Control of Egypt and Iraq made the Umayyad Caliphate the sole southwest Asian superpower; steppe nomads to the north of the Caucasus were the only military force capable of meeting the Caliphate on equal terms. The conquerors inherited a strong fiscal organization, and over 80 percent of government revenues went to support of the army. However, much of that revenue was collected and disbursed locally, severely reducing the power and flexibility of the central authority. Meanwhile, the legitimacy of caliphal rule was eroded by the succession dispute that created the party of Ali, on the one hand, and by the anti-imperialist ideology of the tribal traditionalists in the ulema, on the other. The party ofAli disputed the Umayyad claim to the Caliphate, while the tribal traditionalists objected to the urban, royal style of the governing house.
The professional Syrian army was the bedrock of Umayyad power, but it carried its own problems. It was tied together by personal allegiance to the military leaders who had access to the fiscal resources of the state through their control of the governorships of the provinces. In one way, apparent tribal divisions continued to be important; however, old tribal names provided the rationale for division of the army into factions competing for control of those same governorships, which were appointed by the caliph. The governors, in turn, worked to control the caliph, who was not usually a military leader. The Umayyad Caliphate was therefore characterized by continual factional infighting in the provinces that nevertheless kept the state unified because it was directed at control of the center. The professional army, in effect, played politics to guarantee its continued access to the tax income that was, in the absence of a landed aristocracy, its only support. This also caused most subjects of the Caliphate outside Syria to view the army as a semifor- eign occupying elite. When factionalism spread to the capital, the Caliphate imploded in the third civil war of 747-750.
The Abbasid State Victory in the civil war by the Abbasids, based in Khurasan in northern Persia, resulted in the movement of the capital to Baghdad in Iraq and the replacement of the army of Syria by the army of Khurasan as occupiers of the provinces. The Abbasids elaborated the machinery of state: The bureaucracy grew, and a spy system kept an eye on provincial leaders. The Abbasids should have been stronger than the Umayyads, but their regime still lacked legitimacy in many Muslim eyes and still had a professional army that played politics to guarantee its access to money, outside of any effective civilian control. The Umayyads had at least been able to claim leadership of Arab tribal society with spiritual ties to Arabia. However, the Abbasids could satisfy neither the Shi’ites with their hereditary claims (they were of the Prophet’s Hashimite clan but descended from Muhammad’s uncle al-Abbas, not through Ali) nor the ulema traditionalists with their claims to have saved Islam from Umayyad oppression, as their state apparatus was even more developed. Unable to base their rule on minority Shi’ite support, the Abbasids were forced into the arms of the ulema. But the latter’s hostility to any mechanisms of power that impinged on society—for example, Muslim law has injunctions against enslaving or keeping in servile status fellow Muslims, which made conscription problematic from early in Islamic history—especially its powerful classes, resulted in the failure of the Abbasids to create viable service aristocracies, either civil or military. Unable to draw on the resources of their society and faced with a politically unreliable professional army whose factions fomented trouble, Abbasid leaders turned increasingly to their personal dependents to run the state. And to ensure their reliability, such dependents were increasingly foreign and sometimes servile.
Arab Armies
Though this manuscript illustration of the civil war between Muhammad’s cousin Ali and his adversaries—the origin of the Sunni-Shi’a division—dates to the thirteenth century, it accurately shows the dominance of sword-armed infantry and cavalry in early Arab armies.
The Rise of Foreign and Slave Soldiers
The same process that affected the civil service also affected the army, combined with a growing tendency to recruit outside the empire. Starting in Spain and North Africa in about 800 and in Baghdad by 830, when Persian aristocrats had failed in ruling the capital, the leaders of the Abbasid state turned to foreign soldiers for their military forces.
The key characteristics of these armies were a powerful separation of the soldiers from society and an equally powerful dependence on the ruler for their livelihood. Separation and dependence came from foreign origins and, increasingly in the century after 850, from servile status. (Servile status was comparatively rare among the initial Abbasid recruits.) In the hand of a strong ruler, such armies could be effective instruments of control. And when drawn from steppe nomads, especially the Turks on the northeastern borders of the empire, they could be efficient military forces—it was with the rise of these armies that horse-archery began to supplement the infantry tactics of the early Arab armies as the dominant Muslim military style east of Egypt. (In Africa, and later in Muslim-controlled areas of the Balkans, mamluks, or slave soldiers, usually served as heavy infantry.) But their dependence on the ruler and separation from society and thus from meaningful politics meant that they were not a military-social elite, but instead were elite military automata. Their usual connection to politics was through palace coups designed to ensure their access to monetary support. Many mamluks were manumitted (freed), and many rose to positions of wealth and importance; in extreme cases, as later in Mamluk Egypt, mamluks took over the state. But their separation from civilian Muslim society remained profound.
