chpt. 12 of his Medieval Persia: 1040-1797, Longman, New York, 1988, pp. 112-123.
Safawid rule over Persia is conventionally dated from Shah Isma'ils capture of Tabriz in the aftermath of his victory over the Aq-Qoyunlu ruler Alwand at Sharur in 907/1501. But there was still a very long way to go before Isma'il could be regarded as anything more than a potential successor to the Aq-Qoyunlu in Azarbayjan. Nor, for some years, was the geographical shape of the new state by any means clear. It may be that Isma'il's expectation was that he would be able to set up an essentially Turkmen empire after the Aq-Qoyunlu pattern, consisting of eastern Anatolia, Azarbayjan, westem Persia and Iraq. After all, the military following on which he depended was Turkmen in composition, he had fixed his capital at Tabriz, the now traditional Turkmen centre on the periphery of Persia proper, and he may have seen himself as in some sense the legitimate successor to his Aq-Qoyunlu grandfather, Uun Hasan.
The direction of Isma'ill's early campaigns certainly suggested that it was the Turkmen heritage he was primarily interested in. The defeat of Alwand had not finally ended the danger from the Aq-Qoyunlu. With Azarbayjan secure for the time being, the Safawid forces marched south, where they met and defeated Murad Aq-Qoyunlu at Hamadan in 908/1503. This led to the establishment of Safawid rule in westem Persia. In 913/1507 Isma'il turned his attention back towards the west, to the old Aq-Qoyunlu heartland in Diyarbakr and their former centres in the cities of Amid and Mardin. This advance into eastem Anatolia meant that Safawid power now reached as far as the frontiers of the Ottoman Empire, but for the time being a dash with the Ottomans was avoided.
Next it was the turn of Iraq. Baghdad was taken in 914/1508, and thereafter the Musha'sha', a Shi'ite power in Khuzistan in the south-west of Persia, whose religious notions were dangerously similar to those of the Safawid order, were brought to submission. By 914/1508, then, Shah Isma'il was effectively the ruler of most of the territories that had constituted his grandfather's Turkmen empire. But the year 916/1510 saw what was to prove a decisive shift towards the east. In Khurasan the Timurids were no more: since 911/1506 the real enemy was Muhammad Shaybam Khan, ruler of the Ozbegs, who were to remain dangerous eastern adversaries of the Safawids at least until the end of the sixteenth century.
By the late fourteenth century the tribes of the steppes west and north of the Caspian, in what is now Kazakhstan, had become known to Muslim writers as Ozbegs. This name may have had some connection with the great fourteenth-century Khan Ozbeg of the Golden Horde, though other explanations have been offered such as that they believed themselves to have had a hundred (oz) chiefs (beg). In the aftermath of Temur's defeat of Tokhtamish in 797/1395, factional strife fractured the Golden Horde. Among the Ozbeg groupings that arose at this time was one headed by the descendants of Shiban (the name was later Arabicized as Shayban), a brother of Batu, founder of the Golden Horde. Another Ozbeg group was known as the Kazakhs (the word "Cossack" is derived from this term).
By 831/1428 a certain Abu'l-Khayr, who was the chief Shaybanid prince, had succeeded in unifying a number of Ozbeg tribes. Following the usual practice, he at once embarked on a career of territorial expansion. He attacked the Timurids to the south, invading Khwarazm and sacking the city of Urganj in the 1430s and occupying theJaxartes River area, in the east of Transoxania, in 849-50/1446. He moved his capital to the city of Sighnaq, in this region, and attempted to consolidate his power by centralizing the government of his new empire and reducing the independence of the tribal chieftains who had joined his confederation. After the death of the Timurid Shah Rukh in 850/1447 he took advantage of the political confusion to invade Transoxania, though he was unable to establish himself there. Nevertheless he continued to intervene in the affairs of the Timurid empire, and in 855/1451 Abu Sa'id was able to seize the throne only with his help.
