It was this integration of the Arabian tribes into a single new Islamic state that set the stage for the conquests, which in fact represented the fruit of that integration. The process of state consolidation that began with Muhammad continued unabated throughout the whole period of the early Islamic conquests. As under Muhammad, each tribal group integrated into the state during the conquest period was administered by an agent ( amil), often one of the Quraysh or the ansar, who appears to have supervised the tribe and collected the taxes that weer due from it. There were, as we have seen, such governors or agents over some of the tribes of Quda'a in southern Syria under Abu Bakr, and we read that somewhat later, under Uthman, a member of the Quraysh named al-Hakam b. Abi l-'As was appointed to collect taxes from the Quda'a. Likewise, Sa'd b. Abi Waqqas had served as Umar's agent in charge of collecting the sadaqa tax from the Hawazin tribe in the Najd before being appinted commander of the army that marched to al-Qadisiyya; Utba b. Abi Sufyan was Umar's agent among the Kinana tribe; and the existence of the agents over other tribes in Arabia in Umar's day is well attested by numerous references. Most notable are those passages that show how, under the early caliphs, the sadaqa or tax in camels was levied by the state on nomadic groups -- traditionally the hardest groups to control. This was sometimes accomplished by sending agents to the wells to wait for a specific nomadic group to gather there for water, thereupon levying the tax on them. Tribes outside the Arabian peninsula proper that were newly subjected by the Islamic state in the later phases of the conquest of the Fertile Crescent, furthermore, were reduced to paying taxes just as their counterparts in Arabia itself had been during the careers of Muhammad and Abu Bakr. In the Jazira, for example, the Islamic state seems to have dispatched two agents, one over the non-Arabs ('ajam) of the region --here approximately equivalent to the settled populace? -- and another over the nomads (al-'arab) of theJazira. In some cases, such as that of the B. Taghlib nomads who lived in northernmost Syria and Iraq, the terms for the tax levied on them depended upon whether they decided to embrace Islam or to remain Christians. Two terms for taxes are commonly used in the sources describing such situations: sadaqa and jizya (or jaza-'). The former appears to mean specifically the tax or tribute levied by the state on nomadic groups, taxes paid usually in camels. The jizya, "tax" or "tribute," appears to have been levied on sedentary populations that had chosen to adhere to their Jewish or Christian faith; presumably the sedentary population of Muslims paid zakat, "alms." The existence of a separate term for the tax on nomads (whether they were Muslims or not) highlights the degree to which the state viewed the nomads as members of a different class from settled people -- not a particularly surprising situation, since the ruling elite was, as we have seen, eager to bring the nomads under control by settling them if possible. Indeed, this differential tax system could discomfort nomadic chiefs of considerable stature who preferred to remain Christian, as the case of Jabala b. al-Ayham of the B. Ghassan demonstrates. He is reported to have come to 'Umar and asked, "Will you levy sadaqa from me as you would from the [ordinary] bedouin (al-'arab)?" 'Umar replied that he would collect jizya from him instead, as he did from others of his religion. Jabala's reluctance to pay the "nomads' tax," the sadaqa, however, hints at the negative overtones carried by this levy, itself a reflection of the inferior status that nomadic groups occupied in the new Islamic political order. The jizya, "tribute" or "tax," on the other hand, which was levied on settled peoples who were not Muslims and therefore were, strictly speaking, subjects and not allies of the Islamic ruling elite, could be waived if an individual or group performed some service for the Islamic state; thus 'Umar is said to have ordered that the Persian cavalrymen (asawira) who had been in the Sasanian armies should be freed of the jaza' if they assisted the Muslims in the conquest of Iraq. It is not specified, however, whether or not the asawira embraced Islam in joining the Islamic armies. The entire discussion of taxes, however, is confused and betrays some meddling by later legal scholars who employed such accounts to provide precedents for their own systematizations of tax laws applying to non-Muslims. But though the details of terminology may have been corrupted, there is no question that some form of tax -- whatever its name or rate of incidence -- was levied on tribal groups such as the B. Taghlib that were newly absorbed by the expanding Islamic state only during the conquest period. This supports the view that the political integration begun during the careers of Muhammad and Abu Bakr continued through the age of the early conquests.
Similarly, this continuing process of state integration can be seen in the fact that tribal groups subject to the state were liable to the recruitment of their members into the Islamic armies sent to fight on various fronts during the conquests. As already noted in describing the campaigns in Iraq, 'Umar drew up the armies of conquest by requesting his agents among the tribes to send contingents from the groups for which they were responsible. On the way to the front, the core forces so assembled were also able to raise further recruits as they passed through the territories of various tribes and could contact the tribesmen at the wells and towns they frequented. These recruited tribesmen were not simply a horde wandering aimlessly toward the Fertile Crescent, furthermore, but were organized into contingents of a relatively well-coordinated army whose objectives and general movements were established by the ruling elite. This provides yet further evidence that the process of state integration and the establishment of some meaningful control over the tribes of Arabia by the Islamic ruling elite continued through the conquest period. The various tribal groups, whether nomadic or settled, were no longer the virtually autonomous political entities they had once been. They were, rather, absorbed into the larger framework of a state, which taxed, recruited, and administered them in certain respects more or less at will. If the conquest period saw the continuation of the process of state formation that had begun under Muhammad and Abu Bakr, however, it also saw some changes in the political structure of the state. At the end of the ridda, as we have seen, Arabian society was divided into three fairly sharply defined political strata: the ruling elite on top, a small middle group of loyal tribesmen allied to the elite, and a large population of recently conquered (or reconquered) tribesmen beneath them. The elite itself was composed mainly of sedentary tribesmen from the Hijaz, notably the Quraysh of Mecca, the Medinese ansar, and the Thaqif of al-Ta'if, who had remained loyal during the ridda wars. The cohesiveness of the elite seems to have been reinforced by intermarriage. Associated with the elite was the middle stratum, which consisted mainly of members of some tribes from the Sarat or the Yemen, many of whom may have had close ties of long standing with the Meccans, Medinese, or Thaqif, and may in some cases have been resident in one of the Hijazi towns as allies of one or another group. The middle stratum also included some members of those nomadic groups of the Hijaz that had remained loyal to Medina during the ridda, such as Muzayna or parts of Sulaym; it is possible, however, that many of these individuals had also taken up residence in Mecca, Medina, or al-Ta'if as allies of one of the three main groups of the elite. The conquered tribesmen, who formed the lowest stratum, were integrated into the state in the sense that they paid taxes to it and were dominated by it, but they had no active share in the formulation or execution of state policy, which remained the exclusive domain of the elite.
