Bānyās (with topography).
For any besieger, three fundamental questions dominated the decision to invest a stronghold or not. The first involved the relative strength of the attackers and defenders: did the besieger have sufficient forces and provisions, or access to such, to overcome the defenders and their defences, or to starve the garrison into surrender if necessary? The second was an assessment of value: how much time, money and manpower was the besieger willing to invest to take the stronghold and how aggressively was he willing to commit these resources? The third involved the probability of relief: what was the likelihood that a relief force would attempt to break the siege, how strong would it be, and when might it arrive? Even if a besieger could afford the material costs of a long siege, morale could be hard to maintain; calls to return home regularly arose during any lengthy campaign that failed or ceased to produce profitable plunder or wealth of another kind. On the defenders’ side, there was an inverse equation to consider involving the same variables. How confident were they in their ability to defend their fortifications? How reliable was their potential source(s) of relief? And what terms of surrender were they willing to accept if necessary? While an attacker was often quite eager to offer generous terms at the start of a siege, this generosity typically decreased as investments of time and money mounted.
The threat posed by a potential relief army had a considerable influence on the strategy and tactics that an invading force might plan to employ. If they believed a considerable force could be assembled and brought forward quickly, the besiegers might execute a very aggressive siege, committing large numbers of men to frontal assaults or efforts to compromise the defenders’ fortifications. If they believed the assembling force would be meagre, the besiegers might opt to divide their force, leaving a contingent to oppose the besieged stronghold while the remainder moved to engage the relieving army before it could break the siege or bring supplies to the defenders.
A siege undertaken with the expectation that relief would rapidly arrive carried significant risk. Alternatively, if relief was not expected to materialize quickly, besiegers had more freedom to design their approach to fit their strengths and resources. Despite the planning that went into any campaign, the unknown elements inherent in any estimation of an opponent’s strength and willingness to commit to it allowed for a wide range of strategic options depending on what else was happening at the time.
Early Opportunism
The initial gains made by the Franks came about through expediency. In 1097, as the First Crusade moved through southern Anatolia, Baldwin of Boulogne and Tancred broke off from the main army and invaded Cilicia. The Muslim garrison of Tarsus surrendered, at least nominally, to Tancred and then to Baldwin, who followed with a larger force and compelled Tancred to move on. Tancred, however, found an ally in a local Armenian ruler, quite likely Oshin of Lampron, who encouraged him to capture Mamistra (al-Maṣṣīsa), where a Latin garrison was then installed. This sideshow in Cilicia foreshadowed the way in which the Franks would take advantage of the fractious political environment over the following years, exploiting the isolation of Turkish garrisons and the rivalries among small Armenian lords.
In the autumn of 1097, while elements of the First Crusade besieged Antioch, Baldwin of Boulogne seized Turbessel and Ravandal. He then travelled to Edessa, where he became heir to the ruler of the city through an agreement of support. With the subsequent establishment of the county of Edessa, some Armenians were quick to accept Frankish suzerainty, recognizing the advantages of their support against rivals, be they Turks or fellow Armenians; others were more resistant, preferring to look elsewhere for assistance. This extended to smaller Turkish lords, such as Balduk, who sold Samosata to Baldwin and became a member of his household, though not a particularly trustworthy one, and Balak, the Artuqid ruler of Sarūj, who solicited Baldwin’s help to take a Muslim town that was defying him. Before becoming king of Jerusalem in 1118, one of Baldwin II’s last campaigns as count of Edessa was against the Armenian lords of northwestern Syria, which resulted in the capture of al-Bīra after a lengthy siege.
In some areas, where predominantly Turkish garrisons ruled over primarily Christian towns, some local populations, who probably interpreted the Franks as Byzantine mercenaries, rose up and expelled the Turks during the First Crusade. This took place at Turbessel in 1097 and Artāḥ the following year. Elsewhere, Muslim garrisons fled ahead of the crusaders’ arrival, as took place at Marash, which the crusaders seem to have given to Tʿatʿul, an Armenian, who loosely held it in the name of the Byzantine emperor until he gave it to Joscelin of Courtenay years later. The garrison of Ḥārim abandoned the hilltop castle following Riḍwān’s failed attempt to break the crusaders’ siege of Antioch in early 1098; a few months later, the garrison of ʿImm fled when news arrived that Antioch had fallen. Both were turned over to the Franks by their Christian populations. Rather than uncalculated acts of cowardice, the Muslim garrisons appreciated their prospects of relief were dismal should the Franks turn against them – by fleeing, they secured what were essentially the best possible terms of surrender. Later in the summer of 1098, the ruler of ʿAzāz took a different path, choosing to ally with the Franks against his lord, Riḍwān of Aleppo, and the town was duly relieved by Frankish forces when Riḍwān quickly moved to besiege it. Further south, power was more concentrated and Muslims outnumbered local Christians in many regions, limiting such opportunistic gains.
Relief Forces
Without the prospect of relief, there could be little hope for a besieged garrison. The norms of war placed a body of defenders entirely at the will of their captors if a stronghold fell by the sword, often resulting in death or enslavement for the defenders. Urban populations could expect the town to be thoroughly sacked, at the very least, while some degree of massacre was common and the complete enslavement of the population was not out of the question. Accordingly, a garrison was expected to hold out only as long as was reasonably possible and the vast majority of successful sieges ended with some kind of a negotiated surrender. Dying to the last man in defence of a stronghold might have been viewed by some as honourable or praise- worthy, but this was not an expectation – the stronghold was still lost and with it the lives of experienced fighters, who might otherwise have been able to help retake it.
Ultimately, it was up to a field force to break most sieges. Besiegers were typically obliged to take up vulnerable positions, focusing their attention towards the besieged stronghold and stretching their forces in order to surround or blockade it. The arrival of a sufficiently large hostile army usually compelled the besiegers to reorganize themselves, as they quickly faced threats not only from the besieged, who might launch a sally from the stronghold, but also from the field army to their rear. This pressure was often sufficient to compel the attackers to lift the siege, and either confront the relief force or withdraw so as to avoid an engagement.
In 1101, Suqmān ibn Artuq’s siege of Sarūj was frustrated by the arrival of Frankish forces. Suqmān and his brother, Īlghāzī, had inherited the emirate of Jerusalem from their father, to whom it had been bestowed by Tāj al-Dawla Tutush, and it was they who had lost it to the Fāṭimids in 1098. Following their relocation to northern Syria, Suqmān had taken Sarūjand given it to his nephew Balak, who appears to have given the town to Baldwin I in 1098, at which point a Frankish garrison was installed under Fulcher of Chartres (not to be confused with the contemporary historian of the same name). Eastern Christian accounts note that Suqmān initially defeated the forces of the new count of Edessa, Baldwin II, allowing the Muslims to besiege Sarūj. The town appears to have been captured but resistance continued from the citadel. Although his initial victory had bought him time, Suqmān was unable to take the citadel before Baldwin returned with another relief force, which defeated Suqmān and effectively ended the siege. Afterwards, the sources agree that there was widespread slaughter within the city at the hands of the Franks, suggesting that a faction of the locals may have attempted to assert their independence or preferred the rule of the Artuqids, to whom the town had belonged less than three years earlier. This defiance may have been encouraged by the death of the town’s Frankish lord, Fulcher, who had fallen in the first engagement between Baldwin and Suqmān. A similar siege played out in 1133 when Pons of Tripoli failed to defeat a large force of Turkoman raiders, who then besieged him in Montferrand. Pons was eventually able to slip out of the castle and make his way back to Tripoli, where he rallied a relief force and was joined by an additional contingent under King Fulk. This force then returned to Montferrand and defeated the Turkoman besiegers, compelling them to withdraw to nearby Rafaniyya.
