely misplaced: the majority of those we know as participants in 1096–9 came from lands between the Rhine and the Atlantic, the English Channel and the Mediterranean. However, to equate the ‘Franks’ with the French ignores their wide differences of language, law, landholding, history, tradition and culture as well as the contributions of other regions, from Denmark to Apulia, and England to Austria.

Shared objectives and shared perils created the cohesion of the First Crusade. After receiving a serious fright in the first field battle with the Turks in July 1097, the expedition’s military decisions were scrutinized by a common council; Adhemar of Le Puy enforced a chairman’s control. At Antioch, a common fund was created to fund expensive capital projects such as a siege tower and, briefly, a commander-in-chief was appointed, Stephen of Blois, who promptly ran away. For battle, leadership was agreed beforehand. Some factions were suspicious of the Provençal monopoly on helpful visions and the discovery of the Holy Lance at Antioch, and Raymond of Toulouse remained an isolated figure, perhaps because he spoke langue d’oc (southern French) not, like the rest of the high command, versions of langue d’oil (northern French). Even at Jerusalem, the princes kept a certain distance from each other, preserving their autonomy.2

This reality of frequent regional and ethnic tensions reflected the basic structure of the expedition which revolved around those lords and knights with sufficient means to support an entourage. Initially such groups mirrored the local circumstances of recruitment and travel. At least seven different currencies circulated in the army, perhaps even as late as May 1099.3 As medieval armies shared many characteristics of moving markets, such distinctions cannot have made transactions easier, exacerbated as they already were in Syria by a silver-based coinage operating within a gold currency area. Different contingents voiced different battle cries. Nevertheless, as funds ran out and leaders died, deserted or opted out of the main campaign, patronage became fluid, and not only for well-connected opportunists such as Tancred, who bartered his services between his uncle, Bohemund, Raymond of Toulouse and Godfrey of Bouillon. In the surviving rump before the walls of Jerusalem, allegiance followed sustenance and the provision of horses rather than race or region. This had not been the case at the outset.


By the time Peter the Hermit entered Cologne on 12 April 1096, Easter Saturday, considerable numbers from northern and eastern France, Lorraine and the Rhineland were already mobilized. Walter lord of Boissy-sans-Avoir in the Ile de France was preparing to leave the city immediately after Easter; on 15 April he set out on the traditional pilgrim route up the Rhine and Neckar to Regensberg and down the Danube to Hungary and the Balkan routes to Constantinople. With him was an infantry force, mainly French, led by eight knights, seemingly an advance guard for Peter’s larger army of levies raised on his march from Berry through the Ile de France and Champagne to the Moselle and the Rhine. Already, Peter had attracted a smattering of French nobles and volunteers from the towns he had visited. Some have seen his force as more of a pilgrimage than a military operation, yet, apart from Walter Sans Avoir, Peter established a military command under Godfrey Burel of Etampes, Reynald of Broyes from Epernay, Walter Fitz Waleran of Breteuil in the Beauvaisis and Fulcher, brother of the vidame of Chartres.4 His concentration on the urban centres of Lorraine and the Rhineland was not fortuitous. Arriving at Trier on the Moselle in early April, Peter bullied the local Jewish community into supplying provisions by showing a letter from French Jews urging them to accede to his demands: news of threatened or actual violence against northern French Jewish communities had probably already filtered through. From Trier, Peter headed north, down the Moselle, to Cologne on the Rhine, probably as much in search of funds and supplies as of men. Cologne possessed a large Jewish community, which at about this time was being blackmailed into subsidizing Godfrey of Bouillon’s expedition. A major commercial centre, although hardly on a direct route from Trier to the Danube and Constantinople, the city provided a convenient muster point for Lorraine recruits, including some German knights.

Peter’s movements displayed deliberation and control. He may have threatened the Jews of Trier and elsewhere by anti-Jewish preaching, but his forces abstained from organized attacks on them, unlike the troops of the armies collected in his wake and scavenging local citizens. Even if, as hostile commentators maintained, his followers were ‘the leftover dregs of the Franks’ with children in tow who, ‘whenever they came upon a castle or city, asked whether this was Jerusalem’, Peter, a small but charismatic figure, unafraid and competent later to negotiate in person with the Byzantine emperor and the atabeg of Mosul, displayed neither ignorance nor naivety.5 He was either well briefed or able to extemporize with skill in delegating further recruitment to the priest Gottschalk. He, in turn, raised an effective and well-funded force, 15,000 strong according to Albert of Aachen, with as many knights as infantry, sufficiently impressive and organized for King Coloman I of Hungary to negotiate a truce and the surrender of their arms, providing him with the opportunity, eagerly embraced, to massacre them at Pannonhalma in early July.

Peter may also have encouraged Count Emich of Flonheim, whose followers began killing Jews in Speyer on 3 May, although his army travelled north, down the Rhine, while Peter’s, days earlier, had passed in the opposite direction. Emich’s muster with significant contingents from northern France occurred at Mainz in late May, by which time Peter was far down the Danube. However, the obscurity of the gathering of Emich’s south German and French force suggests local recruitment. Peter, Gottschalk or Urban may have had a focusing effect; so too did local interest, traditions and contacts. As a child, Guibert of Nogent had known one of the knights later killed at Antioch, Matthew from the Beauvaisis, who had served the Byzantine emperor.6 His example may have exerted as much influence as Peter’s evangelism, garbled accounts of Urban II’s call to arms or rumours of a millennial holy war.

The orderliness of Peter’s forces stands in contrast with what followed. In mid-May, his lieutenant Walter Sans Avoir, marching only days ahead of him, negotiated a safe-conduct with the new Hungarian King Coloman, including access to markets, an important privilege as the early summer, before the harvest, were the hungriest months in the middle ages. There was trouble at Semlin on the Hungarian border over purchases of arms. Once they were across the Byzantine frontier, the hazards of early summer campaigns were exposed, Walter being refused market facilities at Belgrade, causing an affray in which sixty pilgrims died. However, the Byzantine military authorities recognized Walter as an ally and, to prevent further pillaging, provided food and an escort to Constantinople, which he reached about 20 July 1096 to await Peter. It says much for Alexius’s involvement in the project that he was so accommodating, not least as he must have been expecting the westerners to arrive some months later, when local provisions would have been more plentiful.

The speed in conveying Water Sans Avoir to the imperial capital shows the Greeks knew that Peter the Hermit’s larger force was only days behind, presenting a potentially dangerous competition for food. Although his regime rested on recent military success against the Pechenegs in the Balkans and some moderate successes in Asia Minor and the Aegean, Alexius I had witnessed too many political coups, one of them his own in 1081, to feel entirely secure. In 1094–5 there was a Balkan invasion across the Danube by Cumans, trouble in Serbia (directly on the crusaders’ line of march), stirrings of a tax revolt and a dangerous conspiracy in the army to replace Alexius by Nicephoras Diogenes, son of the Emperor Romanus IV (1068–71), the loser at Manzikert. Pressure on food in the strategically vital Balkan provinces and, still more, in the capital itself, could erode Alexius’s precarious support.7 Alexius needed western aid but could not allow it to disrupt his delicate political arrangements. A hungry, resentful population in Constantinople would have been very dangerous. Alexius determined to push the crusaders into Asia as quickly as possible to minimize the risk. It was less, as his daughter Anna Comnena claimed half a century later, that the emperor feared a western attack, more that he was wary of food riots or dissident Greeks recruiting the foreigners to overthrow him. From the first Alexius attempted to control his unexpectedly numerous allies through a mixture of hospitality, generosity and firm direction, careful always not to commit too many of his own stretched resources to their cause.