This situation persisted into the caliphate of'Umar and the beginning of the campaigns of conquest in Syria and Iraq. The ruling elite continued to dominate the conquered tribesmen of Arabia, and at the beginning of the conquest it was state policy that former rebels of the ridda should not be recruited into the Islamic armies, both because they were deemed unreliable and because they were, in view of their earlier opposition, not considered worthy to share the spoils of conquest with those groups and individuals that had remained loyal to the state. The former rebels were, then, still excluded from any active participation in the state's activities. The escalating conflicts of the conquest era, however, strained this simple arrangement of things to the breaking point. The ruling elite itself had provided a significant part of the military manpower for the ridda wars and for the campaigns for the first and second phases of the conquest of Iraq as well, but their manpower resources were strictly limited, and when further hostilities demanded the raising of new and larger armies, a manpower crisis arose. It was resolved by a decision, made during the caliphate of 'Umar b. al-Khattab, to begin recruiting former rebel tribesmen for military duty in the conquest armies. This decision, as we have seen, made available the manpower necessary to wrest victory in the transitional and third phases of the conquest in central Iraq, for unlike the armies of the first and second phases, the core of which had consisted mainly of the ansar, Thaqlf, and some nomadic groups that had been consistently loyal to Medina, those of the transitional and third phases consisted mainly of former rebels of the ridda. The recruitment of these former rebels greatly enlarged the intermediate political stratum that lay between the Islamic ruling elite and the conquered tribesmen of Arabia, namely, a class of tribesmen who were allied to, or associated with, the ruling elite, and who were literally employed by the ruling elite to help accomplish the state's goals. This shift in policy toward the former rebels of the ridda, and the growth of the intermediate stratum of tribesmen who were associated with the state but were not part of the elite, however, posed serious problems for the state leadership. There was considerable peril in bringing contingents of such tribesmen into the armies, for to do so meant putting men of dubious loyalty together in sizable groups and allowing them to bear arms -- a situation that could all too easily be an invitation to foment another rebellion against the ruling elite. The fact that no such rebellion occurred during the conquest period is itself the strongest evidence demonstrating that the Islamic state and its ruling elite succeeded in integrating disparate tribal groups to itself. What remains to be considered, however, are the means by which this integration was effected. How did the elite manage to keep these conquered former rebels under control and prevent them from raising another rebellion once they were assembled in large groups as part of the Islamic armies? This becomes the central question around which revolves an understanding of the new Islamic regime's stability during the conquest period, and of the dynamism of the Islamic conquest movement itself.
Tribesmen also became bound to the state when they received shares in the nasib lands, that is, in the abandoned lands taken over by the Islamic state and offerecl to the tribesmen for settlement or exploitation. Since the benefits accruing to the tribesmen from such lands could only be enjoyed if the tribesmen remained loyal, such grants made it further in their interest to be politically quiescent.
In addition, the organization of the army was itself a factor that weakened purely tribal ties and strengthened the bonds between the tribesmen and the state -- a matter of considerable importance because at this early stage the army was in fact so much of the state apparatus. The existence of units (whether for pay or for combat) such as the hundreds or tens that cut across tribal lines by embracing individuals from many tribes doubtless helped to establish new lines of solidarity that helped transcend the narrow tribal identification.
Ideologically and organizationally, then, the Islamic state had resources upon which it could draw to override the tribal loyalties that had traditionally been the stumbling block in the path of successful political integration in prestate Arabia. It would, however, be a serious mistake to conclude that the successful integration of the Islamic state from the time of Muhammad through the early conquest period was solely the result of these rneans of transcending tribal ties. The methods outlined above did contribute much to the state's cohesion, above all by providing organizational goals that were supratribal in the context of a justifying ideology. But the day-to-day stability of the new regime, and the effectiveness with which the rulers were able to control the thousands of tribesmen now under their charge and to bring them to do their bidding, was also the result of the elite's keen awareness of the ingrained strength of tribal ties and of the ways in which these ties could be used to foster, rather than to obstruct, their consolidation of power. Themselves, after all, Arabian tribesmen, the members of the elite realized that the tribal identification was too well rooted in Arabian society simply to be abolished by decree or swept aside by a few measures that tended to transcend the exclusiveness of the tribal bond. The success of their integration of the tribesmen into a state, then, depended as much upon their ability to use tribal ties for their own ends as it did upon their ability to override those ties.
It was not simply by chance, for example, that the tribesmen settling in al-Kufa were organized into quarters by tribe; the tribal identification provided a means -- perhaps, in view of the tribesmen's background, the only means -- by which the state could conveniently classify individuals for its administrative purposes related to pay, military organization, and l~like. It is probably for this reason that tribal and nontribal arrangements for battle and pay appear to have existed simultaneously. The only way the ruling elite could keep track of the thousands of individual tribesmen serving it was by their tribal affiliation, however much it may have wanted eventually to submerge such affiliations in a greater loyalty to the state. After all, a tribesman could readily have deceived others about his membership in one or another military unit -- who was to know, given the rudimentary character of the administrative apparatus at this early date? -- but he could not very well lie about his tribal origins.
The tribal tie, furthermore, was even more vital to the state in cases where the elite needed a certain leverage over the tribesman. If, for example, a military unit's commander was held responsible for seeing all the men in his unit -- drawn, let us say, from many tribes -- appeared in formation when the army marched to battle, what was to prevent the individual tribesman, not of the commander's tribe and perhaps disgruntled with the rigors of state service, from simply leaving and quietly returning to his kinsmen somewhere in Arabia? If, on the other hand, the individual warrior's tribal chief was held responsible for seeing that he showed up in his unit when expected, it became much more difficult for the warrior to vanish in the same fashion. His absence could be reported to the tribal chief, who would know the tribesman, his kin and therefore probably his whereabouts. The tribesman who wished to turn his back on his duties to the state could now do so only by breaking his ties with his fellow tribesmen, not a very inviting prospect.
