386
CHAPTER 6Infantry continued to serve in almost all Muslim forces from the 7th to 13th centuries. Yet this had been true of most armies since war began. Even the nomads of Central Asia had some foot soldiers amongst them, even if they were only poverty-stricken tribesmen who could not afford a horse. Such criteria would oblige one to term any individual who seized a weapon, but did not possess a horse, as an infantryman. Within the world of Islam, however, professional infantry forces of men skilled in specific military tasks continued to play an important, though fluctuating, role despite the increasing importance of cavalry in most regions.
As such, those troops should be distinguished from tribal levies of backward peoples such as the Berbers who, in the early 8th century, fought virtually naked armed only with slings,1 or those peaceful peasantry of ʿAbbāsid Āzarbāyjān who would also fight with slings, should the need arise.2 The local volunteer muṭṭawīʾah who fought on many fronts against various infidels, heretics or pagans, often only for a limited period, should similarly be disregarded. Their duties could be wide-ranging, and included keeping order or collecting taxes.3 They could, however, also be
---------------------
1. Ibn ʿAbd al Ḥakam, op. cit., 129.
2. Minorsky, Studies in Caucasian History, p. 112.
3. H. Kennedy, "The Early Islamic City: Self Government or State Control?" during Colloquium on the Early Medieval City, (Univ. of Edinburgh, Dept, of Extra-Mural Studies 6th May 1978).
387
a source of rebellion if their religious sensibilities were offended by an insufficiently pious local governor.4 Such warriors would continue to play their part in the armies of Islam, particularly against the Crusades of the 12th and 13th centuries.
The same might be said of part-time city militias, known as aḥdāth.5 The origins of such forces are obscure, although they night have emerged from a mingling of various urban military or volunteer elements, including troublesome gangs based upon the quarters of a town. Their military duties were normally to defend their city and although they included some horsemen, the majority fought on foot. First appearing in the Jazīra and Iraq in the 9th century, they could have shared the same origins as the ʿurāt of Baghdād. These latter consisted of a city mob and although they fought with desperation and even some success against the Khurāsānīs of al Maʾmūn, they were quite distinct from the professional abnāʾ and ḥarbīyah troops in Baghdād.6 Unlike the later aḥdāth who seem to have possessed proper, if assorted arms and armour, these ʿurāt used home-made shields, helmets and slings that seem to recall Sassanian or even ancient Babylonian traditions.7
Such aḥdāth militias were also important in Syria from the 10th to 12th centuries where their development was greatly encouraged by the Fāṭimids in the face of Byzantine expansion.8 Many such
---------------------
4. Bosworth, "The Armies of the Ṣaf'f'ārids," p. 538.
5. Cahan, "Djayah", pp. 504-509.
6. Ayalon, "The Military Reforms of Caliph al Muʾtaṣim", pp. 11-12.
7. Al Masʿūdī, op. cit., vol. Vl, pp. 452-453, 461-462 and 470.
8. Beshir, op. cit., p. 53.
388
warriors probably broadened their military experience by participating in annual raids against Byzantium.9 The Syrian aḥdāth was still important when the First Crusade appeared on the scene at the end of the 11th century,10 and remained so during much of the Crusading era. At this time they might have merged with, or been referred to as, remnants of the earlier jund structure.11 Perhaps we have an impression of these ill-equipped 12th century local militias in one of the earliest surviving Syriac Gospels (Fig. 128).
Such local, non-professional or part-time, troops survived in other provinces including Egypt. Here a local Muslim shurṭa or police force played a prominent role in late Umayyad and early ʿAbbāsid times, but appears to have declined as Turkish ghulāms took over most military responsibilities.12 Even under the Ṭūlūnids, however, local levies probably played some part,13 and, as already mentioned, the Fāṭimids once tried to solve their lack of troops by arming the ḥawwālah labour corps. During the Crusades, muṭṭawīʾah volunteers came from Egypt, as they did from other parts of Ṣalāḥ al Dīn's empire,14 (Figs. 150 and 156), although by contrast many Egyptian cities, unlike those of Syria, do not appear to have had their own aḥdāth.15
---------------------
9. Leo VI, op. cit., Inst. XVIII.
10. C. Cahen, La Syria du Nord au Tempa des Croisades, (Paris 1940), p. 195.
11. Ibn al Qalānisī, passim; H. A. R. Gibb, "The Armies of Saladin", Cahiers d'Histoire Egyptienne, III (1951), pp. 304-320.
12. Kennedy, loc. cit.
13. Hassan, op. cit., p. 167.
14. Gibb, "The Armies of Saladin, loc. cit.
15. Ibn al Athīr, "Al Kāmil fī Taʾrīkh", in Arab Historians of the Crusades, F. Gabrieli trans., (London 1969), p. 253.
389
Where professional infantry were concerned, they were generally recruited on an ethnic basis. At first, of course, the most important such groups were the Arabs themselves. During Umayyad and early ʿAbbāsid times, as already stated, cavalry grew both in importance and in numbers. Yet they generally still fought in strict cooperation with the infantry. By the late 7th century, for example, one notably rich Arab tribe could still only field half as many armoured cavalry as it could infantry.16 At first tactics were simple, as they had been under the Rāshidūn Caliphs. Any armoured men (Figs. 122, 123 and 340) stood in the front ranks, usually kneeling behind their shields with rumḥ spears. These were supported by archers (Fig. 127) or men with ḥarbah short spears or javelins. All, apparently, also carried swords (Fig. 141).17
The close-packed, solid ranks of the early days proved ineffective against the fast-moving, but similarly Arab, Khārijīs and hence Marwān II introduced his famous reforms18 by breaking up the old five divisions, or khamis, of van, rear, centre and two wings, into smaller kardus squadrons. These could act in closer cooperation with cavalry. In many cases, throughout the Umayyad era, the lightly armed infantry archers (Fig. 122) seem to have been considered as a separate unit, perhaps to be moved around a battlefield in support of those among the more heavily armoured spear-carrying infantry who were under the greatest enemy pressure.19 Others could also be spread about as skirmishers.
---------------------
16. Al Al Masʿūdī, op. cit., vol. V, p. 140.
17. Al Ṭabarī, op. cit., vol. II, pp. 58, 337, 411, 520, 790, 652, 889 and 959.
18. Morses Farias, op. cit., p. 812.
19. Al Ṭabarī, op. cit., vol. Il, pp. 344,1551 and 1708.
390
Maces of various types (Fig. 115) might also have been used by the infantry, particularly against heavily armoured enemy troops kneeling in ranks behind their shields and with their spears held as pikes.20
An echo of these Umayyad tactics may be heard in the traditional battle-array of the kings of Ḥimyar or Yemen. Here a screen of "scouts" preceded an almost entirely infantry army, itself led by a front line of élite troops. Very small groups of horsemen protected the wings, and then only from some distance to the rear.21
Early ʿAbbāsid infantry tactics and equipment continued in this same Umayyad tradition. Ḥarbīyah heavy infantry with spears, swords and shields, and rāmiyah archers with bows, swords and shields, were soon joined by a small specialist corps of naffaṭīn. These men used Greek fire projectors and grenades, both in open battle and in sieges. Yet the tactics remained exactly the same, even in the early 9th century,22 and during regular Muslim incursions on to the Anatolian plateau infantry remained vital. if it were a small raid, then such infantry would protect their cavalry as it passed through the mountains and then perhaps remain in the passes to guard a line of retreat. If it were a major assault, however, the aim would be to find and destroy the Byzantine army. In such a task infantry played a leading role.23
