Organizational Power I: The Structures of Provincial Government

The scale of the conquests created a logistical challenge. This was partially met by the capacity of the Arabian nomads to exploit the infrastructure of the defeated empires. The use of express riders for communication was already known in pre-Islamic Arabia (band, from Greek beredarion); after the conquests, Arabian, Roman, and Iranian practices were combined.

Another response to the problem of scale was the empire's devolved provincial structure, with power vested in a few governors (amirs) who ruled large swaths of territory. There were other provincial administrative officials, among them the sahib al-shurta (chief of the militia), the muhtasib (market inspector), the qadi (judge), as well as katibs (scribes) and amils (officials) managing administration and tax. Most had analogues in pre-Islamic administration, and uniform practice was only established gradually. However, power tended to be concentrated in the hands of the amir; down to the early Abbasid period, most of these provincial appointments were in his hands.^[1027]

The first imperial structure was built around three large provinces ruled from Medina, each determined by the shape of the conquered administration. Syria and its four (and then five) military districts (junds) coincided approximately with the mil­itary spheres of influence of the senior Roman commanders in the region (duces).^[1028] Unusually, it retained Damascus as its administrative capital—perhaps a reflec­tion of the continued importance of the Romano-Syrians there. Likewise, the basic structure of Roman Egypt was retained, but ruled from a new garrison at Fustat, near the Roman fortress at Babylon, just south of the Delta.^[1029] In contrast, the Sasanian “Western Quarter” (Khvarvaran) was divided in two by the conquests.

When Damascus replaced Medina as the imperial metropolis in 661, the Hijaz be­came a province, with governors of other regions of the Arabian Peninsula subor­dinate to a Medinan governor appointed by Damascus. The same period witnessed the extension of Islamic rule into Ifriqiya (Roman Byzacena), where the new garrison of Qayrawan was built in 670, and into Sasanian Khurasan, where the former cap­ital of Merv had been retained, and was garrisoned with troops sent from Iraq in the following year. Governors of new conquests tended to be appointed from the region that colonized them; Ifriqiya was subordinate to Egypt, and Khurasan to Iraq. Beyond these garrisons, and the territories they controlled, administrative power was limited; for example, it was only in the early Marwanid period that Islamic rule began to be ex­tended into northern Iraq (the Jazira), and into the rest of the Iranian plateau.^[1032]

Gubernatorial appointments reflect the evolution of imperial power. In the sev­enth century, key governorships were granted to relatives of the caliph, or other close tribal allies. Then, under the Marwanid Umayyads, there was a growing ten­dency to appoint men from more obscure origins, who had risen to prominence through their military and administrative abilities—a reflection of the growing institutional strength of the army and the bureaucracy. Nonetheless, this change should not be overstated: these appointments were mixed with those of members of the ruling dynasty, and close associations of kinship, marriage, and patronage still bound a small patrimonial elite together.^[1033]

After the Abbasid Revolution, the major provincial governorships (largely the same as those of later Umayyad times) usually went to the cadet branches of the Abbasid family.

The importance of the revolutionary army and the growth of centralized power also caused changes in relations between the imperial court and the periphery.^[1036] In the absence of primogeniture, two heirs had tended to be appointed to succeed one after the other, in an attempt to balance competing interests at court and in the armies. In the Umayyad period, the devolved imperial structure had meant that the key interest groups were within Syria—the Umayyad family and the Syrian tribes— and so the Umayyad caliphs and their heirs had usually remained in Syria. In con­trast, the more centralized Abbasid Caliphate had a metropolitan elite connected to the frontier armies, and provincial elites with interests at the center. As a result, Abbasid heirs, and the factions that supported them, were appointed to a new, supra-provincial level of administration. The division was usually between two crown princes, each with responsibility for one of the two prestigious frontier zones of Anatolia and Khurasan, and associated adjoining provinces.

Pressure from metropolitan factions and their provincial supporters tended to force the nomination of a second heir upon the caliph, and so violent conflict be­tween heirs took place in three successive generations, before a full-blown civil war erupted in 810-819. After then, regional and provincial appointments tended to be granted to army commanders and provincial warlords in return for loyalty to the incumbent caliph, bringing an end to an Abbasid dynastic presence in the prov­inces and an end to provincial involvement in the succession. With the collapse of the 860s, even indirect power over the provinces was ended; Abbasid administra­tive power was now limited to Iraq and western Iran; after 936, it was confined fur­ther, to a few quarters of Baghdad itself.