THE EFFECTS of sudden changes in the political control of an area are often difficult to clarify even with the advantage of plentiful and reliable sources. How much more difficult it is, then, for historians of the late Roman and early Islamic states to understand the processes which followed the Arab Islamic conquests in Syria, Palestine and Iraq, is well-known, in view of the limited and fragmentary source materials from the Greek, Syriac and Arabic traditions. This affects not only our understanding of the actual course of events, but in particular of the structural transformations which resulted, or were exacerbated, as a consequence of these political changes. Administrative and fiscal institutional arrangements are particularly difficult to perceive through the partiality and the rhetoric of the literary sources; and even with the help of material remains - archaeological and epigraphic data, for example - our knowledge of what went on in these respects is extremely vague.

It is clear that the earliest Arab-Islamic administration of Syria, Palestine and Iraq, as well as of Egypt, relied initially on pre-existing institutional patterns and arrangements, both in respect of fiscal practices (methods of assessing and collecting “tribute” i.e. tax, for example) as well as civil administrative structures. In particular, it has usually been accepted that the early Arab ajnad or army-districts of Filaṣtīn, al-Urdunn, Dimashq and Ḥimṣ reflect pre-Muslim structures. But it has been recently argued that these are not the late Roman provinces as they are known from sixthcentury sources. On the contrary, they reflect the results of a major reform of the late Roman civil and military administration of Syria and Palestine, a reform which is supposed to have involved the establishment in the period 629-34 of military districts or themata (normally associated with the later seventh century and after in Byzantine Asia Minor) by the emperor Heraclius, primarily with the function of countering a future threat from the Sasanid Persian empire in the East.

In this contribution I will accept that some changes in the military and possibly also the civil administrative arrangements of the region were probably introduced at about this time, particularly with respect to the districts later incorporated into the jund al-Urdunn; I will also accept that the ajnād do not represent late Roman civil provincial boundaries. On the other hand, I will argue that there is no connection with any imagined Heraclian theme system, and that the ajnād actually represent nothing more than the regions across which several of the duces operating in Palestine I, II and III and Phoenice I and II commanded their troops, together with the associated ḥāḍirs or encampments which served as the bases for the numerous Arab federate and allied forces of the late Roman military establishment in the region. Furthermore, it will be argued that it is the ajnād and the network of camps and garrisons associated with them which can best explain the absence of substantial new amṣār or military bases, and consequently of major new urban foundations, during the post-conquest period in Syria, in contrast in particular to Iraq and Egypt.