---------------------
20. Ibid. vol. II, pp. 712, 912, 917, 927, 966,1889 and 1927-1928.
21. Mubārakshāh, op. cit., p. 325.
22. Levy, The Social Structure. of Islam, pp. 432-433; Leo VI, loc. cit.
23. Howard-Johnson, op. cit., pp. 225-226.
391
The most detailed information that is available concerns what might have been the most prestigious corps in the ʿAbbāsid army. These abnāʾ were the senior element in Baghdād's resident ahl Baghdād or jund al Baghdād.24 As mentioned earlier, they may have been descended from those Khurāsānīs who took part in the original ʿAbbāsid revolution. Their equipment is unlikely to have differed much from other ʿAbbāsid infantry, whether in the capital or garrisoning other major cities. Not only were they renowned in siege and counter-siege, but also in close country and mountains.25 Abnāʾ were also trained to maintain ranks with their long qanāh spears and broad swords, however hard the enemy pressed them, and then to fight hand-to-hand with khanjar short swords and sikīnah daggers. In attack they used the same weapons, although the mitrād short spear or javelin seems to have replaced the long qanāh. The ʿamūd mace seems also to have been added. Such troops were to be recognized by their long beards and ʿimāmah turbans.26 Although abnāʾ were clearly often armoured, they would also fight without cuirass or even shield, while their ranks similarly included a number of infantry archers.27
Although the abnāʾ and other non-Turkish, non-ghulām, troops of Baghdād declined in importance during the 9th and 10th centuries, there is little reason to suppose that they disappeared entirely. Indeed those "picked shield-bearing warriors of Baghdād" mentioned by Fidawsī may have been their descendants. These latter infantry
---------------------
24. Ayalon, "The Military Reforms of Caliph al Muʾtaṣim", pp. 6-7.
25. Al Jāḥiẓ, Rasāʾil al Jāḥiẓ, pp, 26-27 and 52-53.
26. Ibid. pp, 26-27.
27. Ayalon, "The Military Reforms of Caliph al Muʾtaṣim", pp. 33-34.
392
troops were described as using an early form of the charkh or jarkh crossbow, the penetrating power of which was clearly already recognized.28
By Firdawsī's day, however, another nation of infantry warriors had made, or rather remade, a name for themselves. These were the Daylamīs from the Elburz mountains south of the Caspian Sea. Some had, of course, long ago transferred their allegiance from the defeated Sassanians to the rising power of Islam. Not all became Muslims, nor did some of their descendants who similarly served as mercenaries in the armies of the Umayyad and ʿAbbāsid Caliphs. Large vestiges of pre-Islamic culture also survived in their homeland, even after the bulk of the population had accepted Islam.29 Much the same might be said of their Gīlānī rivals from the narrow Caspian plain. Both peoples served as infantry in many areas, although the Gīlānīs never made quite such a name for themselves and also remained true to orthodox sunnī Islam while the Daylamīs became fervent shiʿīs.30
The reputation of these warlike mountain folk, with their hirsute appearance, liking for garlic, very large brightly painted shields and traditional zhūpīn javelins, was clearly established in the Muslim world by the 10th century.31 At this early stage, and during the first phase of Būyid expansion in Iran, the Daylamīs still fought solely as mounted infantry, with mules to carry their
---------------------
28. Firdawsī, op, cit. p. 1280.
29. Bosworth, "Military Organization under the Būyids of Persia and Iraq," pp. 146-147.
30. Ibid., p. 149.
31. Mininorsky, "New Light on the Shaddādids of Ganja," p. 113; al Tanūkhī, op. cit., pp. 95-96.
393
equipment, javelins and armour.32 By the mid-10th century, leading Daylamī warriors employed shield-bearers, normally younger men or boys, though this might always have been the case. Full equipment now consisted of the large turs shield, dirʿ hauberk, zhūpīn javelin and perhaps a small, slender dagger known as a sakk or "nail".33 Some wore very heavy jawshan cuirasses while the jubbah hauberk was also mentioned. This latter was, however, described as an inadequate protection against the arrows of Turkish ghulāms.34 Yet the Daylamīs chief defence was his large, brightly painted turs shield, and to have this burned after a defeat was a mark of ultimate disgrace.35
As the Būyid state, established by Daylamī arms grew in wealth and power, so the equipment of these troops seems to have grown in variety. Battle-axes and bows are now mentioned,36 the latter apparently using the nāwak arrow-guide to shoot short arrows elsewhere known as ḥusbān or jawāldūz. Such short arrows had, of course, been known in Sassanian times.37 Daylamī tactics seem to have remained the same, with a steady advance in an unbroken line or moving shield-wall. Javelins were then thrown to disrupt the foe, followed by close combat with battle-axes that might have been similar to an example held by a 9th or 10th century (warrior on a north-Persian Gabri-ware bottle Fig. 344).
---------------------
32. Al Tenūkhī, loc. cit.
33. Miskowaihī, op. cit, vol. II, pp. 152-153.
34. Ibid., vol. II. pp. 161 and 336.
35. Ibid., vol. II. p. 205.
36. Bosworth, "Military Organization under the Būyids of Persia and Iraq," p. 149.
37. Al Balādhurī, op. cit., pp. 362-363.
394
This weapon is extraordinarily similar to one carried, perhaps by a mounted infantryman, in north-western India eight centuries earlier (Fig. 71). Some warriors from Daylam had also taken to fighting on horseback rather than operating solely as mounted infantry. These seem to have been heavy cavalry, perhaps influenced by the Kurds or Caucasian Albanians, and they fought with sipar shield and ṭabarzīn horseman's axe.38
During the later 10th and 11th centuries Daylamī infantry seem to have been most successful when cooperating closely with cavalry, usually Turkish ghulāms, in Iran and the east. This was also apparently the case in Fāṭimid Egypt where the Daylamīs became close allies of the Turks in Cairo's turbulent politics.39 Similarly, the Daylamīs of the Fāṭimid Caliphate still fought with zhūpīn javelins and battle-axes, and employed young shield-bearers for their tall, oval or kite-shaped shields that were now known as tāriqah (Fig. 149). Other weapons in the Daylamīs' Egyptian armoury might have included qalāchūr long curved swords, perhaps referred to in Egypt as galjūrī swords,40 plus nafṭ fire-weapons.41
While Daylamī infantry made their greatest impact in Būyid Iran and Iraq, and in Fāṭimid Egypt and Syria, they were also employed elsewhere. They had already served the orthodox sunnī Tūlūnid governors of Egypt,42 and the sunnī ʿAbbāsid Caliphs,
---------------------
38. Miskawalhī, op. cit., vol. II, pp. 336 and 382; Mubārakshāh, op. cit., 262.
39. Beshir, op. cit., pp. 47-49.
40. Ibid., pp. 67-70.
41. Ibid., pp. 47-49 and 74 n. 210.
42. Ibid.; Hassan, op. cit., p. 167.
395
as palace guards and urban police forces,43 though they themselves were shīʿī. The greatest Daylamī impact was necessarily in Iran where, in the late 10th century, their military fashions had become dominant in, for example, the province of Fārs.44 One may assume that the warriors shown on foot, wielding broad, blunt-tipped swords, in an early 11th century Arabic manuscript were typical of their day (Fig. 361).
