iii. Arab Settlements in Iran
Arab settlements in Iran proliferated after the wars of conquest that destroyed the Sasanian empire. Though it is difficult to trace accurately their extent and development, they certainly were critical in making the effects of the conquest long term rather than transient and in facilitating the symbiosis of Iranian and Arab cultures within a mutual Islamic context.
The history of Arab colonization in Iran (understood here to be equivalent to the territorial limits of the Sasanian empire) has been clouded to a great degree by the inherent biases and limitations of the sources and three fundamental misconceptions arising from them. First of all, the term “Arabs” did not denote only the desert tribesmen and peoples of the remoter areas of the Arabian peninsula. It must also be understood to include a number of Arabic-speaking sedentary or semi-nomadic people living within or in near proximity to the two major pre-Islamic empires of the Middle East. Second, this Arab presence in Iran clearly did not begin with the Arab destruction of the Sasanian empire. For centuries, Iranian rulers had maintained contacts with Arabs outside their borders, dealt with Arab subjects and client states in Iraq, and settled Arab tribesmen in various parts of the Iranian plateau. Third, it follows that the “Arab” conquests and settlements were by no means the exclusive work of Arabs from the Ḥeǰāz and the tribesmen of inner Arabia. The Arab infiltration into Iran began before the Muslim conquests and continued as a result of the joint exertions of the civilized Arabs (ahl al-madar) as well is the desert Arabs (ahl al-wabar). Unfortunately, some modern authorities have exceeded even the traditional sources in exaggerating the role of the latter and minimizing the role of the former in the inspiration, conduct, and exploitation of the Arab-Muslim conquests.
Just as these conquests differed in nature and effect from region to region, the Arab settlements in Iran varied greatly from one area to another. Some parts of Iran took so long to subjugate and control (Daylam, Ṭabarestān, the Hephthalite principalities on the eastern periphery of Iran) that the Arab, as opposed to Muslim, settlement in them was negligible and need not be considered here. The situation in Iraq (used here to denote the area from Mosul to Baṣra) was fundamentally different from that found on the Iranian plateau. Beyond that, there were significant differences in the cases of western Iran (the Caucasus, Azerbaijan, Media), southern Iran (Fārs, Kermān, and Sīstān), and eastern Iran (greater Khorasan and Transoxiana).
Iraq. In the Iraqi zone of settlement, the ethnic Iranians constituted only a minority of the existing population and seem to have been largely confined to a narrow social elite. A substantial Arab presence before the conquest included the Christian Taḡleb tribes in the environs of Hatra in the north, a series of tribes (Tanūḵ, Kenda, Ṭayy and Šaybān) around Ḥīra in central Iraq, and still other tribes (ʿEǰl, Tamīm, ʿAbd al-Qays) near Obolla in the south. There were also tribesmen of Bakr b. Wāʾel and Tamīm in Ḵūzestān and Fārs prior to the arrival of the Arab Muslim armies. This Arab population was anything but homogeneous: Some were clearly sedentary while others were nomadic; some fought on behalf of the Sasanians (Taḡleb, Eyād) while others (Lakhmid notables, Šaybān, Bakr b. Wāʾel) began struggling against the Sasanians after the destruction of the Lakhmid vassal state and had already won a celebrated, if limited, victory at Ḏū Qār (ca. 604?). Some local Arabs led by al-Moṯannā b. Ḥāreṯa al-Šaybānī helped to direct the attention of the nascent Muslim state towards Iran by converting to Islam and negotiating with Medina for support in their anti-Sasanian moves. During the wars of conquest, the fighting in Iraq was the most intense and sustained of any of the campaigns on the eastern front. As a result, most of Iraq was taken by conquest and therefore differed in legal status from much of the Iranian plateau (where numerous areas arranged treaties of capitulation).
