11
Of all the dynasties which succeeded the ‘Abbasid caliphate, the Hamdanids have perhaps the highest reputation with posterity. This is partly due to the personality of the amīr, Sayf al-Dawla, who is seen as a representative of all the Arab virtues of generosity and courage, fighting against overwhelming odds on the Byzantine frontier. This romantic image has been much enhanced by Hamdanids’ interest in poetry, notably Sayf al-Dawla’s patronage of al-Mutanabbī, arguably the greatest of the classical Arabic poets, and of Sayf al-Dawla’s cousin, Abū’l-Firās. Contemporaries were not so favourably impressed, however; the geographer Ibn Ḥawqal, who travelled through the Hamdanid domains, was deeply critical of their oppressive policies, painting a grim picture of overtaxation and exploitation. Even the capital of Aleppo seems to have been more prosperous under the following Mirdasid dynasty than under the Hamdanids. After the first generation of Nāṣir al-Dawla in Mosul (d. 358/969) and Sayf al-Dawla in Aleppo (d. 356/967), they produced few leaders of great ability, and while Mosul soon fell to the Buyids, Hamdanid Aleppo was ruled more by their ghilmān than by members of the dynasty. Nor should we imagine that the Hamdanids were bedouin chiefs of the traditional sort. They rose to power originally because of the support they could command among their Taghlibī fellow-tribesmen, but as soon as they established a territorial base, they dispensed with this support, preferring to base their power on taxation and an army of salaried ghilmān. In this way, the Hamdanid state was more similar to Buyid Iraq or Ikhshidid Egypt than to bedouin states like the Mirdasids and ‘Uqaylids which succeeded it.
The Hamdanids of Mosul
The Hamdanids were drawn from the Banū Taghlib, a tribe which had grazed the Jazīra area since pre-Islamic times. During the anarchy in Sāmarrā, if not before, they had come to dominate the city of Mosul, and various Taghlibī chiefs succeeded each other as governors, defending the city against the Khārijī brigands of the steppe lands who had threatened the area since later Umayyad times. Around 266/879, the Taghlibī chiefs were replaced by a Turkish soldier, ‘Īsā b. Kundājīq, in an effort to restore the power of the Baghdad government in the area, with the natural result that the Taghlibī chiefs, among them Ḥamdān b. Ḥamdūn, went over to the Khārijīs in an effort to recover their position, and disorder spread throughout the Jazīra. In 279/892, the new Caliph al-Mu‘taḍid determined to reestablish order in the area, and in a series of campaigns he reduced most of the dissidents to obedience. Ḥamdān was one of the most stubborn in resistance and he was strengthened by his hold over various strong points in the mountains to the north of the Jaziran plain. In the period before al-Mu‘taḍid’s advance, he had acquired a series of strong points, notably Mārdīn to the west and Ardamusht near Jazīrat Ibn ‘Umar (now Cizre) in the east, and he seems to have reached a working alliance with the Kurdish tribes of the foothills. This meant that Ḥamdān was more than simply a bedouin chief. The refuge offered by the mountain fortresses and the relationship with the Kurds were important factors in the survival of the Hamdanids in the years to come.
Ḥamdān’s career ended in failure when, in 282/895, after an epic chase, he was forced to give himself up to al-Mu‘taḍid in Mosul. This surrender might have seemed the end of the family fortunes, as it was for other local leaders in the area, had not Ḥamdān’s son, al-Ḥusayn, made himself useful to the government. In 283/896, al-Ḥusayn was released from prison and allowed to recruit a small army of his followers to fight against the remaining Khārijīs in the Jazīra. The campaign was a success and in return for his services, al-Ḥusayn was able to request, not only that his father be released, but that the taxes on the Banū Taghlib be remitted and that he himself be allowed to recruit a body of 500 Taghlibīs to form a group in the armies of the caliphate and be paid by the government. This secured his position, not only with the caliph but also among the Taghlib. To the caliph, he offered a group of experienced warriors under his own skilled and loyal leadership; to the Taghlib, and other people in the Jazīra, he offered the prospect of salaries and booty; and to his own family, military command and the opportunity of acquiring wealth in government service. It was in fact not as an independent tribal leader, but rather as an intermediary between government and the Arabs and Kurds of the Jazīra that al-Ḥusayn made the family fortune. His troops were conspicuous in ‘Abbasid campaigns, especially against the Qarāmiṭa, the bedouin heretics who were such an intractable problem for the ‘Abbasid government of the time, where their experience made them particularly valuable, but also in campaigns in the Jibāl and during the ‘Abbasid reconquest of Egypt in 292/904–905.
By the end of al-Muktafī’s reign in 295/908, al-Ḥusayn had established his position as one of the leading generals in the armies of the caliphs, while other members of his family were rewarded with office elsewhere, the most important of these being in the old Taghlibī preserve of Mosul, where al-Ḥusayn’s brother, Abū’l-Hayjā’, was appointed the governor in 293/905. These two centres of power, in Mosul and in the army of the caliph, enabled the family to survive storms which destroyed less well-established families. Al-Ḥusayn was naturally drawn into Baghdad politics; in 296/908, he was one of the leaders of the abortive coup against al-Muqtadir launched in the name of Ibn al-Mu‘tazz, but he seems to have lost his nerve and fled to Mosul when he saw that the putsch was encountering resistance. Other leading participants were executed, but Abū’l-Hayjā’ negotiated a reprieve for his brother, who saw government service again in the Jibāl and finally as the governor of the area of Diyār Rabī‘a to the west of Mosul. In 302–303/914–915, he became involved in another rebellion. This time it seems to have been in protest against the attempts of ‘Alī b. ‘Īsā to reduce government spending and his own exclusion from power in Baghdad. It is a measure of his continuing influence in the area that he was able to raise a force of 30,000 Arabs and Kurds, but the rebellion failed when he was captured by the forces of the leading ‘Abbasid general Mu’nis and taken to Baghdad, where, in 306/918, he was executed.
