Текст книги "The Crusades. The Authoritative History of the War for the Holy Land"

Автор книги: Thomas Asbridge

сообщить о нарушении

Текущая страница: 11 (всего у книги 54 страниц)

This view of events is riddled with significant problems. The bulk of contemporary evidence suggests that the pope was aware of Bohemond’s intentions and nonetheless supported him, even dispatching a papal legate to accompany and endorse the preaching campaigns in France and Italy. Even in the unlikely case that the pope was misled, there can be no doubt that a huge number of lay recruits accepted the idea of joining a crusade against the Greeks. In fact, the tendency to sideline Bohemond’s expedition as a perversion of crusading is symptomatic of a more fundamental misconception: a belief that the ideas and practices of crusading had already coalesced to create a uniform ideal. For most people living in western Europe in the early twelfth century, this new type of devotional warfare had no finite identity and was still subject to continual, organic development. As far as they were concerned, crusades did not need to be directed against Muslims, and many readily accepted the idea of waging a holy war against Alexius Comnenus once he had been deemed the enemy of Latin Christendom.

However the background to the 1107–8 ‘crusade’ against Byzantium is viewed, the expedition itself proved to be a shambolic disaster. Crossing the Adriatic in October 1107, the Latins laid siege to the city of Durazzo (in modern Albania), regarded by contemporaries as ‘the western gate of the [Greek] empire’. But, in spite of his military pedigree, Bohemond was outwitted by Alexius, who deployed his forces to cut the invaders’ supply lines while carefully avoiding direct confrontation. Weakened by hunger, unable to break Durazzo’s defences, the Latins capitulated in September 1108. Bohemond was forced to accede to a humiliating peace accord, the Treaty of Devol. By the terms of this agreement, he was to hold Antioch for the remainder of his life as the emperor’s subject, but the Greek patriarch was to be restored to power in the city and the principality itself to be all but emasculated by the cession of Cilicia and Latakia to Byzantium.

As it was, this agreement was not implemented and thus had little bearing upon future events, because Bohemond never returned to the Levant. After sailing back to southern Italy in the autumn of 1108, he appears only fleetingly in historical records, his reputation broken, his grand dreams and ambitions shattered. Constance bore him a son, also named Bohemond, around 1109, but by 1111 the once great commander of the First Crusade was ailing, and on 7 March he died in Apulia. At Antioch, Tancred remained in power, perhaps still nominally as regent, but with his authority uncontested among the Franks. From the perspective of Outremer, one positive did emerge from Bohemond’s later career: his Balkan campaign diverted Greek resources from the Levant, allowing Tancred to assert lasting control over Latakia and Cilicia.⁶⁶

TO RULE IN THE HOLY REALM

Tancred’s drive to expand the principality of Antioch and to augment its wealth and international influence accelerated after 1108, and he showed a ruthless willingness to use any and all means in pursuit of these ambitions, even if that meant fighting fellow Latins while engaging Muslim allies. For the next five years he worked tirelessly, drawing upon a seemingly inexhaustible pool of martial energy to engage in near-constant campaigning. Beleaguering his neighbours and opponents through a mixture of territorial conquest, political coercion and economic exploitation, Tancred came close to forging an Antiochene empire in the Levant.

The counties of Edessa and Tripoli

Between 1104 and 1108 Antioch was the effective overlord of the county of Edessa. Once Tancred assumed control of the principality in autumn 1104, he installed his brother-in-law and fellow southern Italian Norman First Crusader Richard of Salerno as regent of Edessa. Even though Richard proved unpopular, Antiochene influence went unchecked while Count Baldwin II remained in captivity.

Antioch certainly made no effort to orchestrate the count’s release. In the summer of 1104, when Baldwin’s captors first sought to organise the terms of his ransom, even Bohemond demurred. Rather than repay the energy Baldwin had expended to secure Bohemond’s own freedom in 1103, the prince preferred to retain control of Edessa’s considerable agrarian and commercial resources, estimated to value in excess of 40,000 gold bezants per annum. Once at the helm of Frankish Syria, Tancred continued to enjoy these revenues and to ignore Baldwin’s plight.

