‘Desert castles.’
The castles and town defences built by the Franks and Muslims are some of the most iconic symbols of the crusader period. For centuries these impressive stone structures have attracted the attention of pilgrims, tourists and other travellers journeying through the Holy Land. The size and shape of these strongholds varies considerably, as each was designed and built with a different set of priorities and resources.
Although towers and castles may have fulfilled what might be termed an ‘administrative function’ more than anything else, the core of their purpose was to act as anchors of regional control. Expansionist neighbours were compelled to overcome these fortifications if they wanted to annex the surrounding area, while raiders were obliged to take these points of resistance or risk leaving the hostile force of defenders within to their rear. Although networks and patterns of strongholds developed over time, significant castles were far too expensive for a single ruler to undertake a grand fortification strategy akin to those commissioned by modern states. While rulers might add one or more strongholds to the existing landscape, this was a piecemeal process influenced by numerous factors. Whether built by a king or a minor lord, a sultan or an emir, each castle or tower was constructed with a particular set of strategic objectives in mind.
Fortifications have been built in the Near East since at least the Bronze Age. This has left a diverse landscape of defences as sites were developed, abandoned and rebuilt over the centuries. From the Muslim conquests in the seventh century until the arrival of the Franks at the end of the eleventh, most of the fortification efforts undertaken in the region can be divided into four categories: the Umayyad ‘desert castles’; the development of town walls; the construction of isolated castles; and the redevelopment of urban citadels.
The so-called ‘desert castles’, most of which are located east of the Dead Sea in modern Jordan, appear to have been rural palaces built under the Umayyads between the late seventh century and the early eighth. Many, including Qaṣr Kharana, Qaṣral-Hallabāt, Qaṣr Qasṭal, Khirbat al-Minya and Khirbat al-Mafjar were built with a quadriburgium design. This type of square or rectangular plan resembles that of many Roman forts; however, of these sites, only Qaṣr al-Ḥallabāt appears to sit on earlier Roman foundations, which date to the second century. These are similar to Qaṣr al-Ḥayr al-Sharqī and al-Bakhrāʾ (Avatha): the former was a contemporary Umayyadpalace with a pseudo-military plan constructed in the Syrian Desert to the north, between Palmyra, al-Raḥba and Raqqa; the latter, near Palmyra, sits on an earlier Roman fort, like Qaṣr al-Ḥallabāt. The size of these structures varied considerably and some, including Aqaba (Ayla), Qaṣr al-Mshatta, Qaṣr Tuba, al-Bakhrāʾ and the larger enclosure at Qaṣr al-Ḥayr al-Sharq ī, are large enough to be classified as small towns, rather than simply palaces. Along the Mediterranean coast, Kafr Lām and Māhūz Azdūd were built with similar plans.
‘Desert castles.’
Qaṣr al-Ḥayr al-SharqῙ, entrance of the smaller enclosure. (momnnentsofiyria.com)
‘Desert castles.’
Further west, urban defences, many predating the Muslim conquests, were maintained and developed under Muslim rule. Due to the periodic refortification of most cities, it can be hard to discern when certain walls were built, although Umayyad, ʿAbbāsid and Fāṭimid work, especially along the coast, can all be found in various combinations. It was the Muslim defences of Jerusalem and many coastal towns, such as Ascalon, Caesarea, Arsūf and Tyre, that the Franks built upon or developed. This was also the case further north, where existing urban defences, whether constructed under the Muslims or Byzantines, were rebuilt or repaired at sites such as Antioch, Maʿarrat al-Nuʿmān and al-Athārib. At some sites, such as Aqaba and Caesarea, a distinctly new line of Muslim walls was constructed, often within the larger trace of earlier Roman defences. One of the last refortification efforts completed before the arrival of the Franks was commissioned at Cairo by Badr al-Jamālī, adding the city’s iconic late eleventh-century gates.
Isolated fortifications were more commonly built in the Syrian Coastal Mountains, Cilicia and the rougher terrain around the Taurus Mountains. In the sixth century, Justinian, the Byzantine emperor, carried out an extensive fortification programme across much of his empire, strengthening many town walls. In the seventh century, as considerable swathes of Byzantine territory in the East were lost, focus shifted from strengthening town defences to fortifying more inaccessible castles and citadels. The ʿAbbāsid caliphs may have fortified a number of positions along their frontier with Byzantium, north of Antioch, in the eighth century, and another wave of Byzantine fortification efforts accompanied the successful campaigns launched across Anatolia and into Syria under Nikephoras II Phocas and his successors in the second half of the tenth century. Saone, Bourzey and many other Byzantine outposts in the Syrian Coastal Mountains south of Antioch were built during these initiatives. Bourzey, for example, appears to have been constructed by the mid-tenth century, at which point it was taken on behalf of Sayf al-Dawla of Aleppo. The stronghold was retaken under Emperor John I Tzimiskes around 975, but later fell to Seljuk forces following their capture of Antioch in 1085. In 1090, the castle was slighted on the orders of Qāsim al-Dawla Āqsunqur of Aleppo, Zankī’s father, and it probably remained in a dilapidated state until it was occupied and refortified by the Franks in the early twelfth century.
Armenians had taken part in these Byzantine campaigns and a migration into southeastern Anatolia and northwestern Syria followed, encouraged further in the eleventh century by the Seljuk incursions into eastern and central Anatolia. As Byzantine influence waned, the Armenian lords of Cilicia and the Taurus Mountains asserted their autonomy and independence from the castles that they had captured or constructed, and it was from these that they continued to resist Turkish authority.
In the fractious political environment of northwestern Syria in the late eleventh century, a number of small Arab and Turkish lords similarly secured their regional authority through the possession and development of castles. Much like the Armenians, some Arab families, such as the Banū Numayr and Banū Munqidh, carved out and held on to regional authority, along the Euphrates and around Shayzar respectively, through the possession of castles rather than larger urban centres. In one of his anecdotes, Usāma ibn Munqidh (1095–1188) states that Shayzar had no town wall in the time of his father, the community relying entirely on the castle that dominated it.
A new focus was placed on urban defences with the arrival of the Seljuk Turks in the eleventh century. Whereas the ʿAbbāsids and Fāṭimids, like the early Byzantines, had tended to develop town walls, the Seljuks focused instead on citadels. Unlike castles (often rural structures), the walls of which were the defenders’ main line of defence (even if attached to a walled town), citadels were secondary strongholds along or behind the walls of what was often a significant urban community. These provided a place of refuge should the defences of the town beyond be overwhelmed by a hostile army, or if the urban population rose up in revolt.
For the Turks, who were relative outsiders, having only recently migrated to the region they came to dominate, these citadels served a military function but also symbolized the new socio-political hierarchy. The citadel of Damascus was likely commissioned in the 1070s while the Temple of Bel at Palmyra and theatre of Bosra were developed into citadels around the same time. Further north, many ancient tells were refortified under Seljuk rule, providing secure seats of local administration and positions from which the local communities below were dominated.
* * *
The Franks are most often associated with the fortifications built in the region during the two centuries that followed the First Crusade; however, they merely contributed to another chapter of what was already a long history. Many of the town walls they defended predated their arrival, while neighbouring Muslims continued to build and improve urban defences as well as a number of more isolated rural strongholds. Although broad trends can be identified, as different styles and systems of defences suited particular political and administrative structures, a unique set of circumstances, motivations and regional factors lay behind the construction, design and use of each stronghold.
The primary purpose of any fortified tower, castle or system of town walls was to provide a defensible position – an enclave of safety if the surrounding region were overrun. Although some strongholds were attacked on a near annual basis, most did not see action more than once per decade. During these extended periods of relative peace, most strongholds acted as important administrative centres: the seats of local authority and bases from which order was both protected and enforced. As Smail put it when addressing Frankish castles, ‘as in Europe, so in Syria, castles had only occasionally to withstand a siege, but they continuously fulfilled their function as the physical basis of overlordship’.13
While Muslim rule was based primarily in the region’s larger urban centres, Frankish authority was based upon a greater network of fortifications; the first of these they took from local Muslim and Armenian rulers, supplementing them with a number of new strongholds over the following years. This difference reflected the nature of the ruling political systems in Europe and the Near East at the time. Authority in much of Europe was structured according to a ‘feudal’ system: a varying concept whereby land and authority over the people who occupied it were held of a superior lord in exchange for military support. In the Frankish East, a firm system of feudal rights and obligations appears to have been slow to form and a more practical arrangement of support developed ad hoc. Even in the kingdom of Jerusalem, where a system of landholding rights developed most clearly, the monarchy retained a considerable degree of influence in the twelfth century, shaping and reshaping the various lordships as vassals died without heirs or left only young successors, who were of little help on the battlefield. Unlike in Europe, most lords appear to have held their lands directly of the ruling prince, rather than an intermediary lord.
Following the initial spread of Frankish influence, the four Frankish princes (the king of Jerusalem, prince of Antioch, and counts of Edessa and Tripoli) retained large tracts of land for themselves; other areas were granted out as fiefs. In some instances, regions that adventurous figures had come to claim, such as Tancred’s relatively independent acquisition of Galilee, were acknowledged as fiefs. The granting of land in this way was a means to provide members of the nobility (the military class) with a sustainable income and a stake in the collective venture that was the larger principality of which their fief was a part.