Peter the Hermit’s army left Cologne on 20 April. It was large, perhaps as many as 20,000 including non-combatants; the line of march in the Balkans was at least a mile long. Its passage through central Europe was rapid, averaging over seventeen miles per day, with twenty-five miles on good roads.8 Most of the pilgrims walked or rode, Peter apparently on his talismanic donkey, although some travelled down the Danube by boat. At Regensberg on 23 May, Peter’s followers orchestrated a mass forced baptism of the city’s Jews in the Danube. Unsurprisingly in view of the expedition’s propaganda, crusaders adopted a belligerent attitude to any who stood in their way, physically or ideologically. This emerged starkly when Peter’s army sacked Semlin in the second week of June after concerted assaults led by heavily armed knights and Godfrey Burel’s infantry. Again, the trouble arose from disputes over supplies – apparently rumours of the ill-treatment of Walter’s followers and an argument over the purchase of a pair of shoes sparked a riot that led to armed intervention – and anxiety over the prospects of help across the frontier in Byzantium. Although capable of storming a city and accompanied by carts full of treasure, under pressure Peter’s army lacked discipline.

The Semlin affair put the Greeks on their guard, evacuating Belgrade, leaving it open to plunder. After a forced crossing of the river Save, the pilgrims reached Nish, the provincial capital, on 27 June, where the crisis of supplies became critical. The Byzantine governor Nicetas negotiated a market for Peter’s men in return for hostages, significantly including the military commanders Godfrey Burel and Walter Fitz Waleran. When this broke down, Nicetas imposed order by force; after a failed attempt to restore peace by Peter, his forces were scattered by a concerted Greek assault. Chastened, Peter led the survivors along the road to Sofia; at the evacuated town of Bela Palanka they regrouped and gathered the local harvest. At Sofia, on 7 July, Peter was met by an escort fromAlexius that hurried them towards Constantinople, making sure they never stopped anywhere for more than three days. The battles at Nish, which cost perhaps as much as a third of his force, had been caused by Peter and his commanders losing control, particularly, Albert of Aachen recorded, of the young men.9 Communications along the line broke down, a sign of inexperienced leadership faced with such a large and disparate force, lacking the cohesion exerted by wealthy magnates. Exhausting marches; uncertain food supplies; alien territory and people; discomfort, fear and the prospect of hunger soured idealism. Yet, once chaperoned by the Greeks and provided with secure provisions, Peter’s army regained its integrity; Adrianople was reached by 22 July and Constantinople on 1 August, just five months after Peter’s first rallying of pilgrims in the Ile de France over a thousand miles behind.

The shambles in the Balkans served as a prelude to disaster. Alexius advised Peter against pressing forward immediately. Evidently abreast of events in the west, some princes and probably the pope having written to him of their plans, Alexius urged waiting for the arrival of the rest of forces being assembled. Reunited with Walter Sans Avoir and reinforced by some Italian levies, Peter was provided with a well-supplied base that Alexius used for western mercenaries at Kibotos, on the Gulf of Nicomedia just across the Sea of Marmora from the capital. There, the usual difficulty of countering boredom in an army camp was exacerbated by regional rivalries and the proximity of territory controlled by the Seljuk Turks, whose capital in Asia Minor was at Nicaea, only twenty-five miles away. With Peter now reduced to a diplomatic role in negotiating the level and cost of regular supplies with the Byzantine authorities in Constantinople, leadership devolved on to the separate captains in whose interest it was to engage in lucrative pillaging of the locality, regardless of whether the victims were Greek Christians or Muslims. The objectives were food, booty and action. It was a truism of medieval warfare that an armed force was never more vulnerable than when foraging. In September, French raiders penetrated to the walls of Nicaea. Not to be outdone, a contingent of Germans and Italians, under an Italian called Rainaldo, ranged further afield, seizing a castle at Xerigordo near Nicaea. There they were trapped and massacred by Seljuks from Nicaea, allegedly only those who surrendered and embraced Islam escaping to lead lives as captives and slaves, one of them being Rainaldo himself.

Disorderly conduct and confused leadership were not the sole prerogatives of these early crusade armies; a year later the princes fared little better during some of the darker days at the siege of Antioch. The populist nature of the whole enterprise now emerged, not for the last time, as a potent force in tactical decisions. Walter Sans Avoir and most of the other leaders at Kibotos argued against any precipitate response to the Xerigordo disaster, but popular demand for revenge found a spokesman in Godfrey Burel, the majority prevailing over the cautious leadership. The popular agitation provoked the main body of the crusaders to advance from Kibotos towards Nicaea. By now the Seljuk Sultan Kilij Arslan was sufficiently alarmed to take personal direction of his forces. In a series of fast-moving engagements on 21 October, a significant proportion of the Christian knights were isolated and killed, including Walter Sans Avoir, pierced, so Albert of Aachen recorded, by seven arrows, and Reynald of Broyes.10 With the elite of knights broken, the Christians were either massacred or fled, the Turks overrunning the camp at Kibotos three miles away within minutes. Only the arrival of a Byzantine relief force saved the remnants of the Christian army that had found refuge in a deserted castle on the shore; a large proportion of these would appear to have been knights.

Although not directly responsible for the catastrophe, Peter the Hermit’s role as a leader was at an end, his presence during the rest of the campaign receiving distinctly muted acknowledgement in the eyewitness accounts. Yet his contribution, ultimately insignificant militarily, demonstrated that the journey to the east was not a fool’s errand. His troops had held together as a viable force for months despite their difficulties with supplies, which were the result of timing as much as anything. He had accomplished a long march with thousands of ill-assorted followers, negotiated with local rulers and secured the patronage and favour of the Greek emperor. The tragic failure of his army in Asia pointed to the requirements for success: united leadership; significant numbers of knights; respect for the enemy; and, above all, adequate and secure supplies, of food, water, war materials and horses.

Peter the Hermit’s failure looked like modest success when compared with the fate of the other large crusader bands that set out from the Rhineland area in the spring of 1096. Gottschalk’s army was destroyed at the beginning of July by an exasperated King Coloman in western Hungary at about the same time as Volkmar’s force was dispersed at Nitra in the north after a career of Jewish persecution in Bohemia. The problem for the Hungarians was of order and supply. Each successive crusader army seemed less disciplined, more eager to plunder, commandeer markets and coerce locals. Beyond the scrutiny of chroniclers, a steady stream of ordinary pilgrims was flowing east, adding to the pressure on food stocks and forage. These material considerations dictated Coloman’s refusal in late July to allow the passage into his kingdom of Emich of Flonheim and his south and west German followers: with a more favourable supply position three months later, the king allowed Godfrey of Bouillon a negotiated passage. However, beyond provisions, Coloman may also have regarded Emich as a dangerous liability, his reputation for violence and flouting of royal authority preceding him. In the three months since embarking on his crusade, Count Emich had, in the eyes of many, indelibly stained the holy project by the systematic persecution of Jews.


