4

The Umayyad caliphate

The Sufyanid caliphs: Mu‘āwiya and his family: 41–64/661–684

The assassination of ‘Alī b. Abī Ṭālib in 40/661 left his rival Mu‘āwiya b. Abī Sufyān as the undoubted strongman of the Muslim community. Mu‘āwiya was by this time in his fifties, having been born in the first decade of the seventh century. His father, Abū Sufyān, had emerged as a leader of the Meccans in the years that followed the battle of Badr and had conducted the negotiations which brought the city to acknowledge the authority of the Prophet. His young son received a political education in the best traditions of the Quraysh. Like his father, he became a Muslim at the time of the conquest of Mecca, that is, later than many over whom he was subsequently to rule, but his education meant that he became one of Muhammad’s secretaries and part of the new Muslim élite. After Muḥammad’s death, he and his elder brother Yazīd went on the expeditions to Syria, where the family had owned property before the coming of Islam. Yazīd’s premature death from plague meant that Mu‘āwiya came to be the leader of the family and governor of Syria after the death of Abū ‘Ubayda. He remained the governor without interruption or challenge for the next twenty years, thus obtaining an unrivalled opportunity to build up and strengthen his power base in the province.

While his place in the affections of the Syrians was secure, even his enemies attested that the problems he faced in the rest of the Islamic world were formidable. He had to assert not only his personal power but also the credibility of the caliphate and the unity of the Muslim community, which had been so badly damaged in the preceding years, and he faced opposition not just from people who resented his assumption of power, but from people who resented the whole idea of a strong and effective government. Perhaps his greatest achievement was to ensure that despite the stresses his policy inevitably caused, the Muslim world remained united enough to resist the attacks of its enemies and to expand its own borders.

Mu‘āwiya’s character and policies emerge with some clarity from the sources, and even hostile commentators paid tribute to his abilities.

His virtues were those of the successful politician, not of the brilliant general or the religious leader; of these virtues, the most important was the quality the Arab sources describe as ḥilm – the shrewdness, moderation and self-control that the situation demanded. Money was as important a weapon as the support of the Syrian army; he knew that most people have a price and he was prepared to pay it to avoid conflict. The loyalty of the Syrians was assured and their army was the most effective in the Muslim world, but he did not use this military power as an instrument to keep control over other provinces; he used it to attack the Byzantines on land and sea. It was there as a reserve, but a reserve that was never needed after the death of his rival ‘Alī. He ruled rather by making agreements with those who held power in the provinces, by building up the power of those who were prepared to cooperate with him and by attaching as many important and influential figures to his cause as possible. It was not an absolute government in which governors were appointed and dismissed by a central authority but almost a confederation of different leaderships acknowledging one overall authority. From his provincial governors, the caliph demanded acceptance of his authority, that they keep order and that, in some cases, they forward revenues to the central government: it is recorded, for example, that of 60 million dirhams collected in the province of Baṣra, only 4 million was sent to Damascus, all the rest being spent in the province, mostly on paying the local military. Beyond that, governors were allowed to establish their own power bases and ensure the fortunes of their families and friends. Unlike his own relative and predecessor, ‘Uthmān, however, he did not attempt to assert his authority through his own family. The Umayyads were an extensive family and contained a number of figures, notably Marwān b. al-Ḥakam, of great ability and political experience, but few of them were given important roles in the government, either as advisers at the court in Damascus or as provincial governors. They were mostly settled, in luxury, in Medina and the Ḥijāz, far from the real centres of power. Mu‘āwiya’s father had been a leading figure in the Meccan commonwealth, that system of trading agreements and alliances which had brought such prosperity to the city, and his son seems to have used this, rather than the absolutist models of the Byzantine Empire or the religious claims of ‘Alī b. Abī Ṭālib, as the inspiration for his statecraft.

The most serious difficulties he faced were in Iraq. The death of ‘Alī had not solved any of the deep-rooted problems which the conquest and the vast flood of immigrants had caused. The Iraqis were resentful of Syrian dominance, but this feeling did not lead to any sense of unity. There remained the social divisions between the ashrāf and the early Muslim élite, now itself divided between the qurrā’ and their descendants who clung to the memory of ‘Alī, and the Khawārij who rejected any but their own rigorist interpretation of Islam. There were regional differences as well; the problems of Kūfa were not those of Baṣra, and the two cities were more rivals than allies. These tensions and divisions made violence almost inevitable and that it did not break out until after the death of the caliph was a tribute to his own talents and those of the men he chose.

In social terms, Mu‘āwiya’s triumph was the triumph of the ashrāf in Kūfa and the surrounding area. The fortunes of tribal leaders like al-Ash‘ath al-Kindī and his son Muḥammad and Jarīr al-Bajalī prospered, while those who had challenged their status, like Mālik al-Ashtar and his son Ibrāhīm and Ḥujr b. ‘Adī al-Kindī, felt betrayed and resentful. As the governor of this most difficult city, the caliph chose a man of ability and experience if of somewhat disreputable morals. Al-Mughīra b. Shu‘ba al-Thaqafī came from the holy family of Ṭā’if, just as Mu‘āwiya did from that of Mecca. Before the conquest of his city by the Muslims, he had fled to Medina and embraced the new religion in the year 8/629–630 to escape the consequences of a particularly outrageous murder. Islam, as Muḥammad said, cancelled all that had gone before, and al-Mughīra, like Mu‘āwiya, became one of the Prophet’s secretaries. When Ṭā’if fell to the Muslims, al-Mughīra, helped by Abū Sufyān, was the man who destroyed the idol of the town and pillaged the temple treasury. Like many Thaqafīs, he joined in the conquest of Iraq, and his knowledge of Persian marked him out for administrative office. He became the governor of Baṣra under ‘Umar and was removed after allegations of adultery but reappointed as governor of Kūfa at the end of ‘Umar’s reign. In the conflict between ‘Alī and Mu‘āwiya, he had played a moderating role, advising ‘Alī to recognize Mu‘āwiya in Syria. Nobody had more experience of the Iraqis and their problems, and Mu‘āwiya appointed him the governor, a post he held for nearly ten years, from 41 to his death in 50 (661–670). He adopted a policy of keeping the peace and turning a blind eye to misdemeanours which did not threaten his authority. The Kūfans were allowed to keep for their own needs the revenues of the district known as Māh al-Kūfa in the Zagros Mountains, and the regular payment of salaries, something ‘Alī had never been able to achieve, must have done much to induce cooperation.

In Baṣra, Mu‘āwiya turned at first to the obvious candidate, the Umayyad ‘Abd Allāh b. ‘Āmir b. Kurayz, who had already held the office under ‘Uthmān and had begun the conquests of Khurāsān and Sīstān. Now reappointed, he again opened campaigns in the east, notably in Sīstān, and Arab armies even entered distant Kābul. While his fighting qualities were not in doubt, however, ‘Abd Allāh seems to have been either unable or unwilling to undertake the maintenance of law and order in Baṣra and there may have been some resentment against these distant campaigns. He was dismissed in 44/664 and replaced by the most remarkable of Mu‘āwiya’s subordinates, Ziyād b. Abīhi, otherwise known as Ziyād b. Abī Sufyān. Like al-Mughīra, he came from Ṭā’if, from a Thaqafī background, although his exact parentage was obscure, as his name implies (Ziyād son of his father). He joined the Thaqīf in southern Iraq in the earliest stages; because of his education, he was put in charge of the division of the booty and then became the secretary to successive governors, including al-Mughīra, who became the patron of this bright young man. He was one of ‘Alī’s most consistent supporters, refusing to be reconciled to Mu‘āwiya even after ‘Alī’s death and remaining in Fārs until 42/662–663, when, with al-Mughīra acting as intermediary, he was reconciled to the new caliph. Mu‘āwiya recognized that he had both the abilities and the all-important local connections to be his right-hand man in Baṣra and he was duly appointed. Ziyād had a very positive view of the governor’s responsibilities and his conception of the office was deeply influential in the later development of ideas of government. In a widely reported speech, he made to the people of Baṣra on his appointment, he is said to have offered both a carrot and a stick. The good news was that the ‘aṭā’ was to be paid regularly and that they would not be forced on long campaigns in distant regions – he also promised to be accessible at all times to his subjects (he himself never led armies to the east). However, he promised draconian measures to enforce law and order within the city; there was to be a curfew; tribal and clan solidarities were no longer to be used to protect the guilty; and the shurṭa, or police force, was to ensure the maintenance of peace. He ended with the celebrated warning, as he surveyed the people gathered before him, that he saw heads tottering on shoulders and they were to make sure that their own stayed in place. The policies seem to have worked, and Ziyād enjoyed the favour and confidence of the caliph to the extent that he was acknowledged as Mu‘āwiya’s half-brother and was officially known as Ziyād b. Abī Sufyān as a result.

In 50/670 on al-Mughīra’s death, he replaced his former patron as the governor of Kūfa as well as of Baṣra and hence became a virtual viceroy in the east.

In Kūfa, his firmness contrasted with al-Mughīra’s more easygoing ways, but the reliability of his financial administration went far to compensate for this. He reorganized the city into four quarters, each with a chief appointed by and responsible to him in order to control the population. As the governor of Iraq, he also controlled appointments in Iran. Here, it is clear that he encouraged further conquests as a way of relieving the pressure on resources in Iraq. From 47/667 onwards, he sent a series of armies to Khurāsān. Hitherto the Arabs had maintained a small garrison in Marv but otherwise left the administration in the hands of the local princes. Ziyād decided to change this position in a way which was to have extremely important consequences for Islamic history. In 51/671, he sent to the provinces a large contingent of 50,000 troops from both Baṣra and Kūfa, not as a garrison but as settlers, and they established themselves in the Marv oasis. This meant that Khurāsān had a larger number of Muslim settlers than any other area outside the Fertile Crescent, more than Egypt and certainly more than any area in western Iran. For the moment, this emigration solved some of the problems of Iraq, but it was the descendants of these settlers who were, in the end, to destroy the Umayyad caliphate itself.

Iraq and the east then were managed by Mu‘āwiya through what some must have seen as a Thaqafī mafia. Thaqafīs were experienced in the affairs of Iraq before the coming of Islam, were well educated (it is worth noting how important the education of both al-Mughīra and Ziyād was in ensuring their success) and had close connections with the Quraysh and the Umayyads from before Muḥammad’s time. In addition, the Thaqīf had played a major role in the conquest of Iraq; nobody could say they were outsiders, creatures of a distant and unresponsive government. All these factors made for a system which worked for Mu‘āwiya’s reign; when the great Ziyād died in 53/673, he was succeeded a few years later in all his offices by his son ‘Ubayd Allāh, more hasty and given to the use of force than his father, but a man whose devotion to the Umayyad cause could not have been doubted.

Mu‘āwiya’s power base was in Syria but, compared with Iraq, we are fairly badly informed about the position in the province. The Islamic historical tradition was developed in Iraq under ‘Abbasid rule, and the sources tend to be indifferent to the Umayyads and much more concerned with Iraqi and Khurāsānī affairs. It is clear, however, that Mu‘āwiya drew support from a wide cross section of Syrian tribes, both from those who had arrived at the time of the conquest and those who had been settled in Byzantine times. As yet the feud between the Qays and Yaman which was to plague Syria for so many centuries was not a major problem. He gathered around himself what might be described as a cabinet of Syrian leaders who were his right-hand men; some of these were Qurashīs, including ‘Abd al-Raḥmān, the son of the great general Khālid b. al-Walīd, who was powerful in the Homs area, but none of them were Umayyads and they tended to come from the less important clans of the Quraysh, notably from the Fihrīs, like al-Ḍaḥḥāk b. Qays, who had played an important role in the conquest of the Jazīra. The others were mostly from Yamanī tribes, including Shuraḥbīl b. al-Simt. al-Kindī, whose power base was in Homs, and Ḥassān b. Mālik b. Baḥdal, chief of the powerful Kalb tribe. This inner circle assured the loyalty of different groups in Syria and the Jazīra and was occasionally used on difficult missions outside that area. While they were sometimes made governors of the districts within Syria, they were never appointed to Iraq or the east and the Syrian army was kept busy attacking the Byzantines. Rhodes was taken in 52/672 and Crete in 54/674, while during the last seven years of Mu‘āwiya’s reign, the caliph’s son Yazīd led continuous attacks on Constantinople itself (54–60/674–680). These expeditions were mostly raids, and even areas such as the Mediterranean islands that were captured and held were used as sources of tribute but not for colonization. The Arabs of the Jazīra led similar expeditions into Armenia but again there was no extensive settlement of new areas such as the Iraqis were accomplishing in Khurāsān. This suggests that the motives for the warfare were to provide the Syrians with military experience and financial rewards rather than to cope with the problem of overpopulation. Mu‘āwiya managed the Syrians with consummate skill and their loyalty to him was famous; it was only after his death that serious divisions began to appear.

