God's War: A New History of the Crusades Read online

  6. Syria in the Twelfth Century

  7. Palestine and Egypt in the Twelfth Century

  6

  The Latin States

  The principalities created by the western invasion of Syria and Palestine shared characteristics of both east and west. The Near East was familiar with culturally and religiously alien elites indifferent or hostile to indigenous peoples, content to rule a heterogeneous society by means of absentee landownership, control of trade and military coercion. As exploiters, not proselytes or social engineers, Levantine rulers forged contacts across communal divisions from convenience, not conviction. The Latins, or Franks as they were more usually called by onlookers, were no different. Their new aristocracy and settlers could not ignore their neighbours; as one of them remarked in the 1120s, a lingua franca combining many languages soon emerged, at least in the cities.1 Yet the political idiom of Latin rule remained severely western, as did the laws they applied to themselves and the vocabulary of government. Consequently, the vision of Outremer is bifocal. Twelfth-century Latin accounts portray a political society apparently hermetically sealed from its immediate environment, a western drama played out in exotic surroundings, whereas contemporary Arabic chronicles emphasize the normality and familiarity of Latin behaviour, one parvenu military governing elite among many.

  EDESSA

  Nowhere was Christian dependence on the politics of Islamic neighbours more obvious than in the fortunes of the county of Edessa, the first Latin principality established in the Near East, by Baldwin of Boulogne in March 1098. Isolated far in the interior, the county relied upon possession of fortified towns such as Turbessel (Tell Bashir), Ravendan and Edessa, from where the fertile countryside of the Euphrates basin could be exploited and the sensibilities of the Armenian lords influenced. As a bastion defending the eastern flank of Antioch and even a potential base for assaults into northern Iraq, the county assumed great strategic importance, which was also the source of its vulnerability. For warlords from Mosul or Mardin to the east or Aleppo to the south, Edessa presented a tempting target in itself as well as a staging post on any more concerted attack on the Christian holdings towards the coast. The county’s survival depended on unity among its Frankish nobility; cooperation with local Armenian lords, some of whom looked to Byzantium to guarantee their status and authority; alliances with Antioch and some, at least, of its Muslim neighbours; and, crucially, disunity among the rulers of northern Syria, the Jazira and Iraq.2

  In 1100, when Baldwin of Boulogne succeeded his brother Godfrey as ruler of Jerusalem, Edessa devolved on to his cousin, Baldwin of Le Bourcq, who did homage to the new king for the county. After consolidating his position by marrying a local Armenian princess, Morphia of Melitene, in 1102, the new count was joined by another cousin, Joscelin of Courtenay, a veteran of the count of Nevers’s crusade of 1101, to whom he gave all the county west of the Euphrates around Turbessel as a fief, thus making him almost a partner. Over the next two years, while Mosul battled with Mardin after the death of Kerbogha and further internecine feuding distracted the Seljuks of Baghdad, Joscelin as much as Count Baldwin campaigned successfully to extend Frankish rule northwards towards Armenian Marasch and south in the direction of Aleppo. In 1104, this aggression was abruptly ended when Soqman of Mardin joined with his erstwhile enemies of Mosul to crush a substantial combined Edessan/Antiochene force attempting to capture Harran, south-east of Edessa. Both Count Baldwin and Joscelin were captured. The dangers of Muslim unity were starkly underlined.