By 1107 the count’s companion, Joscelin of Courtenay, lord of Tell Bashir, had been ransomed by the populace of that town, and in the following year Joscelin successfully negotiated Baldwin’s release from Mosul. It was the Turkish warlord Chavli, the latest ruler of Mosul, who finally agreed terms; but with an eye to the fragility of his own position and the ongoing internecine struggles within Near Eastern Islam, Chavli demanded not only a cash ransom and hostages, but also a promise of military alliance.

When Baldwin sought to reclaim Edessa in the summer of 1108, a tense standoff ensued. Having enjoyed access to the wealth and resources of the county for four years, Tancred had no intention of simply handing over a territory which he had saved from conquest, and he now sought to pressure Baldwin into taking an oath of subservience; after all, he argued, historically Edessa had been the vassal of the Byzantine duchy of Antioch. The count refused, not least because he had already sworn allegiance to Baldwin of Boulogne in 1100. With neither side willing to give ground, conflict seemed inevitable.

In early September both men raised armies. Less than ten years after Jerusalem’s conquest, Baldwin and Tancred–fellow Latins and veteran crusaders–were now ready and willing to crush one another in open war. More shocking still was the fact that Baldwin marched forth to this struggle alongside his new ally, Chavli of Mosul, and some 7,000 Muslim troops. When battle was joined, probably near Tell Bashir, Tancred, although outnumbered, managed to hold the field. But with some 2,000 Christian dead on both sides, Patriarch Bernard, the ecclesiastical overlord of both Antioch and Edessa, stepped in to calm frayed tempers and adjudicate. When witnesses publicly attested that Tancred had actually promised Bohemond in 1104 that he would relinquish control of Edessa upon Baldwin’s release, the Antiochene ruler was forced grudgingly to back down. The city of Edessa itself may have been repatriated, but the embedded hatred and rivalry remained. Tancred stubbornly refused to hand over territory in the northern reaches of the county and was soon pressing Baldwin to make tribute payments in return for peace with Antioch.⁶⁷

With this dispute still simmering, Tancred’s acquisitive gaze settled upon the nascent county of Tripoli. In the immediate aftermath of the First Crusade his old rival Raymond of Toulouse had sought to carve out his own Levantine lordship centred on the northern reaches of modern-day Lebanon. The challenge confronting Raymond was considerable, for unlike the founders of other Latin settlements he had no crusader conquests to build upon, and the region’s dominant city, Tripoli, remained in Muslim hands.

Nonetheless, Raymond made some progress, capturing the port of Tortosa in 1102, with the aid of a Genoese fleet and survivors from the 1101 crusade. Two years later he conquered a second port to the south, Jubail, resplendent with Roman ruins. Meanwhile, on a hill outside Tripoli, Raymond constructed a doughty fortress, christened Mount Pilgrim, thereby securing effective control of the surrounding region. Yet, despite his tenacious efforts, when the count died in his mid-sixties on 28 February 1105, Tripoli itself remained unconquered.

In the years that followed, two men sought to press claims to Raymond’s legacy. His nephew, William Jordan, the first to arrive in Outremer, continued to pressure Tripoli while also overcoming the neighbouring town of Arqa. In March 1109, however, Raymond’s son Bertrand of Toulouse reached the Holy Land, determined to assert his rights as heir. When he brought a sizeable fleet to reinforce the siege of Tripoli, the two claimants squabbled over rights to the city, even though it had yet to be captured, and William Jordan quit Mount Pilgrim for the north. The emergent county of Tripoli looked as if it might founder amid bitter dynastic squabbling.