Egypt was controlled almost as an appendage of Syria. Arab settlement here was very limited, being confined almost entirely to the cities of Fusṭāt. and Alexandria. In the struggle with ‘Alī, Mu‘āwiya had received the loyal backing of ‘Amr b. al-‘Āṣ, anxious to regain control of the province he had conquered for Islam, and ‘Amr ruled Egypt almost as a partner until his death in 43/663. For most of the rest of the reign, 47–62/667–682, Egypt was governed loyally by Maslama b. Mukhallad, the only member of the Anṣār to achieve a high rank in Mu‘āwiya’s government.

If Mu‘āwiya’s caliphate was decentralized politically, it was decentralized administratively as well. No attempt was made to introduce a general, Arabic-speaking administrative system for the whole Muslim world. On the contrary, each province continued the local traditions of the previous rulers. In Syria, financial administration wasṭalmost entirely in the hands of local Christians, including Sarjūn, son of Manṣūr, who had held the same office for Heraclius at the time of the Muslim conquest. In Egypt, as we know from the evidence of the papyri, the Byzantine system continued with significant changes but with local headmen continuing to be responsible for tax collecting and administration at a local village level. It used to be thought that the Arab conquerors had little understanding of administration and simply carried on existing practices, but recent research has shown Arabic documents being issued from the year after the first conquest, 22/643, and an Arabic-writing administration ensuring that taxes were paid and brought to the capital at Fusṭāṭ. We have already seen that the local magnates in Khurāsān continued to manage their own affairs, only paying tribute to the Arabs. Even in Iraq, the area most affected by the Arab expansion, the native landowners continued to play a role in local administration. This local autonomy was typified by the coinage; there was no Islamic coinage as such, rather each province used and supplemented where necessary the existing money supply, the gold dīnār of the Byzantine areas and the silver dirham of the Sasanian countries. Taxation systems differed as a consequence and there was no uniformity in the dues paid by the native inhabitants of different areas, these being determined by local tradition and the circumstances of the conquest rather than by government policy. Mu‘āwiya supplemented the limited income derived from provincial taxation in two ways; the first was by frontier warfare, and there can be no doubt that at this stage, booty and tribute helped to secure the loyalty of the Syrians. The second way was by extensive agricultural developments. We know most about this in the Ḥijāz around Medina, since his activities provoked considerable opposition, but no doubt other likely areas were developed as well. This large-scale farming, using the labour of mawālī (freedmen, sing. mawlā) or slaves meant a substantial income for the treasury. In Iraq, the extensive ṣawāfī lands which had caused so much anger during the reign of ‘Uthmān were taken over as government property, or rather they ceased to be the māl al-muslimīn (wealth of the Muslims), controlled by the Muslims, but became rather the māl Allāh, controlled by the caliph. The Kūfans were paid their ‘aṭā’, regularly and honestly, but they no longer had direct control over the lands which produced it.

The one serious problem which disturbed the later years of Mu‘āwiya’s reign was the ordering of the succession. Mu‘āwiya was determined that his son Yazīd should succeed him and that he should be formally acknowledged by the Muslim community in his father’s lifetime. Many were scandalized by the idea. No caliph had been so succeeded by his son, and those who opposed the plan were quick to call for a shūra or some more open form of decision and accused Mu‘āwiya of attempting to set up a hereditary monarchy. Everyone understood that the acceptance of Yazīd would mean the maintenance of the sort of Syriā-based regime which his father had set up, and the idea was therefore bitterly opposed, not just by the Iraqis but also by those Umayyads and other Qurashīs in the Ḥijāz who had been effectively excluded from power by Mu‘āwiya. In the end, the opposition was silenced by threats or payments, but when Mu‘āwiya died in Rajab 60/April 680, his son, although able, well liked and of proven military ability, faced a whole series of challenges.

The Kūfans had not lost their enthusiasm for the house of ‘Alī during the long years of Mu‘āwiya’s reign, since they believed that only the coming to power of the Family of the Prophet would restore Kūfa and the Kūfans to their rightful position. ‘Alī’s elder son al-Ḥasan had abandoned his claims to leadership, but his younger brother al-Ḥusayn was prepared to take up the cause. On the death of Mu‘āwiya, al-Ḥusayn left his place of retirement in Medina and travelled across the desert towards Kūfa to seek his supporters, accompanied by only a small band of family and friends. The governor of Iraq, Ziyād’s son ‘Ubayd Allāh, was waiting for him; he prevented the small party from reaching the settled lands and, after some days of confrontation when al-Ḥusayn and his followers suffered increasingly from thirst, there was a short battle at a place called Karbalā’, and al-Ḥusayn and most of his party were killed. The Kūfans had made no effort to support him. It was 10 Muḥarram 61/10 October 680. At the time, the attempt to raise the ‘Alid banner in Iraq was a pathetic failure, but the events of the massacre at Karbalā’ made a deep impression on the Muslim community. Many felt that the grandson of the Prophet, a man whom Muḥammad loved and had played with on his knee, had been brutally done to death by the forces of godless oppression. Al-Ḥusayn might be dead but his memory lived on; the accounts of his sufferings at the hands of Umayyad troops were elaborated in Kūfa to discredit the regime, and al-Ḥusayn became the symbol for the sufferings of all the weak and defenceless. To talk of Shi‘ites in Iraq at this stage is misleading; the development of a Shi‘ite “sect” was a much later development, but the death of al-Ḥusayn played an important part in the attachment of people’s hopes and political aspirations to the family of ‘Alī.

The death of al-Ḥusayn meant the effective end of trouble in Iraq; much more intractable were the problems of the Ḥijāz. Yazīd was well aware of the strength of feeling against him in the Holy Cities, and he made great efforts to be conciliatory, but some refused to take the oath of allegiance, notably ‘Abd Allāh, the son of that al-Zubayr who had died at the Battle of the Camel. An austere and determined figure, he held that the caliphate should not be the plaything of the Sufyanids but that the ruler should be chosen from all the Quraysh. Ibn al-Zubayr established himself in Mecca where he became a rallying point for all those who opposed Yazīd’s claims, asserting that there should be a new shūra. The Medinese, for their part, were motivated by other considerations as well; it is clear that Mu‘āwiya’s agricultural activities in Medina had aroused widespread opposition, especially from the Anṣār, who felt that their city was being taken over by an absentee landlord, and they rallied to the anti-Umayyad party. There were some two years of negotiation and manoeuvre, while Yazīd and a series of governors attempted to defuse the situation. Finally, it came to open hostilities; the Umayyads in the city, led by Marwān b. al-Ḥakam, were driven out to Syria, while a Syrian expeditionary force marched on the Prophet’s city. The Medinese attempted to defend themselves, just as the Prophet had done, by digging a trench around the town, but the Syrians were an easy match for them, and at the battle of the Ḥarra (Dhu’l-Ḥijja 63/August 683), they were defeated and the city sacked. It only remained now for the Syrians to march on Mecca and eject Ibn al-Zubayr, and it was while they were engaged in the unsavoury task of attacking the Ka‘ba itself that news came of Yazīd’s death and once more the whole situation was changed.

The disasters of Karbalā’ and the Ḥarra cast a cloud over Yazīd’s brief reign. Yet, it was not devoid of achievement. Compared with Mu‘āwiya’s ambitious raids, he tried to stabilize the frontiers between Syria and Byzantium; outposts on the Sea of Marmara were abandoned, while in the north of Syria itself, he carved up the overly large jund of Homs to make a new frontier province based on Qinnasrīn which was to include both Antioch and Manbij. Had he lived longer, his reputation might have recovered from the shocks of the early part of his reign; as it was, he died in his desert encampment at Ḥawwārīn in Rabī‘ I 64/November 683, when he was only about forty years of age.

Yazīd was succeeded immediately by his young son Mu‘āwiya. His accession was due to the influence of his cousin, Ḥassān b. Mālik b. Baḥdal of Kalb, and the prince himself proved sickly. Mu‘āwiya II, the last of the Sufyanids, died after only a few weeks.

The foundation of the Marwanid caliphate and the achievement of ‘Abd al-Malik: 64–86/684–705

The death of Mu‘āwiya II led to a deep crisis in the Umayyad regime; Yazīd’s other sons were too young to be generally acceptable and the major provinces of the empire were slipping rapidly from the control of the Syrians. ‘Ubayd Allāh b. Ziyād was forced to flee Baṣra, despite a speech in which he not only recalled his own Baṣran connections but promised to retain the wealth of the province for its people. These concessions were not enough, and almost alone he fled along the direct desert road to the Ḥawrān in Syria. On his arrival in Syria, he found the Syrians in total disarray. Al-Ḍaḥḥāk b. Qays al-Fihrī, one of the last surviving members of Mu‘āwiya’s inner circle, held power in Damascus but wavered between acknowledging Ibn al-Zubayr and finding a suitable Syrian candidate. Many other Syrians were willing to acknowledge Ibn al-Zubayr as well, notably the Qaysī tribesmen of the northern part of the country, and even leading members of the Umayyad family were preparing to go to the Ḥijāz to offer their allegiance to him. Not all Syrians, however, were prepared to surrender their privileged position. The Kalbī chief Ibn Baḥdal was related by marriage to the ruling house and now put his power and that of his tribe firmly behind the continuation of the Umayyad house. ‘Ubayd Allāh b. Ziyād had been taught by his experience in Baṣra how closely his own fortunes were linked to those of the dynasty and on his arrival in Syria he too set about organizing the resistance.

The most serious problem faced by Ibn Ziyād and the Kalbī chief was that of finding a suitable Umayyad candidate. Ibn Baḥdal favoured Mu‘āwiya II’s young half-brothers, but ‘Ubayd Allāh turned to Marwān b. al-Ḥakam. Marwān was by this time an old man. He had served his uncle ‘Uthmān as his right-hand man in Medina and after the caliph’s assassination had remained to become the leader of the Umayyads in the Ḥijāz. He had never moved to Syria and had remained very much in the background during Mu‘āwiya’s reign, but the troubles which led to the battle of the Ḥarra had forced him to leave along with the rest of the Umayyads in Medina and make their way north as refugees. Marwān seems to have been resigned to accepting Ibn al-Zubayr when he was met by ‘Ubayd Allāh, newly arrived from Iraq, who persuaded him to change his mind. Ibn Baḥdal then summoned a meeting of Syrian ashrāf, not in Damascus where al-Ḍaḥḥāk b. Qays remained in control, but in his own territory at the old Ghassanid centre of Jābiya. Here, he presided over a sort of extended shūra at which Marwān was finally acknowledged as the candidate of the Umayyad party, in exchange for financial promises. Marwān had no experience or contacts in Syria; he would be entirely dependent on the ashrāf from the Yamanī tribes who had elected him.

Not everyone in Syria could accept this, especially the leaders of the Qaysī party, who more and more turned to al-Ḍaḥḥāk b. Qays. They persuaded him not to go to the Jābiya meeting but rather to prepare for battle. The Yamanīs marched to meet them, Damascus was brought over to their side by the prompt action of a scion of the old Ghassanid ruling family and, in Muḥarram 65/July 684, just nine months after Yazīd’s death, the two sides met in a terrible battle at Marj Rāhiṭ, on the road from Damascus to the north. The Yamanī tribesmen were victorious and the Qaysīs, although much more numerous than their opponents, were driven from the field with a dreadful slaughter, al-Ḍaḥḥāk himself being killed. While the remnants of the Qaysīs, led by Zufar b. Ḥārith al-Kilābī, fled to Qarqīsiyyā’ on the Euphrates, Marwān was acknowledged as caliph in Damascus.

The war had been won, but the bloody victory at Marj Rāhit. had long-term consequences which were to be fatal for Syria and the Umayyads alike because it divided the Arabs of the province into two separate parties, Qays/Muḍar on the one side and Yaman on the other. The basic distinction between northern (Qays) and southern (Yaman) Arabs was an ancient but hitherto fairly harmless one. In theory, the allegiance of each tribe was decided according to complicated genealogies, but in practice there was considerable confusion and some tribes were claimed by both sides. In Umayyad times, the division was in part geographical, the Qaysī tribes being mostly settled in the northern parts of the province, in the newly created jund of Qinnasrīn, and in the Jazīra, including the Byzantine frontier areas as far east as Armenia. Yamanī tribes occupied the steppes around Homs, the Palmyrena, Palestine and the areas to the east of the Jordan valley. It was also a division arising from the Islamic conquests. Some Yamanī tribes like the Kalb, Tanūkh, Judhām and Taghlib had been in Syria before the coming of Islam, while the majority of Qaysīs like the Sulaym, Kilāb and ‘Uqayl were newly arrived from the Arabian Peninsula, many of them brought to the Jazīra by Mu‘āwiya in ‘Uthmān’s reign. Some newly arrived tribes from south Arabia, like the Kindīs and Haḍramīs settled in Homs, joined the Yaman group. Much of the hostility must have been the result of disputes over grazing rights. Many of the Yamanīs came from tribes that were still in part Christian, like the ruling house of Kalb where Ibn Baḥdal’s grandfather had remained Christian until his death during ‘Ali’s caliphate. They had also in many cases been allies of the Byzantines, and in some ways the battle of Marj Rāhit. was the battle of the Yarmūk refought, with the victory this time going to the old-established Syrian tribes. The confused succession to Mu‘āwiya II had led to the emergence of two coalitions, Qaysīs supporting al-Daḥḥāk b. Qays and Yamanīs supporting Ibn Baḥdal and Marwān – but the battle meant that the split continued beyond the immediate crisis. Qays had many dead to avenge and the feud was to continue for generations. As late as the nineteenth century, battles were still being fought in Palestine between groups calling themselves Qays and Yaman.