The rich agricultural lands of Iraq made it a choice area for Arab settlement after the Sasanian collapse. In fact, the desire to wrest these lands from the Iranian aristocracy may have provided a prime motive for the aggressiveness of the local Arabs. Accounts of this early period of the conquests often contrast the “poverty” of the Arabs with the riches of the Iranians, and the major Iranian landlords are almost invariably said to have been killed or dispossessed and expelled. In the immediate wake of the Arab-Muslim victories, many of the Arabs attempted to seize lands for themselves; throughout the period of the conquests the term “emigrant” (mohāǰer) was used almost synonymously with “warrior.” Caliph ʿOmar (13-23/634-44) discouraged such excesses. Even though the region was taken by conquest, the people were generally granted protected (ḏemmī) status and not enslaved; steps were taken to keep peasants on the land, and the state, rather than the soldiers, took title to as much of it as possible. To prevent the Arabs from dispersing to private estates, ʿOmar ordered them to congregate in new camp cities or “emporiums” (qayrawān) and paid them stipends from the state treasury. The two most famous of these camp cities were Kūfa and Baṣra. By most accounts, Baṣra was founded by ʿOtba b. Ḡazwān in 14/635 and settled primarily by Arabs of the tribes of Tamīm, Bakr, and Azd. Saʿd b. Abī Waqqāṣ founded Kūfa, probably in 17/638. The original Arab population is reported by Balāḏorī to have been 20,000. No one tribe or group of tribes dominated Kūfa, but in contrast to Baṣra it was home to a high proportion of companions of the Prophet Moḥammad (i.e., Hejazi Arabs). Within a generation Kūfa reportedly had 80,000 households (with 60,000 soldiers) and Baṣra had 120,000 households (80,000 soldiers), representing total populations of perhaps, 300,000 and 500,000 respectively. Since it is almost inconceivable that so many people could have migrated from Arabia, they must have included many migrants from the Iraqi countryside.
In any case, the most striking feature of Arab settlement in Iraq was the concentration of large numbers of Arabs in a few urban centers. Large Arab garrisons were also established at Ḥadīṯa and Mosul, but they may have been substantially reduced or replaced by non-Arab “clients” (mawālī); it is also possible that the importance of Mosul as a third camp city in Iraq has been neglected in the extant sources. Much later, the Omayyad governor Ḥaǰǰāǰ b. Yūsof founded yet another city, Wāseṭ, in the heart of the Sawād and garrisoned a number of his Syrian troops there. From almost the beginning of the conquests, the Arabs were joined by segments of the local population, especially members of the Iranian cavalry (asāwera and small landholders (dehqāns), who converted to Islam and accepted the status of mawālī, or “clients” of Arab tribes to hold on to their property or position. It was not long before they too began to congregate in the new cities. These trends, along with the destruction of the local Iranian aristocracy and the presence of a substantial Semitic population, may explain why the Arab cultural influence in Iraq remained relatively undiluted and the area eventually ceased to be “Iranian” in any meaningful sense of the term.
Northern Iran. Information about Arab colonization in northwestern Iran is very sketchy, but the general pattern is clear. After the battle of Nehāvand, some Arab troops, mostly from Kūfa and apparently on their own initiative, fanned out through Azerbaijan. They left behind small colonies of troops and tax officials (usually numbering only a few hundred men) at a number of towns, including Ḥolwān, Dīnavar, Ray, Hamadān, Qazvīn, and Ardabīl. Other troops pushed into Armenia and the Caucasian provinces and established garrisons at towns such as Dabīl. The closest parallels to the Iraqi camp cities were the large garrisons at Isfahan (primarily in the suburb of Jayy), Barḏaʿa, Qālīqala, where as many as 8,000 troops were maintained to prevent the territory from being retaken by the Byzantines, and Bāb al-Abwāb (Darband), where some 24,000 Syrian troops were reportedly stationed to resist Khazar invasions. Many of these troops may have eventually returned to their homes, but in the case of some of the smaller garrisons the sources explicitly state that the Arabs were given grants of land (and/or pensions from the treasury), built homes and mosques, and settled down on a permanent basis in the area (see, e.g., Balāḏorī, Fotūḥ, pp. 197-98, 322, 329).
Since this part of Iran included many groups of nomadic and semi-nomadic tribes (Kurds, Daylamites, Khazars), the central government also often encouraged groups of nomadic rather than sedentary Arabs to migrate there, in order to play off one tribe group against another. As late as the eighth century, the ʿAbbasids attempted to strengthen their hold on the region by encouraging a new influx of Arab colonists from Baṣra as rivals to the earlier settlers (see Yaʿqūbī, II, p. 446). While this policy was not particularly successful, it may help explain the special importance of pastoralism in the development of west Iranian society and history.