In the end, however, the contest for control of Tripoli involved far more than the simple issue of inheritance; it became the centrepiece of a wider struggle for dominion over the crusader states. Realising that he would need an ally if he was to have any hope of claiming Tripoli, William Jordan turned to Tancred, offering to become his vassal. Not surprisingly, Tancred seized this sudden opportunity to expand Antiochene influence southwards; should Tripoli fall under his sway and his designs upon Edessa come to fruition, then the principality might rightly claim to be Outremer’s leading power. Modern historical analysis has persistently underestimated the significance of this episode, the assumption being that the kingdom of Jerusalem was automatically and immediately recognised as the overlord of the Frankish East at the start of the twelfth century. True, the Holy City had been the focus of the First Crusade, and Baldwin of Boulogne was the only Latin ruler in the Levant to assume the title of king, but his realm encompassed Palestine, not the entire Near East. Each of the four crusader states was founded as an independent polity and Jerusalem’s pre-eminent status among them had never been formally ratified. A current of rivalry had coloured relations between Baldwin and Tancred ever since they contested control of Cilicia in 1097; now, in 1109, Tancred’s brash assertiveness offered a challenge to Baldwin’s authority that would determine the balance of power in the Latin Levant.

Over the next twelve months, Jerusalem’s monarch resolved this political crisis with stunning finesse, roundly outplaying his old opponent. To his credit, Baldwin made no attempt to counter Antiochene ambition with direct force of arms, preferring instead to promote and harness the notion of Frankish solidarity in the face of Muslim adversaries. Employing diplomatic guile, he affirmed Jerusalemite supremacy even as he advanced Outremer’s defensive security.

In the summer of 1109 Baldwin called the rulers of the Latin East to assist Bertrand of Toulouse at the siege of Tripoli. On the face of it, this was to be a grand Frankish alliance, dedicated to the subjugation of an intransigent Muslim outpost. The king himself marched north with some 500 knights; Tancred, together with 700 knights, arrived in the company of his new ally, William Jordan; and Baldwin II of Edessa and Joscelin likewise brought a sizeable force. Alongside Bertrand’s Provençal navy and a Genoese fleet, this represented a formidable assembly. And yet, entrenched animosity and fractious suspicion rippled beneath the surface of this coalition.

Of course, the subtext to the whole affair–as all the key players must well have known–was the issue of power among the Franks. Would Baldwin I allow Antioch’s burgeoning influence to go unchecked, and if not, what manner of riposte would the king employ? With the gathering complete, the king enacted his canny scheme. Having already taken Bertrand of Toulouse under his wing, extracting an oath of fealty in exchange for Jerusalem’s support, he now convened a general council to resolve the dispute over Tripoli’s future. Baldwin I’s masterstroke was to comport himself not as a wrathful, overbearing overlord, nor as Tancred’s conniving rival, but rather as an impartial arbiter of justice. In the words of one Latin contemporary, the king listened to ‘all the injuries of both sides’ along with a jury of ‘his loyal men’ and then enacted reconciliation. Raymond of Toulouse’s heirs were ‘made friends’, with Bertrand given rights to the bulk of the county, including Tripoli, Mount Pilgrim and Jubail, and William placated with Tortosa and Arqa. What is more, Baldwin II and Tancred were said to have been ‘reconciled’ on the understanding that Antioch would relinquish control of all remaining Edessene territory. By way of compensation, Tancred was reinstated as the lord of Haifa and Galilee.

The king appeared to have achieved an equitable settlement, restoring harmony to Outremer. The coalition forces were certainly able to prosecute Tripoli’s investment with renewed vigour, bludgeoning the city’s Muslim garrison into submission by 12 July 1109. In reality, however, Tancred had been stymied and humbled. He made no effort to claim his lordship in the kingdom of Jerusalem, not least because this involved an oath of subservience to Baldwin I. The king, meanwhile, despite maintaining a façade of impartiality, had served his own interests, protecting his relationship with Edessa and positioning his own favourite as the new ruler of a Tripolitan county. He cannot have been overly dejected when, soon after Tripoli’s capitulation, William Jordan was ‘pierced through the heart in a secret attack and died’, leaving Bertrand in a position of uncontested authority.