The newly established caliph had a mass of problems to contend with, but he and his supporters, ‘Ubyad Allāh and Ibn Baḥdal, set about tackling them with energy and determination. Marwān’s son Muḥammad was sent to contain the Qaysīs on the middle Euphrates, while ‘Ubayd Allāh prepared for the reconquest of Iraq. An expedition to the Ḥijāz ended in failure, but the caliph, accompanied by Ibn Baḥdal and another of his sons, ‘Abd al-‘Azīz, had a signal triumph when, with the help of the ashrāf of Fusṭāṭ, they seized Egypt from the representative of Ibn al-Zubayr. Already the characteristics of the new regime were becoming apparent. Marwān looked back for his political guidance not to Mu‘āwiya but to the uncle he had served so long and well, the Caliph ‘Uthmān. Like ‘Uthmān, he made the family the basis of his power, just as Mu‘āwiya had neglected it; his sons Muḥammad and ‘Abd al-‘Azīz had been given military commands almost from the beginning, while his eldest son, ‘Abd al-Malik, was soon acknowledged as heir to the caliphate. It was a shaky beginning, but a new regime had been born.

Marwān died, an old man by the standard of the day, in Ramaḍān 65/April 685 and he was succeeded immediately by ‘Abd al-Malik, his son and designated heir, the accession being smoothlykk managed by the Yamanīs. The new caliph had been born in his father’s house in Medina about 26/646–647, near the beginning of ‘Uthmān’s reign, and he was a boy when the old caliph was killed. Like most members of his family, his relations with Mu‘āwiya were not close and he grew up a pious and serious young man in Medina, where he worked as his father’s assistant in the administration. He came to Syria with his father after the expulsion of the Umayyads from the Ḥijāz and was one of his father’s closest advisers during the few months of his caliphate. His political experience may not have been very extensive when, now, in his early forties, he found himself caliph, but he was methodical, careful and determined to make his power effective.

The obstacles the new caliph faced were formidable. The Syrians were far from united and he was heavily dependent on the Yamanī leaders, Ibn Baḥdal and Rawh. b. Zinbā‘ al-Judhāmī, and the support of members of his own family. Egypt was now firmly under Umayyad control again, but while it was an important source of income, the smallness of the Arab population meant that it provided little in the way of military support. Ibn al-Zubayr had consolidated his power in the Ḥijāz, while his brother Muṣ‘ab had moved to Baṣra and was working to win over the whole of Iraq to his side. In Kūfa, on the other hand, Mukhtār b. Abī ‘Ubayd was trying to turn local discontents to his own advantage and was seeking an ‘Alid candidate for the throne. Finally, the Qaysīs of northern Syria and the Jazīra – defeated, bitter but by no means powerless – had established themselves under the leadership of Zufar b. al-Ḥārith al-Kilābī at Qarqīsiyyā’ on the Euphrates and continued to acknowledge Ibn al-Zubayr as their caliph.

‘Abd al-Malik’s forces began to take the initiative immediately after his succession. ‘Ubayd Allāh b. Ziyād, whose arrival in Syria had done so much to ensure the throne for the Marwanids, was now determined to return to Iraq and reestablish control. His progress was delayed by the activities of the Qaysīs in the Jazīra, and by the time he approached Iraq from the north, via Mosul, Mukhtār had consolidated his position in Kūfa. Against ‘Ubayd Allāh, he sent the son of ‘Alī’s old champion, Ibrāhīm b. Mālik al-Ashtar, who was to prove his most effective military commander. After some setbacks, the Kūfans defeated the army and killed ‘Ubayd Allāh himself at a battle on the river Khāzir, near Mosul (Muḥarram 67/August 686). With ‘Ubayd Allāh gone, ‘Abd al-Malik took no further initiatives to retake Iraq for the next five years. He contented himself with consolidating his power in Syria and waited while his opponents destroyed each other.

In Iraq, the social tensions which had led to the murder of ‘Uthmān and the problems which confronted ‘Alī broke out again with renewed virulence. Umayyad rule had tended to favour the interests of the ashrāf, and there had emerged a new landowning aristocracy in the neighbourhood of Kūfa, where families like those of al-Ash‘ath b. Qays al-Kindī had become important property owners. Farther to the south, much land had been reclaimed in the marshes of southern Iraq, by Ziyād and Mu‘āwiya’s mawlā, ‘Abd Allāh b. Darrāj, but the beneficiaries of this reclamation had not been the people of Iraq as a whole but the supporters of Ziyād and the Umayyad government. As in Medina, Mu‘āwiya’s large-scale agricultural developments had proved a major source of discontent. Along with these old disputes about landownership and social status among the Arabs, a new problem had emerged in the generation since ‘Alī had died. Many of the local people, especially in the Kūfa area, had been converted to Islam. While this was no doubt welcome in many ways, it did pose serious problems to the rulers. The real question was how far these new converts, usually referred to as mawālī (clients or freedmen) of the Arab tribes, should be treated as first-class Muslims. While in theory all should be equal, objections were raised on two fronts. The non-Muslims paid a higher rate of tax on their landed property, the kharāj and a poll tax on top of that, while the Arab Muslims simply paid the ṣadaqa. This meant that conversion would result in a substantial loss of income for the government. Naturally, converts tended to be denied the fiscal benefits of their conversion and, equally naturally, they resented it. But the idea of raising the status of the mawālī was also opposed by the ashrāf, who feared that this flood of new converts would escape their power and undermine their privileges. In this way, a whole new set of destructive social antagonisms came into being.

As usual, it was in the turbulent city of Kūfa that these tensions became most apparent. The weakening of the Umayyad grip had led immediately to unrest. After Mu‘āwiya’s death, the Kūfans had looked to ‘Alī’s son al-Ḥusayn to provide leadership, and al-Ḥusayn’s death had reinforced, rather than diminished, the loyalty of the Kūfans to the ‘Alid family, now represented by al-Ḥusayn’s son ‘Alī and his half-brother, Muḥammad b. al-Ḥanafiyya, both in the Ḥijāz and both reluctant to become actively involved in politics in Iraq. The lack of any forceful ‘Alid leader in Kūfa at this moment led to a certain jockeying for position among old associates of ‘Alī until the leadership was gradually taken over by the enigmatic personality of Mukhtār b. Abī ‘Ubayd, who by Rabī‘ I 66/October 685 had driven out the governors sent by Ibn al-Zubayr and made himself master of the city. Mukhtār was by this time an elderly man; he had been born in Ṭā’if around the time of the Hijra but seems to have gone to Iraq when ‘Umar sent his father Abū ‘Ubayd b. Mas‘ūd to command the armies there. Abū ‘Ubayd had been killed at the Battle of the Bridge, but the family remained in the area and acquired property, and Mukhtār himself held minor office during ‘Alī’s caliphate. Although a Thaqafī, he does not seem to have formed part of the Thaqīf group enriched by Ziyād and his son and remained a representative of the old Islamic élite of Kūfa, anxious to regain its prestige. By the time he assumed control in the city, he was very experienced, and few could have had as good an understanding of the city and its problems.

At first Mukhtār seems to have tried to enter into partnership with Ibn al-Zubayr, but the latter distrusted him and negotiations came to nothing. He then espoused the cause of the ‘Alid family, but it was more difficult to find a candidate who would accept the doubtful honour; both ‘Alī b. al-Ḥusayn and Muḥammad b. al-Ḥanafiyya were in the Ḥijāz, under the eye of Ibn al-Zubayr, and neither of them were keen to share the fate of al-Ḥusayn; while ‘Alī refused absolutely, Ibn al-Ḥanafiyya’s replies were ambiguous but allowed Mukhtār to claim that he enjoyed his support. He proclaimed Ibn al-Ḥanafiyya not just caliph but mahdī. This implied that he was a divinely guided saviour, a messiah, who would, with God’s support, establish justice for all Muslims. This seems to have been a novel concept in Islam, but the idea of the mahdī was widely accepted and was to provide the ideological inspiration for many subsequent movements and rebellions. Ibn al-Ḥanafiyya’s candidature marked a new departure in other ways. Al-Ḥasan and al-Ḥusayn were both sons of Fāṭima and hence direct descendants of the Prophet as well as of ‘Alī, but Ibn al-Ḥanafiyya was not; it was only his descent from ‘Alī that gave him a claim to belong to the holy family, and his acceptance is a sign of the increasing importance of the memory of ‘Alī among many in Iraq.

Mukhtār’s supporters came from many different, largely irreconcilable groups. He took power with the support of Arabs from all the different tribes represented in the population and some mawālī, about 500 in number. While it was not originally a movement of mawālī, their numbers and influence seem to have grown as a result of Mukhtār’s success, and one of their leaders, Abū ‘Amra Kaysān, was appointed the chief of police. Mukhtār’s propaganda explicitly claimed that he was acting in the interests of the “weak” and of the mawālī. At the same time, he tried to win over the ashrāf and so create a united Kūfan militia, but the nobles were increasingly alarmed by the radical nature of Mukhtār’s movement and their support was at best half-hearted.

The threat of ‘Ubayd Allāh b. Ziyād attacking Kūfa from the north and the need to raise an effective army against the menace brought the tensions between Mukhtār and the ashrāf into the open. The latter complained that he had given the mawālī horses, paid them the ‘aṭā’ and generally favoured them unreasonably, and this led to fighting in Kūfa itself, at which the ashrāf were worsted. At the same time, Mukhtār’s relations with Muṣ‘ab b. al-Zubayr also became increasingly problematic, and after Mukhtār refused to accept a Qurashī governor from the Ḥijāz, Muṣ‘ab and he became rivals in Iraq. The defeated ashrāf of Kūfa took advantage of this and, under the leadership of Muḥammad b. al-Ash‘ath, they and their followers, 10,000 strong, left Kūfa and joined Muṣ‘ab in Baṣra. With the Zubayrid forces, they then returned to Kūfa, defeated Mukhtār’s supporters and finally killed him after besieging him in the governor’s palace in 67/April 687.

Mukhtār’s revolt had lasted less than two years and his control had never extended beyond Kūfa and its countryside. Nonetheless, it is of great importance in illustrating the problems of early Islamic society, showing clearly the antagonism between the ashrāf and other Arabs and the mawālī, who, for the first time, played a significant role in the politics of the community. The idea of the mahdī too was to have a lasting importance, and many of the ideas which had been developed in Kūfa during Mukhtār’s brief, revolutionary rule survived among small, closely knit groups of Iraqi Muslims and were to provide the inspiration for future rebels. At the time, however, his failure was complete and Muṣ‘ab and his forces extracted bloody reprisals on his supporters in Kūfa. He had tried to put together a coalition of all the groups in Kūfa, a united front which would once again restore the city to a position of influence in the Islamic community. Despite his attempts at compromise, he had been defeated by the internal divisions. He had also been let down by the ‘Alid family, and even Ibn al-Ḥanafiyya’s endorsements were ambiguous. Many supporters of the family were dubious about Mukhtār’s claims to represent them.

The collapse of Mukhtār’s rebellion meant that Ibn al-Zubayr and ‘Abd al-Malik were now the major rivals for power. ‘Abd Allāh b. al-Zubayr had asserted his claim to leadership on Mu‘āwiya’s death and had fled to Mecca rather than acknowledge Yazīd. He represented and was supported by those Qurashīs who resented the way in which the Umayyad family had monopolized power which they felt should have been more generally distributed among the Quraysh, especially members of his own clan of Makhzūm, rivals of the Umayyads since the days of Abū Jahl and Abū Sufyān. He himself was an austere, some said a miserly, figure, who directed operations from Mecca, gaining a certain prestige from his guardianship of the Holy Places but thereby distancing himself from events in the more populous parts of the Muslim world. In Iraq, he was represented by his more dynamic and worldly brother Muṣ‘ab, who attempted to control Iraq. Muṣ‘ab faced two main problems. In Kūfa, there was Mukhtār and in southern Iraq and the areas around the Gulf there were the Khawārij.