The Jews of northern Europe shared in the economic growth of the eleventh century, especially in the revival of urban life. Attracted from the Mediterranean regions, Ashkenazic Jews became established in market towns of northern France such as Troyes or Le Mans by the late tenth century, as well as in the towns of the Rhineland. New communities continued to be established, such as in England after 1066 or in Speyer in 1084; older ones, such as those of Rouen, Cologne or Mainz, flourished under the protection of local rulers or bishops eager to promote trade. Jewish banking became a feature of the expanding markets of the area. As well as direct involvement in trading goods, with increased long-distance commerce and the persistence of varying local currencies, weights and measures, the network of Jewish financiers proved useful. Judging by Rhineland evidence, interest rates were not exorbitant, 8 per cent in one example, Jewish credit being certainly more accessible and in the long term cheaper than obtaining cash from another source of bullion, religious houses.11 With success came dangers. In northern France there had been sporadic outbreaks of anti-Semitic persecution allied to forced conversions, in particular in the years 1007–12.12 As holders of movable wealth, Jews were targets for casual as well as systematic larceny. As a religious minority, Jews remained tolerated if not accepted. A more consistent threat to their communities than persecution lay in conversions of successful and ambitious Jews to the majority faith, as occurred with sons of two famous Mainz rabbis. Privileged and protected status in the confined streets of eleventh-century towns presented its own problems: Bishop Rudiger’s charter establishing Jews at Speyer provided for a walled enclave to protect them from ‘the violence of the mob’.13 Such communal tensions played their part in the tragedy of 1096.

On 3 May 1096, the Jewish Sabbath, Count Emich’s troops attacked the Jews at Speyer, near to his estates, killing a dozen of them who refused baptism, before the bishop came to their rescue. One woman committed suicide rather than submit to the Christians. The persecutors received the help of townspeople, as Bishop John punished some of them by having their hands cut off, a penalty for theft. Those Jews who had fled to the surrounding countryside or had accepted baptism returned under the bishop’s protection, the apostates allowed to revert to Judaism; a new synagogue was begun. Steven Runciman rather astonishingly dismisses this episode as ‘not a very impressive attack’.14 Perhaps the walls prescribed in 1084 proved their use. Over a fortnight later, on 18 May, Emich arrived at Worms where he managed to mobilize more effective local assistance, including peasants from the countryside as well as burghers. Given the proximity of Emich’s own lands, the count was probably exploiting known local tensions. Jews found in their quarter were massacred, the Torah Scrolls desecrated; those who had fled to the protection of the bishop’s palace were besieged and, on 20 May, slaughtered. Some resisted forcible conversion, one of the bishop’s relatives being killed; others may have taken the route of suicide. Hundreds died.

The destruction of the Jews of Mainz attracted the most detailed attention, later held up to Jewish audiences as a model of fortitude under persecution and of holy martyrdom. Mainz was a major centre of Jewish learning and culture as well as business. Jewish leaders were prominent in commerce; the chief rabbi, Kalonymos, on good terms with the archbishop and recognized by the emperor. On Emich’s appearance before their gates, which the archbishop had ordered to be shut against him, some townspeople provoked riots. The Jewish leaders bribed the archbishop to protect them and tried to buy off Emich with a gift of seven pounds of gold, to no avail. The gates were opened on 26 May; the killing and looting lasted two days. The archbishop reneged on his promise of protection and fled; the Jews sheltering in his palace, despite initial vigorous armed resistance, were slaughtered with the rest. The search for money and Jews throughout the city was thorough. The synagogue was destroyed in the mayhem; some Jews apostatized; others chose suicide. The story of the young mother Rachel’s sacrifice of her four children, circulated for the edification of the faithful in the twelfth century, is grim. Her youngest, Aaron, terrified at seeing the deaths of his siblings, begged his mother to spare him, running away to hide under a box.

When this pious woman had completed sacrificing her three children to their Creator, she raised her voice and called to her son: ‘Aaron, Aaron, where are you? I will not spare you either, or have mercy on you.’ She drew him out by his feet from under the box where he had hidden and slaughtered him before the Exalted and Lofty God.15

Surrounded by the still-twitching corpses of her children, Rachel waited to be found by the Christians; before killing her, they demanded, ‘Show us the money you have in your sleeves’. Hers was not the only horrific death. Rabbi Kalonymos and fifty others escaped to seek asylum at the archbishop’s country retreat across the Rhine at Rudesheim. Archbishop Ruthard, pusillanimous and discreditable to the last, tried to exploit the rabbi’s predicament by offering protection only in return for conversion. Kalonymos, so furious at this self-seeking betrayal that he tried to assault the archbishop, was butchered with his companions. The amount of loot gained by Emich’s men and the local Christians is unknown; perhaps about a thousand Jews died.

By the time Emich reached Cologne on 29 May, lessons had been learnt, local Jews having dispersed across the countryside or sought shelter from friendly Christians in the city, hoping to avoid trouble during the following weekend and Whit Sunday (1 June). The synagogue was burnt and the Torah Scrolls desecrated, but casualties among the Jews were light, the quest for booty more obvious: a wealthy Jewish woman, Rebecca, was murdered when found trying to smuggle gold and silver to her husband in hiding with a Christian family.16 The Jews who had fled the city were soon being hunted down, attacks being recorded in Neuss, Wevelinghofen and elsewhere in the neighbourhood. With the best plums picked, Count Emich and his men turned south and east, along the Main towards the Danube and Hungary. Denied entry into Hungary at Wiesselberg in mid-July, Emich discovered that thuggery and bullying cut no ice against an organized armed enemy. Settling down to an elaborate siege, Emich and his French and Swabian allies showed tactical expertise and engineering skill in constructing pontoon bridges and siege-engines but, on a rumour of the approach of King Coloman, morale disintegrated. His men beginning to flee, Emich and his knights were unexpectedly worsted by a sortie from the Wiesselberg garrison, the count only escaping because of the speed of his mount. The army dissolved; the French nobles returned west to seek other routes and new leaders; Emich went home.

The Rhineland pogroms of May did not end with Emich’s departure. Whether pursued by other bands of crucesignati or by opportunist locals, the area around Cologne continued to suffer depredations for some weeks. In June attacks spread down the Moselle to Trier and Metz, where over twenty Jews died. It was high summer, and, if not necessarily as scorching as the dry year of 1095, tempers could have frayed as hunger increased; shortages for crusaders implied shortages for the locals; prices rose in the wake of the levying of armies. A lead had been provided by Count Emich’s butchers, with their sanctimonious, bloodied aprons of righteousness. In late June and July, further attacks occurred in the Cologne region and to the north, at Xanten, Mehr, Eller and Geldern. The descriptions of the assailants are vague. In places such as Mehr, neighbours played a key role in the Jews’ ordeal. By late summer, the outbursts of hate had died away, perhaps as the harvest came in. Followers of Godfrey of Bouillon, recruited from adjacent regions, whose leader had done his own blackmailing of the Jews of Mainz and Cologne, caused no trouble; perhaps they thought the Jews had nothing left worth plundering.

Emich of Flonheim’s campaign against the prosperous Jewish communities was deliberate, far from mindless vandalism. The rhetoric, not least that recorded in the harrowingly full Jewish accounts of the pogroms, was religious, but the motive may have been financial. It was not that the crusaders were in debt to the Jews, merely that many had sold or pledged their patrimonies and still faced further expense. For leaders such as Emich, cash meant power and authority. Locals, including some bishops, erstwhile protectors, exploited the crusaders’ greed by extorting protection money from the helpless Jews as well as looting. The Mainz community offered Emich money to spare them and delayed their own fate by throwing coins at their ravaging persecutors. Albert of Aachen drily commented that the pilgrims had slaughtered the Jews ‘more from avarice than for the justice of God’, a sin to which he attributed their later travails in the Balkans.17 The lust for money alone cannot explain the consistent flouting of canon law and religious teaching witnessed by the repeated forcible conversions. Nothing in official Christian doctrine justified slaying Jews. Pope Alexander II had explicitly prohibited it when drawing a careful distinction between them and Muslims in 1063. No justification of holy war could embrace victimization of those whom the Christians ruled anyway, hence the repeated attempts to blame the Jews of subversion and plotting the destruction of Christendom to excuse persecution. However, the preaching of the cross emphasized meritorious Christian violence, the legitimacy of revenge and religious vendetta and the suffering of Christ Crucified. Christian sources record how such messages were translated into a gospel of indiscriminate religious hate. Crusaders at Rouen thought it absurd to campaign against God’s enemies in the east ‘when in front of our eyes are the Jews, of all races the most hostile to God’. Albert of Aachen noted that recruits to Emich’s army at Mainz insisted that killing the Jews was the first act of their campaign against the ‘enemies of the Christian faith’. The Christian love Urban may have preached to justify his Jerusalem project was reserved for Christians; the obverse of this message of charity was intolerance and violence. According to a disapproving German witness, Ekkehard of Aura, the persecutors were zealous Christians who ‘took pains to destroy utterly the execrable Jews’ either by death or forced conversion.18 Not only were Jewish communities ransacked for money and goods, their synagogues, Torah Scrolls and cemeteries were repeatedly desecrated. Jews believed the incentive for their attackers was religious. In general, Jews were enemies of the church; in particular they killed Christ. When the gates of Mainz were opened to them, Emich’s followers were reported by one Hebrew source as exulting, ‘All this the Crucified has done for us, so that we might avenge his blood on the Jews.’19 All three Hebrew chronicle sources for the pogroms agree on the persistence of the theme of vengeance for the Crucifixion. Thus the mixture of demotic religious propaganda and material greed combined to create an obscene cocktail of butchery and bigotry.