To administer their lands, lords typically built, developed or repaired some kind of stronghold, which served as a seat of local power and symbolized their authority. The seats of many significant lordships were in cities, occupying or building upon earlier citadels. Those whose power was based in a more rural setting often constructed a tower near the main source of the lordship’s wealth, usually the main village or town, or developed a pre-existing stronghold, which often had a supporting community located nearby.
Large or small, the primary obligation of these lords, the price for the privilege of land ownership and the social and economic benefits that came with it, was military service. Each lord was expected to provide to his overlord a number of knights, or the equivalent sum to finance such, relative to the wealth of his lordship. In return, he could expect his overlord to defend his lands if they were threatened by a hostile power. During times of peace, some of these knights would have administered their lords’ estates, acted as castellans of smaller strongholds and served as the leading members of the garrisons of larger castles and towns. Others would have served in baronial entourages.
In the Latin East, there are few mentions of the corvée, the system by which peasants were compelled to work a lord’s lands for a certain number of days, and serfdom was far less prevalent than in Europe. Frankish lords appear to have sourced much of their rural incomes from the rents they gathered from tenant farmers, which replaced the need for forced labour to work the lands they retained. These rents were supplemented by fees to use communal service structures, such as the lord’s mill, oven or baths, or those imposed on any sale of property. Some individuals were even given ‘money fiefs’ rather than lands: sometimes the revenue from a toll or urban tax, or a salary upon which the knight was expected to support himself.
Although power was expressed through the control of land, many significant seats of Frankish authority were located in the towns along the Mediterranean coast, which were both traditional bases of local power and centres of commerce. The largest interior lordships retained their importance; however, their source of revenue, the land, was more vulnerable to raids than the lands along the coast and the duties collected in the ports. As power became increasingly concentrated among certain baronial families, and the threat posed by Nūr al-Dīn became more apparent, smaller interior landholders sold or gave away their lands with increasing regularity to religious orders. But even after the collapse of the county of Edessa and the loss of inland territory that followed the battle of Hattin, which further concentrated Frankish rule in the coastal towns, Frankish authority was never as centralized nor as focused on urban centres as was that of their Muslim neighbours.
Count of Jaffa and Ascalon 100
Jaffa (25), Ascalon (25), Ramla and Mirabel (40), Ibelin (10)
Prince of Galilee 100
Lands to the Jordan (60), lands beyond the Jordan (40)
Lord of Sidon 100
Sidon and Beaufort (50), Caesarea (25), Baysān (25)
Lord of Transjordan and Hebron 60
Kerak (40), Hebron (20)
Lordship of Joscelin of Courtenay 24
Castellum Regis (4), Saint George (10), lordship of Geoffrey le Tor (6), lordship of Philip Rufus (2), the chamberlain’s lordship (2)
Toron and the Maron 18
Toron (15), the Maron (3)
Bānyās, Asebebe, Chastel Neuf ?
Ecclesiastical lands 16
Bishop of Lydda (10), Archbishop of Nazareth (6)
Cities owed to the king 257
Jerusalem (41), Nablus (85), Acre (80), Tyre (28), Dārūm (2), Beirut (21)
TOTAL 677
(Derived from John of Ibelin, ed. Edbury, pp. 607–14.)
The earliest Frankish fortifications were occupied rather than built; they were the product of previous Byzantine and Muslim construction efforts, some acquired through conquest, others by treaty. We know little about the Franks’ early castlebuilding efforts to the north, aside from the increased involvement of the military orders in the second half of the century. Although there are fewer significant castles in Palestine, the documentary evidence is better. Here, it is clear that the monarchy took a fairly proactive role in securing Frankish rule by commissioning a number of the kingdom of Jerusalem’s most significant twelfth-century castles.
Montreal was commissioned by Baldwin I in 1115, during his expedition through Transjordan. The stronghold dominated a fertile area, wetter then than today, known to produce grain, wine and olives. The remains of sugar mills have also been found in the region. The original structure, built in just eighteen days, was probably rebuilt or enlarged soon after. William of Tyre described it as a castle defended by both natural and artificial defences, which included a wall and towers, an outer wall and a ditch, defended by a garrison of cavalry and infantry amply provided with arms, food and machines. There was a settlement near the castle and a community of some size lived within the stronghold itself. Thietmar, who visited in 1217, similarly described the castle as very strong and defended by three lines of walls. Montreal still displays two lines of what appear to be Frankish walls, allowing for the possibility that the innermost structure (the third line of walls) viewed by Thietmar was the original structure built in 1115, which has since disappeared through the course of subsequent renovations by the Ayyūbids, Mamlūks and the later Arab community that occupied the castle until the 1980s.
Montreal. (Courtesy of APAAME)
Royal initiative was similarly responsible for the commissioning of strongholds such as Scandelion (Iskandarūna) in 1117, the castles built around Ascalon during the reign of Fulk, and Castellum Regis (or Castellum Novum). The latter is first mentioned in a document dating to 1160, suggesting it was constructed under Baldwin III or his father, Fulk. In 1178, Baldwin IV financed the castle at Jacob’s Ford, with the understanding that it would be held by the Templars. Although castles were an important means of securing and administering the kingdom, they were expensive to maintain and defend, a cost the monarchy was increasingly willing to relinquish. Like the baronage, the kings of Jerusalem recognized that most commerce, and comforts, were in the cities and towns along the coast.
Castles are most often associated with baronial power and autonomy, so it is perhaps surprising how few significant Frankish castles were built by the nobility. This becomes understandable when considering the costs of constructing and maintaining these great structures. Toron (Tibnīn), among the baronial castles founded in the first decade of Frankish rule, would become an impressive stronghold, but it probably began as a simple keep tower. By comparison, Kerak, one of the kingdom of Jerusalem’s most impressive castles, was planned as a monumental stronghold from the outset. Kerak was established a few decades after Toron, allowing its commissioner, Pagan the Butler, to draw upon the established and considerable revenue streams of his dominion, the lordship of Transjordan.
According to some Muslim sources, Safed was fortified prior to the arrival of the Franks, although a degree of confusion is evident as they suggest it was the Templars who rebuilt the castle in 1101–2. Although the Templar order had not yet been established, construction could have been overseen at this time by Hugh of St Omer, the commissioner of Toron. Marino Sanudo, writing at the start of the fourteenth century, confirms that Safed was built by the end of Fulk’s reign and it was clearly in existence when Baldwin III fled there in 1157, having been defeated by Nūr al-Dīn after relieving Bānyās. The earliest twelfth-century record of the castle dates to 1168, when Amalric gave it to the Templars. The document reveals that Safed had been held by Fulk, constable of Tiberias, before it was acquired by Amalric, and suggests that the Templars may have held it of Fulk for some period before this, perhaps since the reign of King Fulk of Jerusalem. Refortification efforts had recently taken place and may have continued once the Templars received outright rule of the castle.
Beaufort, 18km northeast of Toron, predated the arrival of the Franks in some form. The castle was acquired by the Būrid dynasty of Damascus before it was surrendered to Fulk of Jerusalem in 1139. It was then allocated to the lordship of Sidon, rather than the closer but already quite influential lordship of Toron or principality of Galilee. If this represented an attempt to divide power in the north, it had limited success, as the three ruling families of these lordships were linked by marriage in the 1140s. It seems more likely that this was an effort to give a third major lord a stake in the affairs of the northeastern section of the kingdom. When justifying Baybars’ siege of the castle in 1268, Ibn ʿAbd al-Ẓāhir calls Beaufort a threat to Ṣubayba, and implicitly Bānyās below, but this threat seems to have been symbolic. Beaufort, sitting high atop the western side of the steep valley carved by the Litani River, is visible from Ṣubayba, about 19km away, but these cliffs isolate it and deprive its garrison of the ability to strike rapidly into the Hula Valley, as those of Chastel Neuf (Hūnīn) to the south and Ṣubayba to the east could have done. A modern set of switchbacks, about 2.5km north of Beaufort, provide a route down to the Litani; however, it may have been necessary to travel a distance of 4km north of the castle to find a more natural path down to the river, from where both the Hula Valley to the south and Biqāʿ Valley to the north could be approached. It may have been this combination of isolation and access that allowed the castle to remain a part of a secular lordship until 1260 (although it was under Muslim rule from 1190 to 1240), while the lordship of Sidon’s critical link to the sea and Europe beyond was probably another important factor.
Reynald of Sidon was one of the few Franks who survived the battle of Hattin. Although Sidon was abandoned less than a month after the battle, Reynald was less willing to give up Beaufort. When Saladin’s army arrived at the castle in April 1189, Reynald, who could speak Arabic, went down to talk to him. In what turned out to be a ploy, Reynald offered to surrender Beaufort and join the Muslims if he were granted three months to retrieve his family from Tyre. Saladin was eager to gain the castle without a fight, as Tyre, Tripoli and Antioch remained bastions of Frankish resistance and bases from which crusaders might quickly reverse his recent conquests. When Saladin returned, Reynald again visited his tent to play for more time. Growing impatient, Saladin had Reynald arrested and ordered him to command the garrison to surrender, which Reynald did, genuinely or otherwise, but they refused. Beaufort continued to hold out until 1190, when the defenders finally agreed to surrender provided their lord was released. Reynald’s reputation seems to have suffered in neither Frankish nor Muslim eyes, as he, along with Balian of Ibelin, led Conrad’s delegation to Saladin in the autumn of 1191.