Some Kazakh clans - under the leadership of two Mongol princes, Janibeg and Qarai - had refused to accept the new order, possibly discerning an unacceptable trend towards sedentarization. They fled to the eastern Chaghatai khanate in Mughulistan. Abu'lKhayr was much weakened by these defections from his ranks and was consequently to suffer defeat in 861/1456-7 at the hands of the Oirat Mongols, who were advancing from western Mongolia. He never really recovered from this setback. In 872-3/1468 he died while on campaign against the Kazakhs and their allies from Mughulistan. His death was followed by internecine warfare among the tribes which had formed his confederation, and its power collapsed. For the time being Ozbeg expansion was at an end, and of Abu'l-Khayr's family only his grandson Muhammad Shaybam, born about 855/1451, survived.
After the disaster of 872-3/1468 Muhammad Shaybani lived for a time as little more than a bandit. But he was able gradually to build up a substantial following, and from 891/1486 he began raiding Timurid territory in force. In the 1490s he was able to occupy the Jaxartes territory and to prepare for a fullscale invasion of Transoxania. After marching unsuccessfully against Samarqand in 904/1499, in 905/1500 he abandoned the ancestral steppes for ever. The vacuum on the steppes was filled by the Kazakhs, who remained there and after whom the modern Soviet republic of Kazakhstan is named.
The conquest of Transoxania took several years: Samarqand itself had to be occupied twice, in 905/1500 and 906/1501. A considerable effort to resist the Ozbegs was made by the Timurid Babur, whose lifelong ambition it was to be the ruler of Samarqand. But in 910/1504 he admitted defeat, for the time being, and withdrew to Kabul. In 910/1504-5 Muhammad Shaybani took Khwarazm, and was poised to invade Khurasan. There the last great Timurid, Husayn Bayqara, died in 911/1506. In 912/1507 the Ozbeg forces entered his capital, Harat. By 913-14/1508 they controlled the whole of Khurasan, and had laid Persia waste as far west and south as Damghan and Kirman. They were even able to inflict severe damage on the Kazakhs in the steppes to the north. By 914-15/1509 it must have seemed that the Shaybanid Ozbegs had succeeded to the whole Timurid inheritance in the east: they looked invincible. In fact it fell to the equally triumphant Shah Isma'll to bring Muhammad Shaybani's career to an inglorious end.
Perhaps Ismal might have been prepared to tolerate a Timurid neighbour that offered no real threat to his own power: Uzun Hasan and Husayn Bayqara had contrived to co-exist without undue strain on either side. But Muhammad Shaybam's expanding Ozbeg empire was another matter. Isma'il evidently concluded that to ignore the new, confident and - dangerously for the Safawids - Sunni state to the east was an unacceptable risk. In 916/1510 he marched into Khurasan and met Muhammad Shaybani in battle near Marv. The Ozbegs were defeated and their ruler killed. Isma'il had his skull set in gold and fashioned into a drinking cup - a way of treating a dead enemy that had a very long steppe pedigree behind it. He despatched this gruesome trophy to his other great enemy, the Ottoman sultan.
Harat surrendered to the Safawid forces, but Isma'il seems to have decided that to attempt to carry the war into Transoxania would tax his resources too far. He accepted the River Oxus as the boundary between the Safawid and Ozbeg empires. Khurasan became a Safawid province. Although Shah Isma'il did not intervene directly in Transoxania, he did encourage Babur to take advantage of the confusion which had followed the defeat and death of Muhammad Shaybam. In 917/1511 Babur invaded, and was able, for the last time, to enter Samarqand. But whether because, as a protege of Isma'il, he had been obliged to declare himself a Shi'i, or because of the harsh economic measures he found it necessary to take, Babur received a hostile reception from the populace and was driven out before long by Muhammad Shaybani's nephew 'Ubayd Allah. This time there was to be no return: the future, for Babur and his Timurid descendants, lay in India.
In 918/1512 Ozbeg troops under 'Ubayd Allah invaded Khurasan once more. This time they had more success: at Ghujduwan they defeated the Safawid army, though on this occasion Isma'il was not in command. A few months later, in 919/1513, he marched east again in person. The Ozbegs withdrew without fighting, and the Safawid-Ozbeg frontier was stabilized, at least for a time.