Bānyās (with topography).
In May 1157, Nūr al-Dīn undertook his second siege of Frankish Bānyās, having failed in his first attempt a few years earlier. The Muslims’ ability to break into the town was in part due to Shīrkūh, one of Nūr al-Dīn’s generals and Saladin’s uncle, who defeated a small Frankish force outside Chastel Neuf, just 14km to the west on the other side of the Hula Valley, delaying relief efforts. Following his victory, Shīrkūh remained at Hūnīn (the town outside the castle), where he was able to monitor the crossroads leading away to Toron, Safed and Beaufort. According to Ibn al-Qalānisī, terms of surrender were offered by Humphrey II of Toron, who commanded the citadel’s defence, but they were rejected by Nūr al-Dīn. This proved to be a costly error as Baldwin III soon arrived with the army of Jerusalem, surprising both Nūr al-Dīn and Shīrkūh. Although the Muslims were compelled to break the siege, Nūr al-Dīn was able to defeat Baldwin’s force not long after at Jacob’s Ford, allowing him to return to Bānyās and renew his siege efforts. Hugh of Scandelion, who led the defence of Bānyās after Humphrey II of Toron joined Baldwin’s main army, opted not to defend the town but to concentrate on the citadel. This second siege was brief. Reynald of Châtillon, prince consort of Antioch, and Raymond III of Tripoli were already on their way with additional relief forces and Nūr al-Dīn withdrew as they approached Hūnīn, deciding not to try his luck in the field again.
The risks associated with opting to engage a relief force are revealed by events surrounding the struggle between al-Bursuqī and Baldwin II of Jerusalem for control of the region between Aleppo and Antioch. In 1125, al-Bursuqī and Ṭughtakīn besieged ʿAzāz, which dominated the plain north of Aleppo. They had undermined a section of its outer defences when a relief force arrived under Baldwin II. The Muslims turned to face the Franks but were defeated, compelling them to withdraw back to Aleppo. A year later, Baldwin II similarly broke al-Bursuqī’s siege of al-Athārib, about 30km west of Aleppo near the eastern entrance of the Sarmada Pass. Rather than risk a defeat similar to that of the previous year, al-Bursuqī opted to withdraw before he could be forced into an engagement. Al-Bursuqī appreciated the risks associated with losing his army, one that may have been gathered largely from his base of support east of the Euphrates, and that defeating a relief army was no guarantee that the accompanying siege would end successfully. In 1138, the besiegers of Edessa defeated a Frankish relief force from Samosata, but they remained unable to take the city. As it turned out, Joscelin I of Edessa’s final action was to relieve Kaysū n, which was besieged in 1131. The count was suffering from injuries sustained while undermining a tower months earlier, and died soon after receiving news that the Muslims had withdrawn upon the approach of his army.
The arrival of a relief force was the most common cause for a siege to be abandoned, but not all relief attempts were successful. Nūr al-Dīn’s efforts to take Ḥārim in 1162 had been frustrated by a Frankish relief force; however, he was more fortunate when he returned two years later. A relief army, led by Bohemond III of Antioch, Raymond III of Tripoli, Joscelin III, titular count of Edessa, Hugh VIII of Lusignan, who was on crusade, Toros II of Armenia and Constantine Coloman (Kalamonos), the Byzantine governor of Cilicia, moved to break the siege, but this force was defeated at the battle of Ḥārim. Nūr al-Dīn had initially withdrawn to Artāḥ upon the approach of this force, but then rapidly turned on the Christians, capturing many of their leaders. Nūr al-Dīn then returned to Ḥārim and renewed his siege, taking the stronghold a few days later.
Similar events had played out in the spring of 1105, as Tancred, then regent of Antioch, competed with Riḍwān of Aleppo for control of Artāḥ. The town was a strategic stronghold at the western entrance to the Sarmada Pass, through which the road between Antioch and Aleppo ran as it crossed the Syrian Coastal Mountains. The Armenian garrison declared itself for Riḍwān, causing the forces of Antioch and Aleppo to race towards the town. Tancred arrived first and established a siege before turning the bulk of his force against Riḍwān. In the engagement that followed, Tancred emerged victorious, compelling Riḍwān to flee and allowing the Franks to take complete control of the Antioch Basin.
In 1109, ʿArqa, still in Muslim hands, requested aid from Damascus. The town was either in dire straits due to the pressure applied by the Franks loosely blockading Tripoli, or its governor, a mamlūk of the ruler of Tripoli, believed this to be an opportunity to gain greater independence from his master. In either case, the town offered its submission if Ṭughtakīn would come and take possession. Ṭughtakīn sent an officer ahead to take control of the town but the movement of his main army was hindered by winter rains and snow, delaying his arrival by two months. Sensing weakness, William-Jordan seized the moment to besiege the town. When Ṭughtakīn arrived, he declined to engage the superior Frankish force he found waiting for him, instead withdrawing to Homs as elements of William-Jordan’s army harassed him along the way. ʿArqa continued to hold out, but without any prospect of relief, and its defenders ultimately surrendered when their provisions ran out.
With an army to their rear, besieging forces were typically compelled either to turn and fight or to withdraw. Very infrequently, besiegers were sufficiently strong to hold their position against the relief force that had come against them. This required exceptional focus and often considerable short-term losses: if the besiegers were not immediately distracted, it was not uncommon for the relief force to invade and plunder the besiegers’ territory. The siege of Tyre in 1124 is one such rare example where the besiegers maintained their focus despite the assembly of a hostile force to their rear.
Having failed to take Tyre during an ambitious siege over the winter of 1111/12, the Franks returned at a time when Baldwin II was a prisoner of Balak. The Franks were prompted into action by the arrival of a Venetian fleet, which had come to the Levant with the intention of taking part in the siege of a significant port, for which they would expect commercial privileges once the city was taken. It was decided that they would attack Tyre rather than Ascalon, and preparations were made for a long siege. Upon their arrival, the besiegers dug two lines of defences: one cut across the isthmus to provide protection against sudden sallies by Tyre’s garrison; the other, on the eastern side of the siege camp, sheltered them from the attacks of a potential relief army. The siege was indeed a long one, but the Frankish siege works proved sufficient. Attempting to pull the Franks away from Tyre, Fāṭimid raiders from Ascalon reached as far as Jerusalem. From the other direction, Ṭughtakīn moved a considerable force to Bānyās, sending raiders into Galilee and threatening to cross the Jordan and invade the kingdom in force. Neither, however, was prepared to attack the besiegers’ fortified camp and the Franks were willing to endure the raids in order to continue the siege. When this became clear, and Tyre’s provisions were exhausted, Ṭughtakīn stepped in and negotiated the city’s surrender, although it was technically still under Fāṭimid rule.
