(1) Mamluk ‘askari wearing laminated leather ‘hoop armour’ and a leather helmet. The thigh defences are based upon pictorial sources because there is not yet archaeological evidence for this form of armour.
(2a) The structure of a laminated leather cuirass based upon elements from several incomplete cuirasses.
(2b) The system of straps inside the cuirass which support the horizontal lames, showing how the straps go through to the exterior of the lowest lame.
(3a) The structure of a rawhide lamellar cuirass, based on the remains of two examples found in a tower of the Citadel of Damascus.
(3b) A single lamel showing the shaping of the upper and left edges when viewed from the front, but without any lacing.
(3c) External view of the lamellae overlapping upwards and to the left, with visible lacing.
(3d) Interior of the vertical edge of the lamellae showing the edging strip secured by rivets and additional lacing.
(3e) Attachment point for tassets, which are not included in this reconstruction.
(4a) Laminated leather helmet, based upon several examples found in a tower of the Citadel of Damascus. Part of the side is cut away to show (from outside to inside): the outer layer of gesso painted with several layers of orange-red varnish; a layer of pieces of roughly cut, thin cotton cloth, overlapping each other in irregular shapes; a middle layer of gesso; an inner layer of cloth; an inner layer of gesso; multiple layers of irregularly shaped, very thin leather and, finally, part of the interior. The dotted line indicates the complete outer profile.
(4b) Iron ring, shaped washer, and inner part of rivet beaten over to hold chin-strap.
(4c) Interior of the helmet showing soft leather supporting segments.
(5) Partially gilded iron mace head.
(6a) Straight-bladed sword with an almost rounded tip and a single broad fuller groove, gilded bronze pommel, leather-covered wooden grip, gilded bronze quillons and sleeve over much of the grip.
(6b) Wooden scabbard covered in dark shark-skin sleeves of light brown soft leather around scabbard beneath suspension points, polished or gilded bronze or brass collars for the rings which take straps to the belt, polished or gilded chape with extension up the lower side of the scabbard.
(6c) Leather belt with gilded bronze buckle, buckle-plate, strap-end and stiffeners.
(6d) One of a pair of gilded bronze loops for the scabbard suspension straps, attached to the back of the belt by short leather straps.
(7a) Dagger with gilded or polished bronze grip and guard.
(7b) Leather-covered wooden sheath with gilded or polished bronze mounts.
(1a) Horse armour, late 13th to 14th centuries, consisting of three cloth-covered, scale-lined elements covering the rear, front and neck of the animal. Circular heraldic medallions made of appliqué work cloth are sewn to the covering of the armour. The horse’s head is protected by an iron chamfron consisting of three hinged elements, covered with stained leather secured to the iron plates by a decorative system of lacing.
(1b) Detail of the structure of the corner of scale-lined horse armour with the decorative outer layer of cloth removed so that only the lowest three rows of riveted bronze scales are showing.
(1c) Interior of the corner of the scale-lined horse armour showing the main leather structural foundation, the large copper rivets securing the edging strip, the beaten heads of three of the four rivets holding each scale and the fourth rivet securing a scale that is not beaten flat but goes through a rectangular copper washer before being turned over.
(1d) Demonstration of the method of joining the sheets of horse armour beneath the animal’s neck.
(2a) Laminated 15th-century leather horse armour consisting of five main elements to cover the rear, sides, front and neck, plus a cloth-covered, three-panel steel chamfron for the animal’s head.
(2b) Schematic arrangement of the main elements of the horse armour and their constituent lames. The red box indicates the area shown in detail in 2c and 2d.
(2c) Exterior of the indicated upper corner of the chest armour showing the straps supporting the horizontal laminated leather lames.
(2d) Interior of the indicated upper corner of the chest armour showing a layer of quilted cloth to support this uppermost part of the horse armour.
(3a–3c) Decorated 15th-century steel chamfrons.
Medieval Islamic slavery was not based upon the master’s ownership of the slave’s body and labour, but upon a concept of patronage which, although not between equals, nevertheless imposed obligations on both parties. In somewhat idealized medieval Islamic legal terms, both parties were mawali (sing. mawla), a superior mawla and an inferior mawla, and their relationship was supposedly a sort of alliance. This ethical concept enabled slaves to have legal rights, almost as if they had been adopted as ‘foster sons’ by a master who accepted legal obligations as their ‘foster father’.
Study of the history and culture of the Mamluk Sultanate has suffered from Western scholars’ deeply engrained abhorrence of perceived ‘slave rule’. However, in reality, the medieval Islamic world had a distinctive basis of authority, especially in the Mamluk Sultanate where legitimacy of rule was a prize to be won, like any other. Rarely inherited by ties of blood, it could be confirmed by a ruler’s behaviour and piety, and by the effectiveness of his government. What is even less understood today is the degree to which women of the Mamluk elite played significant cultural and even religious roles as patrons.