Despite the defeat of 918/1512 and the failure of Babur, Isma'il's policy in Khurasan could on the whole be counted a considerable success. He had contained, if he had not destroyed, a formidable enemy, and had acquired a vast province for his empire, one that had not for many years been held by the rulers of western Persia. But it could not be pretended that the Ozbegs were Safawid Persia's most dangerous adversary. A state whose centre was still in Azarbayjan and eastern Anatolia clearly had much more to fear from the Ottoman Empire.
The root cause of the conflict between the two empires lay in the fact that so large a proportion of the supporters of the Safawid cause came from among the Turkmen tribesmen of eastern and central Anatolia, near to, or actually inside, Ottoman territory. This meant not only that the Ottoman state was losing valuable manpower to its neighbour, but also that the loyalty to the Ottomans of many of those who remained could not be relied on. As early as 907-8/1502 there had been mass deportations of Qizilbash to southern Greece, where they would be out of harm's way.
In 917/1511 and 918/1512 major Qizilbash revolts broke out in Ottoman Anatolia, in the latter stages with active Safawid support. The devastation caused was immense. The reasons for these events seem to have been as much social and economic as specifically political or religious. Severe economic distress, plague and famine, had alienated the Turkmen of Anatolia; and they had a further grievance in that the policy of the Ottoman govemment towards its nomadic subjects was to restrict their freedom of movement and bring them firmly under the bureaucratic control of the central administration. In these circumstances the appeal of the Safawids was not simply a religious one. Turkmen tribesmen may have followed Shah Isma'il because they regarded him as divine, but they also found the tribally oriented Safawid political enterprise altogether more attractive than submission to the government in Constantinople. Their prospects in Safawid Persia were good: they had none as subjects of the Ottoman sultan.
The Qizilbash revolts exhausted the patience of the Ottomans. They helped to precipitate the abdication in 918/1512 of Sultan Bayezid in favour of his formidable son Selim ("the Grim"), who was destined not only to deal with the menace from Persia but also to add Syria and Egypt to the Ottoman dominions.
Selim first of all inaugurated a savage repression of the Qizilbash, executing many and deporting others. He then prepared to confront Shah Isma'il directly, marching across Anatolia towards Azarbayjan in 920/1514. It was a long march, and the logistical problems, in the face of a Safawid scorched earth policy, were great. Nevertheless Isma'l chose to accept battle at Chaldiran, an encounter which was to have permanent consequences both for the future of the Safawid empire and for the political geography of the Middle East down to the present day.
Whatever we may make of the figures quoted in contemporary sources, it seems clear that the Ottoman army at Chaldiran was much larger than its Safawid opponent. Perhaps as important as relative size was the nature of the two armies' composition. The Safawids still mobilized a force essentially of the traditional Turko-Mongol type: its strength lay in its cavalry archers. The Ottomans, on the other hand, had, as well as cavalry, the Janissaries, infantry armed with hand guns, and field artillery. The Safawids, on this occasion, had no guns, though they had in the past used them, at least in siege warfare.
The Ottoman victory was total. Shah Isma'il escaped from the field, but his army was crushed and many of his highranking officers killed. The conventional explanation of the scale of the defeat is in terms partly of the numerical superiority of the Ottomans and partly of their guns: Chaldiran is seen above all else as the victory of modern military technology over the outdated steppe ways of warfare. There may be something in this, and it should be remembered that the Ottoman artillery had also made a contribution to the defeat of Uzun Hasan Aq-Qoyunlu at Bashkent in 878/1473. But the superiority of the guns ought not, at this date, to be pressed too far. It may well be that what was really dangerous to the Safawids about the Ottoman field artillery was not so much what came out of the guns' barrels as the fact that, chained together, they formed an effective barrier to cavalry charges and a safe refuge behind which the Janissary musketeers could shelter while loading and firing.
In the aftermath of Chaldiran Selim marched into Azarbayjan and occupied Tabriz. But his troops, never very enthusiastic about the campaign, proved unwilling to winter there, and he withdrew westwards. Isma'il was able to reoccupy his capital unopposed. What, then, were the consequences of the battle of Chaldiran?