Region aro und the Hula Valley
Beaufort. (Courtesy of Jean Yasmine)
Beaufort returned to Frankish hands in 1240, becoming a part of the lordship of Sidon once more. Following the Mongol raid on Sidon in 1260, Julian of Sidon was forced to sell both Sidon and Beaufort to the Templars. Julian had previously sold the Cave of Tyron (Shaqīf Tīrūn), 22km east of Sidon in the southern Lebanon, to the Teutonic Knights, suggesting he was already overburdened before the Mongols arrived. Although compelled to sell his strongholds, Julian held on to them longer than most. Cursat (al-Quṣayr), the patriarch of Antioch’s castle, which was finally taken by the Mamlūks in 1275, was one of the very few inland castles that did not pass to the military orders. Even along the coast some baronial strongholds were sold: Balian of Arsur sold (or leased) Arsūf to the Hospitallers in 1261, a year after Julian sold Sidon to the Templars.
Montreal and Kerak remained in secular hands through the twelfth century, though they never returned to Frankish control following their loss in 1189 and 1188 respectively. Montreal had been expanded to essentially its present footprint by the 1150s, but whether most of this work was financed by the crown or the later lords of the castle is unclear. Kerak, founded in 1142 by Pagan the Butler, lord of Transjordan, came to replace Montreal as the lordship’s seat of power. Kerak appears to have been planned as a mighty castle from the outset, as there is no sign of an early phase with a reduced footprint. Unsurprisingly, work on the castle is said to have continued under Pagan’s successors. Even here, in these two bastions of baronial authority, the military orders had a presence by the reign of Maurice, nephew and successor of Pagan the Butler, who, in 1152, gave the Hospitallers property in both Kerak and Montreal.
The trend towards sharing the responsibility and costs of fortifications with the military orders can be seen easily in the case of Bānyās. The town had been given to the Franks by the Assassins during a period of persecution in Damascus, at which point Baldwin II bestowed it on Renier Brus. The town was retaken by Shams al-Mulūk of Damascus in December 1132, but it ended up back in Frankish hands from 1140, when the Damascenes exchanged it in return for support against Zankī. Bānyās came into the hands of Humphrey II of Toron through his marriage to Renier’s daughter. Although Humphrey was constable of the kingdom and one of its leading barons, by 1157 the costs associated with Bānyās compelled him to give half of the town, and the burden of its defence, to the Hospitallers. The force of Hospitallers that was then marching to reprovision the town was ambushed on its way, leading the order to reconsider its decision to accept this offer.
Along what might be considered the Ascalon frontier, Bethgibelin (Bayt Jibrīn) and Gaza, completed around 1136 and 1150, were entrusted to the Hospitallers and Templars respectively. At the opposite end of the kingdom, the Hospitallers had established a presence in the region north of the Sea of Galilee by 1157, having acquired a portion of Chastel Neuf along with Bānyās. In 1168, the same year that outright ownership of Safed passed to the Templars, the Hospitallers acquired a new estate south of the Sea of Galilee, where the order would construct Belvoir. In addition to Safed, the Templars stood to gain Jacob’s Ford, although construction was interrupted when Saladin seized and destroyed the incomplete castle in 1179.
Around the time of the Fifth Crusade, James of Vitry, bishop of Acre from 1216, judged the success of a crusade by its actions to either besiege a stronghold or repair existing defences. Fortifications had become critical to preserving a Latin presence in the Levant, and as Muslim armies grew, so too was it necessary for Frankish strongholds to become more impressive. To build such elaborate fortifications, the assistance of crusaders was often required.
Crusaders had contributed to the construction of strongholds from at least the Third Crusade, when Richard I oversaw repairs to the defences of Acre and Jaffa and the rebuilding of Ascalon and Casel des Plains (Yāzūr). In the early phase of the Fifth Crusade, while the army waited in Palestine for the arrival of more forces from Europe, a group of crusaders helped strengthen Caesarea and another group assisted with the construction of ʿAtlit, which was to become the Templars’ great stronghold south of Acre. These defences were quickly put to the test. While the bulk of the kingdom’s forces were campaigning in Egypt, al-Muʿaẓẓam ʿĪsā sacked Caesarea, although not its citadel, in either 1218 or 1219, and besieged ʿAtlit, without success, in 1220.
Before the worst of the partisan feuding that accompanied the arrival of Frederick II in 1228, participants of the Sixth Crusade helped construct and refortify certain strongholds. Duke Leopold VI of Austria, who participated in the early part of the Fifth Crusade, had made a significant donation to the Teutonic Knights, still a fairly new order. These and other funds were put towards acquiring lands in western Galilee, where the order was prepared to build a new castle, Montfort, by the time elements of the Sixth Crusade had begun to arrive. While German crusaders helped construct the first phase of Montfort in 1227, English, French and Spanish crusaders helped rebuild the defences of Sidon. In 1190, Sidon’s town walls had been pulled down and its population moved to Beirut. It was reinhabited following the Third Crusade and shared between the Franks and Muslims for a period. By March 1228, the Franks had taken complete control and efforts were under way to refortify the town. It seems to have been at this point that a new castle was commissioned just offshore. Originally consisting of little more than two towers with a connecting wall between them, the sea castle of Sidon was developed over the following decades into an impressive stronghold.
Caesarea. (Courtesy of Michael Eisenberg)
As work at Montfort and Sidon progressed, Frederick seems to have been more concerned with refortifying the citadel of Jaffa, of which nothing now remains. The citadel of Caesarea may also have been improved during this period, as permission to do so was secured in a treaty between the emperor and al-Kāmil of Egypt. In a letter to Henry III of England, his future brother-in-law, Frederick made a point of specifying that while the Franks were permitted to carry out these fortification efforts, al-Kāmil agreed not to build or repair any castles during the period of the ten-year truce they had arranged. Although Frederick could brag about this point, it had little real value considering al-Kāmil’s strength was in his army and the manpower reserves he could raise, not in his fortifications.
Aside from Montfort and some work at Jerusalem, which had returned to Frankish hands under the peace negotiated by Frederick II, Frankish fortification efforts during the early thirteenth century focused on the port towns and their citadels in particular. Much as castles secured territory, these citadels held influence over key harbours, the critical portals to Europe. This policy continued during the Barons’ Crusade with the fortification of Ascalon around 1240, overseen by Theobald I of Navarre and Richard of Cornwall. Rather than rebuild the town walls, in ruins since the Third Crusade, a castle was built at the north end of the town. The destroyed town defences provided plenty of stone and construction proceeded quickly. Although boats could anchor off Ascalon or moor on a stretch of beach, there was no natural harbour here and the castle appears to have been built as a forward base or buffer against Gaza, which had become the main staging point for Egyptian campaigns in Palestine and Syria, much as Ascalon had been under the Fāṭimids in the first half of the twelfth century.
While Richard of Cornwall was overseeing work at Ascalon, efforts began to rebuild the Templar castle of Safed. The castle had been destroyed by al-Muʿaẓẓam ʿĪsā during the Fifth Crusade and it lay in ruins until al-Ṣāliḥ Ismāʿīl returned it to the Franks in 1240. Benoit of Alignan, bishop of Marseilles, led the initiative and the first stone was laid in December 1240. The bishop had made a pilgrimage to Damascus earlier in the year, having secured free passage to do so, and on his way back remarked that Ṣubayba was the only stronghold between Safed and Damascus. Although he calls S afed just a pile of stones, the fact that he was received there by Brother Rain-hard of Caro suggests that some kind of structure was garrisoned before the rebuilding programme was launched. The castle was probably in its final Frankish form by the time the bishop returned to view it in 1260, six years before it was captured by the Mamlūks.
Refortification efforts shifted back to the coast during the Seventh Crusade, following Louis IX of France’s ultimately disastrous campaign in Egypt. This fortification programme began with Acre in 1250 and expanded to include Haifa and then Caesarea from April 1251, at which point the town walls received their iconic continuous talus. From 1252 to 1253, attention shifted to Jaffa, its town walls, twentyfour towers and ditch, which extended around the town to the sea on either side.
When it had been decided that Safed should be built, there was great joy in the House of the Temple and in the city of Acre and among the people of the Holy Land. Without delay, an impressive body of knights, sergeants, crossbowmen and other armed men were chosen with many pack animals to carry arms, supplies and other necessary materials. Granaries, cellars, treasuries and other offices were generously and happily opened to make payments. A great number of workers and slaves were sent there with the tools and materials they needed. The land rejoiced at their coming and the true Christianity of the Holy Land was exalted . . .
It is not easy to convey in writing or speech . . . what number, size and variety of construction of crossbows, quarrels, machines and every sort of arms, and what effort and amount of expense in making them; what number of guards every day, what number of the garrison of armed men to guard and defend and repel enemies who were continually required there; how many workmen with different trades, how much and what expenses are made to them daily . . .
In the first two and a half years, the Templars spent 1,100,000 Saracen bezants on building the castle of Safed, in addition to the revenues and income of the castle itself, and in each following year 40,000 Saracen bezants, more or less. Every day victuals are dispensed to 1,700 or more and in time of war, 2,200. For the daily establishment of the castle, 50 knights, 30 sergeant brothers and 50 turcopoles are required with their horses and arms, and 300 crossbowmen, for the works and other offices 820 and 400 slaves. There are used there every year, on average, more than 12,000 mule-loads of barley and wheat, apart from other victuals, in addition to payments to the paid soldiers and hired people, and in addition to the horses, patrols and arms and other necessities, which are not easy to account.
(Adapted from De constructione castri Saphet, trans. Kennedy, pp. 194–6.)