Yet the anti-Jewish persecutions reflected more than mob violence and hysteria. The chronicle attributed to Solomon bar Simson, a Mainz Jew writing c.1140, recorded that those who set out for the Holy Land ‘decorated themselves prominently with their signs, placing a profane symbol – a horizontal line over a vertical one – on the vestments of every man and woman whose heart yearned to go on the stray path to the grave of their Messiah’. The butchers of the Mainz community are described as raging ‘in the name of the crucified one’ and carrying banners of the cross.20 Just as much as a shared campaign, a shared pogrom cements identity on a group. Crusaders possessed a special sense of identity; already by June 1097 one wrote of ‘the army of God’.21 In the early days of 1096, this uniqueness of purpose and community sought and found expression, warriors of the cross fighting for Christ. Theological niceties were irrelevant and, in any case, the clergy travelling with the crusaders may have encouraged the outrages while those in the towns affected were rarely able to sustain the orthodox line. The massacre of the Jews was just the first of many articulations of the corporate spirit of crusading. There also existed a local political dimension. Henry IV had explicitly and repeatedly forbidden the Jews to be harmed; they were under his protection. Emich’s attacks represented a challenge to Henry’s authority, an assertion of his independence, made easier by the emperor’s absence in Italy. The political dividends of the upheavel of 1096 may not have been confined to the papacy.

For the Jews, the Rhineland pogroms did not mark ‘the first holocaust’.22 There had been assaults before; neither did they mark the opening of a sustained campaign against the Jews. If now more wary and uneasy, the Ashkenazic Jewish communities of the Rhineland and northern Europe survived and thrived for generations, despite further atrocities attendant on the Second and Third Crusades. Jews continued to migrate into the areas of persecution. More conductive to intolerance was the growing exclusivity and general militancy of the western church. With the battles with Islam, in Spain and the east; the conversion of the Baltic; the elaboration of canon law; and the war against heresy, the persistence of a religious minority appeared more anomalous and, to some, more offensive. 1096 was only one part of this process. Ironically, its impact on Ashkenazic memory was testimony to its lack of profound material effect on the victimized communities. It was to renewed and expanding Jewish congregations in the Rhineland itself that the image of the martyrs of 1096 spoke most eloquently, as in the liturgical prayer first mentioned by Ephraim, a twelfth-century rabbi in Bonn: ‘May the Merciful Father, who dwells in heaven, in his abundant mercies remember compassionately the pious and righteous and pure, the sacred communities who sacrificed themselves for the sanctification of the Divine Name’.23


The failures and excesses of the bands travelling east in the summer of 1096 attracted scorn, contempt and ridicule but hardly impinged on the project’s popularity. By the date fixed by Urban II for departure, 15 August 1096, all three German expeditions had collapsed; Peter the Hermit’s troops were perched precariously in their base on the rim of western Asia, soon to be annihilated; none of the princes of the west had embarked. Yet Urban II had not yet returned to Italy and recruitment was gathering momentum across western Europe. Before stopped by their local bishop, the abbot and monks of Cerne in Dorset had invested thirty shillings in a ship to take them to Jerusalem. At the same time, the pope voiced anxieties about indiscriminate enlistment, especially by clergy and young husbands with itchy feet.24 Frustrated veterans of the first armies sought new comrades. By the end of the year, by land and sea, by boat, horse, wagon and on foot, perhaps another 50–60,000 had embarked for the east, casting the earlier efforts into a deep shade.

For each crucesignatus and those left behind, departure was a solemn moment. While most hoped to return, none could guarantee it. As he travelled south to rendezvous with the duke of Normandy and count of Blois in September 1096, Count Robert of Flanders was received at a monastery near Rheims by a procession of monks; there too a local magnate came to pay his respects.25 Most adieux lacked such grand ceremony, although many would have been witnessed by parish priest and villagers and accompanied by ritualized as well as genuine grief. Fulcher of Chartres, chaplain on the march to Count Robert’s companion Stephen of Blois, provided an imaginative, yet universal description:

What sighs, what weeping, what lamentation among friends when husband left wife so dear to him, his children, his possessions however great, his father, mother, brother and other relatives. But however many tears those remaining shed for departing friends and in their presence, none flinched from going… Then husband told wife the time he expected to return, assuring her that if by God’s grace he survived he would come back home to her. He commended her to the Lord, kissed her lingeringly, and promised her as she wept that he would return. She, though, fearing that she would never see him again, could not stand but swooned to the ground, mourning her loved one, whom she was losing in this life as if he were already dead. He, however, like one who has no pity – although he had – and as if he were not moved by the tears of his wife nor the grief of any of his friends – yet secretly moved in his heart – departed with firm resolution. Sadness was the lot of those who remained, elation, of those who departed.26

The first great western lord to set out for Jerusalem, somewhat paradoxically, was the brother of the king Urban II had excommunicated at Clermont. Hugh count of Vermandois was the younger brother of Philip I, the Fat. Distinguished only by blood, Hugh acted as a magnet for some of his brother’s leading vassals, including the king’s constable (Walo of Chaumont-en-Vexin) and seneschal (Gilbert of Garlande). The Ile de France was well represented in Hugh’s entourage, including later William the Carpenter of Melun, Thomas of Marle and Drogo of Nesle. Capetian interest was not entirely ideological. Participation in the expedition was agreed during a council at Paris in February 1096; in July, Hugh’s participation was announced to the pope by King Philip with his own submission to Urban’s judgement over his adulterous marriage (to the wife of the count of Anjou, to whom Urban had presented a golden rose during his preaching tour in March). Thus Urban’s Jerusalem scheme produced immediate and direct political gains for the wider papal cause by allowing Philip to be reconciled without losing too much face. The settlement suited both sides, Hugh receiving a papal banner to carry on his pilgrimage. The numerous recruits from the Paris region indicate another political benefit, this time for the Capetians, by providing a rare opportunity to exhibit practical leadership over their unruly vassals of the Ile de France, although Hugh hardly proved a dominant figure.