In addition to this “official” form of colonization, there was also a great deal of important but apparently unauthorized settlement by land-hungry individuals or tribes which was facilitated by the relative weakness (or disinterest) of the central government in the area. Balāḏorī (Fotūḥ, p. 329) notes that the families of Arabs who had settled in Azerbaijan (presumably in the military colonies) were eager to move there. When they arrived, they attempted to seize landed estates for themselves. In some cases, they purchased land; more often, the Iranian villagers gave up title to their lands and became the tenants of the Arab families in the hope of obtaining protection from other invaders or avoiding exploitation by the government’s tax agents. Various Arab tribal families are thus known to have come into control of lands or “castles” around Dastabā, Karaǰ, Hamadān, Māḏarān, Marand, Urmia, Narīz, Tabrīz, Barza, Sarāt, and Isfahan. One result of this phenomenon was the appearance of a host of virtually autonomous petty dynasties founded by Arab colonists, such as the families of Rawwād al-Azdī (Tabrīz), Abū Dolaf al-ʿEǰlī (near Hamadān), Baʿīṯ b. Ḥalbas al-Rabaʿī (Marand), and Abū Dolaf al-Šaybānī (Naḵǰavān).
As a dependency of the politically turbulent city of Kūfa this portion of Iran was also settled by a number of political dissidents and their families. The best documented example of this is the city of Qom, which was extensively settled by pro-ʿAlid Arabs from the tribe of Ašʿar and Maḏheǰ as well as various Talebid notables following the failure of al-Moḵtār’s revolt in Kūfa. In addition to these pro-Shiʿite groups, some Kharijite tribesmen from the Jazīra also took refuge in Azerbaijan. Their main stronghold seems to have been in the area of Baylaqān. There are also reports of tribesmen and townsmen fleeing from Iraq to Azerbaijan to escape taxation and misgovernment; these groups tended to roam the mountains and countryside, where they caused considerable turmoil.
Though it is difficult to estimate the size of the Arab population in western Iran, it was probably small and was dispersed irregularly throughout the region, with no significant concentrations except perhaps the garrison at Darband and the settlements around Qom and Isfahan. Under such circumstances, the Arabs assimilated Iranian culture fairly rapidly (for example, speaking Persian instead of Arabic), but they preserved their separate ethnic identity. When Yaʿqūbī (Boldān, pp. 269f.) visited the region, he reported finding such Arab communities in Saymara, Ḥolwān, Dīnavar, Qazvīn, Nehāvand, Karaǰ, Qom, Ray, and especially Isfahan (where Arabs retained their tribal affiliations with Ṯaqīf, Tamīm, Żabba, Ḵozāʿa, Ḥanīfa, and ʿAbd-al-Qays).
Islamization, the most obvious aspect of the Arab cultural impact on the Iranian population, seems to have developed somewhat differently in western Iran than in other areas of the plateau, perhaps because of the rather “missionary” oriented attitudes of the Kufans. Iranian notables in several places converted in order to avoid the indignity of paying the poll tax from which they had previously been exempt. The people of Qom originally harassed Arab Muslim worshipers there, but were soon won over to Islam themselves. Ašʿaṯ b. Qays is supposed to have made a register of the Arabs in Azerbaijan and ordered them to proselytize the Iranians. Abū Noʿaym al-Eṣbahānī, (Dekr aḵbār Eṣbahān, ed. S. Dedering, Leiden, 1934, I, p. 75) reports that Arabs migrated to Isfahan to teach the “people” (presumably not just the Arabs) about religious duties and the traditions of the Prophet. Some reports allege that by the time of ʿAlī’s caliphate, “most” of the population of Ardabīl and Qazvīn had become Muslim (Balāḏorī, Fotūḥ, p. 328). It is tempting to assume (but very difficult to prove) that the significant proto-Shiʿite presence in western Iran is not unrelated to the subsequent strength of Shiʿite Islam in this region.