Mamluk armies—armies of slave soldiers—were therefore the ultimate expression of Muslim states’ problematic relationship to Muslim society and the failure of Muslim civilian administrations to effectively control their armies. The institution spread rapidly throughout the Muslim world and no farther: Virtually no one else adopted this military model (see the Issues box “Slave Soldiers”), nor could most Muslim states do without it. It seems an odd result for a religion that makes religious law so central to all aspects of society. But then, Islam is a world religion that began as the exclusive tribal faith of the Arabs, so paradoxical results are not unknown to it.
Fragmentation and Decline
Unfortunately for the Abbasids, the beginnings of mamluk armies coincided with rulers who either were weak or had no field on which to prove their military prowess. The borders of the empire were distant, and the caliph was not preeminently a military leader. As a result, almost as soon as mamluk armies came to Baghdad, they started causing trouble. Political chaos ensued, and within a century, the Abbasid revolution resulted in the breakup of the unitary Islamic empire. An Umayyad restoration separated Spain; the great provinces—Egypt, Syria, Persia—became effectively independent; then even the provinces fragmented into smaller warring emirates. Fragmentation brought a decisive end to Islamic expansion and, in fact, even allowed Byzantium to resume the offensive in Palestine in the 900s (see below) by picking off frontier emirates one at a time. The Caliphate itself receded into feebleness and figurehead status, and an era of smaller Islamic states began.
Military Organization
These smaller Islamic states resembled the post-Roman states of western Europe in drawing on certain common elements of military organization throughout the Muslim world, though with significant regional variations on the theme. At the core of most states’ armies were the personal slave soldiers of the emir, sultan, or other political leader. Indeed, at times (as in Mamluk Egypt), the slave army dispensed altogether with a nonslave leader and took over governance directly. Such armies were replenished each generation with new imports of slaves bought or captured along the frontiers; military service rarely descended within mamluk families.
These cores provided the standing army for each state but were usually relatively small. Local levies of free men and, above all, foreign mercenary groups filled out the ranks in times of emergency. The iqta’ an institution imported first by Turkic slaves and later by Turkic conquerors, increasingly provided a decentralized way of paying for military manpower (as well as for nonmilitary administrators). An iqta’ was a set of rights to revenue of particular lands; these rights were granted to mamluks and nomadic tribesmen in exchange for seasonal military service. While the system bears a superficial resemblance to the system of fief holding developing in Europe at the same time (see Chapter 7), the differences are crucial. The iqta’ holder was a dependent with regard to central authority (not, as a fief holder was, a participant in that authority), and a mercenary with regard to the land (not, as a fief holder was, a steward and, effectively, owner). Iqta’ holders were thus neither an aristocracy of service nor an aristocracy of local power, but rather represented mamluk military organization spread across the land. In short, iqta’ did not remedy the lack of connection between settled Islamic society and the military states that governed it. Only at the edges of the Islamic world, where ghazis, frontier raiders, carried a still-military jihad to the infidel, was there any such connection, and the usually tribal ghazi was a marginal figure to mainstream Islamic society.
The Military and Society
The institutional heritage of the Abbasid revolution thus gave Islamic states after 950 (including those after 1050 such as the Seljuk and Ottoman empires and Mamluk Egypt) the character of permanent conquest societies. That is, the settled, agricultural societies of the Islamic world were consistently ruled by states dominated by tribal or slave soldiers from the nomadic pastoral world. The lack of organic connection between Islamic societies and states meant that the states had little internal opposition when they were strong but also little internal support in times of crisis or weakness. They therefore oscillated between despotism and anarchic infighting.
One crucial result of this character of Islamic states was a serious decline in agriculture in the Islamic heartland of Iraq. Periodically plundered by out-ofcontrol slave armies and ruled in the best of times by armies with little cultural interest in settled agriculture, peasants throughout the Islamic world escaped when they could, with pastoralism advancing at the expense of agriculture. In Iraq, where continued farming depended on maintenance of the complex irrigation system, the decline was particularly steep and hard to reverse. The decline of Iraq shifted power within the Islamic world to Egypt and increased the importance within the larger world of southwest Asia of the Turkic nomads to the north.
In fact, the widespread use of Turkic slaves as the core of Islamic armies opened a pipeline to the nomadic world that could not remain in the control of Islamic states. Periods of weakness or infighting created openings and effectively extended invitations for nomadic conquests. While such conquests often led to the revitalization of Islamic states and to temporary unification, they also reinforced the conquest nature of those states and maintained a distance between rulers and subjects that undermined political legitimacy, leading to the continuation of mamluk armies even after nomadic conquests.