His journey was carefully planned; before leaving, probably in late August, he wrote to Alexius I, informing him of his intended itinerary.27 This took him through Italy, where he may have received the papal banner and blessing, to Bari. By this time Hugh’s small contingent of knights had been swelled by the French lords from Emich of Flonheim’s misadventure led by William of Melun. In southern Italy, his party was joined by one of Bohemund’s nephews, William FitzMarquis, and others, including veterans of Byzantine service.28 Crossing the Adriatic in October, after the indignity of a shipwreck, Hugh was held under comfortable house-arrest in Durazzo by the nonetheless hospitable Greek authorities before being escorted under close guard to Constantinople. Alexius seemed concerned lest Hugh linked up with the large numbers of Italians following the same route along the Via Egnatia from Durrazzo to the capital; or he may have received warning that his old enemy Bohemund was only a fortnight behind the count. Hugh was welcomed at Constantinople in November, only a few weeks after the massacre at Kibotos. Alexius’s treatment of Hugh betrayed nervousness; although well entertained and apparently rather embarrassingly easily flattered by the emperor’s attention, the count’s movements were monitored and some of his followers kept under close arrest. The emperor was beginning to appreciate the scale of his problems. Almost every day, news came of more western grandees bearing down on him while the flow of lesser pilgrims became a flood, swelled by the bumper harvest experienced in the west in the autumn of 1096. Miraculous would not necessarily have been Alexius’s word for it.

Shortly before Christmas 1096, Godfrey of Bouillon, duke of Lower Lorraine, arrived at the Greek capital with a substantial army derived mainly from Lotharingia (Lorraine) and the Low Countries. He proved an awkward guest. His march through central Europe followed the pilgrims’ road which had carried Peter the Hermit’s armies some months earlier. In contrast to his predecessor, Godfrey’s diplomacy worked all the way, a sign of meticulous preparation. Far from the selfless hero of chivalric legend he later appeared, Godfrey struck a number of hard bargains to raise funds for his expedition. Apart from extorting money from Rhineland Jews, he sold some estates; Bouillon itself he mortgaged to the bishop of Liège, with a proviso of restitution if he returned. Although unmarried, perhaps from sexual preference, Godfrey did not regard the expedition to Jerusalem as an excuse for abandoning his status in the west. The younger brother of the wealthy Count Eustace III of Boulogne, Godfrey’s career had flourished as a partisan of Henry IV. Succeeding to the disputed duchy of Lower Lorraine as a teenager in 1076, Godfrey fought for Henry in Italy in 1083. In 1087, his rights as duke were confirmed by a grateful emperor: his army in 1096 attracted many imperialists from the diocese of Liège.29Although before his departure he had minted coins inscribed ‘Godefridus Ierosolimitanus’, and despite apparent political ineffectiveness, he never relinquished his duchy even after becoming ruler of the Christian enclave in Palestine in 1099. With him were two future kings of Jerusalem, his ambitious opportunist younger brother Baldwin and his cousin Baldwin lord of Le Bourcq; the counts of Toul and Hainault; other relatives such as Henry and Godfrey of Esch; and perhaps over 100 other knights. He was later joined by survivors from Peter’s army, such as Fulcher, brother of the vidame of Chartres. One of his strengths on crusade, and as ruler of Jerusalem, lay in the loyalty of his sizeable military household.

Godfrey’s march was prolonged but not turbulent. Leaving Lorraine in August, he negotiated a peaceful crossing of Hungary and access to markets with King Coloman who insisted, as he had with Peter the Hermit, on the security of grand hostages, in this case an extremely reluctant Baldwin of Boulogne and his Anglo-Norman heiress wife, Godehilde of Tosni. Godfrey’s chief spokesman had been Godfrey of Esch, a veteran of earlier diplomacy with the Hungarians, another indication of the scale, depth and complexity of the political as well as material preparations. Reaching the Byzantine frontier in early November, Godfrey quickly struck a deal with the Greek authorities over provisions, promising not to engage in violent foraging in return for secure food supplies, the Byzantines having prepared large food dumps along the route. After a leisurely escorted progress, by the time he reached Adrianople, Godfrey, learning of the treatment of Hugh of Vermandois, become alarmed lest he was walking into a gilded trap. Given his minor role in western European politics, the duke’s pride and self-importance unexpectedly came to the fore as he insisted Alexius release the Frenchmen. As later admirers and perhaps he himself liked to recall, a descendant of Charlemagne, whose mythologized exploits furnished an important corner of the mental world of aristocratic crusaders,30 Godfrey behaved as if he were the emperor’s equal, not a policy designed to endear him to Alexius. Perhaps Godfrey saw himself in some way as representing his lord, the western emperor Henry IV; certainly the chronicler of Godfrey’s campaign, Albert of Aachen, placed the German king at the head of his list of rulers in 1096, above the pope.31 Godfrey’s objections to Alexius’s handling of Count Hugh spilt over into violence, as Alexius cut off aid and the Lorrainers began to pillage the neighbourhood of Salabria, between Adrianople and the Sea of Marmora. Only by sending an embassy of Franks in imperial service to reassure the duke of his reception did hostilities cease, but it was a somewhat prickly Godfrey who arrived at Constantinople on 23 December 1096. Stationed on the Golden Horn, then at Pera opposite the city, for weeks Godfrey resisted Alexius’s attempts, conveyed by Hugh of Vermandois and others, to arrange a meeting. Alexius again withdrew food supplies, forcing Godfrey into an abortive assault on the city (13 January 1097) and further ravaging until diplomacy prevailed. Hostages were exchanged (including Alexius’s son and eventual successor, John) before Godfrey attended an audience with the emperor. The outcome was satisfactory to all concerned. Godfrey swore an oath to the emperor, of vassalage according to Albert of Aachen.32 Alexius became his patron and helped ship his army across the Bosporus by the end of February 1097. For the Greek emperor, the presence of such a large army, even if peaceful, had presented serious logistical and political problems. Godfrey’s initial refusal to reach some accommodation with Alexius or to move forward across the Bosporus to Asia presented dangers as the winter progressed and the capital and its suburbs had to absorb increasing numbers of pilgrims. Both Alexius and Godfrey were exercised by the imminent arrival of the other major commanders of the expedition; the one fearful of the implications to his capital’s food supplies and security; the other eager to consult with his peers as to how best to proceed. Around 20 January 1097 Godfrey apparently received an embassy from Bohemund, then making very slow but careful progress from the Adriatic coast, suggesting a combined attack on the capital. Despite his stand-off, Godfrey rejected Bohemund’s plan; later veterans of his army spoke about the Greeks without hostility or malice.33 At a popular level, relations remained good; equally, Godfrey had not resisted manipulation by Alexius only to become a pawn in Bohemund’s deep-rooted schemes concerning the Greek empire.

Bohemund of Taranto is the most controversial leader of the First Crusade. Of all the major surviving commanders, he alone failed to join the march to Jerusalem in 1099, more concerned with securing his hold over Syrian Antioch. Admired for his generalship, his pious credentials have been impugned in the light of his priorities in 1099 and his career of attempting to carve out for himself a kingdom in the Balkans at the expense of the Byzantine empire. The traditional view sees his motives as basely material, in contrast to the supposedly more elevated inspirations of some of his colleagues. This is untenable. The psychologies of the crusade’s leaders cannot be reconstructed. Each can be shown to have as much avarice or as little piety as the other. The dichotomy between spiritual and mercenary possesses little meaning. Raymond of Toulouse, whose religious sincerity has been widely accepted, proved both scheming and petulant in his earnest quest for an eastern principality, which he finally achieved in the lands around Tripoli in the south Lebanon. The spiritual agonizing of Tancred of Lecce, Bohemund’s nephew, was matched by his alert political opportunism. Godfrey of Bouillon accepted power and lands when offered them in 1099. Baldwin of Boulogne, the most obviously careerist of all, devoted the last twenty years of his life to defending the Holy Places. All the leaders sought to protect their material interests rather than proceed to Jerusalem in the five months after July 1098. Bohemund was not alone in his desire to achieve status, lands and wealth; neither did this ambition automatically contradict the genuineness of his adherence to the cause of Jerusalem. With Baldwin, he undertook a tricky and dangerous journey to fulfil his pilgrimage to the Holy Sepulchre at Christmas 1099, a gesture that, for lack of evidence, cannot be assumed to have been purely for reasons of image or politics.