Once again, the influence of the family appeared to have been destroyed, but as before, the Hamdanids proved too useful and well entrenched to be excluded from power for long, and Abū’l-Hayjā’ was soon back in government service. Like his brother, he distinguished himself against the Qarāmiṭa. In 312/924, he had been captured by the rebels while escorting the Ḥajj but was able to negotiate not only his own release but that of the other captives as well. This apparently close relationship with the Qarāmiṭa, who usually put their prisoners to death, has been seen by Canard, the great historian of the dynasty, as a sign of sympathy between the Ismā‘īlī Qarāmiṭa and the Shī‘ī Hamdanids. Despite the early involvement of the family with the Khārijīs, the Hamdanids are usually said to have been a Shī‘ī dynasty. In practice, this seems to have affected their policies very little; they were, in fact, prepared to acknowledge the sovereignty of ‘Abbasids, Buyids or Fatimids, depending on political circumstances. The Hamdanids flourished before the division between Sunnīs and Shī‘īs had become as clear-cut as it was to be later. Soon after this, we find the Qarāmiṭa actually being recruited into the ‘Abbasid armies, and the occasional signs we find of cordiality between them and the Hamdanids are probably the result of common political interests of the moment rather than of lasting religious sympathy. During the crisis of 315/927, Abū’l-Hayjā’ played a key role in defending Iraq from Qarmaṭī attack.
Abu’l-Hayjā’ left the affairs of Mosul in the hands of his son al-Ḥasan, while he himself became increasingly entangled in the complexities of Baghdad politics. Unlike his dead brother al-Ḥusayn, however, he was on consistently good terms with Mu’nis, with whom he worked on numerous occasions, and it was his alliance with Mu’nis and the policies of ‘Alī b. ‘Īsā which was to cause his downfall; in 317/929, he was a leading supporter of al-Qāhir in his unsuccessful coup and was killed fighting bravely to defend his candidate for the throne.
The ups and downs of Baghdad politics did not destroy the position of the family in the Jazīra and from this time onwards, Mosul and its surroundings are the centre of family activities. After the death of Abū’l-Hayjā’, leadership in the family was divided among his surviving brothers and his son, al-Ḥasan, who gradually established his supremacy. The Hamdanids refused to be drawn into further rebellions against al-Muqtadir and supported him in his final struggle against Mu’nis, who even occupied Mosul for a time. By 324/935, the authorities in Baghdad had been obliged to accept al-Ḥasan as the ruler of Mosul and the Jazīra as far as the Byzantine frontier in exchange for a payment of 70,000 dīnārs a year and supplies of flour for Baghdad and Sāmarrā. From this time onwards, the power of the family was established. It had also changed its nature. Ḥusayn b. Ḥamdān had come to power by raising Taghlibī troops for the caliph, but relations between the Ḥamdānīs and their fellow-tribesmen gradually became more distant and we no longer hear of Taghlibī troops under their command. A key stage in this process seems to have occurred in 323/935 when a rival Taghlibī chief, al-‘Alā’ b. al-Mu‘ammar from the Ḥābib clan in the area of Niṣībīn, took 10,000 of his followers and crossed the frontier to Byzantium to escape Hamdanid control and, above all, Hamdanid taxation. There were several factors involved in this unprecedented move. The Taghlib had been a Christian tribe for many years after the Muslim conquest, and it is possible that many of its members still held to their old faith and this would have made Byzantine authority much more acceptable than it would have been to a Muslim group. It also seems likely that the Taghlib were coming under increasing pressure in their native grazing areas from aggressive tribes like the ‘Uqayl and Numayr, moving into the area from the south, and that the removal to Byzantine territory was forced on them. The movement of such a large number across the frontier is said by contemporaries to have been an important factor in Byzantine successes in the area, and it is certainly ironic that in the very year that al-Ḥasan’s authority was recognized over the whole area, the Taghlib tribe should have disappeared as a political force. The Hamdanids did not, then, base their power on their tribal following. Certainly, Arabs were mentioned in their forces but they came from tribes like the ‘Uqayl and the Numayr, not the Taghlib, and usually figure as irregulars and temporary allies rather than as regular supporters.
In the main, the Hamdanids, like most of their contemporaries, seem to have relied on ghilmān of Turkish extraction for their armed forces, often recruiting those who were dissatisfied with the government in Baghdad. The Kurdish element was probably important as well. The early Hamdanids seem to have established close relations with the Kurds who wintered on the plains of the Jazīra, notably the Hadhbānīs around the Zāb and the Jalālī Kurds farther south around Shahrazūr. The mountain fortresses especially, which played so important a part in Hamdanid survival, were often in Kurdish territory. Members of the dynasty intermarried with Kurds as well; both Abū’l-Hayjā’ and his grandson Abū Taghlib had Kurdish mothers, and these connections may well have brought military support. Certainly when the dynasty’s last stronghold in the Mosul area, Ardamusht, was surrendered to ‘Aḍud al-Dawla, the Hamdanid commanders in the fortress were a Kurd and a Turkish ghulām.