The first was a major and definitive readjustment of the Perso-Turkish frontier. The eastem Anatolian provinces, homeland of the Qara-Qoyunlu and the Aq-Qoyunlu, the principal Qizilbash recruiting ground, became and remained Ottoman territory. The present border between Iran and Turkey is a result of Chaldiran: there was no such frontier before 920/1514. This meant a decisive change in the shape of the Safawid empire. It was no longer the old Turkmen state with Khurasan added on: instead it was something more like Iran as we think of it today, although Shah Isma'il still controlled Iraq, as well as lands in the Caucasus and in the east which Persia later lost. Tabriz, the capital, was now almost a border city, hazardously close to the enemy - a consideration which no doubt influenced the transfer of the capital, under Isma'iI's successors, first to Qazwin and then to Isfahan. Inevitably, although the Turkmen element in the Safawid polity was still of immense importance, this shift of the centre of gravity eastwards also resulted, in time, in the state becoming more "Persian" and less "Turkish" in character.
The second consequence, the one most frequently remarked on, is the effect the defeat had on Shah Isma'il himself and on his status in the eyes of his Qizilbash followers. We read that Ismal sank into a deep depression, that he never smiled again, and so forth, though this is hardly borne out by the testimony of European ambassadors who saw him in the years after Chaldiran. It is indeed true that he did not, in the ten remaining years of his life, take the field again in person, and this may be significant.
So far as his status is concerned, belief among the Qizilbash in the shah's divinity must certainly have been damaged by the defeat: gods are not usually expected to lose battles, and Chaldiran cost Isma'il his previously well-deserved reputation for invincibility. But this should not be exaggerated. Both his father and his grandfather had been defeated and killed in battle, without this having any discernible detrimental effect on the religious appeal of the Safawid cause. As late as the 1540s, well into the reign of Isma'il's son and successor Tahmasp I, the Italian traveller Michel Membre observed practices at the Safawid court which clearly indicated that the old beliefs of the Qizilbash had not as yet been destroyed either by Chaldiran or by the establishment of the much more staid and sober Twelver variety of Shi'ism.
If the shah's loss of personal status was as serious as is sometimes maintained, one might have expected the Safawid state to have collapsed in the aftermath of the Ottoman victory, but in fact nothing of the sort occurred. It would seem that even by 920/1514 the regime was resting on foundations more varied and durable than the mere loyalty of the Qizilbash tribesmen. This was just as well, since the state had to endure, and did endure, two long periods of weak central government and Qizilbash factional struggles before the century was out.
The Safawid state established by Shah Isma'il relied on three elements for the maintenance of its power. The first, without which the Safawid family could never have seized control of Persia, was the loyalty of the Qizilbash tribesmen to the shah in his capacity as head of the Safawid order, descendant of 'Ali, even divine being. The acquisition of political power required above all, in the first instance, the necessary military capability, and this the Qizilbash provided. There were a host of clan and tribal affiliations, but the Qizilbash were organized, up to a point, into a number of broad tribal groupings - Rumlu, Shamlu, Afshar, Qajar, and so forth.
After the conquest of Persia many of the Qizilbash groups were granted grazing lands, especially in the north-west of the country, which not only contained the grasslands traditionally most favoured by nomads but was also the area nearest to the point of greatest danger to the state, the Ottoman frontier. Whole provinces were granted to tribal chiefs as governors, in return for the obligation to provide the shah with troops. The most powerful tribal leaders took office in the central government, near to the person of the shah, the source of all authority and patronage. At court, and increasingly without much direct relevance to original tribal loyalties and bases of power, factions developed and alliances were forged. The Qizilbash leaders long remained a potent force within the state, and not always one that the shah found possible to control as he might have wished. But they, and the military forces they held at their disposal, could not easily be dispensed with.
"The empire", a Chinese bureaucrat is supposed to have said to the Mongol khan Ogedei, "has been conquered on horseback, but it cannot be governed on horseback." As all nomadic and semi-nomadic dynasties had found, this was no less true of Persia than of China. The local bureaucracy, recruited especially from members of the urban population of the cities of central Persia, was an essential part of the machinery of government once the initial period of conquest was over.