In May 1110 Baldwin I seized an opportunity to consolidate further his status as overlord of the Latin Levant. That spring, Muhammad, the Seljuq sultan of Baghdad, finally reacted to the Frankish subjugation of the Near East. He dispatched a Mesopotamian army to begin the work of reclaiming Syria under the command of Maudud, a capable Turkish general who recently had come to power in Mosul. The first target was the county of Edessa. In the face of this threat, the Latins united, and the swift arrival of a large coalition army from Jerusalem, Tripoli and Antioch forced Maudud to break off his short-lived siege of Edessa. King Baldwin I used the opportunity presented by this gathering of the ruling Frankish elite to call a second council of arbitration, this time with the sole focus of addressing the ongoing dispute between Tancred and Baldwin of Bourcq. According to one Christian contemporary, resolution was to be achieved, ‘either by a fair trial or by agreement of a council of magnates’. Knowing that he was unlikely to receive anything approaching ‘fair’ treatment, Tancred had to be persuaded to attend by his closest advisers and, once the council began, his fears were soon confirmed. With King Baldwin presiding in judgement, Tancred was accused of inciting Maudud of Mosul to attack Edessa and of allying with Muslims. These charges were almost certainly manufactured and, notably, no mention was made of either Baldwin of Bourcq’s own alliance with Mosul in 1108 or Baldwin I’s dealings with Damascus. Facing the united opprobrium of the council and threatened with ostracism from the Frankish community, Tancred was once again forced to back down. From this point on, he seems to have stopped demanding tribute from Edessa.

Antioch’s submission had not been formalised and, in the years to come, the principality would make renewed attempts to assert its independence. Throughout the early decades of the twelfth century this secular power struggle was also mirrored by a protracted and embittered squabble over ecclesiastical jurisdiction between the Latin patriarchs of Antioch and Jerusalem. Nonetheless, in 1110 King Baldwin had, for the time being at least, affirmed his own personal authority and established Jerusalem’s position as the pre-eminent secular power in Outremer.⁶⁸

Tancred’s legacy

In spite of the political setbacks of 1109 and 1110, the closing years of Tancred’s life proved to be a triumph. With unabated vigour he pushed the principality’s frontiers to the limit and subdued his Muslim neighbours, fighting for months on end almost without pause. In this period, Tancred confronted a significant strategic quandary that has been largely ignored by modern historians. For Tancred, as for all medieval military commanders, topography was a key consideration. By 1110 the principality had expanded its borders to two natural boundaries. To the east, on the frontier between Antioch and Aleppo, Frankish power now extended to the foot of the Belus Hills, a craggy spine of arid, low-altitude fells. To the south, towards Muslim Shaizar, the principality stretched to the edge of the Summaq plateau and to the Orontes River valley. As it stood, the physical barriers running along these two border zones offered both Latin Antioch and its Muslim neighbours a relatively equal balance of power and security.⁶⁹

Tancred could have settled for this situation, allowing the status quo to be maintained, engendering the possibility of long-term coexistence. Instead, he chose the risks and potential rewards of continued expansion. In October 1110 he crossed the Belus Hills, prosecuting a taxing winter expedition that led to the capture of a string of settlements in the Jazr region (east of the Belus Hills), including al-Atharib and Zardana. This left barely twenty miles of open, undefended plains between the principality and Aleppo. Then, in the spring of 1111, he moved to apply a similar degree of pressure to the south, initiating construction of a new fortress on a hill close to Shaizar. To begin with, at least, Ridwan of Aleppo and the Muslim rulers of Shaizar, the Munqidh clan, responded to this aggression with conciliatory submission, offering tribute payments totalling 30,000 gold dinars in return for peace.