The Būyids' gradual political decline encouraged Daylamī mercenaries to seek their fortunes further afield. They were soon as numerous in late 10th and early 11th century Syria as were unemployed Turkish ghulāms, and similarly sought service under the banners of Fāṭimids, ʿUqaylids and Mirdāsids.45 Eastwards, in Ghaznawid Afghanistān and north-west India they even formed an élite infantry guard with gilded and bejewelled rather than simply painted shields.46 After the fall of the Būyid state, the orthodox sunni Turkish Saljūqs seized power, but the reputation of the Daylamīs persisted to such an extent that they were again recommended as special palace guards, though whether they were ever recruited as such remains unclear.47
But it was in Fāṭimid Egypt that they not only continued to serve in the late 11th and early 12th centuries,48 but left
---------------------
43. Bosworth, "Military Organization under the Būyids of Persia and Iraq," p. 148.
44. Ibn Ḥawqal, Configuration de la Terre, J. H. Kramers and G. Wiet trans., (Paris 1964), vol. II, p. 283.
45. Beshir, loc. cit.
46. Bosworth, The Gḥaznavids, p. 111.
47. Niẓām al Mulk, op. cit., p. 67.
48. Canard, "La Procession du Nouvel An chèz les Fātimides," pp. 392-393.
396
perhaps their clearest impression in surviving art of this period (Figs. 157 and 158).
The Daylamīs were not, of course, the only professional infantry of Iran, nor were they the only such troops to earn a reputation outside their own area. The people of Khurāsān were noted foot soldiers, as well as cavalry. They too had their own long established traditions, particularly in siege-warfare and above all in mining operations.49 Those Khurāsānīs who were sent westward to the Byzantine frontier seem to have been experts in siege and counter-siege, as were their descendants still living in Anṭākīyah, Ṭarsūs and other parts of the province of ʿAwāṣim in the late 10th century.50 These people also manufactured siege equipment and other weaponry.51
The best descriptions of 10th century Khurāsānī infantry are probably to be found in Firdawsī's Shāhnāmah. Most details refer to open battle rather than to sieges, however, and here such east Iranian foot soldiers are described as advancing with sipar shield and bow, supported by spear-men with normal nīzah spears. Elsewhere those with sipar shields, jawshan cuirasses and nīzah spears formed the front rank while archers and men with iron quzār short spears or javelins stood behind them. A third variation had the corps of infantry to the rear of some cavalry and crossbowmen. In this case they were themselves led by men with nīzah spears and shields from GīIān, while archers with
---------------------
49. Cahen, "Djaysh" loc. cit.; C. Cahen, "Ḥiṣār", Encyclopedia of Islam, second edition, vol. III, pp. 469-470.
50. Canard, "Quelques observations sur l'introduction géographique de la Bughyat atʾ-Tʾalab", pp. 46-47.
51. Ibid.
397
similar sipar shields supported them.52 Some such troops appear on Iranian ceramics from the 9th to 11th centuries. These show a variety of warriors, standard bearers, javelin-men with large shields, and armoured troops with sword and buckler (Figs. 343, 351, 354 and 448).
All other eastern dynasties employed infantry to a greater or lesser extent, even the Ṣaffārids who, in addition to their large cavalry forces, had a sophisticated siege-train. This necessarily consisted of infantry,53 perhaps like those appearing on an east-Iranian chess-piece (Fig. 487). Infantry apparently rose in prestige under the Ghaznawids in the 10th to 12th centuries. In addition to Daylamīs, Indian troops were enlisted by these rulers and similarly fought as highly mobile, camel-riding, mounted infantry.54 Though rare in the art of the area, javelin-armed infantry do appear (Fig. 379).
Under the Ghūrids of the late 12th and early 13th centuries, infantry were naturally even more pronounced for, like the Būyids before them, this dynasty sprang from a mountainous region long famous for its foot soldiers. Their most notable tactic was the use of the kārwah, a large mobile mantlet of raw bullock hide stuffed with cotton, to be carried into battle on the shoulders of the leading troops.55 This was proof against most arrows and javelins and could also act as a movable shield-wall to
---------------------
52. Firdawsī, op. cit., pp. 1022,1156 and 1280.
53. Bosworth, "The Armies of the Ṣaffārids", pp. 547-548.
54. Bosworth, The Ghaznavids, pp. 113-114; C. E. Bosworth, "Ghaznavid Military-Organization", Der Islam, XXXVI (1960), pp. 59-60.
55. Bosworth, "Military Organization under the Būyids of Persia and Iraq", p. 151.
398
trap any enemy who broke the Ghūrid line.56
Once established as a power in northern India, the Ghūrids apparently adopted more sophisticated battle tactics and techniques, though apparently still giving high priority to their infantry. Archers were traditionally stationed on the right wing, zhūpīn javelin-men on the left. The placing of large numbers of men wielding ʿamūd and qurz maces, armoured in full kharātagin armour that covered the legs, plus troops armed with dabbūs mace, shamshīr long-sword and nāchakh axe in the centre would, however, seem to have been a new development.57 Otherwise Ghūrid ranks were much as they would have been in any ʿAbbāsid or other Muslim army that had a predominance of infantry. Armoured men with wide shields, siparhā farākh, short ḥarbah spears, tīr andāzān "throwing arrows'" or small javelins, stood in front. A second rank consisted of men armoured in both jawshan cuirasses and khaftān qambesons, having shamshīr long-swords, sipar shields and nizah long spears. Third came a rank of archers, also with shamshīrs and kārdhā buzurg long daggers, perhaps defended by a thicket of spaced wooden stakes thrust into the ground. Behind them all stood junior officers with daraqah shields, shamshīr long-swords and ʿamūd maces. Once again such forces acted in close collaboration with their cavalry, leaving wide spaces so that the latter could manoeuvre and strike the enemy should he break through the infantry ranks.58
Very little pictorial evidence survives from the Ghūrid era, although one of the finest pieces of Muslim metalwork was
---------------------
56. Maulānā Minhāj al Dīn, op. cit., pp. 352-353.
57. Mubārakshāh, op. cit., p. 339.
58. Ibid., p. 330.
399
made in the city of Hirāt at a time when both the Saljūqs of Iran and the Ghūrids of Afghanistān were reeling before a new wave of Ghuzz Turkish nomads. This metal cauldron, known as the Bobrinski bucket, does show a solitary soldier with a curved sword, perhaps a qalāchūr, battling against two horsemen (Fig. 368). He may, however, himself be a dismounted horseman.
Infantry continued to play a vital role under the so-called Slave Kings of Delhi, that mamlūk dynasty which seized control of much of the fragmented Ghūrid state early in the 13th century. They were divided into two groups, the Muslim pāydah and the Hindu paik.59 Ghūrid infantry were also reportedly serving the Khwārazmshāhs in Samarqand when this city fell to the Mongols in 1220 AD.60
Some infantry soldiers even served in Mongol armies, as they did in all nomad forces, though of course their status was low. Traditionally a true Turk loathed dismounting to fight.61 Yet even in their own earliest recorded sagas, probably written down in the 13th or early 14th centuries, a Saljūq hero was prepared to get off his horse, take arrows from his quiver, place them in his belt, roll up his skirts and face his foe on foot.62 Equally traditionally, the battle-plan of early Turkish nomad states like that of the Qara Khitai included small units of infantry among the cavalry on its wings, particularly on the left where there was a danger of being outflanked, and also to protect such vital
---------------------
59. Yār Muḥammad Khān, op. cit., p, 49.
60. Maulānā Minhāj al Dīn, op. cit., p. 274.
61. Leo VI, loc. cit.
62. Anon., The Book of Dede Korkut, p. 145.
400
assets as a herd of spare horses.63 That such traditions had a basis in fact is suggested by the well-documented armies of a later but comparable Turkish nomad state on the Eurasian steppes. The 15th to 17th centuries Crimean Tartars had, for example, a small infantry force raised from settled villagers living on the khān's personal domain.64
Whereas in the east of Islam most professional infantry were drawn from backward mountainous areas or from major cities, in Africa most infantry came from regions that were poor in horses for both economic and climatic reasons.