The Khawārij, who had deserted ‘Alī, had remained largely quiescent under Mu‘āwiya, but his death seems to have sparked new disturbances. Acting in small, fast-moving guerrilla groups, they were a constant threat to the government of the settled areas. The two principal groups which threatened Zubayrid control were the Najdiyya and the Azāriqa. The former were based in Yamāma, on the eastern side of the Arabian Peninsula and recruited almost entirely from the Ḥanīfa tribe, who had supported Musaylima the false prophet at the time of the Ridda. They tried, with some success, to assert their control over ‘Umān and even Yemen, and for a period in 68/687 they had more power in the Arabian Peninsula than either Ibn al-Zubayr or the Umayyads. After 72/691, however, internal differences led to a breakup of their power and they were eventually subdued. Even more threatening were the Azāriqa of Ahwāz and Fārs, against whom the Zubayrids were compelled to wage a long and exhausting struggle. Fortunately for Muṣ‘ab, he found a new and important ally to help him in this unrewarding task. Many Azdīs from ‘Umān had moved to Baṣra towards the end of the reign of Mu‘āwiya and their chief, al-Muhallab b. Abī Ṣufra, established himself as one of the most powerful men in the area and a very successful soldier. He is said to have introduced metal stirrups into Muslim armies. It was he and his men who provided the backbone of the Zubayrid army against the Khawārij and against Mukhtār. Despite his efforts, however, the Azāriqa remained a threat until after the family of al-Zubayr had been defeated.

Throughout this period of confusion and bloodshed in Iraq, ‘Abd al-Malik had consolidated his position in Syria. In 71 (691), he reached an agreement with Zufar b. al-Ḥārith al Kilābī and his Qaysī followers in the Jazīra who agreed to abandon Ibn al-Zubayr in exchange for a privileged position at court and in the Umayyad army. After this, ‘Abd al-Malik was able to turn his attention to Iraq without any fear of disturbances along his lines of communication. The end, when it came, was swift. Muṣ‘ab was beset by troubles and resented by many Kūfans for the bloody way in which Mukhtār’s rebellion had been suppressed while al-Muhallab was entirely occupied fighting the Khawārij in the south. The caliph in person led the army which defeated Muṣ‘ab at Dayr al-Jāthalīq (72/691) and entered Kūfa to receive the homage of the people. In the south, al-Muhallab was reconciled without much trouble to the new regime. The final act was the despatch of a small force under al-Ḥajjāj b. Yūsuf to Mecca, where, after the bombardment of the city and the Ka‘ba itself, Ibn al-Zubayr was finally killed in Jumādā I 73/October 692. The unity of the community had been restored. Once again, the dissensions in Iraq had allowed the Umayyads and their Syrian supporters to dominate the richer and more populous areas to the east. It now remained for ‘Abd al-Malik to consolidate his gains and make his rule effective.

With the elimination of active opposition, ‘Abd al-Malik could pursue his policy of consolidating his authority. It seems that the events of the early part of the reign and the crisis which had affected the Umayyad caliphate since Mu‘āwiya’s death had convinced him that the decentralized system that Mu‘āwiya had introduced was not practical; under the control of a highly skilled manipulator like Mu‘āwiya, it could be effective but it was not a solid foundation for building a state. He began a policy of administrative and political centralization which was to provide a model for many later Muslim governments. This involved the development of a standard Arabic coinage. Up to this time, the coinage had been essentially regional: in Iraq and the east, the Muslims had begun to issue silver dirhams based on Sasanian models with a simple Muslim overstrike from a very early period. As the governor of Iraq under Mu‘āwiya, Ziyād b. Abī Sufyān had minted dirhams bearing his own name. This reflects the early development of the dīwān system of paying the military, which required large quantities of cash. In Syria, by contrast, there is no evidence that gold or silver was minted before the reign of ‘Abd al-Malik. There is some suggestion that copper coins (fulūs) were minted locally but even that remains uncertain. This probably reflects the slower development of systems of military pay in Syria and the absence of any centralized dīwān.

After ‘Abd al-Malik’s reforms, the old distinction between the silver (dirham) areas of the Sasanian Empire and the gold (dīnār) areas of Syria, Palestine and Egypt was maintained but from now on, after a few experiments with figurative designs, including a “standing caliph” portrait, all Arabic coinage was to be of a standard weight and design, bearing inscriptions but no images. The new gold dīnārs were introduced in 77/696–697 and the silver dirhams in 79/698–699. The gold dīnārs, which were probably produced in Damascus, carried pious phrases and the date but neither the place of minting nor the name of the ruler. The dirhams produced in a number of mints in Iraq and the east gave the name of the mints but not that of the caliph or governor. From the reign of the caliph Hishām, dirham production was centralized in Wāsiṭ, where the Syrian troops were stationed.

Along with this standardization of the coinage went an Arabization of the administration. From now on, all the administration in any part of the empire was to be conducted in Arabic, once again eliminating local differences and forcing those Persian and Greek officials who wished to continue to work for the government to learn the new ways. Nor was there any doubt that surplus taxation was to be forwarded to the treasury in Damascus; the Iraqis had lost the battle to retain the revenues of the Sawād in their own hands. All these activities served to introduce a unity and degree of control which had hitherto been lacking.

This administrative centralization reflected a broader political change. ‘Abd al-Malik’s authority was based on his own family and on the Syrian army. His own family had been employed from Marwān’s reign to govern different areas. His brother ‘Abd al-Azīz was the ruler of Egypt for almost the entire reign, from the time of its conquest from Ibn al-Zubayr in 65/685 to the year before the caliph’s death, keeping the province loyal and peaceful. Other experiments were not as successful; his brother Bishr b. Marwān does not seem to have been able to control Iraq, and as we shall see, family rule was not continued there. Another brother, Muḥammad b. Marwān, however, was the governor of the Jazīra from its first subjection in 73/692–693 until 91/709–710, some years after the caliph’s death. Here again, a key province was held by a member of the Marwanid family who secured its loyalty to the regime. One result of this was that the family itself tended to become divided and to represent different interests within the Syrian ruling group; Muḥammad b. Marwān, for example, became closely associated with the Qaysīs of the northern Jazīra and the Byzantine frontier region, while the caliph’s son Sulaymān, living at Ramla in Palestine, became identified with the Yamanī interest. These associations were later to lead to differences and divisions within the ruling house itself.

The Umayyad élite and their allies benefited from the revenues of landed estates known as qaṭā’i‘ (sing. qaṭī‘a), not to be confused with the iqṭā‘ or fief which became common in the Islamic world from the fourth/tenth century onwards. The settlement of the Arab conquerors in the time of ‘Umar b. al-Khaṭṭāb and his successors had envisaged that all the resources of the conquered lands should be helped in common as the fay’ of the Muslims, but from the earliest times members of the élite had sought to carve out private estates of their own. They were enabled to do this by means of a legal device, probably borrowed from Roman or earlier Mesopotamian land law. This allowed men who developed waste land and brought it under cultivation to claim it as their own property. It was heritable (they could pass it to their children), alienable (they could sell it) and defensible at law. Furthermore, these estates would pay only a tithe or ‘ushr rather than the much higher kharāj to which other lands were subject. This encouraged members of the élite to invest in irrigation and drainage systems, especially in the desert margins of the Fertile Crescent. So we see the early Muslim élite of Basra investing in canals to bring the barren lands around the city under cultivation, the sons of the caliph ‘Abd al-Malik creating new agricultural settlements along the Middle Euphrates valley and various members of the élite, including the Caliph Hishām and the future Caliph al-Walīd b. Yazīḍ, constructing farms and palaces along the fringes of the Syrian desert from Qasr al-Hayr al-Sharqī in the north to the plains of the Balqā east of Amman in the south. The desert quṣūr, such a characteristic feature of the Umayyad period, were at least in part a product of this favourable fiscal regime. After the coming of ‘Abbasid rule, new qaṭā’i‘ were rarely established in rural areas: by the third/ninth century, they had largely disappeared and with them the large-scale land reclamation which had been part of them.

The Syrian army also contained different groups. The Marwanids had been brought to power by the Yaman tribes under the leadership of the Kalbīs, who for the first seven years of the reign remained the mainstay of the forces. The agreements of 71/691 between the caliph and the Qaysī leader, Zufar b. al-Ḥārith at Qarqīsiyyā’, however, broke this monopoly of power. From this point onwards, ‘Abd al-Malik tried to balance the interests of these two groups, both at court and in the army. It was not an easy task. The feud between the Yamanī groups and the Qaysīs which had emerged after Yazīd I’s death was exacerbated by raids and counterraids, at first in the Jazīra and Palmyrena but later as far away as the Ḥijāz and Iraq. Tribes which were not previously involved, and whose genealogical affiliations were obscure, were drawn into this conflict whether they liked it or not. The Taghlib, for example, were a tribe which had lived for many years on the middle Euphrates, but in the aftermath of the conquest, they found their position threatened by newcomers from Arabia, the Banū Sulaym. The Sulaym were attached to the Qaysī group at Marj Rāhiṭ, and so the Taghlib, to defend themselves, were forced to look for protection from Kalbī leaders and became attached to the Yamanī cause. In this way, the pernicious feud which had previously been confined to Syria alone was spread through much of the Islamic world. The violence, raiding and counterraiding were almost impossible to stop; the best that the caliph could do was to keep it under control by threats and to pay the blood money demanded to prevent further retaliation. Like the activities of the Khawārij, the Qays–Yaman feud illustrated the problems of transition from a nomad to a settled society and policy.

Iraq of course remained the major problem, and it was here that ‘Abd al-Malik employed his greatest and most celebrated helper, al-Ḥajjāj b. Yūsuf al-Thaqafī. He was born in Tā’if in 41/661 and was thus some twenty years younger than the caliph and still a young man when he was called to high office. He worked his way up through the military police, the shurṭa, and came to the caliph’s attention during the preparations for the Iraq campaign of 71–72/690–691. After the fall of Kūfa, al-Ḥajjāj was sent to fulfil the unpopular task of reducing Ibn al-Zubayr in Mecca, and his efficiency and loyalty were shown to their fullest. ‘Abd al-Malik’s experiments with family rule in Iraq had not been a success, so in 75/694, he turned to al-Ḥajjāj and put him in charge of both Baṣra and Kūfa, as Ziyād b. Abīhi had been before. In 78/697–698, Khurāsān and Sīstān were added to his responsibilities, making him an effective overlord of all the eastern half of the caliphate. He did not administer Khurāsān directly but appointed the Azdī leader al-Muhallab, who had fought the Khawārij so well and now became a major force in the Umayyad east. As soon as he was appointed to Iraq, al-Ḥajjāj demonstrated his fierce determination to rule effectively. In this, he was helped by his Thaqafī connections, and while his origins are usually described as obscure, his mother was actually the divorced wife of al-Mughīra b. Shu‘ba al-Thaqafī, governor of Kūfa for Mu‘āwiya.

From the beginning, he seems almost to have goaded the Iraqis into rebellion, as if looking for an excuse to break them. Under pain of death, he forced the armies of Kūfa and Baṣra into the field, to that most unrewarding of tasks, combating the Khawārij. Furthermore, in an effort at economy, he reduced their ‘aṭā’, so that they were actually paid less than the Syrians. There was bound to be an explosion, and five years after his appointment it came. The issue was an old bone of contention, the reluctance of the armies of Kūfa and Baṣra to go on long campaigns, and in the year 80/699, al-Ḥajjāj obliged ‘Abd al-Raḥmān b. Muḥammad b. al-Ash‘ath to lead an expedition to Sīstān. Ibn al-Ash‘ath was the greatest of the ashrāf of Iraq and a symbol of their pride and independence. His family had done much to undermine the positions of both ‘Alī and Mukhtār. To his army were attached a number of other ashrāf, so much so that it was called the Peacock Army, from the wealth and display of its leaders. The campaign was hard, but more difficult for the leaders to bear were the constant rebukes and instructions from al-Ḥajjāj, and this overbearing attitude finally drove Ibn al-Ash‘ath to rebellion. He led his army west from Sīstān, through Fārs, and defeated al-Ḥajjāj’s forces at Tustar, 81/701, before sweeping on to establish himself in his hometown of Kūfa. Al-Ḥajjāj maintained himself in Baṣra with Thaqafī support and a Syrian garrison, but without massive Syrian aid, he was powerless to take the initiative.

In Kūfa, Ibn al-Ash‘ath was welcomed by all sections of the population, but there must have been many in the city who remembered how he and his family had betrayed the cause of Mukhtār, and the rifts soon began to show again. The caliph was seriously alarmed at the turn of events and, to al-Ḥajjāj’s disgust, offered terms. These included the dismissal of al-Ḥajjāj, the appointment of Ibn al-Ash‘ath to high office and the raising of the ‘aṭā’ of Iraqi troops to the Syrian level. Ibn al-Ash‘ath wanted to accept these terms but his more determined followers among the qurrā’ rejected them and he was forced to continue the fight. With the breakdown of negotiations, al-Ḥajjāj and the Syrians took the initiative again, defeating Ibn al-Ash‘ath in a series of battles, notably at Dayr al-Jamājim (Rabī‘ I 82/April 701), and dividing his supporters by generous offers of amnesty to all those who would lay down their arms. Ibn al-Ash‘ath himself was forced to leave Iraq and fled to the east, where he eventually tried to take refuge with the Rutbīl, the prince of Zābulistān in southern Afghanistan, and the man he had originally been sent by al-Ḥajjāj to fight, but his followers were dispersed and he himself died soon after. Once again, the divisions among the Kūfans and the determined use of Syrian military power by ‘Abd al-Malik and al-Ḥajjāj had frustrated the attempts of the Iraqis to recover their former status.