Tyre (with topography).
The siege of Acre (1189–91) played out in a similar, if more prolonged and desperate, manner. Guy of Lusignan’s inability to quickly take the city led to a protracted siege, in which the Franks, joined by successive elements of the Third Crusade, were in turn besieged by Saladin’s field army. During the Fifth Crusade, the siege of Damietta was also executed with a considerable Muslim force in the region. Al-Kāmil kept the bulk of his army on the east side of the river through the first phase of the siege, pulling back to Manṣūra once the Franks gained the right bank in February 1219. Throughout both phases of the siege, the crusaders and their camps were subjected to frequent attacks.
Instances in which besiegers were not dislodged by the arrival of a relief force were rare, usually corresponding with Frankish sieges that were launched following the arrival of large groups of Europeans. The Mamlūks’ dominance of the region in the years following 1260 provided them with the numbers to conduct similar operations; however, their strength was so significant that their sieges were seldom challenged by Frankish forces.
For an army unwilling to maintain a siege with a relief force to its rear, the decision whether to withdraw, cutting its losses, or to turn and fight, risking even greater losses if defeated, was critical. The defeat of a relief force typically dealt a devastating blow to the morale of the besieged, often leading the defenders to seek terms shortly thereafter, as was the case at Ḥārim in 1164. On the other hand, as came about at Bānyās in 1157, lifting a siege, even temporarily, provided the defenders with a brief respite and quite possibly the chance to bring in provisions – even if the relief force were defeated, the besiegers might be compelled to start their siege all over again.
Although the battle of Hattin drastically upset and reshaped the balance of power, the principle of relief remained the same through the thirteenth century. In 1199, the efforts of al-Ẓāhir and al-Afḍal to besiege Damascus and their uncle, al-ʿĀdil, were foiled with the arrival of a relief army led by al-Kāmil, al-ʿĀdil’s son, bringing to an end a siege that had lasted six months. Ironically, the siege of Damascus had the effect of breaking another siege: al-Kāmil had been besieging Mardin until news arrived suggesting he needed to move to his father’s aid. Al-ʿĀdil had initiated the siege of Mardin almost a year earlier, leaving it to al-Kāmil to complete when he returned to face brewing opposition in western Syria. The ultimately successful defence of Damascus allowed al-ʿĀdil to move on Cairo, from which point he emerged as the most powerful Ayyūbid ruler since the death of Saladin in 1193.
In January 1265, Baybars, who had become the most powerful man in the region little more than four years earlier, moved out of Cairo with his army to relieve al-Bīra. The important stronghold, which dominated a crossing of the upper Euphrates, had been invested by Mongol forces and their Armenian allies from late 1264. Moving with the main body of his army, Baybars sent an advance force ahead to support the besieged garrison and stiffen their resolve. Whether it was the size of this advance force or the foreshadowing of the larger Mamlūk army to come, the Mongols were persuaded to lift the siege and withdraw while Baybars was still at Gaza (700km away).
The Consequences of Battles
One of the ways a besieging force could avoid a relief army appearing to its rear was to engage that force before initiating a siege. Seeking an engagement in this way should not be considered in a Napoleonic sense, where annihilating an enemy’s army was the key to victory; instead, it was a means of temporarily weakening an opponent to gain enough time and space to take a stronghold, and with it the surrounding region. This was the potential prize that an invading army stood to gain if it was willing to commit to battle, but seeking such an engagement was rarely planned from the outset of a campaign.
On 28 June 1119, the Franks suffered one of their greatest defeats in an engagement known as the battle of the Field of Blood. Roger of Antioch had mustered his forces in response to an invasion by Īlghāzī, taking up a position near Sarmada from which he could rapidly move to relieve al-Athārib or any other stronghold that might be invested. Although this spot allowed the Franks to shield the Antioch Basin and quickly relieve a besieged stronghold to the east, the topography limited their ability to withdraw rapidly if necessary and the surrounding hills screened the movements of some elements of the Muslim army. Rather than commit to a siege with a large force mustered against him, Īlghāzī seized the opportunity to confront the vulnerable Frankish army, surrounding it on three sides. The Franks were crushed in the battle that followed and Roger lay among the dead. Īlghāzī then sent a contingent to raid the region around Antioch, providing his Turkoman forces with plunder and tying down the remaining Franks in the city, while he took al-Athārib and Zardanā, which submitted with minimal opposition.
Major battles.
Īlghāzī’s decision not to invest Antioch directly has been criticized, but it was probably quite sound. Had Antioch surrendered without much of a fight, he would then have faced the wrath of the Byzantines, who had never given up their claim to the city. If he were to be drawn into a more protracted siege, he risked the sudden arrival of another Frankish army – as would play out countless times, the defeat of one Frankish force precipitated the arrival of another soon after. As it turned out, Baldwin II soon appeared with the army of Jerusalem. Īlghāzī’s measured reaction following his victory left his forces prepared to confront Baldwin’s army, fighting it to a standstill on 14 August. Both sides then withdrew from the region, preventing Īlghāzī from making any further gains but securing those he had already made.
Similar events may have played out in January 1126 when Baldwin II of Jerusalem invaded the Ḥawrān. According to Ibn al-Qalānisī, the Franks planned to besiege Damascus but were prevented from doing so when Ṭughtakīn intercepted and defeated them on 25 January – Ṭughtakīn prevailing against Baldwin where Roger had failed against Īlghāzī. The reality, however, appears to have been somewhat different. When considering that forces from neither Antioch nor Edessa were involved in this campaign, nor was a significant body of crusaders, it is very unlikely that Baldwin had any intention of besieging Damascus. Revealingly, Baldwin made no attempt to garrison or defend any of the small strongholds that he captured leading up to the battle, indicating he did not intend to claim or otherwise hold the region he had invaded. The aims of this campaign were instead short-term wealth, which was gathered through raiding, and a show of force, perhaps something he hoped the Damascenes would recall when terms of a truce were next negotiated. The engagement that came about, like the battle of the Field of Blood, was probably the result of a series of circumstances rather than strategic planning: neither side had set out with the objective of drawing the other side into a major pitched battle.
One of Nūr al-Dīn’s most decisive victories over the Franks took place before his power had expanded beyond the plain around Aleppo. On 29 June 1149, his army defeated Raymond of Antioch, Reynald of Marash and ʿAlī ibn Wafā, the leader of the Assassins, at the battle of Inab. All three men were dead by the end of the day, allowing Nūr al-Dīn to quickly acquire Ḥārim, ʿImm and Artāḥ. His opponents had mustered against him after he had raided around Ḥārim and then settled down to besiege Inab. When the Franks approached to break the siege, Nūr al-Dīn pulled back his forces, but, learning they had made camp in the open ground around Inab, he opted to risk battle. Despite crushing his opponents in the engagement that followed, Nūr al-Dīn, like I¯lghā zī thirty years earlier, opted not to attack Antioch. He decided instead to join elements of his army that had laid siege to Apamea, solidifying his control over the region north of Hama.