There was a well-established precedent for this form of financial exploitation. In eleventh-century Iberia, the Christian powers of the north had gradually come to dominate the fractured Muslim city-states of the south, establishing complex networks of annual tribute payments. This system famously culminated in the peaceful occupation of the peninsula’s long-lost capital, Toledo (central Spain), in 1085.

Tancred may well have harboured similar plans to reduce Aleppo and Shaizar to the point of collapse, but his policies had a dangerous edge. Apply too much pressure, demand overly exorbitant protection payments, and the quarry might be driven to risk retaliation. In the case of Aleppo, the mixture of intimidation and exploitation proved effective and culminated in a sustained period of submission. But in 1111, Tancred pressed Shaizar too far and the Munqidh clan readily allied with Maudud of Mosul when he led a second Abbasid army into Syria that September. Threatened with an invasion of the Summaq region, Tancred mustered every possible ounce of Antiochene manpower. He also called for aid from his fellow Latins and, despite the tensions which had recently divided their ranks, the armies of Jerusalem, Edessa and Tripoli assembled once more. This composite force took up a defensive position at Apamea, and by patiently holding its ground, blunted Maudud’s attempts to provoke a decisive battle and eventually forced his retreat.

Tancred once again had repulsed a threat to the principality’s survival, but any hopes of securing the conquest of either Aleppo or Shaizar came to nothing when, after years of tireless campaigning, his health failed him at the age of thirty-six. The early twelfth-century Armenian Christian historian Matthew of Edessa lavished elegiac praise upon Tancred when recording his death in December 1112, writing that ‘he was a saintly and pious man and had a kind and compassionate nature, manifesting concern for all the Christian faithful; moreover he exhibited a tremendous amount of humility in his dealings with people’. This panegyric conceals Tancred’s darker traits: his unquenchable hunger for advancement; his gift for political intrigue; and his willingness to betray or battle all around him in pursuit of power. It was these qualities, allied to his boundless dynamism, that lent Tancred his remarkable potency and enabled him to forge an enduring Frankish realm in northern Syria. If justice be done, history should regard Tancred, not his infamous uncle Bohemond, as the founder of the principality of Antioch.⁷⁰

OVERLORD OF OUTREMER (1113–18)

Tancred’s death came at a time of more general change in the shape and balance of power in the Near East, brought on by a mixture of dynastic succession and political intrigue. At Antioch itself, power passed to Tancred’s nephew, Roger of Salerno, son of the First Crusader Richard of Salerno. Roger was soon woven into the fabric of Frankish society as a series of high-level marriage alliances bound together the ruling elite of Outremer. This complex web of familial connections ushered in a new phase of heightened interdependence among the crusader states. Roger himself married the sister of Baldwin of Bourcq, count of Edessa, while Joscelin of Courtenay, lord of Tell Bashir, was wed to Roger’s sister. Bertrand of Toulouse’s death early in 1112 led to the accession of his youthful son, Pons, as count of Tripoli. He soon distanced himself from the traditional Toulousean policy of subservience to Byzantium and antipathy to Antioch and, at some point between 1113 and 1115, married Tancred’s widow, Cecilia of France. Pons remained a dependant of Jerusalem, but Cecilia’s dowry brought him a significant Antiochene lordship in the Ruj valley, one of only two southern approach routes to Antioch itself. The wider significance of these shifts in personnel and allegiance was twofold: on the one hand, they promised to engender a new era of Frankish cooperation in the face of external threats; on the other, they reopened old questions about the balance of power in Outremer and, most notably, the relationship between Antioch and Edessa.

Strength in unity

The bonds of Latin unity were soon tested by the ongoing threat of Iraqi invasion. In May 1113 Maudud of Mosul, now Baghdad’s foremost military commander, led a third Abbasid army into the Near East, and on this occasion he turned away from Syria to invade Palestine. The frequency and ferocity of Frankish raiding upon Damascene lands to the north and east of Galilee appear to have convinced Tughtegin that he must now turn his back on any form of enduring rapprochement with Jerusalem. In the last week of May he led a sizeable army to join Maudud, and together they marched into Galilee.