The use in these ways of tribal ties thus helped the elite consolidate control over the tribesmen to a considerable degree. Because such methods usually worked through the tribal chiefs or lineage heads, their allegiance was in several ways critical to the successful integration of the tribes into the state. These leaders could, if themselves strongly enough tied to the state, bring with them a sizable body of tribesmen over whom they had considerable influence or control. This possibility became atttractive to the Islamic state, as we have seen, when it was faced with a pressing need for more manpower to carry out the large campaigns of later phases of the conquests. On the other hand, these chieftains -- particularly those who had once led rebellions against the Islamic state during the ridda -- were themselves the most serious potential rivals of state's power. It was just these tribal leaders, after all, who would be most likely and most able to break away from the state and to establish themselves as independent rivals; indeed, it was for this reason that the elite had, during the first phase of the conquests, barred the former rebels from any participation in the state's activities. The challenge facing the ruling elite, then, was to tie the interests of key tribal leaders firmly enough to the state that their loyalty, or at least their cooperation, was assured. This done, the elite could use the essentially tribal allegiance of these chieftains' followers to accomplish the ultimate objectives of the state.
We find, consequently, that the elite used on these tribal chiefs not only the methods applied to secure the loyalty of the average tribesman -- stipends, appeals to their religious conscience, and the like -- but also a number of additional, extraordinary incentives to make certain that the chiefs knew where their interests lay. The meager evidence available suggests that 'Umar and the ruling elite during the conquest period resorted to a policy modeled after that pursued by Muhammad in a similar situation: namely, "conciliation of hearts" (ta'llf al-qulu-b). This policy is visible in the elite's arrangements withJanr b. 'Abdullah of the Bajlla tribe. In their moment of greatest need, after the severe setback the Muslims had suffered at the Battle of the Bridge, the elite approached Janr in an effort to raise badly needed troops. Jarir drove a hard bargain; he agreed to put his sizable following of B. Bajila at the service of the Islamic state, but only in return for a promise of extra booty over and above the normal share. Significantly, this extra booty is likened to ta'lif al-qulub. The sources generally state that the extra booty was being granted to Jarir himself, which makes it look as if Jarir, rather than the Bajila tribesmen, were being won over: ''IfJarir wants it understood that he and his tribe only fought for a pay like that of'those whose hearts were reconciled' (al-mu'allafa qulubu-hum), then give them their pay," 'Umar wrote to the Muslims' commander, Sa'd b. Abl Waqqas, after the Iraqi campaigns. Jarir's allegiance to Islam was, however, fairly well assured, as his previous career indicates. It seems probable, therefore, that his demand for extra pay was rooted in a realization that he would only be able to exercise meaningful authority over the tribesmen of Bajila by himself holding out to them the promise of extra booty. Perhaps the regime felt that an added inducement to ensure the loyalty of Bajila was well advised because the number of them that went to fight in Iraq was so great (according to some accounts, one-quarter of the Muslims at alQadisiyya). In any case, it is clear that at least in the case of Jarir and the B. Bailla, the ruling elite employed the policy of "conciliation of hearts" that Muhammad had used in similar circumstances.
In addition, the Islamic elite could tie important tribal leaders to the regime by means of marriages, which thereby cemented an alliance between an individual chieftain and some key figure in the elite itself. This use of marriage for political purposes, already practiced by Muhammad and indeed quite frequent in pre-lslamic Arabia, was thus continued. We learn, for example, that the chief of Kinda, al-Ash'ath b. Qays, who had rebelled during the ridda, was eventually pardoned by Abu Bakr, who bound him to the regime by allowing him to marry his own sister, Umm Farwa. Thereafter, he witnessed the major confrontations in central Iraq and settled in al-Kufa. In other cases marriage may have functioned more indirectly to tie tribesmen to the regime, as in the case of Sa'd b. Ab Waqqas, who married the widow of the Shaybam chieftain al-Muthanna b. Haritha; the objective was perhaps to bind al-Muthanna's former followers to Sa'd.
Key tribal leaders could also be bound to the state by granting them special gifts of land to be held as private estates; these appear to have been larger tracts than the nasb lands distributed to the ordinary tribesmen. This phenomenon has been discussed within the context of migration and settlement.,P. Finally, the ruling elite could bind important tribal leaders to the state by associating them in special ways with the elite itself. The process of currying favor with these leaders, by inviting them to attend a governor's audiences and to discuss with him affairs of government, functioned in this manner. The status of those chiefs who became the governor's intimates must have risen considerably in the eyes of the tribesmen who served under them. Tribal leaders in this category began to see their status enhanced and their enhanced status solidified by their evolving relationship with other tribal leaders; at the same time, they could see that their enhanced position was generated, not by their own actions, nor even by their actions together with others in this evolving group, but solely by virtue of their role as associates of the Islamic regime. If the effective power of certain tribal leaders over their tribesmen was thus increased, it was increased by enhancing the leaders' status in ways peculiarly related to the state, so that it could not easily be used against the state. These leaders emerged from the conquest period as the ashraf; the tribal notables of the garrison towns, who for decades showed themselves subservient to the state and its interests.