Here, Eudes of Châteauroux, a papal legate accompanying Louis, financed the reconstruction of one of the three town gates and a section of adjoining wall. Work extended to Sidon, where Simon of Montceliard, master of the king’s crossbowmen, oversaw the development of the sea castle and town defences, including the old land castle at the southeast salient of the town, ahead of Louis’ arrival in 1253. The French king’s focus on coastal settlements is clear, but so too is his attention to town defences, rather than just citadels, reflecting his appreciation that the success of these communities, not just their survival, was critical to preserving Frankish rule in the Levant.
As Louis was seeing to the kingdom of Jerusalem’s coastal fortifications, Muslim forces made a raid against the outskirts of Acre, destroying Doc (Daʾuk) and Ricordane (Kurdāna), mills of the Templars and Hospitallers respectively. A more concerted attack was then made against Sidon, interrupting refortification efforts there and compelling the Franks to withdraw to the sea castle, allowing the Muslims to sack the town. While Louis marched north from Jaffa to Sidon, a contingent of the force he led attacked Bānyās. Although elements of the army were able to break into the town, they were quickly ejected and forced to content themselves with burning some of the surrounding fields. No attempt was made against Ṣubayba, which overlooked the town from the heights to the east.
There was value in occupying a crusading army with a physically demanding task, such as these refortification efforts. Pelagius, legate to the Fifth Crusade, encountered issues of idleness following the capture of Damietta, when the army refused to march out towards Cairo, and, in 1250, Louis IX had to scold one of his brothers who had already turned to dice and gambling as they sailed for Acre immediately after their defeat in Egypt. Beyond this pragmatic means of keeping discipline and avoiding rowdiness, there was the greater good: the defences that were built contributed to the Frankish principalities and there were spiritual benefits for the penitential service of completing these labours or financing them.
By the mid-thirteenth century, however, the inability of the Franks to re-establish Latin influence was becoming clear. John of Joinville provides an anecdote that in a way sums up European perceptions of the situation in the Levant in the 1250s. While John the Armenian was in Damascus sourcing crossbow materials for Louis IX, he encountered an old man who remarked to him: ‘You Christians must be hating each other very much. For once, long ago, I saw King Baldwin [IV] of Jerusalem, who was a leper, defeating Saladin [at Montgisard, 1177], though he had only three hundred men-at-arms, while Saladin had three thousand. But now, through your sins, you have been brought so low that we take you in the fields just as if you were cattle.’14
Following the established trend of financing fortifications, Prince Edward of England commissioned a tower at Acre before leaving the Holy Land in 1272. This was probably the ‘English Tower’, which is labelled along a stretch of the town’s outer walls near the northeastern salient on some early fourteenth-century maps of the city. A number of other improvements were made around this section of the city’s defences between Louis IX’s contributions in 1250 and the city’s destruction in 1291. Ahead of the Accursed Tower, which anchored the apex of the inner line of walls, the round King’s (or New) Tower stood at the corner of the outer line of walls. This was named after Henry II of Cyprus, who financed the reconstruction of the tower in the 1280s, improving the defences that had been built there by one of his predecessors in the mid-thirteenth century. Alice of Brittany, dowager countess of Blois, who arrived at Acre in 1287 and died the following year, financed another tower in this section of the town’s defences. Nicholas of Hanapes, titular patriarch of Jerusalem, paid for another tower nearby, also in the 1280s. The attention devoted to this part of the city’s defences was warranted: this was the same area where siege efforts had focused in 1191 and where Mamlūk forces would ultimately break into the city in 1291.
From the time of the First Crusade, the principal seats of power in western Syria were Damascus and Aleppo, while Cairo formed the third major Muslim power base in the region. Beyond these, a number of smaller administrative units were formed around the secondary towns and cities, such as Hama, Homs and Shayzar, which at times were able to exert their own authority and at others found themselves subject to the ruler of Damascus or his counterpart in Aleppo. Comparable, if quite different, from European feudal systems, was the Islamic iqṭāʿ system.
Whereas European fiefs were plots of land, the rights to which were typically granted in perpetuity and passed from one generation to the next so long as the reciprocal conditions of support were met, an iqṭāʿ was rather the right to collect taxes from a certain region. Although these sometimes carried administrative privileges, the land involved was not owned by the muqṭaʿ (he who held the iqṭāʿ ). In the late eleventh century, Nizām al-Mulk, a Persian scholar and Seljuk vizier under Alp Arslān and Malikshāh, provided the following advice in his treatise on governance:
Officers who hold assignments [iqṭāʿ s] must know that they have no authority over the peasants except to take from them – and that with courtesy – the due amount of revenue which has been assigned to them to collect; and when they have taken that, the peasants are to have security for their persons, property, wives and children, and their goods and farms are to be inviolable; the assignees [muqṭaʿ s] are to have no further claim upon them.15
Although a feature of Muslim administrations across most of Egypt and Syria through the twelfth and thirteenth centuries, the nature of the iqṭāʿ system and the rights and responsibilities of the muqṭaʿ differed according to the political system, and ruling dynasty, of which they were a part. Whereas the Fāṭimid iqṭāʿ system appears to have been used to pay administrative officials, the Zankid s expected muqṭaʿ s to use their revenues to raise troops – Nūr al-Dīn was known to confiscate the iqṭā[ of any muqṭaʿ who failed to meet his military obligations. When Saladin took power in Egypt, he preserved the existing structure of the Fāṭimid iqṭā[ system, but reoriented it in order to bring it closer into line with the Zankid system of Syria. When the Mamlūks came to power, they developed the Ayyūbid system, which had evolved through the early thirteenth century, to better suit the slave-based nature of their political hierarchy.
Despite these differences, every iqṭāʿ was ultimately granted by the sultan and he retained the right to cancel or confiscate it at will. Accordingly, each iqṭāʿ would need to be confirmed, whether as a matter of routine or more elaborate ceremony, upon the accession of a new sultan. On the flip side, each muqṭaʿ was free to give up his iqṭāʿ at any time if he felt its revenues did not support the obligations expected of him .
The size of an iqṭāʿ could vary enormously. Some were essentially administrative regions, ruled by prominent figures from mighty seats of power; others were quite small and could even be shared. As Saladin came to rule Egypt and much of Syria, he placed his family members in critical positions, using the iqṭāʿ system to do so. He gave his father, Ayyūb, Alexandria and Damietta, the most important ports in Egypt; his brother, Tūrānshāh, was given significant districts in upper Egypt, from which he invaded Yemen; and Syria was similarly distributed among other family members and significant emirs, such as Ibn al-Muqaddam, who received Baalbek, and Saʿd al-Dīn Masʿūd, who received Safed. These grants became a part of the Ayyūbid power struggle that followed Saladin’s death in 1193. For example, although al-ʿĀdil was recognized as ruler of Egypt in 1195, so too did he retain his iqtāʿs in the Jazīra (Edessa, Ḥarrān and Mayyāfāriqīn). As the various parties manoeuvred for control of Syria, those who held iqṭāʿs in the area became influential, throwing their support behind the party who might best serve their interests or who would back their claim to their iqṭāʿs. Although mamlūks had been granted iqṭāʿs from the reign of Saladin, most of wh ich were in Syria, such figures were typicallynot given significant positions until the reign of al-Ṣāliḥ Ayyūb, whose mamlūks would later murder his son and establish the Mamlūkdyn asty.
While muqṭaʿs were relied upon to provide military support, it is easy to confuse these figures with a number of other administrative figures, as various titles tend to be translated in different ways. A ṣāḥib typically held his lands as a muqṭaʿ but ruled more independently. These figures could grant iqṭāʿs but were still required to provide the sultan with military support when summoned. It was in this capacity that Saladin’s son, al-Ẓāhir Ghāzī of Aleppo, resisted his uncle, al-ʿĀdil, and al-Mughīth ‘Umar, ṣāḥib of Kerak, similarly refused to recognize the authority of the Mamlūks.
The nāʾib (provincial deputy) and wālī (governor), for example, were administrative positions. Although a significant town or castle was typically granted as an iqṭāʿ, a wālī was commonly entrusted with the citadel – Saladin gave Homs as an iqṭāʿ to his uncle, Shīrkūh, while governance of the citadel was given to a Kurdish emir, Badr al-DīnIbrā hīm al-Hakkārī. Saladin made Qarāqūsh, one of his most trusted emirs, nāʾib of Egypt, entrusting him with the region’s administrative and political affairs. When Mamlūk influence spread across western Syria in the wake of the battle of ʿAyn Jālūt, the vacant seats of power left by Ayyūbid princes, most of whom had ruled as ṣāḥibs, were filled with nāʾibs and each was allocated an iqṭāʿ equal to the prestige of his position – the nāʾib of Kerak, for example, was assigned Hebron as an iqṭāʿ. As further conquests were made, a nāʾib was typically appointed to rule each district, although these lesser figures were not implicitly granted an iqṭāʿ.
The armies of the realm are partly resident at the Sultan’s court and partly distributed around its provinces and throughout its lands. Some of them are nomads, such as the Arab tribesmen and the Turkomans. The regular troops are mixed in origin, being Turks, Circassians, Rūmīs, * Kurds and Turkomans. The majority are purchased mamlūks. They are ranked as follows. The greatest are those who hold an emirate of a hundred troopers together with a command of a thousand, from which category come the most important vicegerents. At times this figure may be increased for some by ten or twenty troopers. Next are the tabl-khānah emirs, the majority of whom have an emirate of forty, although there may be those for whom that figure is increased to seventy. The tabl-khānah rank is not held with fewer than forty.