The aftermath of Ibn al-Ash‘ath’s rebellion saw the introduction of a new and more rigorous system of government in Iraq. The next year, al-Ḥajjāj established a permanent garrison of Syrian troops at a new military centre, Wāsiṭ, between Baṣra and Kūfa. Syrian troops seem to have developed effective military tactics: when faced by an enemy charge, they would get down on the ground and use their spears to present a bristling wall to the enemy. It required discipline, nerve and willingness to obey orders, but it proved effective against less well-organized opponents. Henceforth, these Syrian troops were to be the real ruling class in the area, and the Iraqis, whether ashrāf, qurrā’ or mawālī, were reduced to subject status. The revolt had been an attempt by the local leadership to assert its control and now that it had been defeated, al-Ḥajjāj seemed determined to break it for ever.

There was also a fundamental change in the fiscal system of the state, at least as it operated in Iraq. Originally, the ‘aṭā’ had been paid according to the role which the recipient, or his ancestors, had played in the Muslim conquests. It was, in fact, a reward for past service, and many in Iraq believed that it was their unalienable right. Al-Ḥajjāj was determined to change this. He saw clearly that this system meant that the caliph and his governor had very little executive authority and few rewards to offer their loyal supporters. He decided that the ‘aṭā’ should be a salary, a payment for serving in the army, rather than a reward for what one’s ancestors had done. People who did not fight would not be paid. Furthermore, this army was now largely made up of Syrian soldiers. This aroused the passionate hostility of many in Iraq, who claimed that the provisions laid down by ‘Umar and supported by ‘Alī were being destroyed by the impious Umayyads, but in the end al-Ḥajjāj’s vision triumphed: by his death, the ‘aṭā’ had become a salary paid for service, not a reward received by ancient right. Thus, the Muslim state developed a professional army and paid in cash salaries from the proceeds of public taxation, a structure familiar from the ancient world but virtually unknown in contemporary western Europe.

Al-Ḥajjāj and members of the Umayyad family, notably the caliph’s son Maslama, began a policy of large-scale land reclamations and development in central Iraq, digging new canals and bringing areas under cultivation, but the fruits of this policy went to the Umayyad family and their supporters and the profits were spent in Syria. Investment tended to be concentrated around the new centre of Wāsiṭ, while the lands of the ashrāf and the dihqāns around Kūfa were neglected by the government. Thus, both politically and economically, al-Ḥajjāj destroyed the power of the Kūfan élite; the family of Ibn al-Ash‘ath and others like them continued to enjoy a modest prosperity from the area but their political power was broken for ever. The people of Kūfa, which in the time of ‘Umar had been the greatest Muslim city of all, were now leaderless and cowed; only a radical and total revolution could reverse this unhappy state of affairs.

The partnership of ‘Abd al-Malik and al-Ḥajjāj lasted until the caliph’s death. Al-Ḥajjāj was not just the ruler of Iraq, but he effectively controlled Khurāsān and all the east, appointing governors from among his protégés. He had made this control absolute when in 84/704 he had secured the dismissal from Khurāsān of Yazīd b. al-Muhallab, who had succeeded his father as the governor in 82/701–702. Despite his affinity with the ashrāf, al-Muhallab had resisted all attempts of Ibn al-Ash‘ath to involve him in his rebellion and held the province for al-Ḥajjāj. Although they had proved their loyalty, however, the Muhallabīs were suspected because of their independent power base among the Azdī tribesmen and their great prestige. By 84/704, there seem to have been growing differences between them and al-Ḥajjāj, differences which the absolutist nature of al-Ḥajjāj’s power could not tolerate.

‘Abd al-Malik died in Damascus in Shawwāl 86/October 705 at the age of sixty. His reign had been a period of hard-won successes; he had established, in place of the decentralized Sufyanid system, a centralized, bureaucratic empire, dependent in the last resort on the power of the Syrian army. The Islamic community was sharply divided between the rulers and the ruled, where the ruled were no longer the conquered people and mawālī alone but included many Arabs, especially in Iraq and Khurāsān, who felt that the regime no longer represented their wishes or needs. It was in many ways an impressive achievement but was to prove something of a difficult inheritance for the later Umayyads. But ‘Abd al-Malik also left one legacy which we can still delight in today: the Dome of the Rock, the earliest great monument of Islamic architecture, constructed in Jerusalem on his orders and at his expense.

The sons of ‘Abd al-Malik and ‘Umar II – consolidation and division in the Marwanid caliphate: 86–125/705–743

‘Abd al-Malik was succeeded by his son al-Walīd, already acknowledged as heir since the death of his uncle, ‘Abd al-‘Azīz b. Marwān, the longtime governor of Egypt, in the previous year. Al-Walīd’s reign was in many ways a continuation of his father’s and a period of prosperity and peace in many parts of the empire. He instituted a system of poor relief and public charity in Syria and began many building projects, notable the great Umayyad mosque in the centre of Damascus. He also attempted to keep the balance, as his father had done, among the different groups in the Syrian élite. The rivalry for power between the Qaysī and Yamanī leaders was as intense as ever, and it is at least possible that the caliph kept it on the boil so that one faction should not acquire a monopoly of power. He himself had a Qaysī mother, Wallāda, and seems to have allowed the Qaysīs certain privileges, but other members of the Marwanid family attached themselves to the rival party; his brother and heir Sulaymān was the governor of Palestine and firmly allied to the Yamanī interest, while his cousin ‘Umar, son of ‘Abd al-‘Azīz b. Marwān, maintained the Yamanī contacts his father had built up in Egypt, and al-Walīd allowed him the prestigious position of governor of the Ḥijāz, which he used to afford a refuge to those who wished to escape al-Ḥajjāj’s power in Iraq.

The reign of al-Walīd (86–96/705–715) also saw the farthest extension of the geographical frontiers of the Umayyad state. In the west, Spain was invaded in 92/711 and was almost entirely taken by 97/716, while parties of raiding Arabs turned their attentions to southern France. In the east, the expansion was under the aegis of al-Ḥajjāj, who appointed governors whose most important function was to extend the frontiers. In Khurāsān, Qutayba b. Muslim launched in 86–96/705–715 from his base in Marv a series of attacks on the area beyond the Oxus, which the Muslims had hardly reached before. Thus, he secured the submission of Bukhārā in 87–90/706–709, Khwārazm and Samarqand in 91–93/711–712 and distant Farghānā in 94/713. This raiding did not, however, result in massive settlements of Muslims in the newly conquered areas; Qutayba made treaties with the existing rulers or, if necessary, replaced them by other local candidates but he did not enforce direct government. The effect of this was to draw into the Umayyad network a number of local kings who, while retaining their own power, paid tribute to and worked in harmony with the Arab governor in Marv. In Sind too, al-Ḥajjāj’s policy of expansion was pursued and he appointed a fellow Thaqafī, Muḥammad b. Yūsuf, to lead the Arab advance in the area. This policy of expansion had, since the time of the Prophet, served to absorb the energies of the Arabs and so preserve internal peace and provide an income and opportunities to those who might otherwise feel deprived. It was not, however, popular among all the participants. Conciliatory governors in Iraq from Ziyād’s time on had promised that the armies of the area would not be sent on distant campaigns against their will, and the revolt of Ibn al-Ash‘ath was at least in part a protest against the policy of expansion. Qutayba led his troops to many victories in the east, but when he tried to persuade them to support him in a revolt after the death of al-Walīd, they demonstrated clearly that these military activities had not won their loyalty and devotion.

Al-Ḥajjāj died in Ramaḍān 95/June 714 aged only fifty-five and the Caliph al-Walīd the next year, 96/715, setting the stage for a reaction against these policies. ‘Abd al-Malik had settled the succession such that al-Walīd would be succeeded by his brother Sulaymān, and despite some efforts to alter this arrangement, Sulaymān, a generous and easygoing prince, now duly became caliph. As has already been noted, Sulaymān had been the governor of Palestine, where he had come in close contact with the local Yamanī leaders, and his tutor and chief adviser was Rajā’ b. Ḥaywa al-Kindī, a religious figure attached to the Yamanī group who had served ‘Abd al-Malik as, among other things, the manager of the construction of the Dome of the Rock. Sulaymān had used his position to make allies among those who opposed al-Ḥajjāj and his policies, including the deposed governor of Khurāsān, Yazīd b. al-Muhallab, who sought refuge with him in his capital at Ramla. Sulaymān’s accession marks the return to power of the Yamanī group and the dismissal of many of al-Ḥajjāj’s nominees, including Qutayba b. Muslim in Khurāsān.

There has been much controversy about exactly what the Yaman Qays–Muḍar parties stood for at this time. They certainly represented rival groups within the Syrian élite and army which claimed a common descent. The Arab sources tend to portray the conflict as essentially a tribal rivalry, ‘asabiyya. It is not convincing to argue that they were really different political parties, the Qaysīs supporting expansion of the empire and an Arab monopoly on political power, while the Yamanīs were less imperialist and concerned to make links with non-Arab Muslims. In reality, the conflict was between two factions based on tribal loyalties, which sought to control access to military power and the privileges that went with it. The prizes were the favour of the ruling caliph and the lucrative governorships in the provinces. What may have begun as a struggle in Syria soon spread elsewhere: in Iraq, the Azd tribe was drawn into the Yamanī group, while their great rivals, the Tamīm, became the centre of the Qays–Muḍar group in the province. These rivalries were, in turn, transported to Khurāsān, where the participation of the Arabs in military forces meant that the rivalries were especially fierce and violent. By the reign of Hishām, if not before, it was virtually impossible for any political or military leader to avoid relying on one faction or the other.

Sulaymān’s reign was only two years (96–99/715–717). He had a reputation for luxurious living and certainly did not cultivate a pious image. His government appointments reflected his Yamanī leanings. Rajā’ b. Ḥaywa al-Kindī remained his chief adviser, while Yazīd b. al-Muhallab was appointed to Iraq, Khurāsān and all the east. He himself continued to live in Ramla in Palestine, a clear sign of his continuing commitment to the Yamanī party. The main business of the caliph was the conduct of the Holy War. Maslama b. ‘Abd al-Malik led the great expedition against Constantinople in 98–99/716–717, and Sulaymān himself died while on his way to lead an expedition against the Syrian frontier of the Byzantine Empire.

Sulaymān’s unexpected death meant a crisis for his supporters, since they felt that ‘Abd al-Malik’s other prominent sons, Yazīd and Hishām, were committed to their opponents. In this crisis, the succession was managed by the caliph’s adviser, Rajā’ b. Ḥaywa, who persuaded him, just before his death, to nominate his cousin ‘Umar b. ‘Abd al-‘Azīz instead.

It was thanks to Sulaymān’s testament and Rajā’s organization that ‘Umar II became caliph. He is the most puzzling character among the Marwanid rulers. The Islamic tradition is unanimous in seeing him as a complete contrast to the other rulers of his family; where they are godless usurpers, tyrants and playboys, he alone is represented as pious and ruling like a true Muslim. Recent scholarship has confirmed the view that his reign marks an important break, although from a rather different point of view.

‘Umar seems to have been determined to reintegrate the Muslim community, to try to break down the barriers between the Arab rulers and their non-Muslim subjects and between different groups of Arabs. Shaban has characterized this as an extreme Yamanī point of view, but this is not wholly borne out by the evidence. ‘Umar seems to have tried to get away from this sort of party involvement and appeal to a broader cross section of the community.

In appointing governors, always a good indication of a ruler’s policies, he made radical changes. In the east, he broke up the vast eastern governorate which had been developed by al-Ḥajjāj into different units, each responsible to him. Ibn al-Muhallab was deposed and arrested and in his place there were a variety of nominees; Kūfa was ruled by a member of the family of ‘Umar I, for example. Few of these people could be described as Yamanīs, and indeed ‘Umar II was responsible for appointing to the Jazīra a man who was to typify the hardline Qaysī attitude in years to come, ‘Umar b. Hubayra al-Fazārī. Rather than favouring one party, he seems to have chosen governors over whom he had control and whom he believed to be competent. His activities suggested that he intended to keep a close eye on provincial administration. He saw that a major cause of friction between the mawālī and the government, especially in Iraq, was the system of taxation, which, under al-Ḥajjāj’s iron grip, had meant that converts to Islam continued to be taxed as if they were unbelievers. This bred resentment, and it was to counter this that ‘Umar produced his famous fiscal rescript, an attempt to reform the financial administration. The most revolutionary aspect of this was that the mawālī were to be taxed as if they were Muslims, that is to say they were to pay only the ṣadaqa, or alms. However, they were not allowed to sell their land to Muslims, and on conversion their land became the property of their villages, that is to say that their lands remained kharāj lands, paying the full rate of land tax. In this way, the anomaly could be removed without doing great damage to the treasury. It was an ingenious scheme and it is not entirely clear how successful it could have been; the reaction which followed his death certainly meant that it was allowed to lapse. It did establish, however, the fundamental point of Muslim taxation: that kharāj land remained kharāj land no matter who owned it. Unbelievers were required to pay the jizya, or poll tax, now clearly separated from the kharāj, and this lapsed on conversion, but the payment of kharāj was unaffected by this.