The association of such key tribal figures with the Islamic state was strengthened by the organization of the military payroll. As noted above, the tribal leaders were apparently in charge of distributing pay for the tribes to the 'anf or head of the 'irafa. This function, that of "distributor of surplus," must also have increased the status of the tribal leaders and their ability to command and control the tribesmen under them. In these ways, then, the ruling elite attempted to tie tribal leaders to the state by special acts of favoritism and to exploit the nexus of tribal allegiances focused on them to the advantage of the state. Yet we should not assume that the Islamic regime, once it had decided to employ former rebel tribesmen in its armies, would willy-nilly shower favors upon their leaders in order to bind them to the state. Ridda leaders were still viewed with suspicion by the ruling elite for obvious reasons, as is revealed by the example of the B. Asad. The predominant figure in the tribe was Talha b. Khuwaylid, a notorious rebel leader who had once claimed prophethood and rallied a large following from the tribes of Asad, Tayyi', and Ghatafan in opposition to Medina. Only a small fraction of B. Asad had remained loyal to the Islamic state during the ridda, and, led by Dirar b. al-Azwar, had fought as part of Khalid b. al-Wahd's force during the ridda wars, and perhaps in Iraq and Syria as well. The great majority of B. Asad followed Talha into rebellion; hence we find little indication that significant numbers of tribesmen from Asad participated in the Islamic armies in Iraq during the offensives of the first, second, or transitional phases there. In the third phase, however, a considerable number of Asadls, including Talha b. Khuwaylid himself, are found among the Muslims at al-Qadisiyya and in the later campaigns in western Iran. Yet traditional accounts record the considerable discomfort that the Islamic ruling elite felt in employing such rebel leaders as Talha. When, during the eastern campaigns, Talha was absent longer than expected on a reconnaissance mission and failed to return after the others who had been sent out had long been back, the Muslims feared that he had "apostasized" and joined the Persians.33 Clearly, his loyalties were suspect. 'Umar instructed Sa'd b. Abl Waqqas, commander of the Muslims at alQadisiyya, not to put any of the ridda leaders in command of a hundred men, for the same reason -- they could not be trusted. In accordance with this general order, we find that the chieftain 'Uyayna b. Hisn of the B. Fazara, who had backed Talha during the ridda and had been captured and sent to Medina, was later sent by 'Umar to Iraq with Sa'd b. Abl Waqqas on the condition that the latter "not appoint him to a position of command." Similarly, Qays b. Makshuh al-Muradl, whose political activities in the Yemen during the ridda had given Abu Bakr reason to doubt his intentions, was sent by the caliph to Iraq, with the proviso that he be consulted in matters of warfare and strategy but not put in command of anything. "Abu Bakr did not ask the former rebels for help in the ridda or against the Persians," we are told; " 'Umar conscripted them but did not put a single one of them in a command post." Passages such as these make it quite clear that the bulk of the former ridda rebels, and (despite the special favors granted them) even the key tribal chiefs among them, participated in the activities of the Islamic state not as members of the policy-making and governing elite, but primarily as simple employees of the state. These arrangements helped assure that the process of state integration would not be wrecked by the secession of powerful tribal chiefs, and that the elite's objectives for the state would not be too greatly distorted by the activities of these chiefs.
,H5.The State and the Nomads,/H5. If the essence of the Islamic state's accomplishment was the integration of all Arabian tribal groups into itself and their domination by the Islamic ruling elite, the real guarantee of the state's continuing ascendancy lay particularly in its ability to keep the nomadic tribesmen under control. As we have already seen, the early Islamic state and its ruling elite took an attitude quite hostile to the nomadic way of life. Early Islam itself appears to have expressed, along with the more strictly religious and ethical notions of God's unity and power and mankind's duty to be faithful and just, the social ideals of the settled life. To a certain extent, this bias in favor of settled life may reflect simple cultural preference, since Muhammad himself and the Islamic ruling elite were sedentary townsmen from the Hijaz. During the conquest period, this bias continued to be characteristic of the elite, despite the great numbers of nomadic tribesmen who eventually came to be associated with the state as employees of the army; it is reflected, for example, in an episode in which some nomads (ahl al-ba-diya) asked 'Umar to provide them with rations, to which 'Umar replied, "By God, I will not supply you until I have supplied the settled people [ahl al-ha-dira]." In a somewhat different vein, 'Uthman is said to have dismissed the opinion of an important tribal chieftain, Ibn Zurara, as the word of an "imbecile bedouin" -- reflecting the general feeling of disdain the settled townsman held for the excitable, undisciplined nomad.
But the ruling elite's bias against the nomad was rooted in more than mere disapproval of the nomadic way of life; it was ultimately founded in the keen awareness that the nomads, above all others, were a potential danger to the integration of the state and the political ascendancy of the elite. For it was the nomadic groups of the peninsula that were traditionally the ultimate source of power there; even in cases of conflict between two rival centers of settlement (as between Mecca and Medina during Muhammad's rise to prominence), the outcome depended largely on which side could most successfully mobilize a coalition of nomadic allies. The new Islamic state's survival, then, depended directly on its continued domination of the nomadic elements in Arabian society.
The leadership of the Islamic state was thus very conscious of the nomads' power and of their ability to obstruct the centralizing tendencies of the state, and their suspicion and fear of the nomads is made quite plain in many instances. Many accounts, for example, make it clear that the challenge facing the fledgling Islamic state upon Muhammad's death was to keep the nomads from rebelling; one of the ansa-r is said to have announced at that time, "By God, I am afraid that the nomadic tribes [qaba-'il al-'arab] may rebel against the jurisdiction of Islam [din al-lslam], and if some man of the B. Hashim or [the rest of] Quraysh does not take charge of this affair, it will be the end [of it].'' During the conquest period, likewise, when the pious companion Abu Dharr (who originally hailed from the nomadic B. Ghifar of the Hijaz) took up residence, not in Medina, but in the isolated village of al-Rabadha, he was urged to make a pact with Medina so that he would not rebel "like the bedouin."
The ruling elite's concern for controlling the nomads, and their disdain for the nomadic way of life, caused them to reserve command positions in the army, governorships, and other important posts in the evolving state apparatus whenever possible for settled people. The nomads seem to have been considered of dubious reliability or of inappropriate background, even if they had a history of loyalty to the Islamic regime. Thus, when 'Utba b. Ghazwan (a very early companion of the Prophet and resident of Mecca) left his post as commander-in-chief and governor of al-Basra, placing Mujashi' b. Mas'ud of the nomadic tribe of Sulaym over al-Basra in his stead, 'Umar's reaction was a thoroughly negative one. Although his exact words are given differently in several accounts, the drift of them is the same in all cases: "Appoint the settled people over the nomads"; "the settled people are more suited to be made commander than are the nomads." Even though Mujashi' was a companion of the Prophet and of unquestionable loyalty, his nomadic background meant that he was deemed unsuited for such commands; 'Umar appointed alMughlra b. Shu'ba, of the sedentary Thaqlf, in his place.