The Hamdanids then, in both Syria and Mosul, asserted their authority by means of a professional army recruited from different sources. Like all their rivals, they were faced with the problem of paying them, and both al-Ḥasan and his son and successor Abū Taghlib were faced by mutinies of their Turkish soldiers. The revenues to pay this force were drawn from the towns and countryside of the Jazīra, and the geographer Ibn Ḥawqal gives a breakdown of the sources of revenue from the Niṣībīn area. The main source of revenue, as everywhere in the Muslim world, was the land tax collected in kind: 10,000 kurr (a dry measure of approximately 2,500 kg) of grain valued at 5 million dirhams (about 300,000 dīnārs). The jizya from the Christians, by contrast, raised only 5,000 dīnārs, a tax on wine brought in 5,000 dīnārs and on food markets the same amount. Urban rents, including inns (khāns), baths, mills, shops and private houses brought in 16,000 dīnārs. This would seem to suggest that the land tax brought in about ten times as much as urban taxation, but this may be misleading. The land tax was collected in kind and valued, whereas urban taxes seem to have been collected in coin. Whether the theoretical value of the grain was ever realized in fact is not clear, and it is probable that urban dues paid in dīnārs were essential for paying troops and other government expenses which had to be settled in cash. Ibn Ḥawqal describes how he was at Mosul in 358/969 when the accounts were presented to Abū Taghlib and criticizes what he considered to be overtaxation. All the Hamdanids had secretaries, but it is not clear that there was any elaborate administration at Mosul, and the ease with which the court could take refuge in the mountains if necessary suggests that it may have been fairly basic.
Among the outgoings of the Hamdanids of Mosul, the payment of tribute to the Buyids was a major factor. In the later years of al-Ḥasan Nāṣir al-Dawla, this tribute seems to have been assessed at between 2 and 4 million dirhams per year, when it was paid at all. Abū Taghlib, in a much weaker position, was obliged to find 6 million dirhams. It is probable that payments of that size made it very difficult for him to recruit and pay enough ghilmān to maintain his position. As always, political and financial weakness went together; rulers were anxious to be independent because it was almost impossible to pay a large tribute and maintain a regular ghulām army.
The establishment of al-Ḥasan as the ruler of the Jazīra by the agreement of 324/935 did not mark the end of the ambitions of the family in Baghdad. The intrigues which surrounded the office of amīr al-umarā’ tempted al-Ḥasan to play a part, and when in 330/942, the caliph, al-Muttaqī, and the leading Turkish officer, Tūzūn, both urged him to try to drive the Barīdīs from Baghdad, he allowed himself to be persuaded. He marched south and was installed as amīr al-umarā’ in the capital with the title of Nāṣir al-Dawla, but despite his good intentions, the attempt was not a success. The problems were, as always, financial, especially as the Barīdīs were still in control of the rich lands of southern Iraq. Nāṣir al-Dawla sent his brother, ‘Alī, with an army to conduct the war in the south, as a reward for which he was given his famous honorific Sayf al-Dawla (Sword of the State). Sayf al-Dawla’s army was composed of Turks, Daylamites and Qarmaṭīs with irregular support from Kurds and Numayrī, Kilābī and Asadī Arab tribesmen. The campaign had only been under way a short time when Tūzūn led a mutiny in the army. Coupled with disturbances in Baghdad, this forced the Ḥamdānīs to withdraw. Nāṣir al-Dawla, however, never forgot that he had once ruled in Baghdad and it made him very unwilling to accept the suzerainty of the Buyids who came to occupy the office.
Nāṣir al-Dawla retired to Mosul where he reached an agreement with the new amīr al-umarā’, Tūzūn, allowing him the Jazīra and northern Syria (which he did not in fact possess) in exchange for an enlarged tribute of 3.6 million dirhams. From this time onwards, Nāṣir al-Dawla had a tributary relationship with the rulers of Baghdad. It was, however, a constantly changing relationship depending on the varying fortunes of each side. Nāṣir al-Dawla paid the tribute when he was forced to but took every opportunity to avoid it. On several occasions, in 334/946, for example, and at the time of Rūzbahān’s rebellion against the Buyids in 345/956–957 (see above, p. 221), the Hamdanids attempted to retake Baghdad, capitalizing on rivalries among the Daylamites and between Daylamites and Turks, but they were never able to maintain themselves in the city. Mu‘izz al-Dawla, in retaliation, tried to take over the Hamdanid lands and on several occasions actually captured Mosul itself. Hamdanid tactics on these occasions were simple but effective; unable to withstand Buyid armies in the open field, they abandoned Mosul and retired to the foothills of the mountains and the fortresses like Mārdīn and Ardamusht, where they kept their treasure, being careful to take with them all records of government properties and taxation. The invading forces found themselves unable to find supplies and harassed by the guerrilla tactics of the Hamdanid supporters and the hostility of the people of the cities to the Daylamite forces. After a short occupation, the Buyids would be forced to make terms again and the Hamdanids could return to Mosul. The most serious of these attacks was in 347–348/958–959, when Nāṣir al-Dawla was forced to flee to his brother Sayf al-Dawla in Aleppo, who received him with honour. It was Sayf al-Dawla who negotiated a new agreement with Mu‘izz al-Dawla and from this time onwards, the Aleppo branch of the family seems to have been more powerful than the Mosul branch, extending its influence over Diyār Muḍar on the east bank of the Euphrates.
