The Principles of Defense in the Medieval Crusades

The Siege of Calais (1346) / Wikimedia Commons

The 13th century was not only a period of revolutionary change in the design of Crusader fortifications, but also in strategic priorities.

By Dr. David Nicolle
Visiting Research Fellow
University of Nottingham

The 13th century was not only a period of revolutionary change in the design of Crusader fortifications, but also in strategic priorities. The main efforts were now focused upon the protection of people rather than territory and in some places the numbers to be defended were very high. Acre was the biggest city, but Crusader-ruled Antioch still had a population of around 100,000 people, mostly Greeks and other non-Catholic Christians. Another feature of this period was the growth of suburbs next to major fortresses, mostly defended by a single wall. Crusader-held territory now consisted of parts of the eastern Mediterranean coast with the sea to the west and hills or mountains to the east. This strip had, in fact, been cut in two by Saladin’s reconquest of some of the coast north of Latakia. Some castles were sited to cover the few east-west routes through which Islamic armies might enter Crusader territory, but it was just as important to locate fortifications on the coast, to hinder movement by invaders who reached it. Furthermore, the Crusader States were now entirely dependent upon contact with, and support from, Europe. The protection of ports and harbours was thus essential. By the 13th century, European fleets dominated the Mediterranean, and without them the Crusader States could not have survived as long as they did. Nevertheless, this domination was not complete and the Mamluks made several efforts to revive the Egyptian navy. Meanwhile, smaller Turkish fleets based along what are now the Mediterreanean and Aegean coasts of Turkey grew increasingly daring.

The perennial problem of ensuring reliable supplies of drinking water resulted in extraordinary care being taken in the fortified Hospitaller headquarters in Acre, with every drop of rain from the wet season being stored. This was not only for drinking but also for hygiene, such as the flushing of the communal toilets; two parallel water systems being kept scrupulously separate. However, Islamic architects always demonstrated a little more sophistication in such matters, and in many fortifications the most impressive water storage cisterns date from Mamluk rebuilding rather than from the Crusader period. One example is in the castle at Safad where a circular large cistern, excavated from rock then covered by a masonry dome, lay beneath the great Mamluk tower at the southern part of the site.

The Crusader States’ shortage of military manpower was also growing more acute. Architectural and engineering skills could help greatly, but could not solve this problem definitively; as a result, most 13th-century Crusader fortifications were designed for small garrisons. Large garrisons existed, though rarely, and they were usually mustered for offensive purposes. In fact many Crusader castles were gravely undermanned when the final crisis came. Belfort is said to have had 22 knights and 400 other men when it was besieged by a Mamluk army in 1268. Even so the Mamluk Sultan Baybars still felt the need to bring 28 powerful siege machines against it.

Whether a reliance on fortification made the Crusader States vulnerable to the Mamluks’ highly developed forms of psychological warfare seems doubtful. This interpretation of events probably reflects the attitudes of 19th- and 20th- century military historians rather than the realities of 13th-century Middle Eastern warfare. Another myth concerns a supposed system of visible communication between Crusader castles. According to this theory, those in Cilicia formed part of an elaborate network. However, most were not in actual or useful line-of-sight with each other and the Crusader States’ hypothetical chain of signal beacons probably never existed. The good visibility enjoyed by such garrisons stemmed from the fact that they were stationed on hilltops for defensive reasons, not for communication.

Height was always sought after, although the emphasis on defending the weakest slope meant that the strongest part of a fortification was not necessarily at its highest point. On the other hand, locating a castle on a high place often limited its internal space. Haruniya, for example, was given to the Teutonic Knights in 1236. Here a largely 10th-century Islamic fort was in the hands of a Latin lord by the late 12th century. It consisted of a cramped central courtyard within a shell keep with two floors of shooting galleries and a rounded tower. The Teutonic Knights probably repaired the tower and perhaps used it as a chapel, but did little else. The northern Lebanese castle of Gibelcar was similarly cramped and rudimentary. Yet the site is so inaccessible that Sultan Baybars found it as difficult to take as the far larger and more sophisticated castle at Crac des Chevaliers. In the event Baybars’ sense of achievement when Gibelcar finally fell is reflected in a letter he wrote to Prince Bohemond VI of Antioch:

‘We have transported the mangonels there through mountains where the birds think it too difficult to nest; how patiently we have hauled them, troubled by mud and struggling against rain.’

The citadels that defended a town or served as places of refuge for the inhabitants faced different problems. They were almost always easier to access than mountaintop or spur castles and could be vulnerable to attack from within the town if it fell to an enemy. In fact urban areas often provided good positions for mangonels to hurl stones against a citadel. A different problem was caused if a suburb extended around or beyond the citadel, leaving the latter as a fortified enclave within the urban area. This happened at Acre, where the Castle of the King’s Constable and the Convent of the Hospitallers lost much of their original value following the fortification of Montmussard.