The elite's concern with controlling the nomads also had far more important repercussions, for it led the elite to elaborate a policy of settlement of the nomads during the conquest period that had a profound influence upon the subsequent history of the entire region. As in many other cases, the precedent or guiding principle for the elite's action seems to have been established by Muhammad himself, for as we have seen, the Prophet placed great emphasis on the importance of settlement (hijra) and the abandonment of the nomadic life for those embracing Islam; it was considered impossible to remain a nomad and to be a Muslim in the true sense of the word. The tax system of the early Islamic state, which had a special tax category (sadaqa) for nomadic tribesmen, appears to support this view; Muslims leading a settled life were subject to alms zakat), and sedentary non-Muslims were subject to some kind of tribute ortax (jizya, jaza'), but the tax on nomads is always referred to as sadaqa. Although this may simply reflect administrative convenience, it may also suggest that even those nomads who had embraced the religious tenets of Islam were in some way not considered to be quite the same as settled Muslims -- or were not really considered Muslims at all. With the completion of the ridda wars and the subjection of most of Arabia by the Islamic state, Abu Bakr, 'Umar, and the rest of the ruling elite found themselves faced with a "nomad problem" of unparalleled dimensions. The nomads were, for the moment, reduced to subject status and supervised by the tax agents sent out among them, but it must have been clear to the elite that something would have to be done to keep the nomads from again asserting their power against that of the state. The solution found by the elite was simple but bold and effective: recruitment of nomadic tribesmen into the Islamic armies and their settlement in garrison towns away from the desert and the home territories of their tribes. By encouraging the nomadic warrior tribesmen to join its armies, the Islamic state not only increased its own military strength, it also reduced the real power of nomadic groups remaining in the desert by skimming off the men of fighting age who would have been able to spearhead potential resistance to Islamic rule. The power of the state over the nomads was thus doubly augmented by this process. It is probable that one of the reasons why the attractions of state service -- stipends, lands, shares of booty, etc. -- were kept high was to assure that the Arabian and Syrian nomads felt it to be in their interest to leave their home territories. Indeed, one passage links the award of the stipend directly to the abandonment of the nomadic way of life, as if the 'ata-' was more a reward for deciding to settle down than it was a reimbursement for military services, for 'Umar is reported to have said, "The sooner one settles, the sooner one receives a stipend." The settlement of conscripted tribesmen, furthermore, was to take place preferably outside the Hijaz or other regions of Arabia where settlement was possible; 'Umar reportedly encouraged some of the tribesmen gathered in Medina to settle in Iraq by reminding them that they were "in the Hijaz, not in a place of settlement...." From the accounts of the conquests in Syria, it is evident that the Syrian bedouin of B. Qays and other tribes who embraced Islam emigrated from the desert and were either sent to the frontiers to campaign or were settled as garrisons in towns such as Balis on the Euphrates; likewise, 'Uthman instructed his governor in Syria and the Jazira, Mu'awiya b. Abl Sufyan, to settle some of the nomads in places away from towns and villages where they could make use of empty lands (presumably as farmers?); others were granted stipends and assigned to the cities, towns, and frontier posts as garrison troops. It was, then, a conscious policy of the early Islamic state through the conquest period to settle nomadic tribesmen. Recruited into the army and dispatched to garrison towns, the nomads became integrated into the state organization as employees and were gradually transformed into sedentary citizens, effectively cutting them off from their former desert life and from the opportunities for secluded opposition to Islamic rule it had allowed.
Indeed, the very placement of the garrison cites may be in part a product of the state's program to break the power of independent nomadic groups. It is generally agreed that the garrisons were established primarily to control the non-Arab populations of the conquered domains, to defend Arabia from invasion by either the Byzantines or the Sasanians, and to function as the springboards for further Islamic campaigns into yet unsubdued areas. Although these considerations are almost certainly valid ones, it is possible that the garrisons were also situated with yet another function in mind: namely, to keep a watchful eye on the Arabian and Syrian nomadic populations. Having reduced the possibilities of revolt among the nomads by winning some to state service and leaving those that remained in the desert weaker in numbers, the elite proceeded to ring the desert with a set of large garrisons. This may explain why the garrison in central Iraq was in al-Kufa, on the desert fringe, and not in al-Mada'in, the old Sasanian capital; the latter would have made supervision of the non-Arab population easier, but the former was more suited to keeping the nomadic tribes of northeastern Arabia under control. The garrisons in Iraq, Syria, and Egypt certainly had important functions as defense points against Sasanian or Byzantine aggression, and the garrison in the Yemen may have been for defense against Abyssinia; but the garrison in al-Yamama ("al-Bahrayn") looks suspiciously as if it were intended to keep a close check on the B. Hamfa, who had been the bitterest opponents of the Islamic state during the ridda and who would ultimately be the first to raise a major rebellion against the state's rule some fifty years later. From this ring of strategic centers, the Muslims could control even the remote central Arabian nomads, many of whom had to move out of the pasturelands of the Najd to those of Iraq, Syria, East Arabia, or the Hijaz during the dry summer months.
The age of the early Islamic conquests, then, saw the successful integration of the fragmented tribal society of Arabia, including the nomadic groups, into a relatively unified state dominated by a ruling elite that was committed to the doctrines of early Islam and to the settled way of life. An appreciation of this political integration suggcsts some new approaches to the debate over the causes of the Islamic conquest, the causes of its success, and the causes of the Arab migrations that accompanied it.