Then follow the emirs of ten, consisting of those who hold an emirate of ten, and sometimes including individuals who have twenty troopers, but who are still only counted as emirs of ten. Next come the troopers of the ḥalqa, whose rights are issued from the Sultan, just as those of emirs are, while, on the other hand, the troops of emirs receive their rights from their emirs. For every forty of these ḥalqa troops there is an officer, one of their number, who has no authority over them except on active military service, when they muster with him and he is responsible for their dispositions.
In Egypt, the iqṭāʿ of some senior emirs of a hundred, close to the Sultan, may amount to 200,000 army dinars, sometimes more. For other emirs of this rank the figure progressively diminishes to around 80,000 dinars. The ṭabl-khānah emirates amount to 30,000 dinars with fluctuations above and below, with a minimum of 23,000 dinars. The emirs of ten have an upper limit of 7,000 dinars. Some iqṭāʿs of halqa troopers reach 1,500 dinars, this amount and those that come near it being the iqṭāʿs of the senior members of the ḥalqa, the officers appointed over them. Then come lesser amounts down to 250 dinars. For the troops of emirs, the value of their iqṭāʿ is at the discretion of the emir.
iqṭāʿs in Syria do not come near these figures, but are worth two-thirds of them, leaving aside what we have said about favoured senior emirs of a hundred, for this is unusual and without normal validity, and I am not aware of anything in Syria that comes anywhere near such a sum, except for what the Vicegerent of Damascus receives.
* This term, meaning ‘Romans’, usually refers to Byzantines; however, it may be used more generally to identify Anatolians, implying Armenians, in this instance.
(Adapted from al-ʿUmarī, Masālik al-abṣār, trans. Richards, pp. 20–1.)
Muslim strongholds were built and developed for many of the same reasons as those constructed by the Franks; however, Muslim authority was generally more centralized and regional seats of power were fewer. The Būrid rulers of Damascus, for example, relied on the support of the men they appointed to rule the secondary centres of Baalbek, Bosra, Ṣarkhad and, at times, Homs. Depending on the nature of the ruler’s relationship with a secondary seat, it might be ruled by an appointed governor, as was Baalbek, or by a hereditary emir, as was Homs during much of this period. To the north, power was more fragmented, increasing the number of secondary seats, often corresponding with the larger urban communities and castles. Although many of these secondary rulers enjoyed considerable autonomy in the early twelfth century, their independence declined steadily as power was consolidated by the Zankids and then Ayyūbids. Due to the urban focus of Muslim administrative structures, castles were comparatively rare.
Many Muslim castles, like those held by the Franks, had been built by the Byzantines; others were initially constructed by regional Muslim and Armenian powers. Castles such as Shayzar, Qalʿat Najm andQalʿat Jaʿbar were among the region’s most impressive and each rests on the remains of an earlier Roman or Byzantine settlement. These, as well as many town defences, were rebuilt or developed during the period of uncertainty from the late ninth century as various ʿAbbāsid, Byzantine, Fāṭimid and Turkish figures fought for control of the region. In the eleventh century, a number of smaller Arab rulers from Syria’s tribal communities, who managed to carve out zones of influence for themselves, not unlike the Armenians in Cilicia, built castles to secure their rule – Shayzar was rebuilt by the Banū Munqidh while Qalʿat Jaʿbar was developed by the Banū Qushayr, and probably the Banū Numayr before them. Many castles and citadels were subsequently refortified in the second half of the twelfth century by Nūr al-Dīn, who used them as bases of power to secure his expanding realm. This work was apparently so significant at Qalʿat Najm that Ibn Jubayr, who spent a night there in 1184, described it as ‘a new-built fortress’.16 Few Muslim castles, aside from a selection constructed by the Assassins, completely post date the arrival of the Franks. Three notable exceptions are ʿAjlūn, Qalʿat Ṣadr and Ṣubayba.
ʿAjlūn. (Courtesy of APAAME)
ʿAjlūn was built on the east side of the Jordan, about midway between the Sea of Galilee and the Dead Sea, possibly on the site of an earlier monastery. Commissioned by Saladin in 1184–85, construction was overseen by his nephew, ʿIzz al-Dīn Usāma. It is often suggested that the castle was constructed to check Hospitaller raids from Belvoir; however, its location, 16km from the Jordan and 36km southeast of Belvoir (as the crow flies), does not support this. A far more sensible position from which to mirror Belvoir would have been at least 25km to the north, due west of Adhriʿāt, from where it could observe Belvoir from the eastern side of the Jordan Valley. This would also have placed it in a position to dominate the main routes passing south of the Sea of Galilee to Bosra and the southern Ḥawrān, which had been raided by Raymond of Tripoli in December 1182. Instead, it was hoped that a castle at ʿAjlun would provide greater influence in the affairs of the local Bedouins, who were known to ally with the Franks when it suited them, and to dominate the local iron industry. It also provided an administration hub in a region largely devoid of significant strongholds and with a history of Frankish influence dating back to the first decade of the twelfth century.
Qalʿat Ṣadr (al-Jundī) was built on a conical hill, about 200m above the surrounding landscape of western Sinai. The castle is around 58km east of Suez and about halfway between Bilbays and Aqaba. It was commissioned by Saladin in the 1170s and developed in 1182–83 under the supervision of his brother, al-ʿĀdil. The castle would have provided Saladin with greater control over the Sinai Peninsula and traffic moving across it, perhaps a comfort to Muslim pilgrims and merchants travelling between Egypt and Syria or Arabia. It was attacked by a party of Franks in 1178 and the timing of its refortification, five years later, suggests Saladin may have feared a new threat from the direction of Aqaba – Amalric’s invasions of Egypt in the 1160s had instead proceeded along the Mediterranean coast. Although Saladin had captured Aqaba in December 1170, Reynald of Châtillon, the former prince-regent of Antioch, had become lord of Transjordan sometime around early 1177.
Jabal ʿAwf.
ʿAjlūn and the region to the immediate south. (Michael Fulton)
Already known for his aggression, it was quite possibly Reynald who had organized the attack on Qalʿat Ṣadr in 1178. Reynald subsequently led an invasion into northern Arabia in late 1181 and orchestrated a daring and unprecedented maritime raid down the Red Sea around January 1183. Having carried boats overland from Kerak to Aqaba, his forces were able to threaten Mecca and Medina, the holiest cities of Islam. It would have been natural to fear that a raid into Egypt might be next. There are reports that in May 1183 a Muslim force heading for Dārūm intercepted a party of Franks, judged to be heading for Qalʿat Ṣadr, on the road towards Aqaba. In September 1184, only months after more provisions had been sent toQalʿat Ṣadr, the inhabitants of Bilbays abandoned their homes and fled to Cairo after reports arrived that a party of Franks had reached Faqus, 40km to the northeast.
Ṣubayba was built from about 1227, during the uncertainty of Frederick II’s visit to the Holy Land. Construction was undertaken by al-ʿAzīz ʿUthmān, a son of al-Ādil, who ruled Bānyās for his brother al-Muʿaẓẓam ʿĪsā of Damascus. Less than a decade earlier, al-Muʿaẓẓam ʿĪsā had ordered the destruction of many castles west of the Jordan during the Fifth Crusade, but appears to have commissioned or approved of the construction of Ṣubayba before his death in 1227. The castle provided an advanced base for attacks into Palestine, should it fall back to the Franks, while also giving Bānyās, which it overlooked from the heights 2km to the east, a much stronger citadel. Regardless of the specific motivations for its construction, the castle was a significant stronghold on the main road between Damascus and northern Palestine.
Reynald of Châtillon’s Arabian campaigns, 1181, 1182–83.
To these might also be added Mount Tabor. Al-ʿĀdil ordered the iconic hill to be fortified in 1211 and a defensive perimeter, complete with towers, was built around the summit, enclosing the Christian monastic complex that had expanded there under the Franks during the twelfth century. The epigraphic evidence reveals that al-ʿĀdil’s son, al-Muʿaẓẓam ʿĪsā, along with ʿIzz al-Dīn Aybak ibn ʿAbdullah and a few other emirs, helped oversee construction. Al-ʿĀdil’s motivations to fortify Mount Tabor have been debated. Most interpretations hinge on a statement made by James of Vitry, who asserted that the stronghold was built to oppose Acre. Although possible, these fortification efforts may have been part of a more general effort to strengthen Ayyūbid authority over northern Palestine. The region had seen considerable action in the 1180s during Saladin’s invasions of 1182 and 1183, and it was here that al-ʿĀdil had moved when elements of the Fourth Crusade arrived at Acre, preparing to counter a possible invasion. The garrison of Mount Tabor successfully repelled elements of the first wave of the Fifth Crusade when an attack was made in 1217, but al-Muʿaẓẓam ʿĪsā found himself doubting his ability to hold the newly built stronghold as the crusaders’ focus shifted to Egypt. To prevent it from falling to the Franks, he ordered it and many other castles slighted before the Fifth Crusade ultimately collapsed in Egypt.
Ṣubayba. (Michael Fulton)
Although these strongholds were significant, Muslim fortification efforts tended to focus more often on the citadels and town walls of the region’s large urban centres. Such defences not only helped protect these centres of wealth and influence, but also acted as conspicuous instruments and symbols of authority.