In the Safawid case the truth of this was accepted right from the start. Before the battle of Sharur, Shams al-Din Zakariya Kujuji, a Persian who was wazir to the Aq-Qoyunlu, had arrived at Isma'il's camp to reveal the parlous state of confusion that prevailed at the Turkmen court and to encourage Isma'il to attack. He received his reward by being appointed to the same post in the new Safawid government, and men of his origin and training proceeded to staff the main administrative- as distinct from military- offices (though the distinction was in practice by no means absolute). In this way Isma'il ensured that there should be a strong thread of administrative continuity between his government and that of his Turkmen predecessors.
This did not mean that no significant changes were made. Some modifications to the system of govemment had to be implemented to take account of the fact that Persia was now, in a sense, being governed as a theocracy, in which the shah was also the head of a sufi order and his military followers his disciples.
One of the principal results of this perhaps anomalous situation was the creation of the office of wakil, deputy to the shah. He was deputy as head of the government - that is, he was a kind of wazir. But he was not only the chief executive of the government: he also in some instances acted as commander-in-chief of the armed forces. It was in this role that the wakil Yar Ahmad Khuzani led the Safawid army to a defeat which resulted in his own death at the battle of Ghujduwan in 918/1512.
What is interesting is that Yar Ahmad Khuzani, like almost all the early Safawid wakils, was a Persian: it is clear enough that to say that all military responsibility fell to the Turkmen chiefs and all civil posts to the Persian bureaucrats is far too simple. Indeed, such appointments to the wakilate are an indication that Shah Isma'il was consciously attempting to shift power away from the Qizilbash towards the more reliable and less militarily dangerous Persians. He could not do without the Qizilbash, but the bureaucracy was in some ways a more secure foundation for his regime. The Qizilbash themselves certainly saw the advancement of Persians to the highest office as an affront to themselves: two Persian waklls were murdered by them.
The Safawid state's third and most distinctive foundation was its new official faith, the Twelver variety of Shi'ism, whose adherents accept a line of twelve infallible imams, beginning with 'Ali and ending with Muhammad al Mahdi, who disappeared (but did not die) in about 264/878 and whose return the faithful expect. Everybody is now well aware that late-twentieth-century Iran is a Shi'i country, and for this the religious policy of Shah Isma'il is directly responsible.
There seems no reason to doubt that a large majority of the Persian people in 907/1501 adhered to the Sunni branch of the Islamic faith. The advent of the Safawid dynasty brought with it compulsory conversion to Shi'ism. Why this should have been so is not immediately obvious: the form of Shi'ism chosen by Shah Isma'il was not the faith of his Qizilbash foilowers. Possibly Isma'il had been influenced by the Shi'i environment in which he had lived in Gilan before marching against the Aq-Qoyunlu; or he may have felt that Twelver Shi'ism was at least nearer to Qizilbash belief than Sunnism was.
It has sometimes been suggested that Isma'il's motives were in reality, in the modem Westem sense, "political": that he saw in Shi'ism a convenient source of identity, a means of differentiating his kingdom from its Sunni neighbours, Ottoman and Ozbeg. Bemard Lewis has some wise words about such reasoning: "When modern man ceased to accord first place to religion in his own concems, he also ceased to believe that other men, in other times, could ever truly have done so, and so he began to re-examine the great religious movements of the past in search of interests and motives acceptable to modern minds.'' There is no evidence whatever to lead us to suppose that Isma'il was motivated by cynical notions of political manipulation, or indeed that to draw such a distinction between "politics" and "religion" would have occurred to him. This, however, is not to say that the adoption of Shi'ism did not, in time, play its part in helping to define the political and cultural identity of "Persia", as against other Muslim countries.
It may be said, with some justice, that the ground had been prepared for the acceptance of Shi'ism during the previous two centuries, a period in which veneration of 'Ali and the other eleven Shi'i imams had become popular and widespread, and was not thought incompatible with adherence to Sunnism. But there was no Shi'l religious establishment in Persia on which Shah Isma'il could call for assistance when, on taking possession of Tabriz in 907/1501, he declared Twelver Shi'ism not only the official but the compulsory religion of his new empire. This did not prevent the speedy enforcement of Isma'il's policy.