The Arab migrations to the Fertile Crescent and adjacent regions that took place during the decades of conquest can best be explained as a result of the state's policy toward tribesmen (especially nomadic tribesmen), whom it recruited and settled in garrison towns, where they could be more easily controlled and could themselves serve as instruments of state control and state expansion. The considerable attractions that the elite offered those tribesmen who joined the army were probably the primary reason why so many tribesmen chose to do so: they would be relatively prosperous because of their regular pay and shares in the revenues from naslb lands, their life would be interesting, and, as warriors for the most wildly successful enterprise Arabia had ever seen, they would be respected. Some, perhaps, were also swayed by the promise of booty to be gained in campaigning. But the realistic bedouin probably knew that booty was not guaranteed -- and those who were really interested in plunder did not necessarily join the army to acquire it, but preferred to extort money from peasants in the provinces already safely under Islamic rule during the absence of the Islamic forces, whose charge included protecting the conquered peasantry from such brigands. A few tribesmen may have wanted to migrate in order to settle on rich new lands, but as we have seen, there is little to suggest that this was a major cause of the migrations, as most emigrant tribesmen preferred to remain clustered in their new garrison towns or in the quarters of established towns that they came to occupy. Nor is the migration to be explained as the result of some natural crisis -- hunger, overpopulation, or the desiccation of pasturelands -- that forced the tribesmen out of the peninsula; 'Umar is said to have complained that he had difficulty locating enough men to conscript into the armies during the third phase of the conquests, which suggests that overpopulation was hardly a problem in the peninsula. How, after all, could any significant "surplus" population have managed to survive in an area of such precarious agricultural resources as were possessed by Arabia? The theories relating an Arab migration to long or short-term desiccation of the peninsula rest on evidence that is tenuous at best, and do not explain why the conquest and migration occurred as a sudden burst of expansive energy rather than as the gradual efflux of the most miserable in Arabian society. Theories relating the Arab migrations to the collapse of the Arabian luxury trade fail to explain why the bulk of Arabia's tribesmen, who played little direct part in this trade, should have been so directly and immediately affected by such a collapse. The collapse of commerce, however, may very well have created a crisis for certain groups that were heavily engaged in commerce -- notably, the Quraysh, Thaqif, and other townsmen -- but in order to link this to a mass migration of tribesmen from all quarters of the peninsula, we must assume that these sedentary groups formed a ruling elite with enough control over the rest of Arabia's tribesmen to use them to "recapture" trade that had shifted to new routes; in short, we are brought back to the notion of a sweeping political integration led by these sedentary groups. It seems, then, that the Arab migrations took place mainly because the political and perhaps economic interests of the Islamic ruling elite were best served by a large-scale emigration of tribesmen into the conquered domains. The migrations were the result of state policy, planned in its general outlines by the state and implemented by the state's offer of various incentives to the emigrants. "The soonel one settles, the sooner one receives a stipend."
The many factors traditionally adduced to explain the military success of the Islamic conquest movement are generally quite plausible and can be accepted without much hesitation. The relative merits of the military organization of various contestants are difficult to assess since practically nothing is known about the tactical or strategic practices of any of them. It seems clear that the Muslims had no technological advantages over their opponents on the battlefield and were in fact inferior to their enemies in the use of cavalry. There can be little doubt, however, that the conquests were made easier by the exhaustion of the Byzantine and Sasanian Empires due to prolonged warfare, the confusion that reigned in the Sasanian ruling house, the disruption caused by recent enemy occupation in Syria and Iraq, the destruction wrought by immense floods in southern Iraq, the disaffection of many of the subjects of the two empires for religious or other reasons, the convenience of inner lines of communication that the Muslims enjoyed, and the like. But to these factors must be added one more that was perhaps the single most important one contributing to the success of the conquests: the remarkable degree to which a new Islamic state with an expansionist policy could harness for its purposes the rugged warriors of Arabia. The rise of the state made it possible to weld into an incredibly effective fighting force those tribesmen whose energies had hitherto been consumed by petty quarrels among themselves and whose political horizons had hitherto usually been limited to their own tribe and its affairs. The success of the conquests was, then, first and foremost the product of an organizational breakthrough of proportions unparalleled in the history of Arabian society until modern times. However important other factors may have been, it is difficult to believe that the conquests could have succeeded without the rise of a state with the capacity to integrate Arabia's fragmented society and draw on it to attain well-defined political and military objectives. It is not even too rash, perhaps, to suggest that the Islamic conquest might have met with great success even had the Byzantine and Sasanian Empires not been reeling from their recent quarrels. The Muslims succeeded, then, primarily because they were able to organize an effective conquest movement, and in this context the impact of the new religion of Islam, which provided the ideological underpinnings for this remarkable breakthrough in social organization, can be more fully appreciated. In this sense, the conquests were truly an Islamic movement. For it was Islam -- the set of religious beliefs preached by Muhammad, with its social and political ramifications -- that ultimately sparked the whole integration process and hence was the ultimate cause of the conquests' success.
Most difficult of all to explain is what caused the conquest itself -- that is, what it was that led the ruling elite of the new Islamic state to embrace an expansionist policy. Several factors can plausibly be suggested, however, any or all of which probably led specific individuals in the elite to think in terms of an expansionist movement. First, there is the possibility that the ideological message of Islam itself filled some or all of the ruling elite with the notion that they had an essentially religious duty to expand the political domain of the Islamic state as far as practically possible; that is, the elite may have organized the Islamic conquest movement because they saw it as their divinely ordained mission to do so. This view coincides closely with the traditional view adopted by Muslims themselves. Skeptical modern scholars have tended to discount the religious factor, but it must be borne in mind that as an ideological system early Islam came with great force onto the stage of Arabian society -- we have seen how it appears to have laid the groundwork for a radical social and political transformation of that society. The pristine vigor of early Islam may be difficult to sense now, after the passage of so many centuries and in the context of an age dominated by social and political ideas very different from those of ancient Arabia, but its revolutionary impact on seventh-century Arabia can hardly be doubted. We should, therefore, be wary of recent attempts simply to dismiss as insignificant what was clearly felt by contemporaries to be a profoundly powerful movement. Furthermore, we must recognize that even in cases where other, more mundane factors were partly responsible for stirring an individual member of the elite to favor the idea of an expansionist movement, it was Islam that provided the ideological sanction for such a conviction. The precise degree to which the purely ideological element may have bolstered the practical resolve of the elite to embark on an expansion that was considered worthwhile for other reasons as well can hardly be estimated, but it would be unrealistic, indeed foolhardy, to dismiss ideology or faith as a factor altogether. Some of the ruling elite, then, may well have believed in expansion of the state and the conquest of new areas simply because they saw it as God's will, and many others were surely susceptible to the psychological comfort of having such legitimation for their actions.