Not all African negro infantry were recruited as slaves. They appeared as free men in the first Muslim armies, and among the puritanical, Kharījīs of later centuries. Nevertheless, the bulk of black troops serving various North African dynasties do seem to have been of slave origin. Such warriors perhaps appeared first in Aghlabid Ifrīqiyah at the northern end of the great trans-Saharan trade routes.65 A little later they are recorded in Ṭūlūnid Egypt where they took a leading role as unarmoured naval warriors, in addition to forming an élite mukhtarah guard. This unit wore black robes over decorated iron armour, black turbans around their helmets, and fought with swords.66 At around the same time black palace guards were also in fashion in Byzantium.67
---------------------
63. Mubārakshāh, op. cit., p. 322.
64. L. J. D. Collins, "The Military Organization and Tactics of the Crimean Tartars during the sixteenth and seventeenth centuries", in War Technology and Society in the Middle East, V. J. Parry and M. E. Yapp edits., (London 1975), p. 260.
65. Beshir, op. cit., pp. 38-44.
66. Ibid.; Haasan, op. cit., p. 171; Rambaud, op. cit., pp. 419-420.
67. Howard-Johnson, op. cit., pp. 77-78 and 83.
401
Although the Fāṭimids were later to rely very heavily on such slave-recruited troops, they at first employed greater numbers of free-born Muslim negroes. Many such warriors from the region of Zawīla in the eastern Sudan marched in that Fāṭimid army that conquered Egypt in 969 AD.68 Other black Saharans included the Maṣmūdah, who were probably Berber in speech though coming from the western regions of the desert belt. Fighting solely as infantry with sword and spear, they first appeared in Fāṭimid Egypt under Caliph al ʿAzīz. Thereafter these Maṣmūdah played a major role in the garrisons of Syria until the rise of the Murābiṭīn in their own homeland stifled recruitment.69
Sudanese slave-recruited troops rose in importance and numbers until they formed almost half of the Fāṭimid army by the year 1000 AD. Thereafter they and free negro mercenaries remained the backbone of Egypt's forces for at least another century.70 Their loyalty and spectacular appearance probably led the Fāṭimids to choose them as guard units, as others had done before them. At the very end of the 11th century some three hundred black troops paraded, each with a pair of specially decorated javelins and shields with silvered bosses, during the New Year celebrations.71 Such splendid infantry guards were clearly still in existence fifty years later.72
---------------------
68. Beshir, loc. cit.
69. Ibid., pp, 28-34 and 38-44.
70. Ibid., pp. 38-44; Canard, "La Procession du Nouvel An chez les Fātimides", pp. 392-393.
71. Canard, "La Procession du Nouvel An chez les Fātimides", pp. 369-370.
72. Usāmah Ibn Munqidh, op. cit., p. 9.
402
Whether or not the dark-skinned infantry archers and javelin-men met by the Crusaders outside ʿAsqalān in 1099 AD and at Arsūf almost exactly a century later were Sudanese slave-troops, or Nubian or Ethiopian mercenaries, is unfortunately unclear. At Arsūf they are, in fact, more likely to have been Arab bedouin auxiliaries.73
Information about these troops in their original homelands, whether Muslim, Christian or pagan, can be found in certain medieval Arab geographies. In the late 10th century the Ahadī people of Darfūr and the Tibesti mountains were rich in iron and had much in common with the inhabitants of the southern Maghrib. They fought only as infantry with large, white leather, daraqah shields similar to the lamṭ shields of Morocco, plus ḥarbah short spears or javelins and poor quality local swords. These Ahadī also wore long, perhaps quilted, mufattiḥah protective garments.74 Quilted armour was certainly widely used in the similar area of Bornu many centuries later, and it may appear in contemporary art from both Nubia ((Figs. 185 and 186) and Ifrīqiyah (Fig. 194). Infantry predominated in much of Ethiopia at this time. In the north, in the Bāja area, the Bāzīn and Bārīya tribes were archers and javelin throwers, but used no shields.75 Pagans living near the source of the Blue Nile fought solely as infantry with ḥarbah short spears or javelins and local pikes of a hard wood known
---------------------
73. R. C. Smail, Crusading Warfare,1097-1193, (Cambridge 1956) p. 85; S. Runciman, A History of the Crusades, (London 1971) vol. III, p, 56; Anon., Itinerarium Pereginorum, The Third Crusade, pp. 87-88; Bahāʾ al Dīn. op. cit., p. 175.
74. Ibn Ḥawqal, Kitāb Ṣūrat al Arḍ, p. 58.
75. Ibid., p. 55.
403
as murrānah.76 Shields from that broad swathe of east African coast known to the Arabs as Zanj were also said to be of elephant hide, though being inferior to those of the Bāja,77
Little seems to have changed by the 14th century except that some cavalry, mostly riding bare-back in the ancient Berber and Nubian fashions, had appeared in central Ethiopia. Most warriors from this part of Africa remained infantry, however, and fought with large bows shooting relatively short arrows. They also used swords, spears, tall narrow shields and, above all, long javelins.78 An impression of these now perhaps better equipped Ethiopians may be seen in some of the earliest Abyssinian manuscripts (Fig. 657).
Negro troops of slave origin served the Zīrid lieutenants and successors of the Fāṭimids in Ifrīqiyah, as they did the Fāṭimid Caliphs themselves. Known as ʿabīd, they formed a number of regiments, some mounted but mostly infantry.79 Others must have served in Sicily but their descendants, unless they were newly recruited mercenaries, appear in Siculo-Norman art both in Sicily and on the mainland. In most cases their weapons, as distinctive as their tightly curled hair, are portrayed as mace and buckler (Figs. 588 and 606). They clearly continued to be recruited by North African rulers, at least until the 14th century in Ḥafṣid Tunisia where those of Guinean origin were known as janāwa and fought with spear, sword and small shield. Whether
---------------------
76. Ibid., p. 57.
77. Al Masʿūdī, op. cit., vol. III, p. 18.
78. Al ʿUmarī, Masālik al Abṣār fī Mamālik al Amṣār, pp. 6 and 25-26.
79. Brett, op. cit., p. 82.
404
or not their white jubbahs were protective hauberks, as they were in 14th century Mamlūk Egypt, or simply a style of tunic is unclear.80
As already mentioned, the Fāṭimid armies contained a high proportion of infantry to the end. These were recruited from a great many sources. In battle they were arrayed by tribe, unit or national origin, with armoured men in the front rank. If attacked, they would make a wall with the bases of their tall shields being thrust into the ground. Their spear-butts were similarly planted to act as pikes, while archers and javelin-throwers supported these pikemen. In attack such infantry would advance either en masse or by sending selected sections of the line to take advantage of any faltering by the foe. Cavalry would cover such attacks and, together with infantry archers, pursue a beaten foe.81
In other words, Fāṭimid tactics were essentially those of the later Umayyad era. Their weapons were similarly traditional, consisting of large leather lamṭ shields, javelins, bows and swords that perhaps included the newer and slightly curved Iranian qalāchūr. Many infantry were in fact armoured. The long pikes with iron butts that were used by men in the front ranks were possibly known as ṣabarbarah. They were five cubits long, of which up to three cubits could form a broad iron blade. As such they might better be described as glaives. Some pikemen also carried small javelins. Crossbows were known, but seem to have been a speciality of naval troops. Some younger soldiers also seem to have been armed with shorter furayjīyah spears which
---------------------
80. Al ʿUmarī, op. cit., p. 82.
81. Beshir, op. cit., pp. 76-79.
405
could also have been glaives or bills.82
This armoured infantry may have included that less than reliable jund miṣr which, living in Cairo, took part in various battles in the declining years of the Fāṭimid Caliphate.83 Certainly there is plenty of pictorial evidence of such infantry in the art of the Fāṭimid era from the 10th to 12th centuries. These show a great variety of weapons, though with little indication of the cultural origins of the troops involved (Figs. 148, 152, 153, 155, 161 and 163). (They are similarly apparent in Zirid Ifrīqiyah (Figs. 193 and 194).