A by-product of any fortification was its symbolism – these were grand structures that typically inspired awe. For the Franks, their rural strongholds were statements of regional control, reminders to the local population of the ruling order and probably a reassuring sight for visiting European pilgrims. The relatively small strongholds first built by the Franks were eventually eclipsed by the great castles, now iconic symbols of the crusades, which were the result of considerable investment. Citadels presented a similar message.
Following the earthquakes of 1156–58 and 1170, Nūr al-Dīn commissioned the refortification of a number of Muslim strongholds across his growing realm. From the town walls of Damascus to castles on the Euphrates, these necessary repairs doubled as tangible statements of his authority. Al-ʿĀdil undertook a similar building programme in the early thirteenth century, adding to the defences of Jerusalem, the citadels of Cairo and Bosra, rebuilding the citadel of Damascus and fortifying the top of Mount Tabor. At each site, his consistent use of extremely large towers with a similar bossed masonry style – essentially his signature – left little doubt as to who built them and who was in control. At Damascus, he went so far as to enclose the citadel, originally built by the Seljuks, behind a new line of walls. While some Ayyūbids (including al-ʿĀdil) and later Mamlūks added bold inscriptions, prominently displaying who was responsible for constructing a certain tower or wall, al-ʿĀdil’s signature building style could be understood by foreigners and the illiterate alike. For al-ʿĀdil, this was not only a way of strengthening important citadels and town defences but a statement of legitimacy – Saladin had intended his sons, not his brother, to succeed him.
Damascus, citadel (after Hanisch and Berthier).
Aleppo, citadel gatehouse. (monumentsofsyria.com)
Although citadels are perhaps associated more with Muslim fortification efforts, the Franks also built and developed urban strong-holds. At some sites, such as Kerak, ʿAtlit and Arsūf, the castle was the focal point of the community’s defences, although each also had town walls. Elsewhere, the Franks occupied and developed existing citadels, as at Jerusalem, Antioch and most other large urban centres.
Even the most isolated castle required some kind of a population centre to support it. The relationship between a stronghold and its supporting community was an important and mutually beneficial one, but the nature of this relationship could vary greatly. At Kerak, for example, the castle was built on a spur that extended south from the existing town, where a Byzantine monastery may once have stood, while the castle-turned-citadel at Shayzar dominated the town to the west. At Montreal, the local community was located to the southeast of the castle, roughly 300m away. Safed was built on a hill that was already occupied by a community of some size, while Beaufort, high above the Litani River, was neighboured by a small supporting settlement. Overlooking a tributary of the Orontes, Montferrand may have relied on the nearby sizeable town of Rafaniyya and Ṣubayba enjoyed the presence of nearby Bānyās. In the Syrian Coastal Mountains, some strongholds, such as Saone, Shughr-Bakās and Bourzey, had large outer enclosures, leaving it possible that a section of the local community may have had access to these areas, although the main population appears to have lived outside each castle.
These supporting communities carried out important day-to-day tasks: these were the farmers, tradespeople and small merchants who supplied the castle with goods and labour when required. In return, they benefited from the protection and consumer demand provided by the fighting men. Larger towns operated on the same principles, although the main commercial interests would have been in trade rather than agriculture.
Strongholds were also convenient places from which to extract and then store tax revenues. In both Frankish and Muslim lands, a variety of taxes were collected, but chief among them was a land tax, which was assessed based on the productivity of a region. Although at times levied on individual landholders, it seems villages were more often assessed as collectives. The rate varied by region, but was typically valued somewhere between a quarter and a half of a village’s agricultural produce. In Muslim lands, the jizya, a head-tax, was levied on non-Muslims and the Franks adopted and imposed a similar tax on non-Christians, while tithes were similarly collected by both Frankish and Muslim rulers.
In areas where local populations were less amenable, strongholds provided secure bases from which revenue could be collected more forcefully. In 1125, Baldwin II commissioned the castle of Mount Glavianus, in the mountains inland from Beirut, because the Muslims of the area were reluctant to pay the local tax. Likewise, a motive behind the construction of ʿAjlun was to impose greater control over the local Arabs, the Banū ʿAwf. Hand in hand with the ability to extract revenue was the need to protect it. Even small towers could safeguard from bandits the agricultural surplus gathered from a region, while larger strongholds might also provide a place of refuge during raids for the local population who worked the land – the same body of fighting men who were responsible for extracting wealth from the local community were typically its defenders.
Whatever the motives behind the construction of a stronghold, a similarly unique set of considerations influenced where each was placed. First, there had to be something to defend. This could be as general as a desire to exercise and secure influence over a region, as may have been a significant motive behind the Hospitallers’ construction of Belvoir. Alternatively, there may have been a specific commercial or administrative incentive. For example, the Templars’ decision to commit considerable resources to fortify and defend Safed reflects the site’s significance as a local administrative centre, while the citadel they constructed in Acre ensured they maintained an interest in the city and the wealth of its commercial activities. The foundation of Montreal seems to have combined both: in addition to extending general Frankish hegemony to the south, it came to dominate the caravan traffic along the desert road south of the Dead Sea and sugar production in the area. The fortified mills in the plain around Acre, including Doc and Ricordane, were foremost economic structures, while many towers built in and around the towns of Palestine would have been primarily seats of local administration.
With sufficient incentives to build a stronghold in a region and a nearby community to support its needs (either pre-existing or subsequently established), there remained the strategic considerations of whether the region could be effectively defended. Questions needed to be asked: How likely was it that the new stronghold might be besieged and how strong might the potential besieging force be? How quickly could relief forces be mustered and how likely was it that they would be able to break a siege? There was no point to building a castle if it could not be defended. This thought process can best be seen when figures opted not to garrison captured castles, although the costs of construction had already been paid. Accordingly, Baldwin II abandoned Jerash in 1121 and Nūr al-Dīn declined to hold Chastel Neuf in 1167. The slighting campaigns ordered by Saladin, al-Muʿaẓẓam ʿĪsā and Baybars similarly reflected fears that the strongholds they ordered destroyed might otherwise fall to large crusading armies and be used against them. By the same logic, the construction of Montreal, 120km from Hebron and more than 140km from Jerusalem (as the crow flies), speaks to the perceived lack of threats that it would face when it was initially founded in 1115.
Regardless of the region in which a stronghold was built, topography was always an important consideration. Castles were typically constructed on hills above the surrounding landscape, although the relative height of these hills varied considerably – most in the Syrian Coastal Mountains were sited on isolated spurs, while many in Palestine sat atop little more than a slight rise.
In Galilee, Castellum Regis was built in the town of Miʿilya on a hill that may have been the site of an earlier strongpoint. This was an ideal place for a mid-twelfth century administrative centre, but when the Teutonic Knights sought to build a larger castle (Montfort) in the early thirteenth century they chose instead a spur along Wādī al-Qarn, almost 4km to the northwest, trading convenience for topographical strength. Further south, the Hospitallers used the topography between the Jezreel Valley (Marj Ibn Āmir) and the Sea of Galilee as best they could, building Belvoir at the edge of the plateau overlooking the Jordan Valley to the east. Like many strongholds built along the coastal plain, Ibelin and Blanchegarde (Tell al-Ṣāfī) were built on small hills, the loftiest positions available.
Castellum Regis, in the centre of Miʿilya. (Michael Fulton)
Montfort. (Michael Fulton)
Ṣubayba, from the east. (Michael Fulton)
North of the kingdom of Jerusalem, strongholds such as ʿAkkār, Crac, Saone, Bourzey, Shughr-Bakās, Margat and Baghrās (Gaston) were built on spurs or otherwise commanding positions. Even where the ground was flatter, as in the Homs– Tripoli corridor, east of the Syrian Coastal Mountains, and south of the Taurus Mountains, hills or old tells were often chosen, as was the case at Arima, Trapessac (Darbsāk), Turbessel and Ravandal. In the environs of Shayzar, the citadel of which was built on a natural spur that rises from the left bank of the Orontes, Abū Qubays (Bochebeis) was built on a hill at the edge of the plain, less than 22km to the west, and between them Tell Ibn Macher sat atop an ancient tell.
Along the coast, headlands and promontories could provide exceptionally strong positions. Seaward fortifications were harder to attack, as it was more difficult to undermine let alone approach them, allowing defences and defenders to be concentrated along landward sides. The city of Tyre, once an island connected to the mainland by a shallow tombolo, was synonymous with strength and impregnability since Alexander the Great’s siege of 332 BC. According to William of Tyre, who was archdeacon of the city from 1167 and then archbishop from 1175, two lines of walls, complete with towers, ran along the city’s seaward fronts and three lines of walls, studded with exceptionally large towers, guarded the narrow landward approach to the east. A single gate controlled passage through the landward defences. A number of other cities, including Acre, Beirut and Tripoli, also made use of the sea for protection on multiple fronts.