Other factors, however, certainly contributed to the adoption of an expansionist policy by the state. Much of the elite -- the Quraysh, Thaqlf, and many Medinese as well -- may have wanted to expand the political boundaries of the new state in order to secure even more fully than before the trans-Arabian commerce they had plied for a century or more, or to recapture routes that had shifted north. There is ample evidence that members of the ruling elite retained a lively interest in commerce during the conquest period and wished to use wealth and influencc accruing to them as governors or generals for new commercial ventures; 'Utba b. Abl Sufyan, for example, who was 'Umar's tax agent over the Kinana tribe, wanted to use the money he made from the post for trade. The Quraysh in particular, perhaps because of their long-standing commercial contacts in Syria, may have been especially eager to see the state expand in that direction. There were also other financial advantages that would accrue to the elite from thc cxpansion of the statc: the acquisition of properties in the conquered areas, the ability of the state to levy taxes on conquered populations, the booty in wealth and slaves, at least some of which would reach the ruling elite even if they did not participate actively in the campaigns that seized them.
Finally, there is the possibility that members of the elite saw an expansion of the state as necessary in order to preserve their hard-won position at the top of the new political hierarchy. The policy of encouraging tribesmen to emigrate, upon which the continued dominance of the elite in part rested, was itself dependent on the successful conquest of new domains in which the emigrant tribesmcn could bc lodged. This view suggests that the conquest of Syria and Iraq was one of the objectives of the ruling elite from a fairly early date. On thc other hand, it is also possible to argue that the conquest of Syria and Iraq were merely side effects of the state's drive to consolidate its power over all Arab tribes, including those living in the Syrian desert and on the fringes of Iraq. This process generated the direct clashes with the Byzantine and Sasanian Empires that ultimately led to the Islamic conquest of Syria and Iraq, but that does not ncccssarily imply that the conqucst of Syria and Iraq was a conscious objective of the ruling elite from the start.
These considerations can serve as plausible guesses as to why the statc and its dominant elite adopted an expansionist policy. In the absence of real primary sources that might illuminate for us the actual motivations of individuals and of the elite as a body, they will have to remain merely guesses. The true causes of the Islamic conquests -- currents in the minds of men -- will probably remain forever beyond the grasp of historical analysis.
Not all of the political changes that came with the rise of Islam proved to be of equal durability, of course. It is perhaps ironic, however, that of the two basic political developments that marked the rise of Islam -- the integration of Arabian society including the nomads into a unified state, and the emergence of a ruling elite that dominated that state -- the latter should prove more durable than the former; that is, the Islamic ruling elite (or a descendant of it) showed itself able to survive long after its original Arabian-lslamic state had disintegrated. Not even the elite weathered the first decades after the conquests completely unchanged, however. Soon after the opening of the conquests, the elite began to undergo a transformation that pitted one branch against another, so that it became increasingly narrowly defined as successive groups were eased out of positions of real influence. At the outset, as we have seen, the elite included tribesmen of Medina (the ansar), the Meccan Quraysh, and the Thaqif of al-Ta'if. But even Muhammad himself had been hard put at times to control the rivalries among these groups, and after his death these rivalries became sharper and eventually broke out in open conflict. The selection of Abu Bakr to be Muhammad's successor as head of state, for example, was made possible only by assuaging the fears of the ansar: that they might be overpowered by the Quraysh. Despite such assurances, however, the Quraysh seem in any case to have risen quickly to a position of practical dominance over other elements in the elite during the conquest period. The ans-r, in Abu Bakr's day, were already worried enough to demand of him, "Who is in charge of this affair? Do the ansar have a share in it?'' A bit later, during the caliphate of'Umar, the governor of southern Iraq, 'Utba b. Ghazwan (a man of B. Qays but a longtime resident of Mecca and ally of Quraysh) complained that the Qurashi commander in central Iraq, Sa'd b. Abl Waqqas, ordcred him about. 'Umar replied to his demand for an independent command by saying, "It is not for you, 'Utba, to be instated with authority over a man of Quraysh who is a companion of the Prophet and a man of honor." 'Utba reminded 'Umar that he, too, was a companion of Muhammad and, as an ally of Quraysh, entitled to be treated as one of them, but 'Umar refused to alter his stand. Similarly, thc general Abu 'Ubayda b. al-Jarrah is said to have opened an address to the Syrians whom he governed by stating, "Oh people, I am a man of Quraysh (appointed) over Syria" -- not a Muslim over Syria, we may note, but a man of Quraysh. His use of this phrase may have been related to the way in which conquered areas in Syria appear to have becn "reserved" especially for the Quraysh, whereas the ansar and Thaqlf were sent more frequently to Iraq. But it is also probably reflective of the general rise of the Quraysh to real dominance within the ruling elite.
The Meccans thus seem to have nudged the ansar out of real power during the period of the early conquests, until the two groups split openly during the First Civil War, when the ansar tried to restore their faded fortunes by backing the faction around 'Ali b. Abi Talib against two other factions representing rival groups within the Quraysh. The ansar lost the struggle, however, and after the First Civil War were for practical purposes no longer a part of the ruling elite. Certainly the caliphate seems to have become the unique preserve of the Quraysh by this time. As for the Thaqlf, they seem to havc avoided a direct clash with the Quraysh, but then they never appear to have posed quite the same challenge to the Quraysh's domination of the elite as had the ansar. Even at the start they seem to have been part of thc elite mainly by virtue of their long and intimate affiliations with the Quraysh, and after the First Civil War they, too, slowly slipped into oblivion, retaining a vestige of their former importance, perhaps, in their accustomed tenure of certain governorships, notably in Iraq.
By the post-conquest period, then, the struggle for dominance within the elite had become exclusively a question of which branch of the Quraysh was to rule. This issue was raised already in the First Civil War in the form of a struggle between the B. Umayya, led by Mu'awiya b. Abl Sufyan, the B. Hashim, led by 'Abi b. Abi Talib, and other branches of the Quraysh, led by Talha b. 'Ubaydallah and al-Zubayr b. al-'Awwam. The Second Civil War (A.H. 60-73/A.D. 680-692) saw a similar struggle between the Alids, branches of the B. Umayya, and an alliance of other Quraysh led by al-Zubayr's sons. The issue was raised yet again in the Abbasid coup of A.H. 132/A.D. 750, when the B. al'Abbas (a lineage of B. Hashim) ousted the B. Umayya from power and had most of them murdered, and yet again in the numerous rebellions of various Alid pretenders against the Abbasids -- that is, in a protracted struggle between two rival factions within the B. Hashim. The ruling elite had thus been successively narrowed to limit leadership of the state first to the Quraysh, and then to a few select lineages of the Quraysh. It is interesting to note that in later years the debate over who had the right to lead the Islamic community eventually emerged in a curiously Arabian formulation, even though the protagonists were by now only in the most attenuated sense Arabians. The arguments used by those groups that were the main rivals for power within the elite (notably the Alids and Abbasids) came increasingly to rest on considerations of genealogy, whereas those of groups outside the elite that wished to gain access to it (notably the pious and the Khawarij) relied increasingly on the importance of virtuous, properly Islamic behavior as justifications for holding power. This dichotomy contains a curious echo of the notions of nasab (nobility of desccnt) and hasab (nobility of action) current among the pre-lslamic Arabian aristocracy as the principles validating their claims to authority and noble status.