The shah is said to have threatened that death would be the penalty for any opposition to his wishes with respect to religion. If anyone had thought this an empty threat, they were soon to be disabused. As the Safawid forces marched across Persia, Shi'ism was imposed at the point of the sword. Sunnis who were reluctant to see the error of their ways were treated with great brutality. Many were executed.
In these circumstances the speed at which Persia apparently became a Shi'l country is hardly a matter for surprise. If Shah Isma'il's motives are unlikely to have been "secular", the same may well not have been true of all those Persians who hastened to declare their allegiance to the official faith. They had that most powerful of motives, the wish to save their own lives and property. But a good many leading Persians seem to have taken a more positive view than this. Before long we see the emergence of a class of what have been called Persian "clerical notables" - wealthy 'ulama, originally Sunni, who adhered to Shi'ism and rose to high rank in the new religious establishment, holding positions as qadis and even the supreme religious dignity, that of sadr. The sadr was appointed by the government to oversee the religious institution, in - particular the administration of waqf property and of the shari'a. Some scholars believe that for a time he had also to supervise the imposition of Shi'ism and root out heresy and Sunnism. On occasion both the sadr and qadis would hold military commands, which indicates that it was not only the distinction between civil and military office that was not always clearly maintained.
But such opportunistic defectors from Sunnism, useful and indeed essential as they were, could not provide the theological and legal backbone for the new Shi'i establishment. No one in Persia could do this. Shah Isma'il had to look elsewhere. He imported Shi'i 'ulama from the Arab lands, from Bahrayn, from Hilla in Iraq, and above all from Jabal 'Amil in Lebanon. Many of the leading theologians and lawyers of the Safawid period were of 'Amili origin, including the most influential religious figure of Isma'il's own time, al-Karak. The religious brain drain to Persia long continued: it was not a merely temporary phenomenon.
The position of the Safawid shah, a ruler who claimed descent from the imams and regarded himself- and was officially regarded - as their representative on earth, was not one that could easily be reconciled with Twelver Shi'i theology and law. But the creation of a new Shi'i state, in which impoverished if leamed theologians and lawyers could expect employment and a sympathetic hearing, was not to be lightly despised. For the most part the immigrant 'ulama kept to themselves any reservations they may have felt, except on occasion in works (in Arabic) on law, intended for their fellow-scholars rather than for a wider Persian readership.
At what stage the majority of the Persian people became actually, as well as nominally, attached to the Shi'i version of their faith is impossible to say without the ability to make windows into the souls of men long dead. The process was probably a slow one, and was never total. Sunni communities, especially among some of the tribal peoples, remain in Persia to the present day. It was still possible for Shah Isma'il II to mount an admittedly abortive attempt to return to Sunnism during his brief reign (984-5/1576-7). But Nadir Shah in the eighteenth century failed to move Persia in a Sunni direction, and it would probably be fair to suggest that Persia was irrevocably Shi'i well before the end of the Safawid period. Certainly the Persian people's allegiance to that faith has remained firm ever since.
Shah Isma'il died suddenly in 930/1524, still only in his thirties. His achievement, despite the failure at Chaldiran, had been very great. He had founced a dynasty which ruled for over two centuries - far longer than any other Persian dynasty in the Islamic period. He had secured its territory within borders which may not have been what he was aiming at, but which his successors did not extend, and which approximate to those of modern Persia before the contractions of the nineteenth century. He had begun the process of organizing his kingdom so as to ensure that it would survive times of adversity. He had promoted the adoption of a new official faith for his country, admittedly by methods that were unsavoury; but the effects were enduring. All in all, he had left more of a mark on Persia than most of its rulers have managed to do. What he had not done, as his successors were to discover, was to solve the problem of how a dynasty which had come to power at the head of an army of Turkmen tribesmen might prevent those same tribesmen from becoming a menace to the stability of the state they had helped establish.