  During Count Baldwin’s captivity (1104–8), Edessa was ruled first by Tancred and then, after Bohemund’s departure for the west in 1105, another Hauteville relative, Richard of Salerno. Tancred, busily extending the principality of Antioch in his uncle’s absence, clearly hoped to annexe Edessa; it made strategic sense.3 Consequently, on Baldwin’s release, relations between Edessa and Antioch were strained beyond breaking, with Tancred trying to assert a spurious legal claim to over-lordship over the county. In 1108–9, both sides to the dispute probably called on Muslim military aid, Baldwin from his former captors at Mosul, Tancred from Aleppo. Baldwin’s relations with the increasingly independent Joscelin of Courtenay also deteriorated. Joscelin was arrested and exiled, Baldwin I of Jerusalem characteristically promptly recruiting him to the lordship of Galilee. However, the major invasions by Mawdud of Mosul each year between 1110 and 1113 and Tancred’s death in 1112 engineered a reconciliation between Antioch and Edessa, sealed by the marriage of Count Baldwin’s sister to Tancred’s successor, Roger of Salerno. At the victory of Tell Danith over Bursuq of Hamadan in 1115, Count Baldwin fought side by side with the Antiochenes. Edessa’s position relative to Antioch improved further after the accession of Count Baldwin to the throne of Jerusalem in 1118, due partly to the intrigues of Joscelin of Courtenay who, despite his exile in 1113, was rewarded with the county of Edessa itself. After Roger of Antioch’s defeat and death at the battle known to westerners as the Field of Blood in 1119, when Baldwin II, as king, assumed the regency of Antioch, Joscelin emerged as the strongest Frankish leader in northern Syria, probably even on occasion acting as regent of Antioch himself. Although briefly held captive by Balak of Aleppo in 1123, Count Joscelin commanded a combined forced from all four Latin principalities in the same year in an attempt to free King Baldwin, who had suffered a similar fate shortly after the count. Until the king’s release in 1124, Joscelin cut the leading secular figure in Outremer. Thereafter he continued to play a prominent role, joining with Baldwin II in attacking Aleppo in 1124–5 and Damascus in 1129. This did not prevent him from asserting his rights against fellow Christians by force, if necessary, or, as in 1127 during a dispute with the new prince of Antioch, Bohemund II, with the assistance of the Turks. The story of Joscelin’s death in 1131 provided Outremer with one of its epic tales. Infuriated by his son’s cowardice in the face of an attack from Anatolia, Joscelin, seriously ill and bedridden, insisted on leading out his troops borne on a litter. Seeing this, the invaders hurriedly withdrew. On receiving the news, Joscelin, ordering his litter to be put down on the road, died giving thanks to God.4

  The failure of the Franks to take Aleppo in 1125 opened the way instead to the new atabeg of Mosul, Imad al-Din Zengi, to resolve the anarchy in the city by occupying it in 1128. The settled union of Mosul and Aleppo posed a serious threat to Edessa, especially after the Franks’ failure in 1129 to take Damascus, which now attracted Zengi’s attention. Although primarily concerned with affairs further east and the politics of the Seljuk Baghdad sultanate, Zengi steadily increased his hold on the eastern frontiers of northern Outremer. In 1137 he captured the Frankish castle of Montferrand (Ba ‘rin), the important Muslim city of Homs in 1138 and the strategically significant town of Baalbek in the Biqa valley in 1140, where he installed as garrison commander a Kurdish mercenary, Naim al-Din Ayyub, Saladin’s father. The failure of an uneasy Byzantine/Antiochene/Edessan army to capture Aleppo and Shaizar in 1138 removed Greek intervention in Syria for a generation but allowed Zengi a free hand, with Damascus compelling its ruler Unur to arrange a treaty with King Fulk of Jerusalem in 1139. This increased concentration on the south left Edessa vulnerable.

  Joscelin II, no great general, continued his father’s active diplomacy with Muslim and Armenian neighbours, but his county appeared fair game for Turkish raiders from north, east and south. The adverse economic effect of this on the county weakened Joscelin’s political hold on his Syrian and Armenian subjects and his capacity to sustain mercenaries. The political chaos in Antioch in the 1130s and the loosening of the intimate ties with Jerusalem on the death of Baldwin II in 1131 left Edessa further exposed, its viability depending more than ever on external military aid. It is perhaps significant that the Franks of Edessa, few in number and reliant on non-Frankish subjects and allies, appear not to have begun a substantial programme of stone castle building. The evidence from the Frankish stronghold of Turbessel is inconclusive, but at Edessa itself existing fortifications seem to hav
e sufficed with local modifications. In 1122, the Armenian governor of the city, Vasil, constructed a new tower, round after an Armenian design.5 The absence of new fortifications did not mean that the counts of Edessa were defenceless, rather that, unlike their peers elsewhere across Outremer, they lacked the wealth to undertake new building projects.