Many of Saladin’s early campaigns against the Franks displayed greater restraint than those from 1182 onward, by which point he had humbled many of his Muslim rivals and his dominance of western Syria was near complete. In late 1178, while Saladin was busy subduing Shams al-Dīn Ibn al-Muqaddam, muqṭaʿ of Baalbek, the Franks began construction of the castle at Jacob’s Ford. The following May, just weeks after the death of Humphrey II of Toron, a capable northern lord and constable to the sickly Baldwin IV, Saladin took the first of a series of actions against the incomplete castle. The attack went on in some form for a week, but was probably limited to sacking the work site. Saladin returned to Bānyās in June, sending raiders towards Sidon. In response, Baldwin IV marched north to Toron and then east towards the Hula Valley, where he could observe Saladin’s position, fortuitously intercepting the returning Muslim raiders on 15 June. Saladin, who had stored his baggage in Bānyās, saw his raiders fall in with the Franks and, noticing the disorganization of the latter, chose to commit his forces to a full attack. The Muslims emerged victorious and many prominent Franks were taken prisoner. Rather than push his advantage, Saladin withdrew and waited more than two months before returning to the area and finally taking the Templar castle at Jacob’s Ford, overcoming its defenders while Baldwin IV was still assembling his forces at Tiberias.
The degree to which Saladin’s victory in June facilitated his capture of Jacob’s Ford is hard to determine, considering the castle was still incomplete and fell to the sword just six days after the Muslims arrived. However, he appears to have recognized the advantage he had gained by weakening the army of Jerusalem, which together with his consolidated power base, persuaded him to launch bolder campaigns against the Franks. It seems that he even sought to provoke the Franks into offering battle during his invasions of Galilee in 1182 and 1183. On their part, the Franks avoided a defeat by shadowing Saladin’s movements, travelling between defensible positions. Small strongholds, including Baysān (Bethsan, Scythopolis), Forbelet (Afrabalā), Le Petit Gérin and others, were evacuated in 1183 ahead of the advancing Muslim army but, with the army of Jerusalem still intact, Saladin made no effort to hold them. He was finally able to bring his superior numbers to bear in 1187. Judging that the Templars’ readiness to engage the raiders that had been sent across Raymond of Tripoli’s lands was an indication that the Franks were more willing to fight, Saladin gave up his siege of Kerak and invaded Galilee, attacking Tiberias to further prompt the army of Jerusalem into action. This precipitated the battle of Hattin.
The consequences of the battle of Hattin were far more dramatic than those that had followed the Field of Blood or Inab, and they almost certainly exceeded any expectations Saladin may have had earlier that spring. His resources had enabled him to field an army great enough to overcome not only the army of Jerusalem but also forces from Tripoli, under Raymond III, and a contingent from Antioch that Bohemond III had sent under his son, Raymond. The weakening of all three remaining Frankish principalities removed the threat that a subsequent relief army might arrive and allowed Saladin to systematically reduce the kingdom’s strongholds. The Mamlūks’ victory at ʿAyn Jālūt had a similar outcome to Hattin: the defeat of the Mongol force left in Syria under Kitbugha, so far from support, essentially gifted western Syria to the Mamlūks.
Battles were fairly unpredictable and sometimes it was parties other than the principal participants that benefited most in their aftermath. In 1104, following the battle of Ḥarrān, during which the forces of Jokermish of Mosul and Suqmān ibn Artuq of Mardin bested those of Bohemond I of Antioch and Baldwin II of Edessa, it was Riḍwān of Aleppo and the Byzantines, neither of whom had committed forces to the battle, who profited most. Riḍwān jumped on the moment of Frankish weakness and seized considerable territory in the Orontes Valley, while many towns in Cilicia, including Tarsus, Adana and Mamistra, turned away from the Franks and looked to the Byzantines for support. Ironically, although the battle had been fought east of the Euphrates and Baldwin II of Edessa, along with Joscelin of Courtney (later Joscelin I of Edessa), was taken prisoner after the battle, it was Bohemond who lost far more territory, losing it to two powers who had taken no part in the defeat of his army. Among the participants of the battle, Tancred might be considered the real victor, despite having been on the losing side. He had acted as regent of Antioch while Bohemond was a prisoner of the Dānishmands, but was left with little when he was released. With Baldwin II’s capture, Tancred was quickly acknowledged as the regent of the county of Edessa and once more became regent of Antioch when Bohemond left for Europe before the end of 1104, designating the administration of Edessa to Richard of Salerno.
Distraction
Seldom was battle sought as a means to gain territory: the circumstances in which it was a practical strategy were extremely limited and the results of battles could be far from predictable. A more popular strategy was to strike while an adversary was distracted by internal issues or an invasion by a third party; alternatively, a similar situation could be created by attacking along multiple fronts. Raids were often launched during such periods of distraction, while Nūr al-Dīn used them more often than most to attack strongholds.
When Nūr al-Dīn approved Shīrkūh’s intervention in Egypt, he recognized that this was not only an opportunity to enhance his wealth and extend his influence, but that it could also weaken the Franks. In 1164, while Amalric moved to counter Shīrkūh in Egypt, Nūr al-Dīn besieged Ḥārim. He had attacked Ḥārim two years earlier but was frustrated by the arrival of a Frankish relief force. Although Amalric was now distracted, a Christian coalition assembled and moved to break the siege. Nūr al-Dīn engaged and defeated this force, in what is known as the battle of Ḥārim, allowing him to take Ḥārim a few days later. Despite the advantage he had gained, Nūr al-Dīn stopped short of attacking Antioch, fearing Byzantine intervention – Bohemond III of Antioch and Constantine Coloman, both Byzantine vassals, were among those captured following the battle but they were quickly ransomed. Word reached Amalric that Ḥārim had fallen while he was besieging Bilbays, a fortified town on the eastern side of the Nile Delta, 790km from Ḥārim (as the crow flies). Learning that Nūr al-Dīn was then moving against Bānyās, the king was compelled to negotiate a truce with Shīrkūh and withdraw from Egypt.
In the aftermath of his victory outside Ḥārim, Nūr al-Dīn must have felt fairly comfortable moving against another Frankish stronghold. He dismissed the Mesopotamian components of his army and, feigning a move against Tiberias, rapidly invested Bānyās. As hoped, the garrison surrendered before Amalric and Humphrey II of Toron, the powerful lord of Bānyās and constable of the kingdom, could complete the more than 900km journey (500km as the crow flies) from Bilbays to Bānyās.
Raymond III of Tripoli was among those captured at the battle of Ḥārim, leaving Amalric regent of the county. When the king was next in Egypt, chasing Shīrkūh up and down the Nile in 1167, Nūr al-Dīn invaded the county of Tripoli and captured a number of strongholds in the Homs–Tripoli gap. It seems he bypassed Crac but that ʿAkkā r was briefly taken; the castle was recaptured by the Franks around the end of 1169 or start of 1170. About a month after this campaign in Lebanon, Nūr al-Dīn launched another into Galilee. Using Bānyās as a staging point, he targeted Chastel Neuf, on the opposite side of the Hula Valley. Although it was an impressive stronghold, some members of its garrison may have accompanied Amalric to Egypt and those who remained probably recognized that they stood little chance of holding out with the kingdom’s army so far away. Accordingly, the garrison set fire to Chastel Neuf and fled. Nūr al-Dīn similarly recognized the difficulties associated with holding the stronghold upon Amalric’s return, leading him to complete the destruction of the castle rather than attempt to repair and then garrison it.