Two towers of hewn and fitted stones, of such greatness that one stone is with difficulty drawn in a cart by two oxen, were built at the front of the castle. Both towers are 100 feet in length and 74 in width. Their thickness encloses two sheds to protect soldiers. Their height rising up much exceeds the height of the promontory. Between the two towers, a new and high wall was completed with ramparts; and by a wonderful artifice, armed horsemen can go up and down within. Likewise another wall slightly distant from the towers extends from one side of the sea to the other, having a spring of living water enclosed. The promontory is encircled on both sides by a high new wall, as far as the rocks. The castle contains an oratory with a palace and several houses. The primary advantage of this structure is that the assembly of Templars, having been led out of Acre, a sinful city and one filled with all uncleanness, will remain in the garrison of this castle up until the restoration of the walls of Jerusalem. The territory of this castle abounds in fisheries, salt mines, woods, pastures, fields, and grass; it charms its inhabitants with vines that have been or are to be planted, by gardens and orchards. Between Acre and Jerusalem there is no fortification which the Saracens hold, and therefore the unbelievers are harmed greatly by that new fortress; and with the fear of God pursuing them, they are forced to abandon these cultivated regions. This structure has a naturally good harbour which will be better when aided by artifice; it is 6 miles away from Mount Tabor. The construction of this castle is presumed to have been the cause of the destruction of the other, because in the long wide plain, which lies between the mountainous districts of this camp and of Mount Tabor, no one could safely plough or sow or reap because of fear of those who lived in it.
(Adapted from Oliver of Paderborn 6, trans. Gavigan, pp. 57–8.)
Elsewhere, towns built against the sea might benefit from its protection on only one side. Some of these, however, had particularly strong citadels, which extended out into the Mediterranean, even if the town defences did not. The citadel of Caesarea and castles of ʿAtlit and Nephin were surrounded on three sides by the sea and water probably also filled the moat that separated each from the mainland to the east. When a new castle was built at Sidon in the early thirteenth century, it was sited on a shoal 120m offshore, connected to the town by only a narrow bridge. The stronghold of Maraclea (Maraqiyya) was similarly built on a shoal slightly further offshore.
Many Frankish strongholds built against the sea were among the last to fall to the Mamlūks: Bohemond VII agreed to destroy Maraclea in 1285; Tripoli fell in 1289; and Acre was famously taken in 1291. After the capture of Acre, Tyre, which had not fallen out of Frankish hands since it was taken in 1124, was abandoned, the towns people of Sidon fled, and Beirut was seized in a ruse. During the following months, the last remaining Frankish strongholds along the coast – the Templar strongholds of ʿAtlit, the sea castle at Sidon and Tortosa – were all abandoned. The small fortified island of Ruad (Arwād), more than 2.5km off the coast of Tortosa, was the last remaining Frankish stronghold in Syria, its Templar garrison holding out until a Mamlūk marine force finally came against it in 1302.
The availability of building materials was another significant factor that influenced where strongholds were constructed and what they looked like. Existing fortifications were typically occupied if available. This mitigated building costs and the incentives that had led to the initial construction of these structures often remained, while many continued to serve as symbols of regional authority. Some strongholds were left largely as they were found, merely repaired to ensure their defensibility. The majority, however, were developed or expanded over time.
Many of the Byzantine outposts built in western Syria were enlarged under Frankish rule, often through the construction of an outer bailey. Where outer defences already existed, as was likely the case at Saone, this is where further fortification efforts usually focused. In the coastal plain and other level regions, town defences were the most common fortifications at the end of the eleventh century. When town walls were strengthened, the line of the earlier walls was typically followed, allowing builders to incorporate new elements, such as additional towers or an outer wall, and rebuilding others on top of existing foundations.
Non-fortified structures might also be used as the core of a castle or citadel. The Roman theatres at Caesarea, Bosra and Sidon were all developed into significant strongholds. While that at Caesarea was fortified before the arrival of the Franks, the first work at Bosra appears to have taken place around the time of the First Crusade. Work continued later in the twelfth century under Muʿīn al-Dīn Unur, but it was under al-ʿĀdil and his Ayyūbid successors that Bosra’s theatre-turned-citadel took its recognizable form, with final touches added by the Mamlūks. Although dilapidated, the plan of the original theatre at Sidon, which the Franks converted into a castle or citadel along the circuit of the town’s walls, can still be discerned. At Baalbek, the Roman temple complex appears to have been made defensible to some degree before significant fortification efforts were undertaken in the twelfth century. It was further developed under the Ayyūbids and subsequently by the Mamlūks, converting the classical site into a formidable fortress. In the Syrian Desert to the east, the Temple of Bel was fortified and functioned as the citadel of Palmyra in the twelfth century, while a structure within the once-thriving Roman city of Jerash had been fortified by the time it was taken by Baldwin II in 1121. Like Kerak and ʿAjlun, Dārūm, built on a slight rise at the eastern edge of the Sinai Desert, may have been built on top of an earlier Byzantine monastery.
The availability of stone and materials for mortar was a consideration whenever a new stronghold was built. At some sites, building materials were provided by existing structures. Scavenging pre-cut stone in particular saved both time and money. At Baysān, no effort was made to repair the expansive Byzantine walls; instead, a simple administrative tower was built among the ruins of the Hellenistic-Roman town. Likewise, Bethgibelin, Ibelin and Gaza were each built near or among the ruined remains of Byzantine settlements. From Kerak to Bānyās and sites further north, ancient column capitals are among the most conspicuous spolia integrated into medieval walls. At sites without previous structures to incorporate or quarry, building materials had to be sourced elsewhere.
Bosra, theatre-citadel (after Yovitchitch).
Unlike in contemporary Europe, suitable timber for wooden fortifications was rare in the Levant. This contributed to a different building tradition in the Near East, one far more reliant on stone. Particularly in the twelfth century, Frankish builders benefited from the involvement of local populations, who presumably provided most of the labour. In the north, friendly groups of Armenians and Syriac Christians may have been relied upon by the first generation of Frankish lords. In Palestine, where large building programmes were uncommon in the early twelfth century, the Franks would have become familiar with local practices over time, as they observed and worked with local communities of Christians and Muslims. European stylistic elements are most clear in the large castles and citadels of the thirteenth century, such as ʿAtlit, Montfort, Caesarea, Acre and Crac. These were financed by the military orders and visiting crusaders, and likely involved higher numbers of European masons, who often employed the gothic style to which they were accustomed.
The availability of natural resources and the will to administer a region were not in themselves sufficient incentives to construct a stronghold; there also had to be a perceived threat. A tower was often sufficient to provide a base for a small policing force that had only to contend with bandits, while a great castle might be necessary to secure a claim over a broader frontier region. Although strongholds were fundamentally defensive structures, they were critical bases from which more offensive activities could also be launched.
Robbers had long plagued the roads of Palestine, preying on pilgrims and other bodies of poorly armed travellers. In 1065, a large group of German pilgrims, led by Bishop Gunther of Bamberg, was ambushed between Caesarea and Ramla. The pilgrims took refuge in a tower inside Kafr Salām until rescued by Fāṭimid forces from Ramla. The small towers built by the Franks, even when provided with an outer wall, were not meant to resist large invading armies, but rather smaller threats such as this. As more of these small strongholds were constructed, each serving the interests of its individual lord, Frankish rule expanded and banditry appears to have declined.
Similar to the administrative towers built by minor Frankish lords, the Templars constructed a number of towers along pilgrim routes and in areas particularly prone to banditry, reflecting the order’s original mandate of protecting Christian pilgrims. Among these was the tower of Le Destroit. Built along the coastal road, 23km north of Caesarea and 11km south of where the Carmel juts out into the Mediterranean, the tower monitored a natural bottleneck between the heights to the east and the sea to the west. It was in this area that Baldwin I was wounded by bandits in 1103. In the words of Oliver of Paderborn, ‘the tower was placed there originally because of bandits who threatened strangers ascending to Jerusalem along the narrow path, and descending from it; it was not far from the sea, and on account of the narrow path it was called Destroit’.17
The Templars constructed a string of similar towers along the pilgrim road from Jaffa to Jerusalem, and onwards to Jericho and the Jordan River. Elsewhere, the Templars showed a similar tendency to accumulate lands along critical roadways. Latrun (Toron des Chevaliers), Castellum Arnaldi (Yālu), Casel des Plains, La Fève (al-Fūla) and Le Petit Gérin (Zirʿīn), as well as Safed and Jacob’s Ford, all dominated a crossroads or important roadway and all came into Templar hands in the twelfth century.
Certain northern castles were similarly built in positions that overlooked critical mountain passes and restricted roadways. At the southern end of the Syrian Coastal Mountains, the Templar castle of Chastel Blanc (Ṣāfītā) was built on a secondary route between Homs and Tortosa, north of the main Homs–Tripoli corridor. Further north, Qadmūs, once a stronghold of the Nizārī Assassins, overlooks the road between Maṣyāf and the coastal town of Valenia (mod. Baniyas). Just south of Valenia, the Hospitaller stronghold of Margat sits 2km inland from the coastal road, leaving a satellite tower, the Tour du Garçon (Burj al-Ṣābiʾ), to police day-to-day traffic along the road. Al-Kahf, Malaicas (Manīqia), Vetula, Balāṭunūs and Saone were also positioned near or along routes across the Syrian Coastal Mountains, while Abū Qubays, Bourzey and Shughr-Bakās were near the eastern ends of roads through the mountains. Between the Iron Bridge and al-Athārib, Ḥārim and Artāḥ observed the western entrance to the Sarmada Pass, a chokepoint on the main route between Antioch and Aleppo, while Sarmada (Sarmadā) was fortified at the eastern end. North of Antioch, the eastern end of the Syrian Gates (Belen Pass), which connected Antioch and Alexandretta on either side of the Amanus Mountains, was observed from a distance by Baghrās and Trapessac, 6km to the south and 12km to the north respectively.