The picture of Bohemund the ruthless schemer derives from the Alexiad of Anna Comnena, Alexius’s daughter.34 Writing half a century after the First Crusade, Anna insists on the deviousness of the westerners, their consistent desire to subvert and occupy the Byzantine empire, and the heroic patience and skill of Alexius, in an attempt to exonerate the emperor from any responsibility for admitting the Franks into the empire and his subsequent failure to establish Greek overlordship over Antioch. Bohemund, who invaded the Balkans twice, in the 1080s and again in 1107–8, is one of the villains of her staunchly anti-western account that has as much to do with the tangled dynastic and imperial politics of the twelfth century as with the events of the 1090s. Seductively vivid, Anna is a confused and misleading source for the crusade let alone the motives of the western leaders. Even her famous description of Bohemund himself – tall, slim, muscular, good complexion, short light-brown hair, forbidding expression – cannot be trusted, still less Runciman’s fancy that the teenage Anna was taken by his good looks, ‘being, like all Greeks down the ages, susceptible to human beauty’.35

Nonetheless, Bohemund’s position on crusade is intriguing. The dominant personality in the expedition’s military leadership from April 1097 until January 1099, he founded a Norman dynasty in Antioch that outlasted the Norman kings of England and of Sicily. Yet in 1096, in contrast to all the other leaders who were at least of comital status (i.e. of a count), Bohemund was technically still a vassal of a count, his half-brother, the ineffectual Roger Borsa (1089–1111), younger son and heir of Robert Guiscard in southern Italy. The fabulous inheritance promised him by his father, Robert Guiscard, in the Balkans had come to nothing after the failure of the Norman invasion of 1081–5. Despite rebelling in 1085 and 1087 against Roger Borsa, Bohemund had failed to establish an independent territorial title for himself in the west, a political frustration that shaped his actions on crusade. Although possessed of political clout and contacts above his formal station – he knew Urban II personally – he lacked the extensive patrimony or dependent baronage that supported his fellow leaders. The army he gathered about him in southern Italy in the autumn of 1096 reflected this. The core seems to have been his close relatives, including his nephew, Tancred of Lecce and cousin, Richard of Salerno, known as Richard of the Principate, as well as his standard bearer, Robert FitzGerald. In addition there were former rebels, such as Robert of Ansa; and vassals of his half-brother, such as Robert FitzTristan, and of his uncle, Count Roger of Sicily, such as Robert of Sourdeval.36 His whole force was small, perhaps in total between 3,500 and 4,000 men. Lacking the power of lordship or the purse, Bohemund had to rely on more overt political and military skills. Even these failed to impose cohesion on his force. One nephew, William FitzMarquis, joined Hugh of Vermandois; another, Tancred, fought under his own banner, refused to accept Bohemund’s authority at Constantinople and thereafter pursued an increasingly independent line.

Bohemund’s army crossed the Adriatic from Bari to the coast of Epirus in late October 1096, probably deliberately avoiding the Byzantine garrison at Durazzo. It then dawdled its way to Constantinople, taking the best part of six months, at an average of just over three miles a day, almost being caught up by the larger force under Raymond of Toulouse, who had landed at Durazzo over three months behind. Yet there was almost no fighting or local resistance. If the story of his approach to Godfrey for an anti-Greek alliance in January 1097 is credited, the initial delay in the western Balkans is explicable as Bohemund would not wish to become too closely entangled with Greek escorts or garrisons near the capital. Godfrey’s rebuff may have inspired a volte face. As his troops neared Thrace, Bohemund left them under the command of Tancred on 1 April and hurried on to Constantinople, which he reached on 9 April. There, so far from fomenting trouble for Alexius, he proved the emperor’s staunchest ally in the often tetchy negotiations with other leaders. Bohemund spent longer with Alexius than any of the other leaders, almost a full calendar month. He eagerly took an oath of fealty and tried to obtain a future role for himself as military commander or ruler of new conquests in the east as the emperor’s vassal. His efforts to persuade Raymond of Toulouse to come to terms with Alexius and to force Tancred to swear an oath of fealty confirmed his alliance with the emperor. In return, Alexius employed him as his agent with the crusade leaders, Bohemund appearing to have acted as the expedition’s quartermaster for the siege of Nicaea, after which he shared the vanguard of the army with Alexius’s representative Tatikios. When he arrived at Constantinople, without his army, Bohemund was the least powerful of the western magnates at the imperial court; when he left Nicaea less than two months later, he was one of its undoubted leaders. Part, at least, of this must be ascribed to his private diplomacy with Alexius.

Instead of the archetype of a bumptious, threatening and deceitful barbarian as portrayed by Anna Comnena, Bohemund provided a medium of contact between east and west. He was not alone. When Godfrey of Bouillon arrived at Constantinople, he was met by a court official, Roger, son of Dagobert, a Norman who had joined the service of Alexius in the 1080s and progenitor of a family of Greek politicians. Peter of Alifa had fought with Guiscard and Bohemund against Alexius in the 1080s but, with many Italian Normans, had entered imperial service after Guiscard’s death in 1085. Accompanying the crusaders after Nicaea, he received the governorship of Comana in eastern Anatolia, captured by the crusaders in the autumn of 1097, ‘in fealty to God and the Holy Sepulchre, and to our leaders and the emperor’. Peter founded a Byzantine dynasty which adopted the name Petraliphas; both he and Roger, son of Dagobert, later fought for Alexius against Bohemund in the Epirus war of 1107–8.37 Another recruit at Alexius’s court was Bohemund’s own half-brother, Guy; in June 1098, when Alexius decided to withdraw from his projected relief of the crusaders at Antioch, he unavailingly begged the emperor to continue to save his kindred. Another imperial servant was Bohemund’s brother-in-law, William of Grandmesnil from Normandy, who travelled with Tatikios’s Greek division that marched with Bohemund in the vanguard across Asia Minor. Thus, viewed in the perspective of Norman experience, not propaganda, the First Crusade appears as part of an existing process of contact, tension and reaction. When Bohemund arrived in Constantinople in April 1097 and swore fealty to Alexius, his former enemy, he was doing no more than his half-brother and brother-in-law had done before him.

Bohemund’s prominence rested on establishing Byzantine credentials. Alexius need not have trusted him; but he could use him to suit his own purpose of controlling the crusade by proxy. He thought he had achieved this, as he had with so many other Italian Normans, by appealing to Bohemund’s ambition and greed. Bohemund was a highly suitable agent not least because he probably spoke Greek. There is evidence that he read Greek; according to Anna Comnena he could pun in Greek; and a number of western sources indicate that he conversed in Greek with the treacherous Armenian who allowed the crusaders into Antioch in June 1098, Firuz, who expected Bohemund’s troops to do the same. Bohemund’s relatives at Alexius’s court spoke Greek; Tancred apparently could speak Arabic (and did so at Antioch); language skills ran in the family. Indeed, Bohemund itself was a nickname, coined by his father after seeing the size of his infant son; it referred to a legendary giant. The boy had been baptized Mark, a Greek name.38

The timing of arrivals at Constantinople exerted a profound influence on the balance, nature and course of the rest of the expedition. That Alexius had managed to extract oaths from Count Hugh, Duke Godfrey and Bohemund, as well as the count of Flanders, who had left his travelling companions the duke of Normandy and the count of Blois behind in Italy at the turn of the year, and had shipped their troops across the Bosporus by 26 April 1097, presented Raymond of Toulouse with something of a fait accompli when he arrived in the last days of April 1097. As his chaplain recorded, it was reported to Raymond that ‘Bohemund, the duke of Lorraine, the count of Flanders and other princes besought him to make a pact with Alexius’.39 His temper can hardly have been improved by what had proved a long, exhausting and increasingly violent and ill-disciplined march.