The decades immediately following the conquests, marked as they were by two civil wars, constituted a period of real political turmoil in the Islamic state. But these quarrels among members of the ruling elite and the turbulence they generated were not caused by the failure of the original process of consolidation by which the elite had integrated Arabia's tribesmen into the state. Indeed, the striking thing about the First and Second Civil Wars is the degree to which the tribesmen remained bound to the state throughout them, even though the leadership of the state was divided against itself. The tribesmen waited out these squabbles in the ruling elite, or plunged in on the side of one or another group in the elite, but it never seems to have occurred to most of them that they should or could raise the standard of revolt in their own name. With very few exceptions, they appear to have accepted their status as subjects or employees of the state and its elite without demur; those tribesmen who tried to evade state control by forming little bands of escapees from the garrison towns, and by raising havoc in the Iraqi and Iranian countryside as Khawarij, were a mere handful of Arabia's population. The victory of the Islamic state over the bodies and minds of Arabia could hardly have been more complete.
Yet, as noted above, this firm integration of the tribes of Arabia by the state was not to be long lived. Arabian society, so long politically fragmented and outside state control, was soon to revert to its original disunity. The Islamic state's victory over the tribes was thus to be a phenomenon unique in the history of the peninsula until modern times. The ultimate disintegration of the Arabian-lslamic state should not be taken, however, as an indication of any decline in Islam's power as an integrating ideology. Indeed, the integrating power of Islam continued to work on an even more extensive scale after the conquests, as it wove disparate communities together to form the rich tapestry of medieval Islamic civilization, linking individuals and groups thousands of miles apart with a sense of common heritage, common values and beliefs, and common goals in life. Rather, the collapse of the Arabian state had more mundane causes. As we have seen, the political integration of Arabia by the early Islamic state was the product of two primary elementsthe integrating concepts implicit in Islam and the vigorous and skillful pursuit of political consolidation by a group of leaders well versed in the techniques of traditional Arabian politics. It was the latter, not the former, that ultimately faltered. Ideologies are like the switch settings and signal lights that control the movement of trains. The signals must be green, the switches open, if a train is to proceed down a certain track; but simply opening the switches will not in itself move the train -- one has also to undertake the difficult chore of firing up the engine and setting it in motion. The Islamic state, first established on the firm domination of Arabian society by the ruling elite, underwent a transformation because the focus of the state moved out of the Arabian milieu, and its rulers gradually lost sight of the fundamental principles of Arabian politics; they forgot, so to speak, how to fire the engine. The caliph Mu'awiya b. Abl Sufyan (A.H. 41-60/A.D. 661-680), though ruling from Damascus, still clearly understood the principles of Arabian politics; he drew on all tribes as much as possible for his military backing, playing them off against one another when necessary, but keeping the interests of all tied in one way or another to the state that he ruled by means of his deft tribal diplomacy. His brilliant lieutenant and foster brother, Ziyad b. Abi Sufyan, kept a close rein on the turbulent tribesmen of Iraq in much the same way; by using one group to control another, and by using the tribal notables to control the masses of ordinary tribesmen, he managed to keep most groups of tribesmen from becoming too powerful and preserved among them at least a modicum of interest in serving the state. But then both Mu'awiya and Ziyad had grown up in tribal Arabia and were familiar with its politics; Mu'awiya, as a boy, had served Muhammad himself as a scribe and participated in the conquest of Syria, and Ziyad had been raised in al-Ta'if by the family that led the conquests in southern Iraq.
As decades passed, however, the caliphs became less and less Arabian. Though descended from the Arabian Quraysh, they were usually raised in new capitals, Damascus or Baghdad, and hence did not grow up with the importance of controlling tribal particularism or nomadic power so constantly before their eyes. Shortly after the Second Civil War, a drastic shift in the basis of the Islamic state's power had already taken place as the result of an intentional change of policy; the caliph 'Abd al-Malik (A.H. 65-86/A.D. 685-705) constructed his new army primarily on tribes of Syrian origin, leaving the bulk of Arabia's tribesmen few ties to bind them to the state. Indeed, he demilitarized the great garrison towns of al-Kufa and al-Basra in Iraq, erecting between them a new garrison at Wasit, which he manned with loyal Syrian troops whose charge was, as much as anything, to keep the Arabians of al-Kufa and al-Basra under control. It marked the end of the process of integration by which the Islamic state had made Arabia's fragmented population a part of itself, and on which the original power of the state had rested. Later Umayyad caliphs, for various reasons, narrowed their base of support even further, relying not even on all Syrian tribesmen but only on one or another faction of Syrians. The integration of Arabia, concurrently, became ever more fragile, and the peninsula slipped slowly but steadily beyond the state's control. By the time of the Abbasid takeover (A.H. 132IA.D. 750), when the new dynasty replaced Syrian with settled Khurasani tribesmen as its base of military power, the Near East had returned to the geopolitical pattern that it has displayed in most historical periods, before the rise of Islam or after it: a powerful state or states centered in the Fertile Crescent and Egypt, supported especially by the rich taxes to be drawn from those regions, relying on standing armies made up of settled soldiers native to those regions -- and struggling with variable success to extend its control over, or at least to keep at bay, the nomadic warriors of the Arabian peninsula who remained generally outside its firm grasp. Arabia reverted to being the land of the nomads. Islam and the Islamic state survived and thrived, to be sure, but the great experiment in the political integration of Arabia by the Islamic state had come to an end.