  Yet the downfall of the county resulted from an opportunist raid rather than systemic collapse, although it occurred through Latin weakness and Edessan diplomacy. The kingdom of Jerusalem was embroiled in internal difficulties following the sudden death of King Fulk in 1143. Antioch, although spared a fresh Byzantine invasion by the death of John II Comnenus in 1143, remained preoccupied, relations between Joscelin II and Prince Raymond of Antioch later described as being of ‘insatiable hatred’.6 Joscelin II’s alliances with Muslims opposed to Zengi gave the atabeg an excuse to attack the eastern frontier of the county in the autumn of 1144 when the count was away from Edessa on campaign towards Aleppo. Edessa itself held out for a bare four weeks before falling to Zengi on Christmas Eve 1144. Despite cursory attempts at reconquest, the county east of the Euphrates was lost. Joscelin retained the western half, based on Turbessel, where his father had begun forty years earlier. Despite the murder of Zengi in 1146, the failure of the Second Crusade (1146–8), civil war in Jerusalem and the defeat and death of Raymond of Antioch at Inab in 1149 left the rump of the county exposed. In 1150, Joscelin was captured by troops of Zengi’s son Nur al-Din and imprisoned for the remaining nine years of his life in Aleppo, allegedly having to endure regular torture. His wife read the signs and sold her remaining forts to the Greek emperor Manuel I in 1150. These were overrun by Nur al-Din the following year. The Christian bulwark to the east that had posed a potential threat to the heartlands of Turkish power was lost for ever, a sign that the political chaos of the early twelfth century that had permitted, even encouraged, the political opportunism of the first two Baldwins and Joscelin I was giving place to an ominous and growing Muslim unity in Syria that challenged all Latin Outremer.

  ANTIOCH

  Like the county of Edessa, the principality of Antioch owed its creation to the secular impulses of the First Crusade. Any talk of Antioch as the first see of St Peter tended to be suppressed by a jealous papacy. Like Edessa, too, politics and society in Antioch followed indigenous patterns: Greek, Armenian and Muslim. Formed out of the ambitions and rivalries of the great western army of invasion in 1097–8, the principality survived by adapting and exploiting local conditions to forge a more pluralist polity than elsewhere in Outremer, embracing Greek and Sicilian institutional practices, in which marcher lords, vassals, tenants and administrators were western European, Armenian and even Muslim. Antioch’s vigorous independent identity sat uneasily between regular Byzantine claims to overlordship and the repeated need for the kings of Jerusalem to rescue the principality from succession crises as prince after prince proved extraordinarily accident-prone and unlucky. Although its politics, self-image and strategic position allied its fortunes with the Holy Land, Antioch could not escape its ties with Byzantium nor its interests in Cilicia, its enforced acceptance of Byzantine suzerainty in 1137, 1145 and 1158–9 in many ways ensuring its autonomy from Jerusalem.7

  Bohemund’s establishment of his control over Antioch in 1098–9 seemed to offer the prospect of the recreation of the pre-1084 Byzantine administrative region, or theme, based on the city. However, he confronted stern obstacles. In Cilicia his influence faced challenge from the Byzantine emperor and the local Armenian nobility keen to win independence by playing off Greek against Latin. On the Syrian coast and to the south and east of Antioch towards the frontier with Aleppo, Raymond of Toulouse and the Byzantines competed for dominion. By the time Bohemund was captured by the Danishmends in August 1100 attempting to relieve Melitene, he had lost control of Cilicia and Lattakiah to the Greeks and failed to exert clear authority over al-Bara and Ma ‘arrat. Thereafter, its charismatic founder exercised very little influence on the formation of the principality. In a Danishmend prison between 1100 and 1103, disastrously defeated at Harran in 1104, Bohemund left the east for good early in 1105 to chase his destiny against Byzantium.

  The real founder of the principality of Antioch was Bohemund’s nephew, Tancred of Lecce, regent 1101–3 then effectively prince 1105–12. Despite numerous reversals, by the time of his death Tancred had recovered Cilicia; extended Antiochene overlordship over Armenian princes to the north; incorporated the Ruj valley and the Jabal as-Summaq after defeating the Aleppans at Artah in 1105; effectively annexed Edessa between 1104 and 1108; occupied the ports of Lattakiah, Baniyas and, briefly, Jubail; pushed Antiochene frontiers east of the Jabal Talat and south to Apamea to threaten Aleppo and Shaizar respectively, both cities at various times paying tribute to the princes of Antioch. Despite his failed attempt to defy King Baldwin I over Edessa in 1109–10 and coming off worse in the succession dispute in Tripoli in 1109, Tancred’s Antioch dominated northern Syria, sufficiently strong to withstand the invasions of Mawdud of Mosul (1110–13); he was confident that the tactic of avoiding pitched battles would not destroy the inner cohesion of his territories. A network of marcher lordships strung along the borders afforded protection to the central areas of the Orontes valley, even when the frontiers themselves were breached. After one such incursion in 1115, Roger of Salerno won a crushing victory at Tell Danith over Bursuq of Hamadan, commander of an army sent by the sultan in Baghdad, to resecure the vulnerable south-eastern frontier. Lasting security received attention with Prince Roger’s capture of the castles at Saone, Balatonos and Marqab. In 1119, Roger’s luck ran out when Il-Ghazi of Mardin annihilated the Antiochene army at the Field of Blood. However, even this revealed the principality’s strength. Roger had foolishly not waited for reinforcements from the south before equally rashly seeking a pitched battle. Yet Baldwin II contrived to retrieve the situation through the continued resistance of the frontier garrisons buying him time and the efficiency of the general mobilization he ordered at Antioch, and not, as some contemporaries suggested, because the victorious Il-Ghazi was an irredeemable bingeing sot.8