Although probably the first magnate to take the cross, and the only one certainly to have had prior warning of Urban’s message at Clermont, Raymond had started late, in October 1096. His army was probably the largest and best funded; his preparations had been meticulous; his entourage filled with the eminent from the Limousin, Languedoc and Provence, including the counts of Orange and Montpellier, the viscounts of Béarn and Turenne, as well as the pope’s designated leader of the enterprise, Bishop Adhemar of Le Puy and his Monteil brothers from the Auvergne. It is possible Raymond’s planning lay behind the Genoese fleet despatched to the Levant in July 1097; Urban had sent a legation to the city led by Bishop William of Orange, who later accompanied Raymond east.40Yet Urban’s grant of a papal banner to Hugh of Vermandois and legatine authority to the chaplains of the duke of Normandy and the count of Blois at Lucca in October 1096 indicated that, despite Raymond’s early involvement and Urban’s tour of his lands in June and July 1096, he had no claim to overall command of the enterprise beyond his seniority (he was about sixty); possibly his experience in fighting in Spain; his association with Bishop Adhemar; and his money.41 He proved a difficult colleague; it is hard to determine whether his repeated displays of ill-temper were a cause or effect of his political isolation. However, his journey to Constantinople might have tried a saint.

Avoiding the Adriatic crossings from Italy, presumably because of the lateness of season, Raymond laboriously led his large army around the head of the Adriatic and down the Dalmatian coast across very difficult terrain. The natives proved unfriendly, provoking reciprocal atrocities. On reaching Byzantine territory at Durazzo in January, Raymond’s troops discovered a resentful populace, suspicious authorities and hard-pressed escorts. It was mid-winter; food supplies were beginning to present a problem, not least because of the recent passage of Bohemund’s army. There were increasing confrontations with locals and the Pecheneg police escort. Raymond of Aguilers caught the bitterness of the crusaders’ reaction:

we were confident that we were in our own land, because we believed that Alexius and his followers were our Christian brothers and confederates. But truly, with the savagery of lions they rushed upon peaceful men who were oblivious of their need for self defence.42

In one incident, Adhemar of Le Puy was quite badly wounded; although he recovered after recuperating at Thessalonica, he appears hardly at all in the chroniclers’ accounts of the negotiations at Constantinople. The problem was food. The Provençals sacked Roussa and, after Raymond had left his troops to parley with emperor in April 1097, were dispersed by imperial soldiers as punishment for ravaging. When he heard, the count was unamused and in no mood to place himself under the lordship of a ruler whose conduct had to date appeared either incompetent or mendacious.

The last army to reach the Byzantine capital contained the contingents led by Robert of Normandy and his brother-in-law Stephen of Blois. Initially, they had travelled with Count Robert II of Flanders, whose father, Robert I the Frisian, having undertaken a pilgrimage to Jerusalem, had served with Alexius in the Balkans in the late 1080s and had later sent the emperor a force of 500 knights. Robert II of Normandy’s grandfather, Robert I the Devil (or Magnificent according to taste) had died on pilgrimage to Jerusalem in 1035; his father, William the Conqueror, had been asked to assist the Byzantines against the Turks in the 1060s.43 His own crusade owed as much to his political difficulties in his duchy as to the advice of his spiritual advisers, who were credited with persuading him to join the march to Jerusalem. A poor politician, Robert was an effective military leader and warrior and a popular companion. Supported by the 10,000 marks provided by his younger brother, William II Rufus, king of England, Robert cut a finer figure on crusade than he had at home. At the head of a substantial force of Anglo-Norman nobles, including representatives of the families of Montgomery, Grandmesnil, Gournay and Percy, he picked up more followers on the journey. Eustace III of Boulogne, eldest brother to Godfrey of Bouillon and Baldwin, and a major landowner in England, probably travelled with him; in Italy Norman émigrés such as Roger of Barneville joined their ancestral lord.

Duke Robert earned glittering fame on crusade, playing prominent roles in the crucial encounters at Dorylaeum (July 1097), Antioch, Jerusalem and Ascalon (August 1099). In 1097–8, he assumed control of the vital Syrian port of Lattakiah. The second Latin bishop appointed by the Franks in the east, at Ramla in June 1099, was a Norman, Robert of Rouen; Duke Robert’s own chaplain, the foul-mouthed philanderer Arnulf of Chocques, was elected Latin Patriarch of Jerusalem in August 1099. There was even talk of Robert being a candidate for the crown of Jerusalem, which he was supposed to have rejected, characteristically, ‘out of fear of the work involved’.44 This prominence was in part a function of Robert’s wealth, which enabled him to maintain his independence and a sizeable retinue of knights; even as late as January 1099, he appears to have been able to maintain about 100 knights in his army, the same as Godfrey of Bouillon and twice the number supported by Robert of Flanders.45 On his return to the west, Robert found himself the hero of instant legend, his alleged deeds enshrined in stained glass at the great royal abbey of St Denis within a decade of his death. His reputation formed an acute contrast with his totally disastrous political career, which ended in twenty-eight years’ incarceration (1106–34) by his youngest brother, King Henry I of England.

His brother-in-law, Stephen count of Blois, left an even more equivocal reputation. Famously hen-pecked by his tough wife, Adela, the Conqueror’s daughter, Stephen may have been a reluctant crusader, but he was one of the wealthiest. Perhaps this explained why, at a crisis at Antioch in 1098, he was chosen by the other leaders as, in his own words, ‘lord, director and governor’ of the enterprise, perhaps implying a chairman’s role in the council of the high command.46 He scarcely exercised any authority, deserting the siege of Antioch the day before its capture in June 1098. His presence with Robert of Normandy confirmed an intimate dynastic network that underpinned their expedition. The counts of Flanders and Boulogne were closely related; Duke Robert’s mother was of the Flemish comital house; Count Stephen was the duke’s brother-in-law. With them was Duke Robert’s uncle, the worldly and acquisitive but now disgraced Odo bishop of Bayeux: he was to die in the winter of 1096–7, the guest of another successful Norman opportunist, Count Roger of Sicily, who provided a fine tomb for him in Palermo. The sense of family business was reinforced when the army reached Apulia, where Duke Roger Borsa’s wife was Robert of Flanders’s sister.

These northern French lords left for the east in late September or early October 1096, travelling across the Alps to the Po valley. They met Urban II at Lucca in late October, before visiting Rome and Monte Cassino on their journey south to Bari. There, Robert of Flanders left them to cross the Adriatic despite the late season. Duke Robert and Count Stephen stayed in southern Italy for the winter. This delay presented some less affluent, self-funded crusaders with acute problems. Unable to forage freely in friendly territory, their costs rising, those who lacked the patronage of lords or knights faced both hunger and ruin. Many, one of the chaplains attached to Count Stephen recalled, ‘sold their weapons and again took up their pilgrims’ staves and returned home’.47 However, it was still a considerable force that was shipped from Brindisi to Durazzo in early April 1097. Shipwreck and flash floods reduced the ranks, but by then supplies in the Balkans presented fewer problems. All the other armies had crossed or were in the process of crossing to Asia when the northern French reached their destination. Arriving at Constantinople on 14 May, the leaders were deeply impressed by Alexius’s lavish welcome; the other ranks received guided tours of the fabulous city in select groups of five or six. None of them had seen anything like it.