The next wave of North African conquerors relied on infantry to a similar if not even greater extent. These were the Murābiṭin, but unfortunately they are not apparently to be seen in any surviving pictorial sources. The main difference between the Murābiṭin and the mid-11th century and other Berber warriors was their refusal to indulge in karr wa farr, attack and retreat, either by infantry or by cavalry. Rather they obliged the enemy to attack them and then refused either to retreat or to advance. Their religious commitment, of course, enabled them to accept unusually high casualties. Nor would they pursue a beaten enemy. Their name probably recalls such tactics, since they made a ribāt fortification or strongpoint, out of their own closely packed and immovable battle-array.84
Although the Murābiṭin had small cavalry and camel-mounted
---------------------
82. Ibid., pp. 67-70; Canards "La Procession du Nouvel An chez les Fatimides", pp. 374 and 392-393; M. Canard, "Le Cérémonial Fatimite et le Cérémonial Byzantin", Byzantion, XXI (1951), p. 397.
83. Usāmah ibn Munqidh, op. cit., pp. 8 and 32.
84. Moraes Farias, op. cit., pp. 815-816.
406
forces even in the mid-11th century, they never changed these tactics and such steadfast discipline certainly appalled their Spanish enemies at the battle of Zallaca in 1086 AD. In equipment, however, they differed little from other North African and Egyptian armies. Long spears or pikes were used by those in the front ranks, with javelin throwers, apparently carrying more of these weapons than was normal elsewhere, standing behind them.85 Their use of large leather lamṭ shields probably made armour largely unnecessary in such a static style of warfare.86
The subsequent Muwaḥḥidun of the 12th and 13th century Maghrib are only slightly better illustrated. They, however, did not rely to such a large extent upon infantry. Nevertheless, even they fought in a somewhat static fashion that gave primacy to the foot soldier. Their most original contribution to these early wars of attrition was a chained palisade to protect their chosen rallying point. This was clearly a defence against cavalry. They themselves, coming from the richer and largely settled regions of Morocco, could afford both cavalry and infantry recruited from the plains and the mountains. In addition to the Muwaḥḥidun chained palisade, both cavalry and infantry appeared in the final battle of Las Navas de Tolosa in Spain in 1212 AD. A little earlier, Muwaḥḥidun weapons seem to have been daraqah shield and rumḥ spear for cavalry, khanjar large dagger or small swords rumḥ spear and sikīnah dagger for infantry.87
---------------------
85. Al Bakrī, Description de l'Afrique Septentrionale, De Slane trans., (Paris 1913), p. 314.
86. H. T. Norris, "The Hauberk, the Kazāghand and the ʿAntar Romance", Journal of the Arms and Armour Society, IX (1978), pp. 98-99.
87. Abū Bakr ibn ʿAlī al Ṣanhājī al Baidhaq, "History of the Almohades", in Documents Inédits d'Histoire Almohades E. Lévi-Provençal, edit, and trans., (Paris 1928), pp. 124, 172, 177, 194 and 199.
407
There is little reason to suppose that such weapons were not interchangeable between foot and horse. As far as the daraqah shield was concerned, this had almost always been of leather, but by the 12th century a specific kidney-shaped version had appeared in the Maghrib. This would later be adopted by the Spaniards and other Europeans as the Adarga and it is perhaps first seen in the hands of a Muslim foot soldier from the late 12th century (Fig. 532).
A possibly earlier version of this small leather shield with its indented top may be seen in Norman Sicily, perhaps in the hands of those originally Arab and Berber troops who were the kingdom's most dependable warriors in the 12th century. Their ancestors, who conquered Sicily in the 9th century, clearly fought in traditional ʿAbbāsid style. During the invasion this largely involved infantry forces, drawn up beneath the banners of their leaders and covered by cavalry, provoking the enemy into making the first attack.88 In those early years most references to items of equipment deal with booty won from the Byzantine foe. It also seems likely that the Christian population of Sicily was not immediately demilitarized following the Muslim conquest. A large Christian contingent helped the Fāṭimids enter Egypt, and this probably came from Sicily or Sardinia which were then both under Muslim rule. Since a fleet from Amalfi perhaps also took part in this operation, such participation should not be too surprising.89 It has also been suggested that a number of Russian-Scandinavian Varangians, plus some Armenians, captured while serving under the
---------------------
88. Amari, Storia dei Musulmani di Sicilia, vol. I, p. 397.
89. C. Cahen, "Une texte peu connu relatif au commerce oriental d'Amalfi au Xe siècle", Archivo Storico per le Province Napoletane, XXXIV n.s. (1955), pp. 64-65.
408
Byzantine flag in Sicily, were later employed by the Fāṭimids during their conquest of Egypt.90 The majority of such troops would almost certainly have been heavy infantry, perhaps operating as mounted Infantry.91
The evidence of the last decades of Muslim rule in Sicily suggests that the warriors of this island were very mixed, both infantry and cavalry, Most were lightly armed but there was also a substantial minority of armoured troops while the heavy hauberk or cuirass was not unknown.92 Numerous carved ivory boxes and oliphants that might have come from these, or from early Norman, decades illustrate infantry wearing mail or scale hauberks of various shapes, but almost all paralleled in Muslim al Andalus (Figs. 597, 598, 599, 600, 601, 602 and 607). The Norman conquest of Sicily had at first little impact on its military organization and, where indigenous Muslim infantry were concerned, this seems to have remained true for a long time. Existing territorial junds provided the new rulers with infantry and cavalry, although in an emergency the entire adult male population, Muslim and Christian, appears to have had a military obligation. Nevertheless, the paid Muslims of the junds led by their own qāḍīs remained the real military backbone of Sicily under the Normans and their immediate successors. They provided half the available troops and fought as siege-engineers, archers and heavy infantry. Their importance also apparently increased as the Norman state grew more centralized,93
---------------------
90. Blondal and Beneditz, o2., cit., p. 39.
91. Ibid., p. 47.
92. Gabrieli, "Gli Arabi in Spagna e in Sicilia", p. 718.
93. C. Cahen, La Régime Féodal de l'Italie Normande. (Paris 1940), pp. 76 and 118; E. Curtis, Roger of Sicily, (London 1912), pp. 365 and 371-372; F. Chalandon, Histoire de la Domination Normande en Italie et en Sicile, (Paris 1907), vol. Il, p. 535.
409
so highly regarded were such troops that they formed a guard of infantry archers for the Royal Treasury.94
Records of their achievements in siege warfare are numerous, though only against Christian princes in Italy or the Balkans, not against their co-religionists in North Africa.95 Yet Muslim Sicilians also served in these war-zones as both heavy and light infantry as well as cavalry, while others may have been included among those volunteers who fought only for booty.96 Such varied troops certainly appear on the paintings and carvings of the Norman kingdom. Many may, however, wear costume and carry equipment that were identical to those of Christian infantry from these same areas. Thus they may be impossible to identify with certainty (Figs. 577, 580, 603, 604, 606 and 609).