Southern Iran. The situation in southern Iran was not dissimilar to that in the northwest except that Basran rather than Kufan troops provided the balk of the colonists. As in all other cases, the initial settlements took the form of military garrisons. One of the first and most important of these was Tawwaǰ (Tavvaz), settled after its conquest by the ʿAbd-al-Qays and other tribes from the Gulf area (Ṭabarī, I, pp. 2694-95; Ebn al-Balḵī, p. 135; Balāḏorī, Fotūḥ, p. 386). Other garrisons were probably maintained at Eṣṭaḵr, Shiraz, Kermān, Bost, and Zaranǰ. This form of settlement was again followed by other types of colonization. Some Arab families took the
opportunity to gain control of private estates. One family from the tribe of Tamīm is known to have held land and “castles” in the environs of Fasā. Balāḏorī (Fotūḥ, p. 392) reports that other Arabs took possession of the homes and lands of Zoroastrians who had fled Kermān and became peasants there. There was also an influx of nomadic tribesmen, described as Kharijites or impoverished marauders (ṣaʿālīk) into the more remote areas, especially in Sīstān. The central government often had more trouble controlling these groups than in governing the local population. Perhaps because of Sīstān was remote and much of the Arab population was heterodox, these Arabs seem to have been absorbed into the local population fairly quickly and to have virtually disappeared as a distinct ethnic community by the beginning of the fourth/tenth century.
Eastern Iran. Whereas Arab settlement in western and southern Iran tended to be relatively small, scattered, sporadic and unorganized, the colonization of eastern Iran was both extensive and systematic. Tribesmen of Bakr b. Wāʾel were established in Qohestān and garrisons of troops were certainly quartered at Nīšāpūr and Marv (and probably Herat and Balḵ) following the campaigns of ʿAbdallāh b. ʿĀmer. Later, garrisons numbering from a few hundred to four or five thousand men were maintained in many towns on a more or less permanent basis. (These, however, were not necessarily purely Arab in composition: Balāḏorī, Fotūḥ, p. 407, states that the forces of Aḥnaf b. Qays in Ṭoḵārestān comprised 1,000 mawālī as well as 4,000 Arabs.) These early colonists were soon joined by successive and ever larger waves of new troops. Unlike the independent settlers in western Iran, these new arrivals in Khorasan had the backing of the government for their colonization of the area. This policy can be explained in part as an effort to relieve the surplus population pressures (and to get rid of political troublemakers) in the Iraqi camp cities, and also as a means of defending the Khorasani frontier and dealing with remaining pockets of fierce resistance to the Arabs in areas such as Bāḏḡīs. At least as early as 45/665 the governor ʿOmayr b. Aḥmar had encouraged Arabs to settle in Marv. In 51/671, the governor Rabīʿ b. Zīād is said to have sent 50,000 Kufans and Basrans with their families to settle in Khorasan. Ten years later, several thousand more Arabs, mostly Azdī tribesmen, were posted to Marv by Salm b. Zīād. In 112/730 Jonayd b. ʿAbd-al-Raḥmān sent 20,000 Arabs (half from Baṣra and half from Kūfa) to Khorasan. At the time of Qotayba b. Moslem’s governorship (1st/early 8th cent.) there were 40,000 Basran, 7,000 Kufan, and 7,000 non-Arab Muslim troops in Khorasan, the Arabs coming from the tribes of Bakr, Tamīm, ʿAbd-al-Qays, and Azd (Baḷʿamī, Chronique IV, p. 211; Balāḏorī, Fotūḥ, p. 423). Although the evidence is open to different interpretations, it suggests that throughout most of the Omayyad period there were frequent movements of new Arab soldiers and colonists to Khorasan and that the Omayyads tried to keep a force of 40,000 to 50,000 Arab warriors (moqātela) in Khorasan at all times. Because of the distance from Iraq and the attractiveness of the country, large numbers of these soldiers acquired lands in villages throughout Khorasan, married local women or brought their families from Iraq, and settled permanently in the province. The repeated infusions of fresh troops were intended to replace these losses and at the same time bolster the authority of the central government. This in turn implies that the Arab population in Khorasan must have been huge in comparison to that in western Iran. Even if the primary component of the Arab colony in Khorasan was limited to just the 50,000 families settled there by Rabīʿ b. Zīād, the total Arab population would have to be estimated at close to a quarter of a million people.