Finally, the Abbasid heritage marginalized military power in Islamic society. Warriors were either geographically marginalized as frontier ghazis or socially marginalized as foreign slaves in the standing armies of Islamic states. The core of Islamic society was civilian, but unlike the civilian government and society of China, which kept close control over its armies, this was a civilian society with little control over or even connection to its wielders of military power.
A last flare-up of resistance to the Umayyad victory in the second civil war resulted in a battle that is well documented and illustrates some of the key features of early Islamic warfare. A small (perhaps 5000 men) army of rebels from southern Mesopotamia (Iraq) moved up the Euphrates valley toward the routes into northern Syria. They paused at the old Roman fortified city of Qarqisiya (Circe- sium) to buy supplies at a market outside the city. Sulayman bin Surad, leader of the rebel force, was advised by the governor of the city to make the most of the small size of his army and their many horses to beat the approaching Umayyad force, larger (perhaps 20,000) and mostly on foot, to Ra’s al-Ayn. He could then place the town and its water supplies (as well as a potential escape route) at his back. He was also advised to deploy his army in small units and a flexible order so that units could dismount or mount at need to support each other and to avoid a set-piece battle with the more numerous enemy, as he risked being surrounded by doing so.
Sulayman managed to occupy the appropriate ground first but disregarded the rest of the advice. He drew his army up in the standard formation of a line divided into a center and two wings and prepared for a set-piece battle, dismounting all his troops. When the Syrian army arrived, divisions in their command seem to have delayed decisive action for a time. But two days of skirmishing escalated by the third day into a full-scale attack in which the numbers of the Syrians began to tell. Sulayman had his men break their sheaths and advance in a final desperate attack, but the Syrians formed a spear line, sent in reinforcements that included a number of archers, which the rebel force lacked, and indeed surrounded the rebels. Sulayman died, and the banner of command of his forces passed down a predetermined chain until the last designated leader led a successful breakout that night and fled with the remnants of the rebel army back home.
While of little significance in any larger strategic picture, the battle does illustrate a number of features common to Arab battles, including infantry dominance, stalwart defensive tactics, and combined arms ability. It also stands as another example of the professionalism of the Syrian army that buttressed Umayyad rule.
Even the most powerful figures in history have at times been frustrated by limitations on the legitimate use of their power and so have turned to personal dependents outside the system to enact their will. Witness the eunuchs of Chinese emperors, used to circumvent the rules of the Confucian bureaucracy, or President Ronald Reagan’s similar use of Colonel Oliver North to circumvent congressional restrictions on aid to Nicaraguan contras. Such figures tended in the traditional world to be slaves, eunuchs, or women, because these groups stood outside the usual bounds of political legitimacy. Incapable of exercising authority on their own, they were utterly dependent on (and therefore in theory completely loyal to) the ruler who employed them. They also tended merely to supplement the operations of legitimate governments run by free males of the polity, however “free” was defined. Likewise, slaves have fought in wars throughout history, from Roman gladiators to African American slaves during the American Revolution. They appeared in large numbers, however, only in emergencies and were almost universally manumitted either before or after their service. Only in unusual circumstances did governments feel so restricted in their legitimate use of power that they depended exclusively on outsider groups to enforce their will through the creation of slave, or female, armies that excluded free men.
This condition was most pervasive in the Islamic heartland, where the conquests effaced pre-Islamic sources of political authority without providing an adequate substitute, as traced elsewhere in this chapter. Mamluk armies became a standard feature of this civilization, the only major civilization in which slave armies are known. But the effect of the conquest and the problem of maintaining a paid professional army without an established bureaucracy or service aristocracy—and not anything intrinsic to the Muslim religion—were vital in the emergence of the mamluk institution.
The institution was never directly copied but arose independently outside the Muslim world a number of times, especially in the slave-trading societies of west Africa. The Wolof kings of Senegal, for example, enslaved their own subjects (much as the Ottomans would do to sustain their Janissary Corps) and relied almost exclusively on slave soldiers. A state that sells its own subjects and collects protection money from those it doesn’t sell clearly exercises little claim to legitimate use of power. In Islamic armies, foreignness complemented servile status in alienating soldiers from society. Gender served the same function in some other societies that used female slaves as palace guards—for example, Mauryan India, the kings of Dahomey, and a fourth-century ruler of Hunnic invaders of China. In all these cases, something that fundamentally alienated the ruler from the society, whether the foreign origins of imperial institutions or a crisis of ethnic identity, prevented an appeal to the usual sources of political legitimation, and insecure rulers turned to unorthodox sources of support.
Thus, no matter how effective particular slave armies were—and mamluk armies of nomadic Turks were often very effective—their mere presence signified a deep problem of political structure and legitimacy in the states where they appeared.