Regular and auxiliary warriors were involved, not only in the collapse of Norman rule,97 but also in attempts to resist the subsequent Hohenstaufen destruction of Islam in western Sicily.98 Even after the Muslims were exiled to the Italian mainland they were still recruited by their new masters, the crossbowmen of Lucera being particularly highly regarded.99 This weapon had, of course, long been known in Sicily where it
---------------------
94. E. Jamison, Admiral Eugenius of Sicily, (London 1957), p. 39.
95. Ahmad, op. cit., pp. 66-67; F. Gabrieli. "Le Politique Arabe des Normans de Sicile", Studia Islamica, IX (1958), pp, 92-93.
96. Buckler, op. cit., p. 376; Arnold of Lūbeck, "Chronica Slavarum", in Amari, Storia dei Musulmani di Sicilia, vol, III/2, p. 547.
97. Jamison, op. cit., p. 114.
98. E. Lévi-Provençal, “Une héroine de la résistance Musulmane en Sicile au début du XIIIe siècle", Oriente Moderno, XXXIV (1954), p. 286.
99. Ahmad, op. cit., pp. 105-106.
410
(was used alongside a variety of bows and slings (Fig. 610).
While Christians might have played a small role in the armies of Muslim Sicily, they certainly played an important one in those of Egypt and the other central Muslim lands from the 10th to 13th centuries. Not all eastern Christian communities were, of course, involved, Many had by this time lost most of their military traditions. The Armenians, by contrast, had not. Although many accepted Islam, the majority of this nation remained faithful to Christianity even when serving in Muslim armies.
Outside their own region, which remained under the distant but effective control of the Caliphs until the 10th century, Armenian mercenaries were not very apparent until the later 9th century when they were employed in small numbers by the Ṭūlūnids. Most such troops would seem to have been Muslim, and at one time formed a militia in a quarter of the Egyptian capital known as al Ḥusaynīyah.100 In the 10th century, when the Armenian mountains were divided between various quasi-independent Christian and Muslim rulers, a few Armenian troops reappeared in Ḥamdānid, Mirdāsid and Fāṭimid forces, and among the Byzantines.101 Some early Armenian manuscripts of this period show armour to have been similar to that of the Byzantines, which was itself, of course, not so different from that of the neighbouring Muslims (Fig. 242).
A major change seems to have taken place with the Byzantine occupation of almost all Armenia in the early 11th century. Thereafter a great many Armenians, mostly perhaps soldiers,
---------------------
100. Beshir, op. cit., pp. 49-51; M. Canard, "Notes sur les Arméniens en Egypte à l'époque Fatimite", Annales de l'Institut d'Études Oriantales de la Faculté des Lettres d'Alger, XIII (1955), p. 144.
101. Beshir, op. cit., pp, 49-51.
411
migrated to Fāṭimid Egypt where they formed an important military contingent which was best known for its infantry archers. Whether the bulk of such troops were Muslim, as would seem likely, or Christian, their leadership certainly was Muslim.102 Unfortunately, the limited available pictorial material from Armenia itself does not show archers. Rather it illustrates infantry with large shields and perhaps pikes (Fig. 243). The turbulent politics of the later Fāṭimid Caliphate even enabled those Armenian troops to seize military dominance in Cairo in 1073/4 AD. Their leader, Badr al Jamālī, became vizier and thereupon encouraged the recruitment of even larger numbers of his countrymen.103 It in interesting to note that one Armenian manuscript of around this period show the same flat-bottomed, kite-shaped shield with pseudo-Kūfic decoration that also appears in various Coptic and other eastern sources (Fig. 246).
The Saljūq and Crusader conquests of eastern Anatolia and the Syrian coast caused Egyptian recruitment of soldiers from the Armenian heartland to be severely limited in the late 11th and 12th centuries.104 With the Crusader occupation of the Cilician plain in Lesser Armenia in 1109 AD, however, many Armenian warriors apparently offered their services to Fāṭimid Egypt. There they were to form the bulk of the juyūshi troops under the leadership of a Christian Armenian, Bahrām, who subsequently became the Caliph's vizier.105 In fact, most such Armenian infantry
---------------------
102. Ibid.; M. Canard, "Un Vizir Chrétien l'Époque Fāṭimide; l'Arménien Bahrām", Annales de l'Institut d'Études Orientales de la Faculté des Lettres d'Alger, XII (1954), p. 97.
103. Beshir, loc. cit.; Canard, "Notes sur les Arméniens en Egypte à l'Époque Fatimite", p. 145.
104. Beshir, loc. cit.
105. Canard, "Un Vizir Chrétien", pp. 94-97.
412
archers in Egypt were now apparently Christians.
These Armenians were clearly catholic in their tastes for employers, serving Nūr al Dīn of Syria in both siege and wider operations, the Munqidhites of Shayzar, the Crusader states where they often fought under their own lords, and the Saljūqs of Rūm.106 The influence of those varied employers may be reflected in Armenian manuscripts showing warriors. Muslim as well as persistent Byzantine styles mingle in various mid- or late 13th century sources from eastern Armenia (Figs. 248, 250 and 251), while Frankish influence is obvious in a manuscript dated 1318 AD, (Fig. 654). Few other Christian communities in the Middle East had retained such martial traditions, although the Maronites of Mount Lebanon remained an exception. They, like the Armenians, were effective infantry archers and although they are best remembered as occasional allies of the Crusaders,107 they probably served locally before the Franks arrived in Syria. The partially, or in some later cases wholly, Syrian turcopoles who played such an important role in Crusading warfare consisted of both infantry and light cavalry. On foot they chiefly seem to have been archers.108
At this same time there was a revival of infantry in those Muslim states opposing the Crusades. Yet this was not, to any great degree, a result of the Franks' own reliance on heavy infantry. Crusader emphasis on close cooperation between horse and foot
---------------------
106. Elisséeff, op. cit., p. 733; Usāmah ibn Munqidh, op. cit., p. 106; Smail, op. cit., p. 47; S. Vryonis Jr., "Byzantine and Turkish Societies and their sources of manpower", in War Technology and Society in the Middle East, V. J. Parry and M. E. Yapp edits., (London 1975) p. 141.
107. Smail, op. cit., pp. 52-53.
108. Usāmah ibn Munqidh, op. cit., p. 51; Smail, op. cit., p. 112; Prawer, op. cit., pp. 340-341.
413
in open battle was essentially no different to those tactics employed earlier by both Byzantines and Fāṭimids. The revival of infantry in those minor states that took over Syria and the Jazīra from the crumbling empire or the Great Saljūqs in the 12th century was more a revival of tradition. It occurred as those nomadic tribes of Turkish horse-archers, upon whom the Saljūqs had largely depended, were relegated to the frontiers.109
In addition to Armenians such as those mentioned above, the Zangids of the Jazīra employed large numbers of archers, crossbowman, siege-engineers, naffaṭīn naptha-throwers and heavy infantry with shield, spear and pike. Among those specializing in siege warfare, Khurāsānīs and men from Aleppo were particularly notable.110 When Nūr al Dīn chose to face his enemies in open, set-piece battle, such foot soldiers seem to have fought in a traditional manner, again cooperating with their cavalry as they had long done.111
Comparable infantry armed with sword, turs shield, quntārīyah or rumḥ, spears, qārūrah grenades of naptha, sikīnah dagger, zardīyah mail hauberk and khūdhah helmet, apparently with a mail avantail across the face, are all mentioned in the memoirs of Usāmah ibn Munqidh.112 In fact, they probably served in most armies of this area, large or small. Such infantry may also be reflected in art from the Crusader states, particularly where "enemy" warriors or figures symbolizing the Sins are concerned (Fig. 267). They are, however, almost certainly to be seen in
---------------------
109 Cahen, "Djaysh", loc. cit.
110. Elisséeff, op. cit., pp. 733-735.
111. Ibid., p. 742.
112. Usāmah ibn Munqidh, op. cit., pp. 74-75 and 124.
414
many pictorial sources from the Muslim side of the frontier, ranging from Azarbayjān and the Caucasus to the Jazīra itself. In most cases their equipment is traditional, though with curved swords now shown more frequently (Figs. 310, 372, 373, 376, 380, 386, 420, 422 and 423).