The negotiations between Alexius and the military leaders of the Jerusalem expedition formed a pivot around which the nature and future perception of the campaign revolved.48 Both sides understood the importance of what was agreed, even though they later chose to interpret events very differently. Alexius wished to use the westerners to exploit divisions among the Turks of Asia Minor and Syria to restore a measure of Byzantine control without risking a full commitment of his own military reserves. His dilemma lay in the extent to which he imposed his authority over the crusaders as paymaster and beneficiary while remaining essentially a sleeping partner in the operation. Fully aware of the obsession with Jerusalem, Alexius needed to encourage the idea that he shared the crusaders’ strategic goals while being more interested in undermining Seljuk power in Anatolia and opening stronger lines of political communication with sympathetic Armenians in Cilicia and Syria. Although he may not have been unhappy that Syrian Antioch had been taken from one of his Greek opponents in 1084–5, the fact that this great city had been lost to Islam in his own reign did not look good.49 On the other side, while Urban II had clearly envisaged the closest cooperation with the Greek emperor, of greater urgency were immediate logistical considerations. The western armies required Byzantine advice and material aid before they headed out across hostile Muslim territory. If they had doubted it, the disasters at Xerigordo and Kibotos would have persuaded them. However, the westerners lacked unified leadership, a coherent political strategy or an agreed military plan. They knew something of what problems to expect across the frontier with the Turks; they also had no clear view as to how to deal with them. Thus, Alexius was eager to assert demonstrable but indirect leadership over the expedition, while the crusaders were equally keen to accept Byzantine assistance. What needed to be resolved were the terms of that control and the conditions of that assistance.

Traditional Byzantine foreign policy, derived from the techniques of the Roman empire, outlined the best course of action when dealing with barbarians, those outside the empire or those, like the Normans in Italy and Sicily or the Armenians and Turks in northern Syria who, in the timeless Byzantine view of the world, were squatting on former imperial lands. If such tribes threatened the empire, or the emperor wished to use them, the tactics remained much the same: smother them with hospitality; learn their customs and exploit these; divide and rule; forge links of dependence based on profit, golden chains as it were; employ them; Byzantinize them. These were Alexius’s methods in the early months of 1097, to which he added a high dose of flexible opportunism. He was welcoming to all who accepted his hospitality; some, such as Godfrey of Bouillon or Tancred of Lecce, who avoided Constantinople and tried not to meet the emperor, required some small element of coercion; for the rest nothing was too much, as Alexius imposed the authority of fabulous wealth on his bucolic visitors. The oath he wished them to swear to him was, according to Anna, a ‘customary Latin oath’; whatever its details, the reactions of the leaders suggest that they recognized it.50 Alexius used Hugh of Vermandois to persuade Godfrey to come to heel and ensured Godfrey and the rest witnessed Bohemund’s oath. Bohemund, Godfrey and Robert of Flanders were cited as wishing Raymond’s adherence to Alexius’s contract. Bohemund was employed to extract agreement from Raymond and to force Tancred to fall into line. Once agreement and acquiescence had been obtained, Alexius lavished gifts on the westeners, whom he now regarded as his servants. The only aspect of the Greek formula that failed, and did so disastrously, was the inability of most of the westerners to become Byzantines. Although they could meet over mutual self-interest in the deals struck at Constantinople, there was a fundamental gulf not so much of understanding but of aspiration.

Alexius saw his interests as eternal: the benefit of the empire. Anything else was peripheral or secondary, not least remote Jerusalem. He probably minimized the importance of the reciprocal nature of his agreement with the crusaders, seeing them effectively as mercenaries; they regarded him as a lord with contractual obligations to preserve the interests of his vassals. When he was persuaded that the crusaders were doomed at Antioch in June 1098, Alexius preserved his strategy by withdrawing his own army from danger. For the crusaders, this withdrawal was inexplicable treachery from a lord whose help had been sworn. They, who had risked all so many times, failed to appreciate his caution. The shadow of Antioch fell deep over Graeco-Latin relations in the twelfth century, nowhere more black than in the pages of the eyewitness chroniclers who felt and experienced the betrayal in which dim light they re-evaluated all that had transpired between Alexius and the crusade leadership. It is small wonder that Anna Comnena was so frenetic in her attempts to exonerate Alexius from any suggestion of culpability over Antioch, for he had been caught out by that most politically damaging agent: events. If the westerners had been annihilated at Antioch, as common sense dictated they should have been, Alexius would have been vindicated. Unfortunately not only did they survive, they proceeded to win Jerusalem and return to tell their tale.

At the heart of the dispute lay the oaths sworn, which were solemn and serious. Despite the contrasting sensitivities locked into the descriptions of events at Constantinople, it appears that Alexius demanded and received from all the leaders except Raymond of Toulouse homage and fealty. They became the emperor’s vassals, promising to restore to imperial rule all lands, towns and castles they captured which had formally belonged to the empire. This is effect meant lands lost in the relatively recent past: even Raymond of Toulouse, who became most protective of the relationship with Alexius, considered towns in Syria beyond Antioch, such as al-Bara, beyond the remit of the agreement.51 In return, Alexius promised help for the crusaders. Some tried to argue that he had promised to join the march to Jerusalem, but this probably represents a post-Antioch gloss. Raymond of Aguilers, a very hostile source, stated that Alexius ruled out his personal involvement. More probable was a guarantee of military aid, supplies and advice, as well as promises to protect the crusaders’ rear and assist reinforcements. The importance of these arrangements was underlined by Alexius’s insistence at the crusaders’ Asiatic base at Pelekanum just before they left for the march across Anatolia that even lesser lords took the oath. The one exception was Raymond of Toulouse, who insisted on swearing an oath more acceptable to Provençal practice, that ‘he would not, either through himself or through others, take away from the emperor, life and possessions’.52 Yet he abided by his obligations more faithfully than his colleagues, perhaps because Alexius had taken special pains to establish good relations with him after their sticky introduction.

The legal aspects of the agreements struck by Alexius and the western leaders were less important than the political implications. Only by becoming Alexius’s vassals could they extract necessary help. For Bohemund, subservience offered an opportunity for self-advancement; he proposed to Alexius that he be created Domestic of the East, effectively commander of the imperial forces in Asia and, in consequence, commander-in-chief of the crusade.53 Alexius temporized but did not reject his offer of service outright. That Bohemund came up with the idea exposes the nature of his ambition; he wanted leadership; he wanted land. Alexius was not to know the extent to which Bohemund was determined to have both without any ties of vassalage. His whole career to date had made the Norman wish to dispense with overlords. Yet for the present, Alexius suited his case.

Initially, whatever the details, the treaties of Constantinople worked. Relations between western leaders, not least Bishop Adhemar, with the Greeks were good. Nicaea reverted to imperial control after its capture in June 1097 despite Alexius’s absence. A Byzantine division accompanied the army eastwards towards Antioch, under an experienced commander, Tatikios, a safe choice as a Turkish eunuch of unavoidable loyalty to the emperor rather than a Greek nobleman who could have harboured imperial longings of his own. Cities captured on the way, such as Comana, were indeed restored to Greek lordship. Optimism, if judged by a cheerful letter Stephen of Blois wrote to his wife from Nicaea on 24 June 1097, was high: ‘the army of God’ was looking forward to reaching Jerusalem in five weeks – ‘unless Antioch holds us up’.54 Urban’s plan seemed to be working.


2. Asia Minor and Syria 1097–99

If you find an error please notify us in the comments. Thank you!