Le Destroit. (Michael Fulton)
Templar towers on the road from Jaffa to Jericho
River crossings were also frequently defended. The fortified communities of Qalʿat al-Rūm (Ranculat, Hṙomgla), al-Bīra, Qalʿat Najm, Bālis, Qalʿat Jaʿbar, Raqqa and al-Raḥba each stood at a crossing over the upper Euphrates, while the Templar castle at Jacob’s Ford was built at a natural crossing point over the upper Jordan. These strongholds and their garrisons could no more obstruct the crossing of a large army than could other castles block roads or mountain passes; however, in the absence of a hostile army, each could control traffic moving across the river. By occupying chokepoints, be they roads, passes or river crossings, the garrisons of these strongholds could intercept or block small raiding parties and launch reciprocal incursions into enemy territory.
Small strongholds might also be positioned to secure natural resources, most often water. Population centres usually developed naturally in areas where water was available, providing a number of incentives to fortify such sites. The springs at ʿAyn Jālūt and Ṣaffūriyya, rallying points for armies in the Jezreel Valley and lower Galilee, were both commanded by Frankish towers in the twelfth century. Neither of these strongholds would have been able to prevent a large army from accessing the springs, but their domination of these water sources was a practical and symbolic extension of Frankish influence over the surrounding area.
Many large cities owed their prosperity to their positions. Coastal cities like Acre, Tyre and Tripoli were major ports, while inland urban centres, such as Damascus, Homs, Hama and Aleppo were on the main inland road that ran north–south to the east of the coastal mountains. Each owed much of its wealth to the trade that moved through its gates. Collectively, these formed a transition area where goods from the East and West were exchanged, Christian and Muslim traders relying on their counterparts to facilitate the movement of their wares. Edessa was in a particularly strategic inland position, on the western side of the main route across northern Mesopotamia, along which traffic passed on its way to and from northwestern Syria, Cilicia and the Mediterranean. The advantages brought by its position also came with drawbacks: armies raised in the Jazīra often passed below its walls on their way to wage jihad against the Franks and the sultans of Rūm to the northwest regarded it as an appealing prize. From its acquisition by Baldwin I in 1098, Edessa was attacked no fewer than seven times before Zankī’s forces besieged the city in 1144, finally taking it from the Franks. Jerusalem, by comparison, occupied a position of minimal strategic importance; its value was instead its religious significance.
The benefits enjoyed by urban communities along the Mediterranean saw the Templars invest significantly in three coastal sites in the thirteenth century – ʿAtlit, Tortosa and Sidon. With the influx of labour and donations that accompanied the Fifth Crusade, the Templars tore down Le Destroit, which had come to include an outer bailey, and built a far larger castle (ʿAtlit) on a peninsula that juts out into the Mediterranean 1km to the southeast. They named this new castle Castrum Perigrinorum (Pilgrims’ Castle). Further north, the order gained control of Tortosa in the 1150s. Here, they developed a tower into the heart of a much larger castle. The Templars did not acquire Sidon until 1260, from which point they developed the sea castle. The sea provided a measure of protection for each stronghold and direct access to Mediterranean trade networks. Although the Templars had no navy to speak of, they benefited from a close alliance with the Venetians through much of the thirteenth century. The only comparable coastal possession that the Hospitallers came to hold was Arsūf, which they acquired in 1261 and lost four years later.
Some castles were built with the primary objective of isolating another stronghold. These were often employed like permanent siege forts, assisting with a distant blockade against a particular urban community. The fighters of the castle could both raid around the targeted town, depriving it of local resources and supplies, and intercept raiders who emerged from it, increasing security and promoting friendly economic activity to their rear. Castles were most often used in this way during the early twelfth century, when Frankish figures sought to apply pressure against cities they could not yet invest directly.
One of the first blockading castles was Mons Peregrinus (Sandjīl). Constructed by Raymond of St Gilles following the Crusade of 1101, the castle was built more than 2km from Tripoli, which Raymond hoped would eventually become his seat of power. Over time, a small town developed outside the castle, which was sacked in 1104 during a particularly successful sally led by Fakhr al-Mulk ibn ’Ammār, leader of the Tripolitans. Raymond was injured during the attack, possibly falling from a burning roof, and died the following year. Despite the success of this raid, the castle effectively isolated Tripoli from the interior of Syria, and the city enjoyed only sporadic Fāṭimid naval support. By 1108, the circumstances had become so desperate that Fakhr al-Mulk left to seek support from the caliph in Baghdad. In 1109, following the arrival of Raymond’s son, Bertrand, the weakened city was attacked and taken by a combined Frankish force.
William of Tyre claims that Hugh of St Omer, who had succeeded Tancred as the second prince of Galilee, constructed Toron in order to apply additional pressure against Tyre. Hugh chose to build the castle at Tibnīn, about halfway along the main road between Tyre and Bānyās, the westernmost stretch of the route from Tyre to Damascus. The presence of the castle and its garrison would have been felt in Tyre as Frankish influence was more generally extended into the extreme north of Galilee. Hugh’s choice of this position, conspicuously midway between the Jordan and the Mediterranean, may also have been a fairly open claim to this region, some distance from his seat of power at Tiberias. Something of a marcher lord in the European sense, Hugh died only months later fighting against a Damascene force in the summer of 1106. His successor, Gervais of Bazoches, was captured two years later and died a prisoner in Damascus. Galilee was then administered by royal officials until it passed to Joscelin of Courtenay in 1112.
When Baldwin I of Jerusalem led a force against Tyre in 1108, he ordered the construction of a small castle on Tell al-Ma‘shū qa, probably 2–3km east of the city’s walls. This castle is not mentioned thereafter, suggesting it was probably a siege fort rather than a permanent stronghold. The Franks spent a month outside the city before the Tyrians bought their withdrawal for 7,000 dinars, at which point the army moved on to attack Sidon. Baldwin led another unsuccessful siege of Tyre through the winter of 1111/12, before commissioning the castle of Scandelion in 1117. This stronghold was constructed about 14km to the south of Tyre and less than 4km inland from the Mediterranean, allowing the garrison to monitor traffic along the coastal road between Tyre and Acre. By the time Ibn Jubayr passed Scandelion in 1184, a walled town had developed around the castle. The construction of Toron and Scandelion contributed to the spread of Frankish influence, leading William of Tyre to observe that the Muslims of the city controlled little beyond its walls long before Tyre was taken in 1124.
With the fall of Tyre, Ascalon became the last Muslim foothold on the coast. It was reasonably cut off from the remainder of Palestine from the time of the First Crusade, during which the Franks occupied Jaffa (47km to the north), as well as Jerusalem and the settlements of the Judean hills to the east. Despite its relative isolation, Ascalon was regularly supplied by sea and raids from the city posed a significant threat to those living under Frankish rule in southern Palestine. In order to increase security between Jaffa and Jerusalem, a communal effort was undertaken to rebuild Castellum Arnaldi in the early 1130s. The small castle, which sat on a rise overlooking the road to Jerusalem as it leaves the coastal plain and enters the Judean hills, had been destroyed by a Fāṭimid force in 1106 or 1107. Once rebuilt, the stronghold and its garrison increased security in the area and made the journey from the coast to Jerusalem much safer for pilgrims and merchants, as the Judean hills were a regular haunt for bandits. By 1179, the castle had passed to the Templars.
During the reign of Fulk, three castles were built to apply pressure against Ascalon and contain its raiders. Construction of Bethgibelin (Bayt Jibrīn) was undertaken by the patriarch and a group of nobles while Fulk was on campaign, and it was granted to the Hospitallers upon its completion in 1136. The site was a strategic one: well watered and at a crossroads on the edge of the Judean hills where roads reached out towards Hebron and Jerusalem to the east as well as Gaza and Ascalon to the west. Over the following years, a community developed around the castle. In 1141, Fulk, joined by the patriarch and notables of the kingdom, built a second castle at Ibelin. Whereas Bethgibelin was built to the east of Ascalon, on the main route to Hebron, Ibelin was built to the north, 7km from the coast, on the road to Jaffa where it forked and the eastern branch proceeded to Ramla and Lydda. The castle and land around it was granted to Balian the Elder, progenitor of what would become the influential Ibelin family, which took its name from the castle and accompanying lordship. The chosen site had an abundant supply of water and the small town nearby, which had been abandoned during the early twelfth century, saw new life following the establishment of the castle. The apparent success of the first two castles inspired the construction of a third. Blanchegarde was built the following spring (1142), and once more the leaders of the kingdom collectively set out to initiate the project. This time the king retained possession of the stronghold once it was complete. The castle was constructed between the other two and, like Bethgibelin, it was built at the edge of the Judean hills. Collectively, the three castles formed a line, Ibelin and Blanchegarde about 29km away from Ascalon and Bethgibelin a few kilometres further. By no means did they function as a wall that cut Ascalon off from the rest of Palestine; instead, they dominated the main roads running to and from the city, thus obstructing only the main arteries of travel.
There were considerable similarities between these castles. Scandelion, Arnaldi, Bethgibelin, Ibelin and Blanchegarde were all planned according to a quadrangular or quadriburgium design, typically with towers at each corner; all but Bethgibelin were positioned on a small hill; and a good water supply is noted near most. The quadrangular design was simple and efficient, providing a very functional internal space. It is often assumed that Fulk had a special affinity towards castles, as he is known to have commissioned some before travelling to the Levant and he hailed from Anjou, a region where some of the earliest known stone castles in Europe were built. However, eleventh and early twelfth-century Angevin castles, often built around a central keep, were quite different from these quadrangular strongholds. Likewise, Fulk appears to have been present at the foundation of only two of the four castles built during his reign.