In the end, family conflicts damaged the Hamdanid cause more than Buyid attacks. In 353/964, Mu‘izz al-Dawla launched another attack; once more, Nāṣir al-Dawla was driven out of Mosul and fled to the hills, but this time the counterattack was led by his son, Abū Taghlib, who eventually came to a new arrangement, a sign that the old man had ceased to be an effective leader. In 356/967, his brother, Sayf al-Dawla, with whom he had maintained good relations, died. Not only was this a personal loss for Nāṣir al-Dawla, but it also meant that the Hamdanids of Aleppo were unable to offer the Mosul branch any further support against Buyid aggression. Nāṣir al-Dawla’s increasing feebleness was shown by his failure to take advantage of the death of his old rival Mu‘izz al-Dawla in the same year, and he was confined by his sons to the fortress at Ardamusht, where he died two years later.
Nāṣir al-Dawla was succeeded as the head of the family by his son Abū Taghlib, whose leadership was accepted by most of his brothers. Family unity had been a major source of strength for the Hamdanids in earlier generations but this advantage was now dissipated. It seems that Nāṣir al-Dawla had considered passing the succession not to Abū Taghlib, but to another son by a different mother, Ḥamdān, and in his last years he had established him in an apanage around Niṣībīn, Mārdīn and Raḥba. Ḥamdān now refused to accept his brother’s authority, and open warfare developed between the two. This rivalry was complicated by relations with the Buyids. While Nāṣir al-Dawla had been a deadly rival of Mu‘izz al-Dawla, Abū Taghlib generally remained on good terms with his son and successor, Bakhtiyār. Unlike his father, Abū Taghlib had no claims to authority in Baghdad, while Bakhtiyār had too many problems maintaining himself in Iraq to spare much time for aggressive moves against Mosul. These good relations did not, however, prevent Bakhtiyār from giving Ḥamdān and other dissident members of the family a place of refuge when necessary, and Ḥamdān was able to maintain himself as a constant threat to his brother’s authority. In 362/973, Bakhtiyār actually took Mosul, and one of the conditions of the subsequent peace was that Ḥamdān be reestablished in all his old properties. In 367/977, it was Bakhtiyār’s turn to seek help against his aggressive cousin ‘Aḍud al-Dawla, who had driven him out of Baghdad, and Abū Taghlib promised to support him in exchange for Ḥamdān, who was handed over and executed. But the family unity which resulted had come too late. ‘Aḍud al-Dawla did not overlook the support Abū Taghlib had given his defeated rival and he soon began to advance on Mosul. This time the Buyids were much more organized, ‘Aḍud al-Dawla brought with him men who were familiar with the administration of Mosul, and the old Ḥamdānī tactics failed. Mayyāfāriqīn, Āmid and even the fortress at Ardamusht were taken by siege. Abū Taghlib was forced eventually to flee across the Syrian Desert to Palestine, where he tried to persuade the Fatimid authorities to grant him the governorate of Damascus. In doing so, he made powerful enemies, notably the Ṭayy chief Daghfal b. Mufarrij, and was finally killed at Ramla (369/979).
Abū Taghlib’s surviving brothers, al-Ḥusayn and Ibrāhīm, took service with ‘Aḍud al-Dawla, who was thus assured of peaceful possession of the Jazīra. Hamdanid rule in the area had effectively come to an end, but there was to be a postscript. Under ‘Aḍud al-Dawla’s weaker successors, Buyid control in the area was challenged by the Kurds under Bādh, the effective founder of the Marwanid dynasty, and the then Buyid leader, Bahā’ al-Dawla, hoped to use al-Ḥusayn and Ibrāhīm to form a coalition against the invaders. In 379/989, the brothers reestablished Hamdanid rule in Mosul, basing their power largely on the ‘Uqaylī tribesmen, who were now influential in the area. But while the ‘Uqaylis succeeded in restricting the Kurds to the mountainous areas to the north, they were not prepared to accept Hamdanid leadership for old time’s sake. The ‘Uqaylī leaders established themselves in Mosul and killed al-Ḥusayn, while Ibrāhīm was released by the Kurds and took service with the Fatimids in Egypt and Syria. Though there seems to have been some pro-Hamdanid feeling in cities like Mosul and Mayyāfāriqīn, based on hostility to the Daylamites, the last Hamdanids of the Mosul branch lacked the support of either a powerful tribal backing or a strong central government, and their attempt was doomed to failure.
The Hamdanids of Aleppo
The establishment of the Hamdanids in Aleppo and northern Syria came later than at Mosul. While the Banū Taghlib had long been influential in the Jazīra, they had no traditional interest in the lands to the west of the Euphrates, and the Hamdanids had not shown any concern with the area until after Nāṣir al-Dawla’s failure to maintain himself in Iraq as amīr al-umarā’ in 330/942. Nāṣir al-Dawla himself retired to Mosul, but his ambitious younger brother, ‘Alī Sayf al-Dawla, who had been his chief military commander in the unsuccessful Iraqi campaign, sought a new area in which to establish himself. Aleppo was at this time more or less under the authority of the ikhshīd, the ruler of Egypt who governed mostly through Turkish officers who had left Baghdad to seek their fortunes elsewhere. As far as the ikhshīd was concerned, Aleppo itself does not seem to have been of paramount concern; he was able to accept a stable buffer state against both Byzantines to the north and Iraqi generals from the east, but he was determined to keep Damascus and Palestine under his control. In 333/944, Sayf al-Dawla was able to take over Aleppo with the support of the leading bedouin tribe of the area, the Banū Kilāb who inhabited the steppe lands to the east and southeast of the city.