In this way, ‘Umar attempted to resolve the tensions which threatened the community. He also forbade any further foreign wars and adventures. His character is difficult to assess. When he was in the Ḥijāz, he seems to have had a reputation for luxurious living, but on becoming caliph he adopted something of the austerity traditionally ascribed to his namesake ‘Umar I. This piety may well have been genuine, but it was the mark of a skilled and determined politician trying to find new solutions to long-standing problems.

This programme of reform was brought to an untimely end by his early death in 101/720. His policies had threatened too many interests and the Qaysīs were determined to reestablish the old order. His successor, Yazīd II, was a caliph in the old tradition, more interested in amusement than the day-to-day business of government and prepared to let the governors of the provinces have their own way. He was content to be ruled by Qaysī advisers, but not everyone in the empire was prepared to accept the change gracefully. Yazīd b. al-Muhallab was allowed to escape from prison and immediately raised the standard of revolt in Iraq, pledging a Holy War against the Syrian rule. Ibn al-Muhallab was perhaps the only indigenous Iraqi leader to have survived al-Ḥajjāj’s rule, and he immediately attracted a large following in Baṣra, where his own tribe was based and the provinces dependent on it. Then, he took the Syrian garrison town of Wāsit. and advanced on Kūfa, attracting some support from that city too. But not all Kūfans rushed to the standard, and there is no evidence that he received any support from the Yamanī party in Syria. He was opposed by one of the greatest soldiers of the age, Maslama b. ‘Abd al-Malik, a hero of the Qays and of numerous wars against the Byzantines. In Ṣafar 102/August 720, Yazīd b. al-Muhallab was defeated and killed on the field of battle, the last of the old-style Iraqi champions. For the final time, the divisions among the Iraqis had allowed the Syrians to triumph and to assert their control, once more, over the east.

Maslama remained only a short time as the governor, apparently because he had failed to send surplus revenue from the province to Damascus. His replacement was the Qaysī thug ‘Umar b. Hubayra, whom ‘Umar II had promoted to be the governor of the Jazīra and who proved violent and brutal in his pursuit of the Muhallabīs and the Yamanī leaders.

Perhaps fortunately for the stability of the Umayyad caliphate, Yazīd himself died in 105/724 after a reign as caliph of only four years, and he was replaced by his much more able, austere and conscientious brother Hishām. Hishām had been born in Damascus in 72/691 and was in his thirties when he became caliph. Little is known of his early career beyond his outspoken opposition to the accession of ‘Umar II, but his reign showed him to be shrewd, cautious and on the whole conciliatory. Like all the wiser and more successful rulers of his family, he attempted to stay above the Qays–Yaman feud and to avoid confrontation if possible. It is a measure of his success that the nineteen years of his reign were among the most peaceful of the entire Umayyad century, at least as far as internal opposition was concerned. In the all-important governorate of Iraq and the east, which now again took in Khurāsān and Iran as it had under Ziyād and al-Ḥajjāj, Hishām appointed a new man from the neutral tribe of Bajīla, Khālid b. ‘Abd Allāh al-Qasrī. Khālid does not have the reputation of his great predecessors for statesmanship or ferocity, but he was thoroughly competent and totally loyal, conscious that he owed his whole career to the Umayyads and considering that his interests were totally bound up with theirs. For fifteen years, he ruled Iraq and it was a period without any serious disturbances.

Hishām’s main problems were external invasion and the finances of the government. During his rule, the Islamic empire suffered three major challenges. The first of these threats was from the Caucasus area, where the Arabs’ control of the passes to the west of the Caspian was seriously threatened by the attacks of the Khazars, Turks from the steppes of southern Russia. To combat these threats, the Qaysīs of the Jazīra were mobilized, as well as local Armenian troops. Command was given first to the veteran general Maslama b. ‘Abd al-Malik and then from 114/732 onwards to another member of the ruling family, Marwān b. Muḥammad b. Marwān, a grandson of the Caliph Marwān. This command gave Marwān a unique opportunity to build up support among the Qaysīs and others in Armenia and the Jazīra and he became almost the only member of the ruling family with an independent military command. Armenia at this time was somewhat similar to parts of Khurāsān, in that the Arabs ruled through the native princes, or nakharars, who collected tribute and some of whose men were recruited for the army against the Khazars. Marwān took great care to remain on good terms with these local nobles, and ensured that their men received the pay owed to them. He also ensured the accession to leadership of the nakharars of his own protégé, Ashot the Bagratid, the ancestor of a long and famous line of Armenian kings, by expelling his enemies, the Mamikonian family. The campaign was difficult but in the end triumphant – the Khazars were driven back from the Muslim lands and in 120/738 Marwān raided as far as the Khazar capital on the lower Volga. After 121/739, he, with the help of his loyal Qaysī assistant, Isḥāq b. Muslim al-‘Uqaylī, concentrated on consolidating his power and influence in the area.

The second threat came from the Berbers in North Africa. The conquest of the Berbers had been very superficial, and while Arab garrisons had been established in some centres of what is now Tunisia, like Qayrawān, there was very little organized government. The Muslim presence was essentially predatory, slaves and booty being the main concern of Arab-raiding parties. As with the mawālī in Iraq, it seems that the beginnings of conversion to Islam made the Berbers’ grievances more acute, and this was brought home to them by missionaries from the Khawārij of Iraq who pointed out how the Umayyad government was exploitative and tyrannical. In 123/740–741, there was a massive rebellion which effectively drove the Arabs out. An expedition sent to retrieve the situation was heavily defeated and it was only with difficulty that the remnants of what had been a large Syrian army fought their way through to Spain, where they settled. In 124/742, another force, under a Kalbī leader, Ḥanẓala b. Ṣafwān, succeeded in restoring the position to some extent, and the rebellion ceased to be a danger. However, this disturbance did considerable damage to the state; Spain was now effectively cut off and the Umayyads maintained no military presence west of Qayrawān. What was more, a large number of Syrian troops had been lost and a further strain put on the Syrian military machine.

The third area of threat was Khurāsān. Here, as in the Caucasus area, the threat came from Turkish nomads, this time a confederation known to Western writers as the Turgesh, who were active in the areas beyond the Oxus and increasingly threatening not just the Arabs but the native people of the settled areas around Bukhārā and Samarqand as well. As the governor of the east, Khālid al-Qasrī first sent his brother Asad to attempt to do what Marwān later achieved in Armenia: to persuade the local people to fight alongside the Arabs against their joint enemies. The problem was more difficult here because the Arabs were reluctant to share any of their privileges with the native population, and in the end the government had to turn to more traditional means to restore the position. In 113/732, another 20,000 men from Iraq were sent, and in the end Asad was able to lead a combined force of Arabs and Iranians who defeated the Turgesh decisively in 119/737. The threat of invasion was lifted, but the immigration of more Iraqis and the mobilization of the local Iranian population were to lead to problems later.

The external threats were defeated but at considerable cost in money and manpower. Expansion had effectively ceased, except in France, where Arab-raiding parties from Spain continued to expand their range. On the Byzantine frontier, the regular summer raids continued but on a limited scale, and no more sieges of Constantinople were undertaken.

Apart from defence, Hishām’s other major concern was money. The financial administration of the Umayyad caliphate is very obscure. The sources tell us a good deal, not always clearly or comprehensibly, about who paid what taxes on what land. It is much less certain how far this money was used to finance central government. There remained a widespread body of opinion, especially in Iraq, which held that the revenue from taxation and the ṣawāfī, the government lands, was the property of all the Muslims and should be spent in the province where it was raised – the sending of taxes to Damascus was a very contentious issue. Governors who were in a weak position or anxious to curry favour with the local people, as had ‘Ubayd Allāh b. Ziyād at the death of Yazīd I in 64/683, always promised that this would not be done. As far as we can tell, even when it was done, only a very small proportion of the money collected was forwarded. ‘Umar II had agreed that the revenues of Khurāsān were to be spent in the province, while Maslama had been sacked from his position as the governor of Iraq for not forwarding any revenue. There seems to have been no elaborate central accounting organization of the sort that developed under the ‘Abbasids; the dīwāns or offices in Damascus dealt only with the collection of Syrian revenues. The caliphs were partly peripatetic, spending little time in Damascus itself. ‘Abd al-Malik and his entourage usually wintered at his palace of Ṣinnabra at the south end of the Sea of Galilee. In spring, he would go to Jābiya on the Golan Heights, then to Dayr Murrān on the western outskirts of Damascus. High summer would be spent in Ba’albak before he returned to Damascus and Ṣinnabra for the winter.

To supplement this meagre income, the Umayyad rulers continued large-scale irrigation and development projects on the sawāfī, and Hishām was one of the foremost organizers of this in both Syria and Iraq. Anecdotes in the sources portray him as having a real enthusiasm for horticulture, and a keen interest in the profits it could bring. In the city of Mosul, Hishām ordered that vast sums be spent on canal digging and erecting eighteen water mills. In southern Iraq, the development was undertaken by both Khālid al-Qasrī, who is said to have made 20 million dirhams out of his agricultural lands, and Maslama b. ‘Abd al-Malik, whose schemes were on an equally vast scale. The remains of numerous Umayyad palaces and castles in Syria show the extent of their developments in rural areas. As always, these activities benefited the government and the élite but were not always appreciated by the local people. A Syrian Christian commentator, Agapius of Manbij, reckoned that Hishām’s profits from his landed estates exceeded the profits from taxation of the empire.

The last years of the caliphate were clouded by concern over the succession and saw a slight but important shift of policy. By previous sworn agreement, the succession was to pass to Yazīd II’s son al-Walīd on Hishām’s death rather than to his own children. Hishām resented this, not just because he favoured his own line but because he had sons who were competent, sober and hard-working like their father. This al-Walīd was not; the playboy of the Umayyads par excellence, he was a talented poet, an enthusiast for architecture (palaces, not mosques) and a heavy drinker. Hishām’s attempts to change the arrangement were widely opposed within the Umayyad establishment, even by his closest political friend, Khālid al-Qasrī, and they came to nothing. It was perhaps in order to ensure stability after his death that Hishām seems to have come to rely more on the Qaysīs. Khālid al-Qasrī was deposed from Iraq in favour of Yūsuf b. ‘Umar, a protégé of al-Ḥajjāj’s and a member of the Thaqafī tribe, which had played such a large part in Iraqi politics. Khālid al-Qasrī may have had Yamanī sympathies; certainly, he was later claimed as a Yamanī. Yūsuf, however, was a determined and outspoken Qaysī. The change in Iraq was followed by one in Khurāsān, where again a committed Qaysī, the veteran military leader Naṣr b. Sayyār, was appointed (both changes in 120/738). Yūsuf made many enemies – at least in part because of his vindictive attitude towards Khālid and his family – and his brief spell as the governor saw the only real disturbance in Iraq during this period. In 122/740, there was an armed revolt in Kūfa led by Zayd b. ‘Alī, a grandson of al-Ḥusayn who had arrived in the city from the Ḥijāz. The Kūfans promised large-scale support and Zayd agreed to put right all the old grievances, the Qur’ān and Sunna would be respected, the ‘aṭā’ would be paid to all, the fay’ (i.e. the revenues of the Sawād) would be distributed to the local Muslims, the weak protected and there would be no more distant campaigns. And of course there would be a caliph from the house of the Prophet with his rule based in Kūfa. Despite this popular appeal, the revolt came to nothing; the Kūfans deserted Zayd and he was easily killed by the Umayyad forces, but the revolt shows how all the old grievances were still alive.

The collapse of the Umayyad caliphate: 125–132/743–750

Hishām died in Rabī‘ II 125/February 743. He left a caliphate prosperous and secure. It is true that the differences between Qays and Yaman were not healed and that the majority of the people in Iraq and points east, both Arabs and mawālī, remained unenthusiastic about the regime. It is true too that the recent wars in the West had put an additional strain on the Syrian army. It should not be thought that collapse was inevitable or that another Hishām could not have sustained the empire. Such, however, was not to be the case. Al-Walīd II fulfilled Hishām’s worse fears. His conduct was increasingly irresponsible; he seldom went to Damascus and paid little heed to normal administrative affairs, while squandering vast sums on ambitious if seldom finished palace projects. He left the administration the Qaysī advisers Hishām had bequeathed to him, and they took the opportunity to assert their position. Besides the caliph’s generally negligent conduct, as offensive to many Muslims as Hishām’s had been pleasing, two major affairs brought resentment to a head.