In much the same way, the loyalty of leaders and troops could not simply be inherited by the successor of a preceding ruler. It had to be earned, or at least purchased, as political and military loyalty was much more fluid than in medieval European societies – hence the Mamluks’ largely undeserved reputation for disloyalty amongst many pre-modern European commentators. Loyalty was almost democratic, being based upon respect, obligation, fellow feeling and rewards. The Islamic chronicles also make clear that emotion played a major role as friendship was undoubtedly a major feature in group solidarity within mamluk society.
Scholars still wrestle with the question of whether the translation ‘slave’ – with all its associations for modern readers – is suitable for mamluk soldiers, even in the early years of their individual careers. In a mamluk context, the term ‘slave’ normally referred to an originally unfree status, not a current one. Furthermore, the word’s most common associations were with obedience and faithfulness, characteristics that reflected a patrimonial rather than a feudal society.
In practical terms this meant that elite military units within the Mamluk army might better be described as households rather than regiments. The great medieval Arab historian Ibn Khaldun (1332–1406) echoed these positive attitudes when writing about mamluks:
It was by the grace of Allah … that He came to the rescue of the True Faith, … restoring in Egypt the unity of the Muslims, guarding His order and defending His ramparts. This He did by sending to them, out of this Turkish people and out of its mighty and numerous tribes, guardian amirs and devoted defenders who are imported as slaves from the lands of heathendom to the lands of Islam… They embrace Islam with the determination of true believers, while retaining their nomadic virtues, which are undefiled by vile nature, unmixed with the filth of lustful pleasures, unmarred by the habits of civilization, with their youthful strength unshattered by excess of luxury.
Who, then, were these troops? Christian Europe was regarded as unsuitable as a source for large numbers of ‘slave’ recruits, and until the 14th century European recruits were commonly regarded as physically ugly, habitually disloyal and generally unreliable. Sub-Saharan Africa had previously been a source of recruits for Egypt and North Africa. However, these were used almost solely as infantry because, as yet, Africans had virtually no cavalry tradition. Non-Islamic India could also be a major source, but Indian mamluks or ghulams were rarely seen beyond the Islamic states of India itself, neighbouring Afghanistan and eastern Iran.
In contrast, the steppes of south-eastern Europe, and central and inner Asia offered a suitable and substantial pool of potential recruits. As Sa’id al-Andalusi wrote of the Turks during the first half of the 12th century: ‘The Turks are also a nation having a large population (second only to the Chinese)… They distinguish themselves by their ability to wage war and by the construction of arms, and by being the ablest horsemen and tacticians.’
In the 10th century these Turks were well known to Muslim scholars such as al-Khwarazmi, who noted that the Turkic peoples of the western steppes had long been under Iranian cultural influence. In reality, this influence had been mutual, as there had been a prolonged Turkish influence upon at least the military aspects of Iranian civilization. At the same time, there may have been some differentiation between truly ‘enslaved’ recruits from more distant tribes who came via the Oghuz and Kipchaq states, and Oghuz or Kipchaq recruits who accepted the status of mamluk more willingly in the knowledge that it could open up glittering career prospects.
Kipchaqs had been widely employed as soldiers even before the Mongol conquest of the Kipchaq tribal states, having been seen in the Byzantine Empire, eastern Europe and the Islamic world. Following the Mongol catastrophe of the early 13th century, however, Kipchaq recruits were far more numerous and probably formed the bulk of those enlisted by the early Mamluk Sultanate. They came from an area known to Persian speakers as Dasht-i Kipçak (the Kipchaq Desert), steppe grasslands with a little agriculture and some urban centres. These steppes were crossed by several trade routes and the natives were thus in contact with a large part of the known world. Kipchaq culture, like that of the other Turkish-speaking steppe peoples, was characterized by a social system where patterns of loyalty and individual attachment were both personal and communal. These people were not territorial, and so it was relatively easy for a mamluk recruit to transfer his loyalties to a new person or community.
Although the mamluk phenomenon could be found across virtually the entire medieval Islamic world, the version that emerged in Egypt and produced the remarkably successful Mamluk Sultanate, might have had its primary origins in the Fatimid Caliphate. The Fatimids had ruled Egypt and dominated much of North Africa and Syria for several centuries before being overthrown in the late 12th century by Saladin, the founder of the Ayyubid dynasty. The Ayyubids took over and then refined the Fatimid military system, abandoning the recruitment of African mamluk infantry but greatly extending their recruitment of Turkish cavalry. Saladin’s successors thereby established a mamluk army that then replaced the ailing Ayyubid Sultanate with its own Mamluk Sultanate. To understand the Mamluk army one must first understand the fragmented Ayyubid armies from which it emerged. The term ‘askar could refer to the unit garrisoning, or paid, by a town, as it would have done for centuries, and these early ‘askaris would not normally be considered part of the military elite. The ‘askar of an Ayyubid ruler, however, consisted of professional, full-time ‘askaris. The most highly regarded of them were by this time largely of mamluk origin, though their numbers could still be remarkably small