For the first decade of his reign, Sayf al-Dawla was preoccupied with reaching agreements to establish his position with the ikhshīd and his successors and with the Arab tribes. After his success in Aleppo, Sayf al-Dawla made an attempt to take Damascus as well, but the ikhshīd himself came to put a stop to this and a peace was arranged the next year (334/945) which allowed Sayf al-Dawla control over Aleppo, Antioch and Homs but not over Tripoli and Damascus. Despite a renewed attempt on Damascus by Sayf al-Dawla after the ikhshīd’s death, these areas were to be the continuing centres of the Hamdanid state in Syria. The political division between Aleppo and the frontier areas on the one hand and Damascus and Palestine on the other was to persist into the fifth/eleventh century, when the Mirdasids to the north and the Fatimids in the south divided the area between them.
Sayf al-Dawla’s other major problem in the early years of his reign was his relations with the bedouin tribes. The key to his policy in this area was his alliance with the Kilāb, who had played an important part in establishing him in Aleppo in the first place. This was not always an easy alliance, but the strategic position of the tribal lands near Aleppo and the ability to cut communications with the east made it essential to handle their leaders with care. Farther south, the Kalb of the Homs region, who had played so important a part in Umayyad politics, were more hostile; many of them had been connected with the Qarmaṭī revolts in the area and probably resented the Hamdanid alliance with their Kilābī rivals. In 344/955, Sayf al-Dawla was faced with a major rebellion of all the tribal groups, including the Kilāb in the north, and the Kalb, ‘Uqayl and Numayr in the Salamiyya region to the east of Homs. He crushed it by dividing his opponents. The Kilāb were allowed to make peace on easy terms; the others were ruthlessly chastised, many being driven out to the desert to die of thirst. This campaign altered the nomad geography of the Syrian steppe considerably. The Numayrīs were encouraged to move across the Euphrates to the Jazīra, while the Kalb were forced to move away from their traditional centres around Homs to the Jawlān (Golan) further in the south. The Kilāb, however, were allowed to expand south into the old Kalb lands, and it seems that this increased territory led other members of the tribe to migrate to the area from Iraq, thus confirming their preponderance in northern Syria.
The crushing of the bedouin revolt probably marks the high point of Sayf al-Dawla’s success and power. His state was based on the area of central Syria from Homs to Aleppo and on the eastern part of the Jazīra where Mayyāfāriqīn became a sort of second capital, increasingly important in later years when Byzantine pressure from the west increased. The coastal areas of Syria as far south as Ṭarṭūs but excluding Tripoli were also in Hamdanid hands. It would also seem that Sayf al-Dawla exercised some suzerainty over Antioch and even Maṣṣīṣa and Tarsus in the Cilician plain, although these seem to have been independent for practical purposes. Canard remarks that the state was Mesopotamian rather than Syrian in character; certainly, none of the Hamdanids seem to have made any effort to develop a fleet and conduct a Mediterranean policy as the Fatimids were to do. Sayf al-Dawla’s power rested on the members of his family, his alliance with the Kilāb and other bedouin tribes and the force of ghilmān he recruited. The Hamdanid family were important in the establishment of Hamdanid power and were employed in governing key centres like Homs and Manbij, where Sayf al-Dawla’s cousin and “court poet” Abū’l-Firās was the governor. In addition, his nephew Muḥammad, son of Nāṣir al-Dawla of Mosul, often acted as his deputy in Aleppo. Bedouin leaders are also found in his entourage and as governors, but as the reign progressed, the ghilmān became more and more influential. In 338/950, he had 2,000 of them, led by a Turkish officer, when he consolidated his power in the Jazīra, and later in the reign they came to hold many key posts. After Sayf al-Dawla’s death, his ghilmān were an important factor in the politics of the area, some of them trying to become independent or accepting Fatimid overlordship. Hamdanid Syria, like Hamdanid Mosul, was not basically an Arab tribal state; it was rather based on the principles of late ‘Abbasid government, with its extensive use of ghilmān, and it is useful to remember how closely Sayf al-Dawla had been involved in Baghdad politics before he ever came to Syria.
Much of Sayf al-Dawla’s reputation, both with his contemporaries and with posterity, rested on his role as a champion of Islam against the military might of the Byzantine Empire, and before discussing his contribution, it is appropriate to consider the history of the frontier throughout the early Islamic period.
After the initial success of the Arab conquests, the Byzantines retained control of the highlands of Anatolia as far as the passes through the Taurus Mountains. At first, the Muslims seem to have preferred to keep the land to the south of this as a no man’s land, a sort of cordon sanitaire, between them and the enemy. Muslim forces tended to concentrate on bases to the rear like Antioch and Dābiq. The Cilician plain, which had been rich and prosperous in classical antiquity, was left almost uninhabited. In the early second/eighth century, however, this position began to change. After the reign of the Umayyad Caliph Sulaymān b. ‘Abd al-Malik (96–99/715–717), there were no more major attacks on Constantinople, and it became clear that the Byzantine Empire and the Muslim caliphate were going to coexist for some time. This change was confirmed at the time of the ‘Abbasid revolution when Byzantine forces led by Constantine V (741–775) began to take the initiative and Muslim cities like Malaṭiya, which had been in Arab hands since the conquests, were threatened by the enemy. Under the later Umayyads and early ‘Abbasids, the frontier provinces began to be settled and fortified in depth. In the Cilician plain, Adana, Maṣṣīṣa (Misis) and above all Tarsus became great Muslim cities, while farther to the east, Mar‘ash, Ḥadath and Malaṭiya became centres of population and military activity. These Muslim outposts were mostly urban in character, and all were situated in the plains or river valleys, while the Byzantines were restricted to the highlands and the cities of the plateau to the north. The Muslims continued to raid Byzantine territory, and sometimes, as in the later years of Hārūn’s reign (187–193/803–809) and the reign of al-Mu‘taṣim (218–227/833–842), there were full-scale military campaigns, but the objectives were booty and prisoners and the demonstration of the caliph’s role as the defender of Islam, rather than the acquisition of new territory.