Infantry remained important under Ṣalāḥ al Dīn and the Ayyūbids. They may, indeed, have increased now that Islam was on the offensive against a string of Crusader states that relied, above all, on the defences of their massive castles. Ṣalāḥ al Dīn's armies varied in their constitution, but at different times included Arab infantry and cavalry of the large Kinānah confederation, plus ʿasāqilah and other troops inherited from the Fāṭimid Caliphate. Junior mamlūks were also trained to fight on foot, as well as mounted, while other tribal levies, jund local militias from various Syrian cities, the highly regarded siege troops of Aleppo and Mosul, plus some comparable specialist men from Khurāsān itself, are all recorded.113
Open battle with infantry facing infantry, cavalry facing cavalry, was an issue that both Franks and Ayyūbids tended to avoid. Yet, according to al Ṭarsūsī, Muslim foot soldiers were still trained to draw themselves up in ranks ahead of the cavalry, behind a wall of januwiyah and ṭāriqah tall shields. Thereafter, cooperation between horse and foot remained as it had for centuries, except that the infantry could now add the jarkh crossbow to their existing arsenal of sword, spear, javelin and bow.114 Such tactics were clearly more than merely theoretical and seem to have been
---------------------
113. Gibb, "The Armies of Saladin", loc. cit.; ʿImād al Dīn, op. cit., vol. I, pp. 22 and 144.
114. Al Ṭarsūsī, op. cit., pp. 125-126.
415
used by Ṣalāḥ al Dīn's garrison at ʿAkkah during one major sortie.115 Comparable troops may well be illustrated in a Coptic Gospel from early Ayyūbid Egypt (Fig. 166).
Those Muslim archers and javelin-throwers who opened the battle of Arsūf in 1191 AD may have included such trained, professional infantry.116 Generally, however, the role of Ayyūbid infantry was limited to siege-warfare. This, of course, could mean open battle during the siege or blockade of a fortified place. Arabic and Latin sources agree that the Muslim troops involved in such fighting varied greatly in their arms and armour, from lightly equipped jarīdah warriors to heavily protected thaqlāh infantry and dismounted, but still armoured, horsemen. Among the items of equipment mentioned are swords, spears, daggers, double-edged axes, winged or knobbed maces, light spears, jarkh crosbows, turs shields, qārūrah naptha grenades, zarrāq naptha "tubes" or flame-throwers, bows, yalab large leather shields, dhubalah spears, sābighah mail hauberks, qunṭārīah lances, dirʿ hauberks and majann leather shields.117 These various shields, plus other forms of mantlet, were in fact often used to build an effective shield-wall in what virtually become trench-warfare.118
One pictorial source from just after Ṣalāḥ al Dīn's death, and from the region where some of the best siege troops were recruited, shows an infantryman with a short spear (Fig. 284). Similarly equipped troops appear in both Christian and Muslim art
---------------------
115. Anon., Itinerarium Pereginorum, p. 26.
116. Runciman, op. cit., vol. III, p. 56.
117. Bahā al Dīn, op. cit., p. 77; Anon., Itinerarium Pereginorum pp. 64 and 78; ʿImad al Dīn, op. cit., pp. 39, 75, 133, 150,164 and 191.
118. ʿImād al Dīn, op. cit., pp. 135 and 191.
416
of this area in the following century and probably indicate that these men and their reputations lasted right up until the Mongol invasions (Figs. 288, 289, 292, 294, 296, 297, 298, 299, 302, 305 and 306).
Many of the same troops, including the ex-Fāṭimid Kinānah, continued to serve Ṣalāḥ al Dīn's Ayyūbid successors,119 and to appear in Ayyūbid art (Figs. 170 and 172). Their equipment also seems to have remained the same, consisting of swords, spears, kabūra cuirasses, mail hauberks, turs, januwīyah, ṭāriqah and other shields, bows, crossbows and a great variety of siege equipment. Their tactics and training similarly did not change.120 While later Ayyūbid art shows many warriors on foot, most seem by their equipment and foot-wear to be dismounted horsemen. Only rarely are they certainly infantry (Fig. 177) as indicated by bare feet and a sort of puttee.
The armies of Islam also attracted troops from Christendom, both Orthodox and Catholic. Much the larger proportion would seem to have been infantry which was, of course, the area in which specialized professional mercenary troops first emerged, even in Europe. Most would probably have been heavy infantry or siege engineers, crossbowmen, archers and the like.
Byzantium was relying on its infantry forces to a great extent throughout this period. They were early divided into light skirmishers, javelin throwers, slingers or archers (Fig. 206), and heavily armoured men trained to fight in ranks (Fig. 199).
---------------------
119. Jamāl al Dīn ibn Wasīl, "Muf arrij al Kurūb fī akhbār Banī Ayyūb", in Arab Historians of the Crusades, Gabrieli edit., p. 285.
120. Al Harawī, "Les Conseils du Šayḫal Harawī. à un Prince Ayyūbide", J. Sourdal-Thomino trans., Bulletin d'Etudes Orientales, XVII (1961-1962), pp. 232-239.
417
Even during Byzantium's offensive phase in the 10th century, heavy and light infantry continued to play a vital role. The former provided a secure square from which the cavalry could deliver its charges, while light infantry continued to fight as skirmishers and play an offensive role in broken country.121 Once again they constantly appear in Byzantine art, light (Figs. 211, 217 and 220B) and heavy (Figs. 212 and 220A). Such troops were almost certainly included among those Byzantine prisoners and mercenaries recruited by the Ṭūlūnids of Egypt during the 9th century.122
While the decline of Byzantine military effectiveness in Anatolia chiefly affected local aristocratic cavalry and to a lesser extent also infantry levies from peaceful regions far from the frontiers, it does not seem to have applied to the troubled border provinces. Here a new system of limites was, in effect, set up. Such frontier troops were backed up by powerful central forces that also included large numbers of heavy infantry among whom the battle-axe became an increasingly popular weapon. In combat the same tactical concepts persisted.123 Light infantry of those provincial limites are probably to be seen in the art of such frontier regions as Cappadocia (Figs. 207, 214, 215 and 216).
Following the disasters of the Saljūq invasions during the later 11th century, Byzantium sought to revive its infantry forces.124
---------------------
121. Howard-Johnson, op. cit., pp. 282-286.
122. Bosworth, "Armies of the Prophet", p. 203; Hassan, op. cit. p. 168.
123. Howard-Johnson, op. cit., pp. 142 and 144; Psellus, op. cit. pp. 30, 47 and 141; G. Schlumberger, "Deux Chefs Normands des Armées Byzantines au XIe Siècles", Revue Historiques, XVI (1881), p. 299.