When Nūr al-Dīn died in May 1174, the various parties who scrambled for control of his realm used the preoccupation of others to their advantage. Saladin, who emerged the eventual victor, was aided by the rivalry between Nūr al-Dīn’s nephews. Nūr al-Dīn had mediated the succession of Mosul following the death of his brother, Quṭb al-Dīn Mawdūd, in September 1170: he recognized Quṭb al-Dīn’s second and preferred son, Sayf al-Dīn Ghāzī II, as ruler of Mosul; granted Sinjār as compensation to Quṭb al-Dīn’s eldest son, ʿImād al-Dīn Zankī, namesake of their dynasty’s progenitor; and took Raqqa for himself. As Sayf al-Dīn settled into his new position, he would have become increasingly aware that Mosul, not Aleppo, was the traditional seat of Zankid rule – upon Zankī’s death, Mosul had gone to his eldest son, Sayf al-Dīn Ghāzī I, Aleppo to his second son, Nūr al-Dīn, and Homs to his third son, Quṭb al-Dīn Mawdūd, who later gained Mosul upon his eldest brother’s death in 1149. Accordingly, Sayf al-Dīn was quick to offer support to his young cousin, Nūr al-Dīn’s son, al-Ṣāliḥ Ismāʿīl, who had fled from Damascus to Aleppo a few months after his father’s death. This, Sayf al-Dīn hoped, would pull northwestern Syria back under Mosuli influence and away from Saladin (then ruler of Egypt), who lost little time extending his influence into western Syria.
After taking control of Damascus in November 1174, Saladin moved quickly against Homs, Hama and Aleppo. He was compelled to lift his siege of Aleppo at the end of January 1175 when Raymond III of Tripoli, who had ransomed himself in early 1174, seized the opportunity to attack Homs – Saladin had taken the town but not the citadel before moving on to Aleppo. Raymond withdrew with Saladin’s approach, and this time Saladin remained there until the citadel was captured on 17 March. To the east, Sayf al-Dīn Ghāzī II tasked his brother, ʿIzz al-Dīn Masʿūd, with confronting Saladin; meanwhile, he set out to besiege Sinjār and subdue their older brother, who had declared his support for Saladin. ʿIzz al-Dīn Masʿūd was defeated by Saladin near Hama on 13 April 1175, allowing the latter to briefly renew his siege of Aleppo before a truce was concluded. The defeat of his brother led Sayf al-Dīn Ghāzī II to lift his siege of Sinjār and move to support Aleppo in person, although he too was defeated by Saladin, who had been reinforced, on 22 April 1176 at Tell al-Sulṭān.
While Saladin fought with Nūr al-Dīn’s kin, he remained focused on securing control of western Syria. Following his agreement to a truce with Aleppo after defeating ʿIzz al-Dīn Masʿūd in 1175, Saladin besieged Montferrand (Baʿrīn), an iqṭāʿ of one of Nūr al-Dīn’s men. In the wake of his victory the following year, he shiftedhis focus northwards, investing and taking Buzāʿa, Manbij and ʿAzāz before returning to Aleppo and renewing siege efforts in late June 1176. A month later, a general peace was concluded between Saladin, al-Ṣāliḥ of Aleppo, Sayf al-Dīn Ghāzī II of Mosul, and the rulers of Mardin and Ḥiṣn Kayfā, supporters of Sayf al-Dīn. With his northern front secure, Saladin turned against the Nizā rī Assassins, who had recently attempted to kill him while he besieged ʿAzāz.
Multidirectional Attacks
By the early 1180s, Saladin had consolidated his position and amassed a degree of power unseen in the region since the arrival of the Franks. Whereas Nūr al-Dīn’s opportunistic invasions of Frankish lands in the 1160s were often facilitated by events in Egypt, Saladin was sufficiently powerful to launch a multidirectional attack against the kingdom of Jerusalem in August 1182 with no such outside assistance. The first prong was a naval element: an Egyptian fleet would move up the coast and blockade Beirut from the sea. The second was Saladin’s main army, which he kept poised in the Biqāʿ Valley until the fleet arrived, at which point it would attack Beirut from the landward side. As these forces pressed the city with aggressive attacks, a force from Egypt, the third prong, invaded the kingdom from the south. Unfortunately for Saladin, the southern element was too weak to distract the Franks. Baldwin IV, who had assembled his army at Ṣaffūriyya when news arrived that Saladin was on the march, remained focused on the northern elements, moving to Tyre to coordinate a relief force as soon as Saladin’s intentions became clear. Perhaps hoping that he could take the city quite quickly, Saladin had not brought siege engines and instead relied on his sappers and relays of frontal assaults. After three days, having made little headway and with news that a relief force was approaching, Saladin withdrew his army from Beirut and sent the fleet away. A Frankish flotilla arrived shortly after the Muslims had departed, while Baldwin IV, taking no chances, marched the army back to Ṣaffūriyya in case Saladin turned south and struck into Galilee.
Baybars’ mastery of logistics and ability to orchestrate multifaceted campaigns, facilitated by the wealth of the Mamlūk treasury, brought things to a whole new level, as seen in his elaborate push against Safed. Baybars left Egypt on 7 May 1266 and arrived at Gaza four days later. From there he arranged for the army of Homs to invade the Homs–Tripoli corridor, where it raided and captured the minor stronghold of Tuban as well as Ḥalba (Castrum Album), ʿArqa and Coliath, destroying the latter three. A part of this force then returned to Homs while the remainder joined Baybars, who had moved to Acre. From Acre, Baybars ordered a portion of his Syrian army to blockade Safed and another to oppose Beaufort, while detachments of the Egyptian army were dispatched to raid the coast around Tyre, Sidon and ʿAtlit, and another contingent was sent to attack Montfort. These arrangements prevented the Franks from concentrating or even coordinating their forces, limiting resistance while disguising the Mamlūks’ ultimate objective. Baybars remained in front of Acre for eight days before finally showing his hand. When his raiding parties returned, he moved his army to Safed, where a blockade was already in place. Despite the considerable force brought against it, the mighty Templar castle held out for six weeks.
Baybars’ campaign against Safed, 1266.
Baybars similarly divided his forces when moving against Antioch in 1268. Setting out from Homs, he split his army into three groups: one was sent towards Trapessac, another to Qalʿat Simʿān (Church of St Simeon Stylites), and the third he took to Jisr al-Shughr via Apamea, reuniting them around Antioch. By dividing his forces, Baybars allowed his army to move more freely while concealing his ultimate target. Bringing his forces together again, they were able to overwhelm Antioch’s defenders, storming the city’s walls four days after arriving.
Opportunism
It required considerable resources to launch attacks against distant fronts, as Saladin’s failed diversion in 1182 revealed. A more practical if less reliable means of achieving the same end was to take advantage of an alternative distraction. This could include an invasion by another neighbour or even an internal dispute, anything that might weaken the defending power’s ability to assemble and dispatch a relief force.