For 200 years, the Muslim cities of the frontier provinces enjoyed security and growing prosperity. It has been suggested that they were major centres of Arab–Byzantine trade, but there is no evidence for large-scale cross-border commerce, and we hear mostly of exchanges of prisoners and ransoms. The wealth of the area was derived from local commerce and industry, especially the textile workshops of Tarsus. They also enjoyed a very favourable fiscal status, since rates of taxation were kept low on the frontiers in order to encourage settlement and because the frontier cities received continuous subsidies from other parts of the Muslim world on account of their role in the jihād (Holy War). Ibn Ḥawqal, writing in the mid-fourth/tenth century, describes how there were hostels in Tarsus for ghāzīs, warriors for the Faith from every corner of the Muslim world, where these volunteers could come and take part in the struggle against the infidel; these hostels were supported by pious donations from the provinces whose ghāzīs were lodged there. The prosperity and security of the frontier areas then were largely based on continuing support, both financial and in men, from the rest of the Muslim world.
In the fourth/tenth century, this happy state of affairs began to change. The Byzantine emperors, encouraged by the great military landowning families of eastern Anatolia, began systematic campaigns of conquest. At the same time, the political fragmentation of the Muslim world meant that the traditional support from the rest of the Islamic world was greatly reduced. The Byzantine advance began in earnest in 322/934 when the general John Curcuas took Malaṭiya, the first large Muslim city to fall into Greek hands, the Baghdad government of al-Muqtadir and al-Qāhir being quite unable to send support to the threatened outpost, despite repeated appeals for help. It is against this sombre background that Sayf al-Dawla’s role in the jihād must be considered.
Even before he became the ruler of Aleppo, Sayf al-Dawla had shown his commitment to the jihād. In 328/940, he led a foray to the Van area and on to the Byzantine territory to loosen the enemy’s grip on the town of Qālīqalā (Erzerum). After he took over Aleppo in 333/945, he could not avoid facing the Byzantines almost every year. Until 345/956, he was moderately successful, holding his own in a long series of raids, ambushes and brief treaties. Sometimes, his expeditions were on quite a large scale: in 339/950, he led 30,000 of his own men and 4,000 men of Tarsus into the mountains on a raid which proved disastrous. He also took care to restore frontier fortresses like Mar‘ash and Ḥadath in 341/952. From 345/956, however, he faced increasing Byzantine pressure, which he was quite unable to resist effectively. The main reason for this change seems to have been the arrival on the eastern frontiers of the Byzantine Empire of a new commander, Nicephorus Phocas, later to be the emperor, and his brother Leo, and a new army, composed largely of Armenians, like the young officer John Tzimisces, who was to be Nicephorus’ successor on the imperial throne. The policy of raids was replaced by one of territorial expansion: in 346/957, the rebuilt Ḥadath was sacked, in 347/958 Sumaysāṭ. (Samosata) fell and in 349/960 Sayf al-Dawla suffered a major defeat when his army was attacked in the mountains by the troops of Leo Phocas. He himself escaped, but his capacity to resist the Byzantines on anything like equal terms was lost. In 351/962, the Byzantine army attacked Aleppo itself. Sayf al-Dawla had been unable to offer any serious opposition, retired beyond the Euphrates and abandoned the city to its fate. While a small garrison in the citadel continued to resist, the rest of the town and the Hamdanid palace outside the walls were devastated. After this blow, nothing could stop the inexorable advance of Byzantine armies: in the same year, Mar‘ash fell and in 354/965 Maṣṣīṣa and Tarsus were taken and Byzantines established control over all the rich Cilician plain.
One reason for the success of Byzantine arms at this time was the massive superiority in numbers which they enjoyed. Although the figures are no doubt exaggerated, it is said that Byzantine armies numbered up to 200,000, whereas Sayf al-Dawla could never call on more than 30,000 and usually very much fewer. Furthermore, he was unable to call on resources from without his own limited domains. When the Byzantines had attempted offensives in the early ‘Abbasid period, Hārūn al-Rashīd or al-Mu‘taṣim had been able to call on the resources of the whole Muslim world and raise up to 100,000 men. But since the fragmentation of the caliphate, this was impossible. Even his own brother, Nāṣir al-Dawla of Mosul, was unable or unwilling to offer Sayf al-Dawla effective support – still less would the Buyids of Iraq or the ikhshīd of Egypt. It was not a question of the Byzantine Empire against the Muslim world but rather the Byzantine Empire against Sayf al-Dawla’s small, northern Syrian principality. Furthermore, the disputes of the rest of the Muslim world were often reflected on the frontiers. In the years before the Byzantine conquest, Cilicia had seen bitter strife among supporters of the Ikhshidids, the Hamdanids and local leaders. Faced with the final crisis, the governor of Tarsus, Rashīq al-Nāsimī, received no outside support at all in his struggle.