One consequence of this influx of large numbers of Arabs into Khorasan was an outbreak of “tribal” rivalries (ʿaṣabīya) involving antagonistic competition for territory and influence between the earlier colonists, many of whom had become assimilated to the Iranian population, and the new arrivals. It is possible that the decision of Qotayba b. Moslem to build up an army of Arab and Iranian troops and to invade Transoxiana may have been motivated in part by the desire to defuse this situation by encouraging cooperation between the Arabs and their Iranian counterparts, reducing the need for new levies of soldiers from Iraq, and opening up new areas for settlement. One certain result of Qotayba’s campaign was the establishment of new Arab colonies in Transoxiana in the areas as remote as Farḡāna and Šāš. The problem of conflict between the established colonists and the new arrivals never disappeared and was one factor contributing to the eventual collapse of the Omayyad government in Khorasan at the time of the ʿAbbasid revolution.
Exactly where in Khorasan such large numbers of Arabs chose to settle remains controversial. M. A. Shaban (The Abbasid Revolution, Cambridge, 1970) has argued that the settlers were at first exclusively and always predominantly concentrated in the Marv oasis. There is no doubt that Marv was required by its treaty of capitulation to provide quarters for Arab forces, that it was always the site of the largest Arab military garrison in Khorasan, and that large numbers of Arabs settled in the surrounding villages but there is solid evidence that the Arab population was in fact diffused throughout the entire province. Of necessity there were Arab garrisons in cities other than Marv immediately after the conquest. Ebn ʿĀmer himself is known to have stayed in Nīšāpūr and to have built a mosque there (presumably for use by Arab residents), and there was also a garrison maintained from an early date at Ḵom near Balḵ. Because of frequent rebellions in outlying areas, these garrisons soon appeared in virtually every town of any consequence in Khorasan: Among the known garrisons, for example, were those of Būšanǰ, Ṭalaqān, Jūzǰān, Āmol, Marv-al-rūd, Termeḏ, and Nasā. From incidental details mentioned in the historical sources and other information (such as accounts of burial places of notable religious traditionists), it can be deduced that many of the warriors established permanent homes in these places, as well as in Marv. In the important case of the Arab families sent to Khorasan by Rabīʿ b. Zīād, Balāḏorī explicitly noted that they were settled in numerous colonies on the Iranian side of the Oxus, not just in Marv. Consequently, it is not surprising that Yaʿqūbī (Boldān, p. 280) reported finding groups of Arabs, still claiming tribal affiliation with Ṭayy, Tamīm, Azd, and others in most towns in Khorasan and much of Transoxiana. Over time, many of these outlying localities came to be dominated by virtually autonomous Arab leaders, often with backing from the Iranian population, while support for the central government was limited more and more to the recently arrived moqātela in the garrisons of Marv and Nīšāpūr. This, too, was a factor in the ʿAbbasid revolution which, contrary to Shaban’s arguments, involved a reaction of these peripheral areas and their leaders against the forces of the central government in Marv.
The original Khorasani response to the Arab presence was one of indifference, probably inspired by the reasonable assumption that the Arabs, like most barbarian invaders, would soon go away. When that did not happen, and the Arabs insisted on payment of taxes and tribute as specified at the time of the conquests, indifference changed to hostility and there were revolts in many part of Khorasan (in fact, all over the Iranian plateau). Ironically, these revolts probably contributed to the decision to encourage large scale emigration to Khorasan. The basic cause of friction, however, was economic rather than ethnic, stemming from opposition to taxation rather than hatred of Arabs. The Arab settlers and the Iranians soon found that they had many common interests and concerns.