Crusader urban defences usually had an outer ditch, often with a counterscarp wall. In many places all or part of the city wall was revetted with a sloping talus. Outer walls and barbicans were not universal in the Crusader States and do not seem to have been used in western Europe before the 13th century. In most places the walls themselves consisted of the previous Islamic defences, more or less improved, as was the case at Arsuf, though here the existing city walls were considerably strengthened. The main changes were usually the addition of larger towers along the curtain wall, and occasionally the building of a second wall. At Acre the resulting defences were particularly impressive. So much so that during the final siege of 1291, the Mamluk army needed a massive and prolonged bombardment using a very large number of the most powerful trebuchets before they could break into the city.

These new-style towers were much more formidable than those built in the 12th century, clearly impressing pilgrims like Wilbrand von Oldenburg, who visited Acre in 1212:

‘This is a fine and strong city situated on the seashore in such a way that, while it is quadrangular in shape, two of its sides forming an angle are girdled and protected by the sea. The other two are encompassed by a fine, wide and deep ditch, stone lined to the very bottom, and by a double wall fortified with towers according to a fine arrangement, in such a way that the first wall with its towers does not overtop the main (second) wall and is commanded and defended by the second and inner wall, the towers of which are tall and very strong.’

This meant that arrows shot from the inner wall could be aimed over the outer wall, which was about one-third lower than the inner. The towers were also staggered so that those in front did not obstruct archers in the rear towers.

The top of a rock-cut cistern next to the fortified manor house at Khirbat Rushmiyah, on Mount Carmel above Haifa. The carefully carved edge of the opening shows that it was designed to have a lid, perhaps of wood or stone. The chamber inside expands into a large bottle shape.

The great castle of Crac des Chevaliers was similarly ringed by an outer wall, which is generally believed to date from the early 13th century. The normal interpretation of Crac des Chevalier’s fortification maintains that the inner defences were strengthened after this outer wall was added, while the southern side of the castle was given massive new-style towers around this time, almost certainly serving as artillery emplacements.

The defences of Atlit castle were similarly designed with defensive artillery in mind. Here three rectangular gate towers were placed approximately 44m apart. They had two floors and were surmounted by a platform enclosed by a parapet. All three projected about 12m from the curtain wall. Behind them was an inner wall with two huge towers approximately 28m long by 18m deep, both of which were originally over 34m high. Their great size and height reflected their role as artillery bastions to bombard enemy artillery, or at least keep it at a reasonable distance.

The greatly increased number of archery embrasures, niches, machicolations and other wall features indicated that crossbows played a very significant role in the defence of 13th- century Crusader fortifications. Some sources refer to ‘underground vaults’ where ‘great- crossbows’ could be sited; these being found in Louis IX’s city walls of Caesarea, probably in the citadel of Arsuf, and perhaps forming a continuous line of niched embrasures in curtain walls and towers. It is also interesting to note the similarity in some details of design and construction, which almost suggest a conscious programme of refortification in the mid13th-century Kingdom of Jerusalem.

Comparable details are found in Crac des Chevaliers, where the outer wall and towers had archery slits to minimise the area of dead ground. These were staggered to avoid weakening the wall and to enable archers to command the area in front of the walls. Similarly a stone-vaulted chemin-de-ronde gave access to box machicolations. However, the cramped interiors of these machicolations meant that crossbowmen squatted or knelt to shoot. The entrance to Crac des Chevaliers was greatly strengthened, resulting in a highly developed bent entrance system, the whole length of which had ‘murder holes’ overlooking it. Although Crac des Chevaliers was a large castle, the space between its inner and outer walls remained so narrow that it could not be used as an outer bailey. The vulnerable south-eastern side of this gap consisted of an open water tank fed by an aqueduct from the neighbouring hill, both as a water supply and perhaps to inhibit mining. The massive inner walls followed those of the 12th-century castle but were built slightly outside the earlier fortification, leaving a narrow passage that was developed as a shooting gallery on the western and southern sides.

However, the southern and western walls were the most impressive, rising from a sloping glacis from which huge round towers emerged. There was even a shooting chamber within some parts of this glacis. Unfortunately, most of the wall-head defences have gone, though traces along the southern wall showed no machicolations. Instead there were arrow slits and larger rectangular openings, perhaps for great-crossbows or espringals.

The design of the walls of Acre meant that any part of the curtain wall that was breached was still covered by crossbowmen in the neighbouring towers.