The cave castle of al-Ḥabis Jaldak, perched above the Yarmūk River in the Sawād, was one of the more exposed strongholds in the frontier between the kingdom of Jerusalem and emirate of Damascus, and as such was often attacked during moments of distraction. Ṭughtakīn seized the stronghold in the winter of 1111/12, while the Franks were busy besieging Tyre. It was later retaken by Frankish forces, following the death of Baldwin I, while Ṭughtakīn was preoccupied attempting to coordinate an attack against the Franks with Fāṭimid elements at Ascalon. A more orchestrated diversion contributed to the stronghold’s capture in 1182, when it was taken by Saladin’s nephew, Farrukhshāh, who led a secondary force against the cave castle during Saladin’s invasion of Galilee. Later the same year, while Saladin was away campaigning east of the Euphrates, the Franks launched a raid into Damascene territory, during which they captured the cave castle after a siege of three weeks. Whether the acquisition of al-Ḥabis Jaldak had been an objective of this campaign from the outset or not, Saladin’s absence provided the Franks with a significant window of time during which they could besiege the stronghold without fear that a large relief force would suddenly appear to their rear.
In late 1132, Fulk, who had been king for only a few months, found himself occupied in the north, first addressing the regency of Antioch and then confronting an invasion by Sayf al-Dīn Sawār, Zankī’s deputy in Aleppo. This preoccupation provided Shams al-Mulūk Ismāʿīl, who had come to power in Damascus that June, with an ideal opportunity to invest Bānyās, a Frankish possession since 1128. The town wall was rapidly undermined and the citadel, which briefly held out following the town’s capture, surrendered no more than five days after the start of the siege. When Fulk was again called north in 1137, Damascene forces sacked Nablus, which lacked a town wall. Zankī similarly capitalized on the absence of Joscelin II, who was assisting his Artuqid allies, when he besieged Edessa in 1144. Uncharacteristically, the count of Edessa found himself isolated and without support when he needed it. Fulk had died the previous year, leaving Jerusalem under the rule of his widow, Melisende, who acted as regent for their young son, Baldwin III, and Raymond of Antioch was more interested in extending his influence across Cilicia following the death of the Byzantine emperor John Comnenus the year before.
Whereas Nūr al-Dīn had a hand in controlling events in Egypt through the 1160s, using them to make gains in western Syria, the Franks had benefited from conflicts in Cairo, of which they had no part, a decade earlier. Baldwin III led the army of Jerusalem against Ascalon in January 1153, accompanied by a number of crusaders who had remained in the Levant following the Second Crusade. Relief efforts were hindered when the Fāṭimid vizier, Ibn al-Sallār, was murdered in April, quite likely with the support of Usa¯ ma ibn Munqidh, who was at that time in Cairo. The eventual arrival of an Egyptian fleet at Ascalon was not enough to discourage the Franks, who maintained the siege until it was brought to a successful conclusion in August. While the Franks inadvertently benefited from the civil strife in Egypt, Nūr al-Dīn took advantage of the army of Jerusalem’s preoccupation to move against Bānyās. Although a portion of the town’s garrison was probably with the army at Ascalon, Nūr al-Dīn did not press the siege very hard and withdrew after only a few days – Damascus remained his main objective.
One of the very few offensives launched by the Franks against a Muslim power in the second half of the thirteenth century was led by Prince Edward of England in 1271. Having recently arrived on crusade, the prince took advantage of Baybars’ distraction with a Mongol raid against Ḥarrān to attack the small stronghold of Qā qūn, a Mamlūk administrative centre that had once been part of the lordship of Caesarea Although Baybars believed that the Franks coordinated their attack with the Mongols, the move was probably more opportunistic than collaborative. In any case, this was more an attack or raid than a siege.
Truces
Peace treaties and alliances could be concluded for a number of reasons. For rulers looking to go on the offensive, they were a means to deny a third party an opportune moment to attack while their attention and army were diverted elsewhere. Alternatively, they could be an effective way of preventing a third party from sending relief or assistance to those who were invaded. While the former may have been more common, Muslim rulers turned to the latter with increasing frequency from the late twelfth century when planning campaigns against their Frankish neighbours, who repeatedly showed a willingness to put aside their differences and march to each other’s aid.
Peace during this period was something secured between individuals, acting as rulers of specific polities. Accordingly, peace agreements often ended with the death of one of the negotiating parties and, as the kings of Jerusalem and emirs of Aleppo/ Damascus came to dominate more than one seat of power, agreements could be limited to certain regions. For example, in the year following the battle of the Field of Blood, Baldwin II of Jerusalem, acting as regent of Antioch, confronted a force led by Īlghāzī; after which, the two men concluded a truce. That winter, Joscelin I of Edessa remained active, raiding Aleppan lands and recovering lost territory. This led Īlghāzī’s deputy in Aleppo to appeal to Baldwin, citing their truce. Baldwin, however, was able to claim that his authority, as king of Jerusalem and regent of Antioch, did not extend to the county of Edessa, and thus he had no authority over Joscelin and their truce did not extend to him; the raids continued. Although Joscelin had been Baldwin’s vassal while the latter had been count of Edessa, this ended when Baldwin became king of Jerusalem and Joscelin succeeded him as count. Despite the increasing hegemony of the kings of Jerusalem, which became a serious issue of contention during the early reign of Fulk, the county remained an independent principality.
In the same way that the truce between Baldwin and Īlghāzī did not include Edessa, neither did it include the emirate of Shayzar. The small Arab lordship had previously agreed to pay Antioch an annual tribute; however, this lapsed when Roger of Salerno died at the Field of Blood. Antiochene forces thus moved against Shayzar in force, compelling the ruling Banū Munqidh to renew their tribute, for which they were granted a year-long truce.
In 1186, when the monarchy of Jerusalem was particularly weak, Reynald of Châtillon, former prince-regent of Antioch and now lord of Transjordan, went so far as to claim the right to conduct his own peace agreements. Upon his coronation, Guy of Lusignan had renewed a peace that had been in place between Baldwin V and Saladin. When Reynald, Guy’s most powerful supporter, subsequently raided a caravan under Saladin’s protection, Guy chastised him, but Reynald rebuked the king in turn, declaring ‘he [Reynald] was lord of his land, just as he [Guy] was lord of his; and that he [Reynald] had no truce with the Saracens’.21 Reynald, however, was not the only lord of the kingdom to claim the right to exercise his own foreign policy.
In addition to his own autonomous county, Raymond III of Tripoli held the principality of Galilee, a lordship of the kingdom of Jerusalem, on behalf of his wife, Eschiva Bures, following their marriage in 1174. Unlike Reynald, Raymond was on good terms with Saladin and in open opposition to Guy of Lusignan. In early 1187, Raymond allowed an armed party of the sultan’s raiders to pass through his wife’s lands, and thus into the kingdom, under the conditions that they withdraw back across the Jordan by nightfall and disturb no property. It was these raiders that encountered and defeated the force of Templars heading from La Fève to Tiberias near the springs of Cresson. Although a small battle took place, the Muslims nevertheless withdrew from Frankish lands before nightfall, keeping to their agreement with Raymond.