In this respect Arab settlement in Khorasan constituted an intermediate pattern between those found in Iraq and in western Iran: As in Iraq, the numbers of Arabs involved were quite large, but instead of being concentrated in a few urban camp cities they were, as in western Iran, dispersed throughout the area. They relied heavily on the Iranian aristocracy, which had survived largely intact in Khorasan, to assist them in matters of taxation and administration and soon came to accept the Sasanian example of statecraft, taught to them by the Iranian dehqāns, as normative. Because of the common dangers on the Khorasani frontier, Iranians and Arabs overcame their initial antipathy and cooperated extensively in military operations. The Arab population in Khorasan was not comprised of hordes of Bedouins from Arabia or just professional warriors. In addition to the military garrisons, the Arabs included merchants, artisans, religious scholars, landlords, peasants, beggars, vagabonds, and bandits. It was natural that in time these groups blended in with their Khorasani counterparts. This development was no doubt facilitated there, as elsewhere, by the process of religious conversion, which is especially difficult to trace in Khorasan but was certainly well underway by the time of the ʿAbbasid revolution. The special circumstances in Khorasan, which integrated Arabs and Iranians into a common social fabric, facilitated the assimilation of Iranian culture by the Arabs (see the classic account in J. Wellhausen, The Arab Kingdom and Its Fall, trans. M. Weir, Calcutta, 1927, pp. 492-97) and the gradual acceptance of much of Arab culture, above all the religion, by their Iranian subjects and peers. It was no accident that this distinctive flowering of Perso-Islamic civilization reached its fullest development in Khorasan.
No traditional Muslim source, Arabic or Persian, deals directly with the subject of Arab settlement in Iran. Various incidental details may be gleaned and pieced together from the standard chronicles and narrative histories, including Ṭabarī, Yaʿqūbī, Dīnavarī; Masʿūdī, Morūǰ; Ebn Aʿṯam al-Kūfī, Ketāb al-fotūḥ, Hyderabad, 1968-75; Ḵalīfa b. Ḵayyāṭ, Tārīḵ, ed. S. Zakkār, Damascus, 1967-68; and Baḷʿamī, Chronique. The most valuable source, and the closest to being a coherent account of this subject, is undoubtedly Balāḏorī, Fotūḥ. Some important information is preserved in the Persian local histories, in particular Ḥasan b. Moḥammad Qomī, Tārīḵ-e Qom, ed. S. J. Ṭehrānī, Tehran, 1313 Š./1934, pp. 240-305 which gives a very detailed account of Arab settlement and activities in Qom.
Also of use are Ebn al-Balḵī, pp. 112f.; Naršaḵī; ʿAbdallāh b. Moḥammad Nīšāpūrī, Tārīḵ-e Nīšāpūr, Tehran, 1339 Š./1960, pp. 124-30; Abū Noʿaym al-Eṣbahānī; and the anonymous Tārīḵ-e Sīstān. Geographical works often mention Arab settlements in reference to various localities; Yaʿqūbī, Boldān, pp. 269-308 and Yāqūt, Moʿǰam al-boldān are especially valuable in this regard. It is possible that close examination of the names preserved in the voluminous Muslim biographical dictionaries could lead to a detailed reconstruction of the patterns of Arab settlement in Iran, but as yet these sources have not been thoroughly and systematically analyzed.
Modern studies include S. al-ʿAlī, “Estīṭān al-ʿarab fī Ḵorāsān,” Maǰallat kollīyat al-adab, Baghdad, 1958, pp. 36-83.
C. E. Bosworth, “Iran and the Arabs Before Islam,” in Camb. Hist. Iran III/1, pp. 593-612.
N. Ḥasan, The Role of the Arab Tribes in the East During the Period of the Umayyads, Baghdad, 1978.
I. Lapidus, “Arab Settlement and Economic Development of Iraq and Iran in the Age of the Umayyad and Early Abbasid Caliphs,” in A. Udovitch, ed., The Islamic Middle East 700-1900, Princeton, 1981, pp. 177-208.
M. Morony, “The Effect of the Muslim Conquest on the Persian Population of Iraq,” Iran 14, 1976, pp. 41-59.
ʿAbd-al-Ḥosayn Zarrīnkūb, “The Arab Conquests in Iran,” in Camb. Hist. Iran IV, pp. 1-56.
Moḥsen ʿAzīzī, La Domination arabe et l’épanouissement du sentiment national en Iran, Paris, 1938, and M. A Shaban, The Abbasid Revolution, Cambridge, 1970, are also worth consulting but should be used with great caution.
A recent and useful study is M. G. Morony, Iraq after the Islamic Conquest, Princeton, 1984, pp. 214-53.
(E. L. Daniel)
Originally Published: December 15, 1986
Last Updated: August 10, 2011
This article is available in print.
Vol. II, Fasc. 2, pp. 210-214