This was not because there was no popular enthusiasm for the jihād. Byzantine successes, the deportation of Muslim populations and the conversion of mosques into stables provoked popular indignation in the Muslim world. In 353/964, a group of 5,000 ghāzīs from Khurāsān arrived to help defend Maṣṣīṣa against the Byzantines, but they were thwarted by famine conditions and retired to Antioch, where they formed another party in the already divided city. In 355/966, a large number of Khurāsānīs, 20,000, arrived at Rayy on their way to the Byzantine frontier and demanded vast sums from the Buyid ruler, Rukn al-Dawla, to support them in this enterprise. When this was not forthcoming, they attacked the city, pillaged the palace of the wazīr Ibn al-‘Amīd and almost killed Rukn al-Dawla himself before being dispersed by his Daylamite soldiers; none of them seems to have reached the frontier. In 361/972, refugees brought news of further Byzantine successes to Baghdad and the populace demanded that the caliph and the Buyid ruler, Bakhtiyār, take action to defend the Muslims, but the politicians diverted the enthusiasm for their own purposes – much of Baghdad was pillaged, but the army never set out. Compared with the inaction or indifference of other Muslim rulers, it is not surprising that Sayf al-Dawla’s popular reputation remained high; he was the one man who attempted to defend the Faith, the essential hero of the time.
The Byzantine successes had important long-term consequences for the history of the area. Towns like Tarsus and Malaṭiya had for two centuries been Muslim Arabic-speaking centres. The Byzantine conquest changed that forever. When the Greeks took these cities, they did not simply replace one government with another; they drove out or enslaved all the existing inhabitants, and mass deportations to central Anatolia and expulsions to other areas of the Muslim world effectively depopulated the cities. These empty areas required inhabitants, and in order to people them, Nicephorus Phocas encouraged Christians living under Muslim rule to come and settle. The historian Yaḥyā b. Sa‘īd, who left Egypt to escape the uncertainties of the Fatimid Caliph al-Ḥākim’s rule and settled in Byzantine Antioch, was one such. The most extensive settlement was around Malaṭiya. Nicephorus invited the Monophysite (Jacobite) patriarch of Antioch, who normally lived around Raqqa or Ḥarrān, to come north with his followers. The settlement of Christians is reflected in the expansion of the ecclesiastical administration. In the century between 950 and 1050, ten new bishoprics appeared around Malaṭiya, four around Mar‘ash and five in the area of Sumaysāṭ, and the Chronicle of Michael the Syrian mentions fifty-six new monasteries in the area between 936 and 1047. This influx of Jacobite Christians, who were regarded as heretics by many in the Constantinople establishment, was to cause problems in the next century, and in the 1030s the Patriarch Dionysius IV preferred to live under Muslim rule in Āmid than under Byzantine authority in Malaṭiya. On the coast, the cities of Antioch and Laodicea seem to have been settled by Greek Orthodox Christians, and as well as refugees from Egypt like himself, Yaḥyā b. Sa‘īd records in 404/1013 that large numbers of Christians from Palestine fled the depredations of the Banū Ṭayy to settle in the area. In Cilicia, the most important settlement was the fifth/eleventh-century migration of Armenians from eastern Anatolia. By the year 1050 then, the Muslims had been eliminated from a large swathe of territory along the frontiers and in northern Syria, and Arab civilization had been almost entirely uprooted. When these lands again passed into Muslim hands, they were settled not by Arabs but by Turkish-speaking tribesmen.
It was not only on the Byzantine front that Sayf al-Dawla’s last years were beset by problems. The last six years of Sayf al-Dawla’s life (350–356/961–967) were marked by an almost complete collapse of his government. This was to some extent because he was afflicted by partial paralysis, which made it difficult for him to play an active role. The conquest of Aleppo by the Byzantines in 351/962 was a terrible blow to his prestige and power. When the enemy approached the city, he was unable to put up any serious resistance and abandoned the people to their fate. It was the citizens who manned the city walls to resist the attackers and the shaykhs of the city who finally negotiated a treaty with the enemy. It is said that 10,000 prisoners were taken to Byzantine territory, most of them young people, and Aleppo must have been something of a ghost town. Sayf al-Dawla abandoned his capital almost completely and retired across the Euphrates to Mayyāfāriqīn, leaving Aleppo in the care of his Ḥājib (chamberlain) Qarqūya.
These misfortunes and Sayf al-Dawla’s manifest inability to deal with the problems which was faced by him led to a number of rebellions. In 354/965, Marwān al-‘Uqaylī, an ex-Qarmaṭī leader long in his service, led a rebellion in the Homs area and was able to enter Aleppo itself before he died of wounds received in the battle for the city. Later the same year, there was a major disturbance in Antioch, now directly threatened by Greek advances in the Cilician plain, and led originally by the exiled governor of Tarsus, Rashīq al-Nasīmī. Once again, the issue must have been Sayf al-Dawla’s failure to protect the area. A force of 5,000 men, led by a Daylamī commander, Dizbar, attacked Aleppo, and Sayf al-Dawla himself came west to confront them. His prestige was not entirely dead; the Kilābī tribesmen who had joined the rebel cause returned to him and he was finally victorious. Severe reprisals were taken against the people of Antioch, only four years before it fell to the Byzantines.