When news of this threat reached Baldwin I at Acre, he dispatched an urgent call for reinforcement to his new neighbours, Roger and Pons. The king now had a difficult decision to make. Should he wait for the full strength of the Frankish alliance to assemble, leaving Maudud and Tughtegin free to ravage the north-eastern reaches of the realm, or risk an immediate move to counter their incursion with only limited military resources? In mid-to late June he settled upon the second course of action. Baldwin’s precipitous behaviour was widely criticised by contemporaries–indeed, even his chaplain noted that the king was denounced by his allies for ‘rush[ing] against the enemy in a rash and disorderly manner without waiting for their advice and aid’–and Baldwin has been similarly condemned in modern historiography. In the king’s defence, he does not seem to have acted with the same damaging impetuosity shown in 1102. Details of events in the summer of 1113 are sketchy, but it would appear that Baldwin advanced from Acre to establish an advanced base from which to patrol Galilee and not with the express intention of confronting the enemy in pitched battle.

Unfortunately for the king, on 28 June his army was battered by a surprise attack. Normally so assiduous in his use of scouts and the garnering of intelligence, Baldwin appears to have camped near the al-Sennabra bridge, a crossing over the River Jordan just south of the Sea of Galilee, without realising that his foes were stationed nearby, across the eastern shore. When Muslim foragers discovered his position, Maudud and Tughtegin launched a lightning assault. Pouring across the bridge, they quickly overran the shocked Franks, killing 1,000 to 2,000 men, including some thirty knights. Baldwin himself fled in disgrace, losing his royal banner and his tent, key symbols of his regal authority.

Chastened, Baldwin retreated to the slopes of Mount Tabor, above Tiberias, where he was soon joined by the armies of Antioch and Tripoli. He now adopted a far more cautious strategy, holding his forces in this defensible position, policing the region but avoiding direct confrontation. For nearly four weeks the two sides remained in the area, testing one another’s resolve, but in the face of such a large Latin force Maudud and Tughtegin could not afford to march south en masse to Jerusalem and were only able to launch a series of wide-ranging raids. In August, the Muslim allies crossed back over the Jordan, leaving, in the words of one Damascene chronicler, ‘the enemy humbled, broken, defeated and dispirited’. As evidence of their triumph they sent a gift of plunder, Frankish prisoners and the heads of the Christian dead to the sultan in Baghdad. Baldwin had survived, albeit with considerable damage to his reputation.⁷¹

Maudud fatefully elected to spend early autumn in Damascus. Having attended Friday prayers with Tughtegin at the Grand Mosque on 2 October 1113, the Mosuli commander was walking through a courtyard when he was ambushed and mortally wounded by a lone attacker. The assailant was summarily decapitated and his corpse later burned, but neither his identity nor his motive was ever precisely ascertained. The suspicion was that he had been an adherent of a secretive Nizari sect. This splinter faction of the Isma‘ili branch of Shi‘a Islam, originally from north-eastern Persia, had begun to play a notable role in Near Eastern politics at the start of the twelfth century. With limited resources, they gained power and influence by murdering their enemies and, because it was rumoured that their adherents were addicted to hashish, a new word emerged to describe them–Assassins. During Ridwan ibn Tutush’s life they gained a significant foothold in Aleppo, but after his death in 1113 they were driven out of the city. The Assassins then found a new ally in Tughtegin, and for this reason the atabeg was suspected of having been complicit in Maudud’s assassination. The true extent of Tughtegin’s involvement is unclear, but the rumour alone was enough to isolate him from Baghdad and to promote a new rapprochement between Damascus and Jerusalem.⁷²

For the Franks, the crisis of 1113 proved beyond doubt the necessity for unified resistance to Muslim aggression; it also reaffirmed the wisdom of a cautious defensive strategy. Taken together, the events of 1111 and 1113 established a pattern of Latin military practice that was to persist for much of the twelfth century: in the face of a strong invading force, the Franks would unite; mustering at a defensible location, they would seek to police the threatened region and to disrupt the enemy’s freedom of movement, all while staunchly avoiding the unpredictability of open battle.