Reynald of Châtillon is typically vilified and blamed for bringing about the battle of Hattin, while Raymond III of Tripoli is often presented as a champion of peace. But by allowing raiders to roam his lands, Raymond established the conditions that led to the engagement at the springs of Cresson. It was this that persuaded Saladin to give up his blockade of Kerak, which he had undertaken in response to Reynald’s raid on the caravan, to seek a greater prize further north. Although it can be hard to see these events independent of the broader context leading up to the battle of Hattin, both Reynald and Raymond were claiming a considerable degree of baronial autonomy during a period of monarchical weakness.
Although most peace agreements were respected, sometimes opportunities were simply too good to pass up. For example, in the spring of 1110, Mawdūd of Mosul led an army from the Jazīra against Edessa, compelling King Baldwin I to march his army to its relief as soon as he completed his siege of Beirut on 13 May. Mawdūd pulled back with the advance of the Franks, but then struck as they turned to withdraw, ambushing the Frankish army as it re-crossed the Euphrates. Riḍwān of Aleppo evidently thought that this was a greater defeat for the Franks than it turned out to be. Seizing what he believed to be an opportune moment, he broke his truce with Tancred and raided lands in the principality of Antioch. The move backfired. With their peace broken, Tancred besieged and took al-Athārib before demanding a considerable tribute from Riḍwān to renew their truce.
For figures whose lands were particularly exposed, due to the expanse of their realm or strength of their neighbours, tactical peace agreements were particularly valuable. Īlghāzī’s acquisition of Aleppo in 1117 forced him to split his focus between the Jazīra and western Syria. In 1121, he ordered his son, Shams al-Dawla, who administered Aleppo on his behalf, to conclude a truce with the Franks according to whatever terms they dictated, ratifying this agreement in person before returning east of the Euphrates. Īlghāzī used this interval to gather his eastern forces and returned to western Syria in June 1122, at which point he besieged Zardanā. While Īlghāzī had orchestrated this peace to buy time, Nūr al-Dīn was the main beneficiary of a later truce he had no part in arranging. Whether coincidental or intended, a two-year peace concluded between Damascus and Jerusalem in the spring of 1149 allowed elements of the Damascene army to join a push by Nūr al-Dīn (then ruler of Aleppo) against the principality of Antioch. The campaign that followed would be distinguished by the battle of Inab and Nūr al-Dīn’s acquisition of Ḥārim.
Until it collapsed in the decade following Zankī’s capture of Edessa in 1144, the county of Edessa, which straddled the Euphrates, proved a thorn in the side of Muslim rulers who held Aleppo as well as lands in Mesopotamia, notably Īlghāzī, who also held Mardin, and Zankī, whose principal seat of power was Mosul. Travelling via Manbij and the Euphrates crossing at Qalʿat Najm, Mardin and Aleppo are separated by a journey of 400km, regardless of whether the northern or southern route is taken, via Edessa or Raʾs al-ʿAyn respectively. From Mosul, the southern route to Aleppo, via Sinjār and either Qalʿat Najm or Raqqa, is almost 600km, slightly shorter than the northern route, via Nisibis, which is better watered but closer to the Frankish sphere of influence.
Even after Frankish rule withdrew west of the Euphrates, Nūr al-Dīn and later Saladin were often compelled to make peace with the Franks before campaigning in the Jazīra or southeastern Anatolia. In May 1155, Nūr al-Dīn negotiated a one-year peace with Antioch and concluded a similar peace with Jerusalem a few months later. Although the latter obliged him to pay a tribute of 8,000 Tyrian dinars, it freed Nūr al-Dīn to campaign against ʿAyntāb and the region that had formerly belonged to the county of Edessa. In the summer of 1175, Saladin similarly made peace with the Franks, allowing him to focus on Aleppo and Sayf al-Dīn Ghāzī II of Mosul. Five years later, Saladin arranged another peace in order to move against Qilij Arslān II of Rūm, assisting Nūr al-Dīn Muḥammad of Ḥiṣn Kayfā, and then campaign against the Armenians.
Saladin’s decision not to arrange a peace with the Franks before campaigning in the Jazīra through 1182–83 was significant. As he attempted to strengthen his rule east of the Euphrates, the Franks were almost compelled to strike in his absence, in order to show Saladin that they were still a threat worthy of consideration. Two invasions were launched from the kingdom of Jerusalem: the first was a raid towards Damascus, following which the Franks paused to take al-Ḥabis Jaldak on their return; the second was a more ambitious and complex series of actions. In December 1182, Baldwin IV took the main army and raided towards Damascus, as had been done weeks earlier. Further south, Raymond III of Tripoli led a secondary force against Bosra, which the Franks had not threatened since Nūr al-Dīn had acquired Damascus. The third, and by far the most famous and damaging to Saladin’s reputation, was undertaken by Reynald of Châtillon. Having constructed prefabricated boats at Kerak, Reynald arranged for them to be moved overland to Aqaba. From there, raiders preyed on the unfortified ports along the coast of the Red Sea, threatening Mecca and Medina, the holiest sites in Islam. Despite stretching his hegemony across most of the Jazīra and acquiring Aleppo at the end of his campaign, Saladin was unable to effect a blow to the Franks comparable to that which Reynald had dealt his image and perhaps pride. He invaded Galilee but the Franks declined battle, shadowing his force from a series of strong and well-watered positions. The effect Reynald’s raid had on Saladin is evident in the focus the sultan subsequently devoted to him, besieging Kerak twice in the twelve months following his withdrawal from Galilee in October 1183. A peace was later concluded between Saladin and the kingdom of Jerusalem following the death of Baldwin IV in March 1185, allowing Saladin to make one more (unsuccessful) attempt to take Mosul. It was the renewal of this peace, following the coronation of Guy of Lusignan in the late summer of 1186, that Reynald later disregarded, contributing to the events leading up to the battle of Hattin.
Although the battle of Hattin transformed the political landscape of the Levant, the utility of truces remained. Saladin crippled the principality of Antioch during the summer of 1188, but he was unable to complete its conquest. At the start of the campaign, Saladin had arranged for his son, al-Ẓāhir Ghāzī, and nephew, Taqī al-Dīn ‘Umar, to move into a position to threaten Antioch, tying down Bohemond III as Saladin led his main army through the county of Tripoli and then north into the principality of Antioch. Reluctant to invest the city of Antioch directly through the winter – his army was already growing restless – Saladin forced Bohemond to agree to a truce. This allowed Saladin to disband most of his army for the winter and freed him to return south to stamp out remaining resistance in Palestine.
By the time Mamlūk authority spread across Syria, the nature of royal power in the kingdom of Jerusalem had changed. The monarchy, often represented by a bailli (an administer who acted on behalf of a minor or absentee monarch), negotiated truces that extended only to royal lands, which were concentrated around Acre. This left the kingdom’s most powerful lords, including those of Beirut, Jaffa and Tyre, the latter two lordships having passed out of royal control, free or obliged, depending on the circumstances, to secure their own diplomatic arrangements.