In Safar 356/February 967, Sayf al-Dawla died aged fifty-one. He had ruled Aleppo for twenty-three years and died in the city, but his body was embalmed and taken to Mayyāfāriqīn for burial. He was succeeded without opposition by his only surviving son, Sharīf Sa‘d al-Dawla, but the disasters of the last year of his father’s life meant that the inheritance was not an easy one. Sayf al-Dawla’s chief minister in Aleppo, Qarqūya, accepted Sa‘d al-Dawla as the ruler but advised him to stay away from the city. Many of his father’s companions went to join his cousin, Abū Taghlib, in Mosul, while he himself with only 300 followers was refused entry to Ḥarrān and was turned away from Mayyāfāriqīn by his own mother. For the next ten years (357–367/968–977), he had neither state nor capital. Aleppo was ruled by Qarqūya, supported by Sayf al-Dawla’s ghilmān, who were determined to preserve their position now that their old master was dead.
The main problem Qarqūya faced was in relation to the Byzantines. In 358/969, they took Antioch and began setting up the Byzantine province of Syria, which was to include the coastal hills as far south as Ṭarṭūs. Faced by this formidable power, Qarqūya had to make terms. In 359/969, the Byzantines temporarily occupied Aleppo, and Qarqūya made a treaty which effectively made Aleppo a Byzantine tributary state. It also made provision for the free passage of commercial caravans between Aleppo and Antioch. This truce laid the foundations of Byzantine policy towards Aleppo for the next half-century. The Byzantines seem to have had no wish to annex the city to the empire. Instead, they sought to keep it as an independent, tribute-paying buffer state. Normally, they do not seem to have interfered in internal affairs, but they were prepared to take military action if any outside Muslim power threatened to take over the city. In the same year, the position was further altered by the Fatimid conquest of Egypt. This meant that the feeble government of the Ikhshidids to the south of the Hamdanid state was replaced by a determined and energetic administration, prepared to make serious efforts to gain control. From 358/969 onwards, the weakened and truncated principality of Aleppo was to be the scene of constant rivalry between the two superpowers. These years saw the high point of Byzantine power in Syria during the reign of John Tzimisces (969–976), who in 361/972 led a great expedition which reached as far south as Caesarea and Tiberias in Palestine, although the southern boundary of Byzantine Syria remained approximately on the modern Lebanese–Syrian border.
Sa‘d al-Dawla’s position improved in 367/977 when two of his father’s ghilmān, who had established themselves in the country to the south of Aleppo, invited him to join them in a bid to take over the city. By this time, power in Aleppo had passed from Qarqūya into the hands of a Circassian mamlūk (military slave) of his called Bakjūr. (It seems that the Hamdanids were the first dynasty to recruit ghilmān/mamlūks from Circassia at the Black Sea end of the Caucasus Mountains. Circassian mamlūks were later to be extremely important in the Middle East, notably in Egypt, where they provided most of the sultans from 784/1382 to the Ottoman conquest in 922/1517.) Despite some setbacks, Sa‘d al-Dawla was able to set up a siege of Aleppo and, with the aid of the Banū Kilāb who had served his father so well, to take the city. Bakjūr had to be placated with the governorate of Homs, which he used as a base for intrigue with the Fatimids.
Once established in Aleppo, Sa‘d al-Dawla released Qarqūya to serve him until his death three years later (380/990–991). After this, Sa‘d al-Dawla continued his policy of dependence on the Byzantines, who supported him with up to 6,000 Greek and Armenian troops against the attacks of Bakjūr and his Fatimid supporters.
Sa‘d al-Dawla died in Ramaḍān 381/December 991. Towards the end of his reign, he seems to have regained something of his father’s position, but he was heavily dependent on the Byzantines and his father’s ghilmān, while his principality was confined to Aleppo and its environs. He was never wealthy enough to build up his own corps of ghilmān. All the coastal area was now in Byzantine hands – the Jazīra, including the second capital of Mayyāfāriqīn, had been lost to the Kurdish Marwanid dynasty, while the southern towns around Homs were usually in the hands of the Fatimids or their supporters. His death caused another rash of defections. His young son Sa‘īd al-Dawla was dominated by his military adviser Lu’lu’, who continued the policy of alliance with the Byzantines. Other ghilmān resented him and went over to the Fatimids, who launched a series of attacks on Aleppo between 382 and 384 (993–995), and in 995 the city was only saved by the arrival of the Byzantine Emperor Basil II in person. In 392/1002, Sa‘īd al-Dawla died, and Lu’lu’ soon exiled the rest of the family to Egypt and took over power in his own name. He tried to break away from overly close dependence on the Byzantines and acknowledged the authority of the Fatimids. After his death in 399/1008, his son, Manṣūr, took over, but he was now little more than a Fatimid governor. The city state of Aleppo was unable to sustain itself against the pressure of the bedouin tribes, notably the Banū Kilāb. Manṣūr tried to solve the problem by one of those macabre dinner parties which were a feature of Middle Eastern political life. He invited the leaders of Kilāb to a feast in the citadel of Aleppo and had them murdered. One leader, Ṣālih. b. Mirdās, escaped and it was he who roused the tribe to vengeance. Manṣūr was forced into exile in Byzantine territory, where he was given some castles in the mountains near Antioch whence he could survey the affairs of the city he had once ruled. With his departure in 406/1016, the last traces of the Hamdanid state perished. Only the desire of the Hamdanid ghilmān and the Banū Kilāb to remain independent of the Fatimids and the support of the Byzantines had kept it alive so long.