The entrance complex of Crac des Chevaliers, mid 13th century. The original name of the Crusader manor house at Khirbat Rushmiyah is uncertain. The complex consists of a tower whose basement is partly groin vaulted and partly barrel vaulted. A rectangular forebuilding was added later, perhaps in the 13th century, and may have contained a staircase. Meanwhile the main door was protected by arrow slits.

Some of the smaller or less important Crusader castles had small resident garrisons, and in some cases no permanent garrisons at all. However, major fortified locations like Crac des Chevaliers housed a considerable number of people and animals, and this number could reach a remarkable level in times of crisis. Consequently even Crac des Chevaliers could get crowded. Small postern gates in the outer walls of such castles were not normally used for entry and exit, so at Crac des Chevaliers everything had to use the main east gate. Behind this was a long, covered, dog-leg entrance ramp (shown here) leading to the centre of the castle. It also went past what are believed to have been the main stables. Here, two war-horses are being brought out of the inner stables by their grooms; having been inside the stables for some time, one of the horses has reared up, as its owner watches, to the right. Behind the horse, a column of baggage donkeys coming down the exit ramp with their handlers has been held up by the commotion.

The entrance complex of Crac des Chevaliers, mid 13th century

Considerable emphasis was also placed on the fortification of gates; so much so that these were rarely attacked. They could include a drawbridge, portcullis and panels in the doors, plus embrasures to increase the number of shooting positions within the gateway. Limitations of space may have been why bent- gates were not always employed. At Atlit the three gate towers had straight-through entrances, two with a portcullis and one possibly having a slit machicolation. In contrast, two of the three gates at Caesarea were certainly bent. Each had a slit machicolation and a portcullis protecting the doors while the inner doors were defended by a slit machicolation. Most town gates, however strongly fortified, were less complicated than those of castles or citadels. The latter tended to have adjacent towers while access to the interior was usually via one or more right-angle bends, often through gate chambers that could be sealed off by the defenders.

Postern gates allowed sorties by the defenders if any attackers came too close. One of the last Hospitaller constructions at Crac des Chevaliers was a postern, built between 1254 and 1269. It had a tower on each side plus a portcullis and probably machicolations. The small castle of Cursat may have included an unusual postern, which seems to have been associated with a cistern, several underground apartments and a vertical shaft cut into the rock. This led from a subterranean complex in the eastern part of the castle, down to a lower chamber at the level of the base of the ramparts. The associated masonry probably dates from the late 12th or early 13th century and the vertical shaft may have been excavated after an upper firing platform was already in place. Perhaps archers climbed down the shaft to cover a small postern gate. Elsewhere some posterns opened several metres above the base of the wall, and were only reached by ropes or a ladder.

This complex at Cursat apparently made use of an existing crack in the rock that was subsequently enlarged; Crusader military architects often made use of available natural features. On the coast these included the sea itself. At Atlit a small promontory was cut by a moat, and although the available technology did not permit the excavation of deep sea-filled moats, the rock could be cut away to just below sea level. Since then, however, part of the coast has sunk, leaving some ‘sea moats’ deeper than they were in the Middle Ages.

The few safe harbours along the Crusader-held coast were vital for the survival of the Crusader States, and so were given special attention. Most consisted of small bays, sometimes sheltered by reefs or rocks. Man-made harbours (moles) existed at Acre, Arsuf, Atlit, Caesarea, Beirut, Sidon and Tyre in the Kingdom of Jerusalem, while others served the County of Tripoli and the Principality of Antioch.

In several places towers stood at the ends of such moles; one such being the Tower of Flies at Acre. This was garrisoned by guards, who checked the identities of ships, the arrival of which was indicated by the tolling of a bell, perhaps in this tower. Towers could also be armed with anti-shipping siege machines, which threatened to sink vessels that attacked the boom. Other towers served as anchor points for chains or booms to close a harbour entrance, like the floating wooden ‘chain’ constructed across Acre harbour by the Genoese during their quarrel with the Venetians in 1258.

Harbours were themselves usually separated from a town by a wall. However, quays were rare and small boats usually carried goods between the beach and ships moored in the harbour. Elsewhere, Jaffa had a very exposed port, which had been refortified by the early Crusaders. It was strengthened during the following century, but still fell to the Mamluks in 1268. The Citadel of Arsuf is sometimes said to tower above a small harbour enclosed by two moles and a breakwater. However, the area in question may actually have been just above water during the Crusader occupation. So perhaps the moles and breakwater enclosed a flat area of flimsy buildings serving as a sort of foreshore beneath the castle and city.

From Crusader Castles in the Holy Land, 1192-1302, by David Nicolle, Illustrated by Adam Hook (Osprey Publishing, 07.12.2005), published by Erenow, public open access.