When Baybars travelled to Damascus in May 1261, he entertained Frankish emissaries and showed himself willing to secure peace with Acre, Jaffa and Beirut, allowing him to solidify his position and ensuring Mamlūk trade continued to flow through Frankish ports. Five years later, as he arranged his forces around Safed, Baybars received representatives from the lords of Jaffa, Beirut and Tyre, as well as the Assassins, reproaching them all on various pretexts. After taking Safed, he travelled to Damascus, where he received a Hospitaller delegation and tentatively agreed to extend the peace in place with the order. It was not until the following year that the ten-year peace agreements negotiated with the Hospitallers and Tyre were confirmed, by which point Baybars had taken Chastel Neuf and Toron, which had both become parts of the lordship of Tyre, and destroyed the Hospitallers’ mill outside Acre. A truce with Beirut was also confirmed around this time, while overtures of peace from Sidon were rejected.
Jaffa was conspicuously left out of the peace agreements confirmed in 1267. The town was by this time isolated and exposed: to the north, Arsūf and Caesarea had been captured in 1265, and the death of John of Jaffa in 1266 left the town without the protection of a peace agreement. The circumstances were thus ideal when, in early 1268, Baybars found himself in Palestine, having mobilized his forces to confront a Mongol attack on Aleppo that failed to materialize. When envoys from Jaffa moved out to greet the Mamlūk army, Baybars had them detained, ordering his forces to launch sudden attacks against the town. The citadel surrendered later the same day.
In a move that has puzzled historians, Baybars agreed to surprisingly generous terms when negotiating a peace with envoys from Acre in April 1272. Once more marching his army through Palestine to confront a Mongol threat to the north, Baybars relinquished a number of estates that the Franks had not held under their previous peace agreement. Baybars’ power had increased since that time, so why he felt the need to make these concessions in order to secure his rear is unclear.
Rapid Attack
The main aim of a rapid or surprise attack was to catch the defenders of a stronghold off-guard. By preventing the defenders from preparing themselves sufficiently, attackers hoped to shorten the potential length of a siege, either by overcoming the stronghold’s defences relatively quickly or by forcing the defenders to seek terms of surrender before a relief force could be assembled. An example of an attack like this was Baybars’ capture of Jaffa in 1268. The Mamlūks’ overwhelming superiority was such that the suddenness with which they attacked Jaffa was probably more of a precaution than a necessity, conceived with a desire to avoid the costs of a longer siege. Baybars launched similar, if far less successful, sudden attacks against Acre in 1263, 1267 and 1271. On each occasion, he found the city prepared and his forces had to content themselves with raiding the surrounding gardens. For adversaries on a more equal footing, shortening the period of time that defenders could hold out was one of the easiest ways that attackers could try to avoid the complications associated with the arrival of a relief force.
In 1115, while Baldwin I and the army of Jerusalem were to the north confronting the Seljuk army led by Bursuq ibn Bursuq of Hamadan, an Egyptian force, supported by a fleet, attacked Jaffa. The Fāṭimids attempted to take the town by surprise, rushing its defences with ladders they had brought. Although the attackers managed to burn parts of the town gates, the defenders were able to keep them back from the walls for the most part. Foiled, the besiegers withdrew after a few days, the land force returning to Ascalon while the fleet proceeded to Tyre, which was still under Muslim rule. Another attack was carried out ten days later, but this lasted only a few hours and also failed. In October of the same year, Ṭughtakīn led a similar attack against Rafaniyya, finding greater success. Marching through the night without any baggage, the Muslim army surrounded the town, which had only recently fallen into Frankish hands, and successfully broke in before dawn.
Zankī successfully took Edessa in 1144 after a concerted siege, but his death two years later, while besieging Qalʿat Jaʿbar, provided Joscelin II with an opportunity to reclaim his patrimony. Leading a small force, Joscelin scaled Edessa’s walls one night and retook the town, but was unable to gain entrance to the citadel. Nūr al-Dīn, who had inherited Edessa along with Aleppo following Zankī’s death, wasted little time in bringing a force to relieve the citadel’s garrison. Even with the support of the Armenian population, Joscelin had little chance of holding the town and fled with his force upon Nūr al-Dīn’s arrival.
During Saladin’s multipronged attack against Beirut in 1182, the main army approached the city by crossing the Lebanon. The rugged path they took over the mountains prevented Saladin from bringing any heavy baggage, but allowed him to appear suddenly from an unlikely direction. Without siege equipment, he was limited to using his miners and frontal attacks, but Saladin’s sappers proved unable to repeat the success they had achieved against the far less impressive defences of Jacob’s Ford, which they had compromised in less than a week three years earlier. Having achieved little, and with a Frankish relief force approaching, Saladin cut his losses after three days and withdrew.
Ibn al- Athīr: Nūr al-Dīn’s capture of Munayṭira, c.1166
This year Nur al-Dln Mahmud ibn Zankl conquered the fortress of Munayṭira in Syria, which was in the hands of the Franks. He made no great mobilization, nor assembled all of his forces. He just marched lightly equipped and took them unaware. He knew that, if he assembled his army, the Franks would be on their guard and concentrate their troops. He seized the opportunity, marched to Munayṭira and put it under siege, attacking energetically. He took it by storm and killed or made captive those within and took large amounts of booty. The defenders had felt safe but God’s cavalry overwhelmed them suddenly before they were aware. The Franks gathered to repel him only after he had already taken it. Had they known that he was lightly equipped with a small number of troops, they would have hastened against him, but they imagined that he led a large host. After he had taken it, they dispersed and despaired of recovering it.
(Adapted from Ibn al- Athīr, trans. Richards, 2:161.)
In 1167, Nūr al-Dīn had made a similar move against Munayṭira, taking advantage of Amalric’s absence in Egypt. Leaving his baggage behind, Nūr al-Dīn besieged and took the castle, on the mountain road between Baalbek and the coast, before a Frankish relief force could be assembled. Baybars also employed this tactic in the spring of 1268. After taking Beaufort, he sent detachments out in different directions to conceal his objective, as he often did, leading his main force to Bānyās, where he sent his baggage and siege equipment to Damascus. He then continued on to Baalbek, from where he set out across the mountains to surprise Tripoli. Like Saladin’s attack on Beirut, Baybars had little chance against Tripoli’s stout urban defences, even with the element of surprise, so avoided a siege of the city and spread his forces through the area, raiding and taking a number of towers and minor cave castles. Baybars repeated this manoeuvre to the north when he set out to attack Margat in January 1270. Leaving his baggage on the eastern side of the Syrian Coastal Mountains, he led a cavalry force towards one of the mountain passes. The season proved to be his undoing. Although an attack would not be expected in the middle of winter, the rains were severe and he was forced to turn back. Undeterred, he made a second attempt about twenty days later but was again forced to turn around after entering the territory of the Nizārī Assassins.
* * *
Far more went into a siege than simply overcoming a stronghold’s defences. Although this was the defining part of a siege, success or failure was often determined by more distant elements. No stronghold was designed to hold out indefinitely; each was reliant on the prospect of relief – the least predictable and greatest challenge for besiegers. Avoiding the interference of a relief force could be done in a number of ways, which might include: striking fast, denying the defenders and a potential relief force the chance to prepare themselves; choosing a moment when relief would be weak or slow to assemble; or securing a peace, which might limit the size of a potential relief force and minimize the chances of a counterattack elsewhere. If a relief force arrived, besiegers were typically left with the option to turn and engage it, hoping to defeat it so that they could return to the siege free from this threat, or to cut their losses and withdraw, avoiding the uncertainty of battle.