The first of these was the question of the succession. Spurred on perhaps by the evil genius of Yūsuf b. ‘Umar, he was at once determined to seek the bay‘a for his two young sons, al-Ḥakam and ‘Uthmān. There was immediate opposition in Syria, and al-Walīd took the most drastic action against members of the Marwanid family; some were executed, while Hishām’s son Sulaymān was flogged and exiled to ‘Umān. Khālid al-Qasrī, now in retirement in Damascus, refused to swear allegiance to the children and for this he was handed over to his enemy Yūsuf b. ‘Umar, who tortured him to death.

According to the historian al-Ṭabarī, al-Walīd arrested his cousins and turned the Yamanīs against him and so was responsible for his own downfall. Certainly a bare year after his accession, he had made himself so unpopular that there was a serious discussion of removing him by force. It must be emphasized that this was very unusual in the Umayyad family. Despite their differences, the princes usually displayed considerable family solidarity, and Hishām’s respect for his brother Yazīd’s testament is simply one example of this. The initiative was taken by a son of al-Walīd I called Yazīd. He developed the radical line, normally taken by Khārijīs, that al-Walīd’s sins were so great that he was no longer fit to be caliph. He was joined by some members of Yamanī tribes anxious to regain their lost position, notably Manṣūr b. Jumhūr al-Kalbī, described as an uncouth bedouin. Many more Umayyads and others were extremely reluctant to support him, however great al-Walīd’s crimes. The caliph’s negligence and the drunkenness of the chief of police allowed the rebels to take Damascus, but it was only with difficulty and considerable expenditure that they were able to persuade an army to search out al-Walīd. True to form, he was in a desert castle at Bakhrā’, about 20 kilometres south of Palmyra, almost defenceless. It was there that he was killed by the rebels in Jumādā II 126/April 744 – murdered, like the Caliph ‘Uthmān, as he sat reading the Qur’ān.

Just as the murder of ‘Uthmān led to the destruction of the Medinese regime of the Rāshidūn caliphs, so the death of al-Walīd led directly to the breakup of the Umayyad caliphate. It re-opened, in a much more vicious form than ever before, the feud between Qays and Yaman, which successive caliphs from ‘Abd al-Malik on had tried to control. It would be wrong to imagine that all members of these two groups were implacably hostile; it would seem that the violence was begun by extremists like Yūsuf b. ‘Umar for the Qaysīs and Manṣūr b. Jumhūr for the Yamanīs, but once it had begun, it was very difficult to stop and came to involve the whole Syrian army and political élite. It was this fatal division, more than anything else, which destroyed Umayyad government.

On his accession, Yazīd III made a number of commitments to redress grievances, almost identical to those that Zayd b. ‘Alī had made in Kūfa, but including the provision that he was to be deposed if he did not fulfil his promises. He also recalled Sulaymān b. Hishām from exile and appointed the Kalbī leader Manṣūr b. Jumhūr as his governor in the east, but he soon replaced him with ‘Abd Allāh, son of ‘Umar II. It is difficult to know whether Yazīd III was a genuine reformer in the spirit of ‘Umar II or simply an adventurer trying to win popularity. In Iraq, ‘Abd Allāh b. ‘Umar seems to have tried to win over powerful Kūfans by promising them salaries and a position in the army in exchange for their support. Yazīd III died after a reign of only six months, and his reforming ideals, if such they were, hardly had time to mature. His place was briefly taken by his unconvincing brother, Ibrāhīm, but he was unable to withstand the new strongman of the Qays, Marwān b. Muḥammad. Marwān had remained the governor of Armenia and Āzarbayjān, grudgingly accepting the changing rulers but keeping his power base intact. He now proclaimed himself the avenger of al-Walīd II and came south with his seasoned troops; the attempts to resist him by Ibrāhīm and Sulaymān b. Hishām, who seems to have recruited a private army of his own from among the mawālī, were futile, and in Ṣafar 127/December 744, he was proclaimed caliph in the mosque in Damascus.

In many ways, he was an excellent choice. Known as al-Ḥimār (the Ass) for his strength and perseverance, he was a tough military veteran, who might have been able to restore the position, but there were too many factors against success. He was not a member of the ruling branch of the Marwanids and furthermore his mother had been a Kurdish slave, not a freeborn Arab. Whether this was really important is difficult to say, but it gave his enemies a reason to attack him. On top of this, he had come to power with the support of only a small part of the Syrian élite, the Qays of the Jazīra and the Byzantine frontier. He had seized the capital by force and driven the Yamanī leaders out of Syria. That there would be no place for them in the new regime was confirmed when he moved himself, the court and the treasure north to Ḥarrān in the Jazīra, to his own territory. At first, he was conciliatory in Syria, allowing the junds to choose their own leaders, but after a series of rebellions, especially in the Yamanī centre of Homs, he resorted to bloody repression. The Kalbīs were finally humbled and by the end of 129/746, he had established a kind of peace.

In Iraq, the situation was extremely confusing and the details do not immediately concern us. In general, two main opposition movements appeared, both of them familiar in many ways from the earlier days. In the Jazīra, there was a rebellion of Khārijīs led by the Shaybānī leader al-Ḍaḥḥāk b. Qays, while in Kūfa, there was a rising in favour of a member of the ‘Alid family, ‘Abd Allāh b. Mu‘āwiya. What distinguished these movements from similar ones in the past was not that they were more powerful or threatening but rather that the Umayyad response was confused and divided. ‘Abd Allāh b. ‘Umar had tried to forge an alliance of Yamanī groups from Syria and the Yamanīs of Iraq and Kūfa, and this group was bitterly opposed to Marwān and all he stood for. So hostile were they that their leaders were prepared to abandon the Umayyad cause altogether and side with their enemies. Sulaymān b. Hishām joined the Khārijīs, while Manṣūr b. Jumhūr and ‘Abd Allāh b. ‘Umar ended up with Ibn Mu‘āwiya. Nothing could illustrate more graphically the breakup of the Syrian élite.

Marwān was determined to press ahead with his conquest of Iraq, despite troubles in Armenia, where his protégé Ashot the Bagratid had been deposed by his rivals. Armenia was held by his devoted lieutenant, the Qaysī leader Isḥāq b. Muslim al-‘Uqaylī, while Marwān sent another leading Qaysī, Yazīd b. ‘Umar b. Hubayra, son of Yazīd II’s governor of Iraq, to attack Kūfa. Marwān’s campaign showed the effectiveness of his troops and his own military skills. In a final last-ditch resistance, all the elements opposed to his rule – Ibn Mu‘āwiya, the Khawārij, Syrian and Iraqī Yamanīs and even ‘Abbasid supporters – came together for a combined stand. But Marwān’s general Ibn Ḍubāra was too strong for them and they were scattered. By the spring of 130/748, peace was restored and Marwān II was caliph, while Yazīd b. ‘Umar b. Hubayra in Wāsit. was his governor of the east, but it was the peace of exhaustion, the peace of the desert. Plague and famine and a terrible earthquake followed in Syria in the wake of the wars. It was a devastated country. Power now lay in the hands of men from the Jazīra and the mountains of Armenia. Recovery might have been possible even yet, but there was a cloud on the horizon in the shape of a rebellion in distant Khurāsān, led in the name of the Family of the Prophet by an obscure mawlā called Abū Muslim. By Jumādā II 130/February 748, Abū Muslim had established himself in the Khurāsānī capital at Marv, having driven out the Umayyad governor, Naṣr b. Sayyār, and was sending his armies to the west to challenge the Umayyads. Marwān was loyally supported by his Qaysī army. From his base in Wāsiṭ, Ibn Hubayra despatched armies under Nubāta b. Hanẓala al-Kilābī and ‘Amir b. Dubāra al-Murrī to check the advance of the rebels on the Iranian plateau, but both were defeated by the opposing general, Qaḥṭabā b. Shabīb. In Iraq itself, the resentment against the Qaysī government of Marwān once more made itself felt; in Baṣra, a member of the famous Muhallabī clan, Sufyān b. Mu‘āwiya, made an unsuccessful attempt to take the city, while Kūfa was held for the rebels by Khālid al-Qasrī’s son Muḥammad. Alienated members of the Umayyad élite played a significant role in the triumph of the rebels in Iraq. By Rabī‘ I 132/October 749, the rebel army had occupied Kūfa, and a rival caliph, the ‘Abbasid Abu’ al-‘Abbās al-Saffāḥ, had been proclaimed. Only in Wāsit. did Ibn Hubayra continue to hold out.

Faced with this new crisis, Marwān summoned his most loyal Qaysī supporters to march in Iraq but the ‘Abbasid armies were ready to meet him. It seems that the ‘Abbasids were heavily outnumbered but they still defeated Marwān at a battle on the river Zāb in Jumādā II 132/February 750, a battle which effectively broke Umayyad power forever. Marwān fled west, but the Syrians refused to rally to a man who had inflicted so much damage on their land. In Damascus, in fact, the Yamanī party opened the gates to the pursuing ‘Abbasid forces. Marwān fled on to Egypt, perhaps hoping to reach north Africa, but at the end of Dhu’l-Ḥijja 132 (August 750), just six months after the defeat on the Zāb, he and his followers were surprised in the village of Būṣir, south of Fuṣtāt, and the last Umayyad caliph was killed as he had lived, fighting.

The fall of the Umayyads can be explained in many ways. On an ideological level, they failed because they could not offer the sort of leadership which many Muslims wanted. It used to be accepted that the Umayyads claimed only secular authority, but Crone and Hinds in their book God’s Caliph have demonstrated that the Umayyad caliphs did claim a religious authority; the ruler was God’s Caliph (khalīfat Allāh) and had the authority to make decisions about Islamic law and practice. However, there were many Muslims, especially in Iraq, who felt that charismatic, truly Islamic leadership was necessary to establish the rule of the Qur’ān and Sunna. By the end of the Umayyad period, it had become an article of faith among such people that only the Family of the Prophet could supply this authority.

There were also regional problems. From ‘Abd al-Malik’s reign onwards, Umayyad government had increasingly meant Syrian government. Despite attempts by ‘Umar II and others to broaden the base of the regime, the Muslims of Iraq were entirely excluded. This narrowness of support became even more pronounced with the Qaysī triumph under Marwān II; at the end, even Syria and Palestine were conquered territories and Damascus had been replaced by Ḥarrān in the Jazīra as the Umayyad capital. This restricted nature of support for the regime was made more serious because neither Syria nor the Jazīra was as rich or had such large Muslim populations as Iraq. In the second half of the eighth century, the revenues from the alluvial areas of southern Iraq amounted to four times those from Egypt and almost five times the revenues from the whole of Syria and Palestine. Constant warfare had certainly drained the resources of manpower in Syria. The wars of Hishām’s reign against the Berbers and the internecine disputes which followed his death must have placed a considerable strain on the manpower of the Qaysī tribes who supported the last Umayyad. In addition, Marwān’s policies had spread disaffection, not just among elements traditionally hostile to the regime but among people who had previously been loyal servants, like the family of Khālid al-Qasrī and even members of the Umayyad house itself, like Hishām’s own son Sulaymān. In these circumstances, it is hardly surprising that the Umayyad state was swept away.

In the final judgement, however, it would be wrong to imagine that the fall of the dynasty was inevitable. The Umayyad regime had never been as strong as it had been under Hishām only a decade before the final collapse. It was only the failure of leadership and murderous conflicts which followed his death which led to disaster, and even at the end Marwān’s Qaysī supporters could raise very formidable armies to oppose the ‘Abbasids.

Julius Wellhausen, in the title of a famous book, described the Umayyad regime as Das arabische Reich, the Arab empire, and held that its passing marked the moment when the Arabs lost control of the movement they had created and the leadership of the umma passed to a new élite, some of Arab origin, many others of Iranian or, slightly later, of Turkish descent. Despite the challenges to this picture, it remains basically intact; no one who could not claim Arab descent played a leading role in Umayyad politics or court life, although the talents of non-Arabs in financial or agricultural administration were certainly used. But it was an Arab kingdom in another important sense as well; it was the period when the Arabic language came to dominate the Near East, not in the sense that the majority of the population became Arabic-speaking but in the sense that it became the language of bureaucracy, high court culture and, above all, the religion of the ruling class. The dominance of Arabic was bound up with the dominance of Islam, which retained its identity and separateness in a society where there were numerous ancient and highly developed religious traditions. To see how important this achievement was, we have only to compare the fate of the Arabs and their civilization with that of the Germanic conquerors who came to rule most of the Western Roman Empire from the fifth century onwards.