  The survival of Antioch after the disaster of 1119 revealed the character of the regime built by Tancred and Roger. The administration displayed continuities with its Byzantine predecessor, as in the office of duke, dux, in the city of Antioch, while the princely household offices – chancellor, seneschal, chamberlain – were reminiscent of similar positions in southern and northern Norman courts in the west, perhaps unsurprisingly as many of the lords enfeoffed in the principality can be traced to Normandy or southern Italy and Sicily. Some may have gathered around Tancred during his adventurous career on the First Crusade and his territorial forays in Judea and Galilee. Others may have been supporters of Bohemund in 1098. Most importantly, the Antioch baronage appeared consistently loyal to their princes in the formative period of Frankish rule and thereafter to the principality’s independent integrity. In 1135, the barons rejected overtures made to Byzantium by the wilful dowager Princess Alice. In 1161–3 they forced her flighty daughter Constance to install her son Bohemund III as prince.9 The constant threat of invasion and dispossession; the vigorous personal support provided by the princes; and the lack of central interference in the workings of their lordships encouraged baronial loyalty. Rainald Masoir built up a strong lordship in the south of the principality, centred on Baniyas and Marqab. Despite the uncertainties and chaos after 1119, he associated himself with the regency government of Baldwin II and, after the arrival of the young Bohemund II in 1126, rose in princely favour to become constable in 1127, uniquely as a substantial landowner holding a household office. In the early 1130s, after the death of Bohemund II in battle (1130), Rainald acted as regent for a few years. The rewards were obvious. Rainald’s origins are obscure, yet his son was considered grand enough to marry the daughter of the count of Tripoli and his wife, Cecilia, was the widow of Tancred and illegitimate daughter of King Philip I, the Fat, of France.10

&n
bsp; Just as they relied on cooperation with the prince, Antiochene marcher lords, as elsewhere in Outremer, could not afford to adopt an inflexible siege mentality towards their Muslim neighbours. Robert FitzFulk, known as the Leper, held, among other properties, the fortress of Zardana on the frontier with Aleppo. Unsurprisingly, he established alliances with those Muslim rulers hostile to Aleppo, including Il-Ghazi of Mardin and Tughtegin, atabeg of Damascus, joining them in a military compact in 1115. Tughtegin was even remembered as being Robert’s friend, although this did not prevent him personally decapitating Robert in 1119.11 Less fraught were relations between Alan, lord of al-Atharib, another frontier fort between Antioch and Aleppo, and his Muslim physician, the chronicler Hamdan Ibn Abd al-Rahmin (c.1071–1147/8) whose reward for healing Alan was the grant of a village and its revenues. Hamdan assisted in regional administration, at one point presiding over the diwan (writing office) at Ma ‘arrat al-Nu ‘man. However, Hamdan’s opportunism matched that of any Frank. In 1128, he transferred his allegiance to Zengi at Aleppo, returning to administer the same border region he had previously managed for the Christians after its conquest by his new master.12 Hamdan was unusual but not unique. In 1118, Prince Roger granted three villages to a local Muslim sheikh.13 The rhetoric of holy war, so favoured by clerical observers such as the Antiochene chancellor Walter in his account of the vicissitudes of Prince Roger, concealed inter-faith cooperation and mutual self-interest, as in the joint campaign by Prince Roger, Tughtegin of Damascus and Il-Ghazi of Mardin against the Seljuks in 1115.14