It was precisely this approach that Roger, prince of Antioch, adopted initially in 1115 when facing the first real threat of his reign. The only difference was that, on this occasion, he enjoyed the support not only of his Latin compatriots, but also of the Muslim potentates of Syria. With Aleppo now in a state of some disarray, the sultan of Baghdad saw an opportunity to take control of the city and thereby reassert his authority over the Near East. To this end, he sponsored a new expedition across the Euphrates, this time led by a Persian commander, Bursuq of Hamadan.

The prospect of such direct intervention prompted an unprecedented reaction from the feuding Muslim rulers of Syria. Tughtegin allied with his son-in-law, Il-ghazi of Mardin, the leading member of a Turcoman dynasty known as the Artuqids, who held sway over the Diyar Bakr region of the Upper Tigris River. Together, Tughtegin and Il-ghazi took temporary control of Aleppo and dispatched an embassy to Antioch to request peace talks. At first, Roger greeted this approach with some suspicion, but he was soon won over, perhaps by the entreaties of one of his leading vassals, Robert fitz-Fulk the Leper, who held a major lordship on the principality’s eastern frontier and had developed a close friendship with Tughtegin. A treaty of military cooperation was duly sealed early that summer and preparations for Bursuq’s invasion began.

Upon reaching Syria and discovering that Aleppo was now closed to him, Bursuq followed the example of Maudud of Mosul in 1111 and sought support from Shaizar for an attack on Antioch’s southern frontier. Roger, meanwhile, responded in kind by marching 2,000 troops to a holding position at Apamea, probably in the company of Baldwin II of Edessa. There the extraordinary pan-Levantine alliance assembled. Tughtegin, true to his word, joined Roger with some 10,000 men, while Baldwin I and Pons of Tripoli arrived later in August. These arrayed forces, so often themselves combatants, held their ground throughout the summer, successfully intermingling Latin and Muslim troops without apparent difficulty.

Facing such a sizeable and entrenched opposing force, Bursuq did his best to provoke open battle, sending skirmishers to harass the allied camp and leading raids into the Summaq plateau. It is testament to the difficulty of maintaining discipline in the face of such provocation that Roger threatened to blind anyone breaking ranks. The Latins, alongside their Damascene fellows, duly held to their position. Thwarted, Bursuq retreated from Shaizar and, with the danger to Syria now apparently passed, the grand coalition broke up.

Roger returned to Antioch, but in the first days of September Bursuq’s withdrawal was revealed as a ruse. Having fallen back towards Hama to await the dissolution of the defending army, he now circled around, cutting a swathe through the northern reaches of the Summaq. With the principality in real danger of being overrun, Roger found himself in an unsettling predicament, isolated from his allies. Only Baldwin of Edessa remained, having held troops in the principality throughout the summer as something akin to a client ruler of Antioch. Should Roger dutifully await the reassembly of the Latin–Muslim coalition, leaving Bursuq to roam the Syrian countryside with impunity, or risk swift, independent action? In essence, his dilemma replicated that faced by Baldwin I two years earlier and, in spite of the evident lessons of that encounter, on 12 September 1115 the prince of Antioch gathered his army at Rugia and marched to intercept the enemy. This was a rather foolhardy act of bravado. Leading some 500 to 700 knights and perhaps 2,000 to 3,000 infantry, he stood to be outnumbered by at least two to one. The Latins seem to have put their faith in an Antiochene relic of the True Cross carried in their midst by the bishop of Jabala and to have undertaken a series of purifying spiritual rites, but even so Roger must have recognised that he was gambling the future of Frankish Syria.