This linguistic and religious ascendancy was not achieved by mass conversion. Even at the end of the Umayyad period, Muslims still formed a very small proportion of the population of the Near East, and it was probably only in parts of Iraq and Khurāsān that conversion to the new religion had made significant progress. Nor was it achieved by forcible repression. One of the most striking features of the history of the period is the almost total absence of anti-Muslim rebellions. Christians, Jews and Zoroastrians seem to have accepted their role as subject peoples with comparative equanimity, and it was not until the pace of conversion began to increase in the early ‘Abbasid period that violent protests developed in parts of Egypt and Iran. Much of this passivity can be put down to the role of these communities before the coming of Islam. The Monophysite Christians of Syria and Egypt, as well as all the Christians of the old Sasanian Empire and the Jews everywhere, had been a demilitarized subject population under the old dispensation. Much of the previous ruling class, the natural leaders of any opposition, had disappeared. At the conquest, many of the Byzantine officials had emigrated to Constantinople or further west to Sicily and Italy, while others had either become converted to Islam, e.g. the scions of the Ghassanid clan, or reached important administrative positions in the government, e.g. the family of St John of Damascus. The disappearance of the élite of Sasanian Persia is more puzzling, since, apart from the mountainous areas of northern Iran and Khurāsān – remote and uninviting districts – they had nowhere to flee to. Many may have lost their lives in the battles of the conquests, many more lost their castles, their lands and their lives in the long struggle to subdue the province of Fārs, the native territory of the Sasanian regime. Others must have been assimilated and become Muslim, perhaps making careers in the bureaucracy, even while remaining conscious of their descent from the Iranian aristocracy. In the fourth/tenth century, under the Buyids, it once again became a sign of status to claim descent from great Iranian families, and when we find in the same period a Fārsī bureaucrat with the Sasanian name of Sābūr (Shāpūr) b. Ardashīr, it is hard to believe that he was not of aristocratic or even royal lineage.

In some ways, the Arabs were in a similar position to the Germanic ruling minorities in western Europe, who equally dominated much larger populations whom they allowed to maintain their own customs and beliefs as long as they paid the required taxes. Perhaps we can see in St John of Damascus a parallel with those Roman officials (like the philosopher Boethius) who made it their business to serve the Ostrogothic kings of Italy, providing a link between the old regime and the new. In Iraq, the dihqāns, who had administered much of the country for the Sasanians, now did so at a local level for the Muslims, and once again bureaucratic practice passed from one empire to another; princes and generals might come and go, but those who understood the financial administration could always find employment. Visual evidence of this continuity and adaptation can be seen in early Muslim architecture. The Dome of the Rock in Jerusalem and the Great Mosque in Damascus build on and develop Byzantine architectural techniques, and the Umayyad palace at Khirbat al-Mafjar in the Jordan valley looks back to the Roman villa, just as Theodoric’s work at Ravenna develops the styles employed when the city was a Roman capital.

This said, however, the Umayyad state was vastly different in many ways. The most important feature of this was a strong cultural self-confidence, born of the possession of Islam and the Arabic language in which God had communicated the Qu’rān. The Germanic rulers of early medieval Europe craved Romanitas, sought to integrate themselves in the culture, language and beliefs of the empire they had entered; Franks, Ostrogoths and Lombards alike abandoned their traditional beliefs and languages to acculturate to the Roman norm. It is strong evidence for the early maturity of the Arab-Islamic tradition that the Muslim conquerors felt no need to do this. The Germanic conquerors of the west abandoned their pagan or Arian beliefs in the face of the opposition of the subject peoples. Despite the vitality of Monophysite Christianity in Syria, Egypt and northern Iraq, the Muslims felt no need to accommodate themselves to the beliefs of the majority – indeed all the traffic was the other way, and it would seem that the more dynamic and ambitious elements in the Christian population soon began to be converted. Nor was there any linguistic imitation; the Arabic the Muslims used was a high-culture written language, the language of the Qur’ān, of a classical poetic tradition and very soon the language of administration and prose narrative. No one wrote Ostrogothic; when Paul the Deacon wrote a history of the Lombards or Gregory of Tours a history of the Franks, they naturally did it in Latin; when Ibn Isḥāq or Abū Mikhnaf wrote a history of the Prophet and of the caliphs in late Umayyad times, they would have not considered doing it in any other language than Arabic. Arabic could never have supplanted the indigenous high-culture and bureaucratic languages unless it had a highly developed written and literary tradition of its own. Where the language of government and culture led, the vernacular soon followed; Greek, Aramaic, Coptic soon ceased to be spoken languages and only in Iran did the language of the conquered, greatly altered in script, grammar and vocabulary, survive to reemerge in the tenth century as a written tongue.

This cultural and linguistic self-confidence was perhaps aided by the fact that the Muslims were heirs to not one but two cultural traditions, the Byzantine and the Sasanian, to say nothing of the Syriac tradition of Monophysite Christianity and the highly developed intellectual and cultural life of Babylonian Jewry. For the invaders of the Western Roman Empire, there could be only one language of learning and writing, and one religious tradition, Catholic Christianity, and it was very difficult to escape from the spell of this all-pervasive culture. For the Muslims in the Near East, however, there was no such universal norm; indeed, the multiplicity of languages and cultures meant that Arabic, far from being overwhelmed, actually became essential; like English in India, it was the only common language in which the peoples of a vast empire could communicate with each other.

The establishment of this culture and civilization owes everything to the Umayyad caliphs. If the Islamic world had broken up, as seemed possible after the death of ‘Uthmān, it would have lost not just its political unity but also its cultural and perhaps even its religious identity as small, isolated groups of Muslims became absorbed into the societies of those they had conquered. The Umayyads seized the high ground. They made Islam the religion of a cultured court and an imperial administration. They made Arabic the language of literature and commerce. After the reforms of ‘Abd al-Malik, the Copt, Aramaean, Kurd or Iranian only had to look at the coins in his purse to see the proclamation of the religion and language of the élite. Many Muslims, at the time and later, have criticized the Umayyads for betraying true Islam. To an outside observer, it seems equally clear that Arab-Islamic civilization could never have developed without the achievements of great politicians like Mu‘āwiya, ‘Abd al-Malik and Hishām in holding the Muslim world together during the crucial formative century.

It is difficult to discern any clear lines in Umayyad relations with the world beyond the Muslim frontiers or to talk with any confidence of a “foreign policy”. Early Muslims divided the world into two zones, the Dār al-Islām (House of Islam) and the Dār al-Ḥarb (House of War), which the Muslims should attack whenever possible. Rather than peace interrupted by occasional conflict, the normal pattern was seen to be conflict interrupted by the occasional, temporary truce (hudna). True peace (ṣulk.) could only come when the enemy surrendered and accepted Islam or tributary status. This vision of the world was not in conflict with observable reality. Apart from the distant world of Tang China, the Muslims did not come into contact with any powers which were their equals in terms of wealth or military power and, like the Romans, they could afford to regard all others as barbarians. It was only towards the end of Umayyad rule that the Byzantines were able to face the Muslims on anything approaching an equal basis. There was therefore little incentive for sending diplomatic missions to establish a modus vivendi with other powers or to try to establish permanent relations with peoples who were soon to be defeated and incorporated within the Dār al-Islām.

The Umayyad period was a bleak time for the Byzantine Empire, when the iconoclast emperors concentrated all their energies on ensuring the survival of the state. The frontier provinces, devastated by war, were lands of ruined cities and deserted villages where a scattered population looked to rocky castles or impenetrable mountains rather than the armies of the empire to provide a minimum of security. The position was very similar at sea, where the Mediterranean was almost entirely a war zone rather than a commercial highway. There are some small signs of cultural and commercial interchanges – the mosaic workers from Byzantium who came to decorate the Great Mosque in Damascus for the Caliph al-Walīd, for example. But by and large, fruitful diplomatic and commercial exchanges were limited before the more settled conditions of the ‘Abbasid period.

With the contemporary Christian powers of Europe – the Lombard kingdom of Italy and the late Merovingian and early Carolingian rulers of France – the Umayyads had no communication at all. It is true that Muslim raiders attacked the Mediterranean shores of both these areas and that Arab expeditions from Spain, which was largely conquered by the Muslims from 92/711 to 97/716, penetrated deep into France, where they were finally defeated by the Carolingian Charles Martel at the battle of Poitiers in 114/732. Muslim raiders maintained themselves in their strongholds in the south of France for two centuries after this. But these were essentially unofficial raids; they led to no government or commercial contacts and the two societies existed in almost total ignorance of each other. Only the occasional pilgrim narrative, like those of Arculf in the first/seventh century and St Willibald at the beginning of the second/eighth would give any indication of the existence of the early caliphate to western Europe, and the writers were more concerned with the Christian Holy Places than with the political condition of the country through which they passed.

More positive were the relations between the Muslims and the Christian people of Nubia, to the south of Egypt. At first, the Muslims had attempted to conquer this area, but resistance proved very fierce and in 31/651–652 they concluded a treaty which allowed the Nubians to live at peace with their Muslim neighbours and obtain food from them on several conditions, notably that they paid a tribute of slaves. This treaty seems to have remained in force throughout the early Islamic period and proved something of a puzzle to Muslim jurists, since Nubia was clearly part of the Dār al-Ḥarb and yet the Muslims had signed a permanent and lasting peace treaty with its people, a situation unknown elsewhere on the frontiers.

In contrast to the Mediterranean, the Indian Ocean saw the extensive development of commercial links in the Umayyad period. In Sasanian times, ships had certainly sailed from the Gulf to India and probably to China, but under the Umayyads this traffic seems to have expanded greatly and ports like Ubulla and Baṣra in southern Iraq grew rich on the products of the trade. This, in turn, led to Muslim commercial settlements overseas, first in Sind (southern Pakistan), which was incorporated as a province of the Umayyad and early ‘Abbasid caliphates, then later in east Africa and even as far as distant China, where in 141–142/758, clashes between Muslims and the local people in Canton resulted in the destruction of much of the city. Not only did these Indian Ocean voyages introduce the Islamic world to new products, such as spices and aromatics from Indonesia and Zanzibar and porcelain from China, but they also provided a setting for imaginative adventure, from which the Sinbad the Sailor stories were to grow.

The northeastern frontiers of the caliphate were essentially a war zone. In antiquity, the overland “silk route” had been the main means of communication between the West and China, but in early Islamic times, it seems that this trade had largely dried up. This was partly because silk was now produced in the Near East – and transporting porcelain thousands of miles on the backs of pack animals was hardly a practical proposition – and partly because the area was now disturbed by Turkish nomads raiding the settled areas. It was just after the close of the Umayyad period that central Asia saw the first and only clash between Muslim forces and the army of the Chinese empire under the strong and expansionist rule of the Tang dynasty. In 133/751, Muslim and Chinese armies met at Talas beyond the Jaxartes river (Syr Darya) and the farthest reaches of Muslim settlement. The Muslim armies were victorious but did not pursue their enemies into the fearsome wastes of central Asia. Like the battle of Poitiers twenty years before, the battle at Talas has symbolic importance, marking the end of Arab Muslim expansion in the northeast of the Iranian world. As in the west, the military encounter was not followed up by any diplomatic interchange, and Muslim caliph and Tang emperor remained entirely ignorant of each other.

In a real sense, the Umayyad caliphate, like the Roman Empire, was a self-contained social, cultural and economic unit. The distances within it were so vast and there was such diversity within its frontiers that Muslims did not have the opportunity or the incentive to concern themselves with matters beyond those frontiers. There in the twilight zone of the Dār al-Ḥarb were numerous obscure and barbarous peoples like the Franks and Anglo-Saxons, but their lands were too poor, their merchants too few and their armies too weak to attract the attention of the rulers of Islam. This self-sufficiency sometimes makes early Islamic history seem rather remote to outsiders, as if all the events were somehow happening on a different planet, and this impression of separateness is probably correct. Nor was it confined to the Umayyad age. Under the ‘Abbasids, closer relations were developed with Byzantium, but it was not until the Fatimids opened up the Mediterranean for shipping from the end of the fourth/tenth century that the affairs of western Europe and the Near East began to interact once again. It was the custom of the Arab Christian annalist of Alexandria, Sa‘īd b. al-Biṭrīq to mention the names of all the patriarchs of the great ancient sees of Christendom, Alexandria, Jerusalem, Antioch, Constantinople and Rome. The popes of Rome were duly recorded up to the time of Agatho (678–681), when information was no longer available, and it was not until the early fifth/eleventh century that Sa‘īd’s continuator, Yaḥyā b. Sa‘īd, was once again able to give an up-to-date pope’s name. In antiquity, and again in the High Middle Ages, the voyage from Italy to Alexandria was a commonplace; in early Islamic times, the two countries were so remote that even the most basic information was unknown, and if the affairs of Rome were entirely obscure to the patriarch of Alexandria, they were much more so to the caliphs of Damascus. The early Islamic world was essentially self-contained and it was against this background of cultural self-sufficiency that Islamic civilization developed.