124. Chalandon, Les Comnènes - Etudes sur l'Empire Byzantin, vol. I, pp, 277-279 and vol. II, p. 21.
418
Again light and heavy corps were organized (Figs. 222 and 225), though circumstances seem to have forced a concentration on the former. These troops, the light armed with bow and small shield, the heavy with axe, sword and either large or small shield, clearly enjoyed a good measure of success in the 12th century. Paradoxically, however, Byzantium now found itself less skilled in siege warfare than many of its foes.125 This would seem to have remained the military situation in Byzantium, though with additional domestically recruited and foreign mercenary cavalry, throughout its long decline during the 12th to 15th centuries. Such troops continue to appear in the art of this twilight age (Figs. 228, 230, 231, 233, 237, 238, 655 and 656). No doubt it was infantry, skilled in the defensive, almost guerrilla, warfare of Bithynia and the other mountainous provinces along Byzantium's new frontier against the Turks, who were now so eagerly recruited by those same Turks. Most were captives who chose to serve the Saljūq sultans of Rūm as garrison troops and siege engineers or as jāndār guards of the ruler himself. Some, however, seem to have been mercenaries126 (Figs. 257, 258 and 262).
The Saljūq rulers of Rūm also employed Georgian troops,127 as had earlier been recommended to the Great Saljūq rulers of Iran. Some Georgians apparently fought as infantry using javelins heavy enough to dismount a horseman. Others fought with bow and lassoo in Turkish style. In general, however, these Georgians seem to have been armoured troops, mostly mounted, but well able
---------------------
125. Ibid., vol.II, pp. 618-619 and 620-622.
126. C. Cahen, Pre-Ottoman Turkey, (London 1968), pp. 230-234.
127. Vryonis, loc. cit.
419
to fight on foot when the need arose.128 As such they seem to have had more in common with their Arab, Persian and Kurdish neighbours than with the more specialized Turks or Byzantines (Fig. 421).
Other foreign troops serving in Islam included the saqālibah of supposed Slav but probably more varied origin. They were apparently acquired as slaves via Spain, Venice and elsewhere, and were trained from youth in the military styles of their new homes. Such saqālibah served in the armies of both Ṭūlūnid and Fāṭimid Egypt, Aghlabid Ifrīqiyah and ʿAbbāsid Iraq.129
Finally there were the Franks. European mercenaries, many perhaps from in or around those Italian maritime republics that were already in close commercial contact with the Middle East, are known in Fāṭimid Egypt at least as early as the late 11th century.130 They may, in fact, have been numbered among rūm al murtaziqah mercenaries recorded in Cairo in 1005 AD. In this case, comparable troops may also have served in Ḥamdānid and earlier Ṭūlūnid force.131 The infantry of Italy was, at this time, not dissimilar to that of the Muslim side of the Mediterranean (Figs. 560, 561, 562, 565, 568, 569, 570, 571 and 574).
---------------------
128. Psollus, op. cit., p. 36; M. Canard, "Les Reines de Gorgie dans l'histoire et Ia lgende musulmanes", in Revue des ?tudes Islamiques, XXXVII 1969, p. 17; Rustʾhaveli, op. cit., passim.
129. Bosworth, "Recruitment, Muster and Review in Medieval Islamic Armies", pp. 66-67; Beshir, op. cit., 34-38 end 63-67; Imamuddin, "Commercial Relations of Spain with Iraq, Persia, Khurasan, China and India in the Tenth Century AC", p. 179.
130. Canard, "La Procession du Nouvel An chez les Fāṭimides", pp. 332-393.
131. Beshir, op. cit., pp. 51-52.
420
After the confusion of the First Crusade and the establishment of relatively fixed frontiers between the Sajūqs of Rūm and their Christian neighbours, west European, Frankish, warriors soon appeared in Saljūq service.132 Whether those of the 12th century were cavalry or infantry is unknown, but they probably included both. Others may also have served in Ayyūbid Syria, though these would have been regarded by their co-religionists as renegades which those in Saljūq Rūm, for some reason, were not. Some of these latter were, in the late 12th and early 13th centuries, recruited from released Crusader prisoners while others, particularly sailors and crossbowmen, were apparently Italian mercenaries. Such adventurers were already serving far from their homes in many European armies (Figs. 579, 582 and 587).133 Their numbers were quite high in the Saljūq sultanate before the Mongol invasions, while after this date others also fought for the Ilkhāns of Iran134 and perhaps even for the Sultans of Delhi.135
The infantry of the Mamlūk state in Egypt and Syria seems to have been either locally recruited or drawn from the ranks of junior mamlūks. Unlike their aristocratic Frankish foes, even the most senior mamlūks were quite prepared, and indeed trained, to dismount and fight on foot. Most of the earliest available information comes from the first half of the 14th century, but there is little reason to suppose that there were any major changes
---------------------
132. J. Richard, "An Account of the Battle of Hattin referring to Frankish Mercenaries in Oriental Moslem States", Speculum, XXVII (1952), pp. 169 and 171; Cahen, Pre-Ottoman Turkey, loc. cit.
133. Richard, op. cit., pp. 172-174,
134. Ibid., p. 174.
135. Yār Muḥammad Khān, op. cit., p. 22.
421
during the first war-torn century of the Baḥrī Mamlūk state. Some of the first furusīyya manuals still declare infantry to be superior to cavalry in "cautious" or defensive warfare,136 and that they were vital to defend an army's encampment while in hostile territory.137 A great deal of attention is given in these works to combat between infantry and cavalry, and how the former can best defend themselves against the latter. Nor were infantry always considered to be on the defensive in such encounters, for they were expected to check, scatter and pursue cavalry.138 Similarly, they still apparently acted in cooperation with horsemen, protecting their own cavalry from hostile infantry in both advance and defence. It was, however, recognized in these sources that infantry were no longer quite as effective as they had been in earlier centuries.139
These various sources list a great variety of arms and armour to be used by infantry, many of whom seem to have been archers drawn from the settled communities of Syria, Palestine and Lebanon.140 For defence there were small daraqah and larger ṭāriqah shields, though the armour to be worn is rarely described as specifically for infantry or for cavalry. In attack there are sword, tabar, axes, dabbūs al muḥarrafah winged maces, bows,
---------------------
136. Al Aqṣarāʾī, op. cit., pp, 325 and 336-337.
137. ʿUmar ibn Ibrāhīm al Awsī al Anṣarī, Tafrīj al Kurūb fī Tadbīr Al Ḥurūb, A Muslim Manual of War, G. T. Scanlon trans., (Cairo 1961), p 87.
138. Al Anṣarī, op. cit., p. 72; al Agṣarāʾī, loc. cit.
139. Al Anṣarī, op. cit., pp. 104-105 and 107; al Agṣarāʾī, op. cit., pp. 325 and 337-338.
140. A. N, Poliak, Feudalism in Egypt, Syria, Palestine and the Lebanon, 1250-1900, (London 1939), pp. 11-14.
422
rumḥ spears, zāriq javelins, khanjar short swords and ʿamūd maces.141 Similarly, the art of the area, including that from the Crusader states, shows late 13th and early 14th century Muslim infantry to be as varied in their equipment, offensive or defensive, as were Muslim cavalry of the period (Figs. 273, 274, 301, 311, 312, 649 and 650A).
Whether they participated primarily as siege troops or as mobile infantry forces in open battle, infantry clearly had a major role to play in Muslim warfare in the late 13th century, even in the Mamlūk state of Syria and Egypt.
---------------------
141. Al Anṣarī, op. cit., pp. 107-108; al Aqṣarāʾī, loc. cit.