Fashioning state and subject in late antique and early medieval North Africa (AD 500--800)

This dissertation uses the material evidence of landscapes, built space and architecture to examine how the Byzantine, Umayyad and `Abbasid states established imperial authority in North America, and the related question of how subject peoples experienced empire.;Chapter 1 introduces the scholarly background of the project, particularly as it relates to our understanding of the Byzantine and Arab empires in North Africa. It argues that in Late Antiquity, scholars have tended to evaluate states not empires.;Chapter 2 outlines a framework for imperialism that explores how state authority is generated and transmitted through spatial practices and what kinds of new subjects each empire produced in late antique North Africa. It suggests that current archaeological scholarship on empire is characterised by divides between representational and material practices, local and global spaces, and state and individual. It argues that we ought to take the spatial implications of imperialism seriously and examine how expressions and negotiations of empire are imposed, worked around and worked out through discursive and material spatial practices at different scales.;Chapter 3 lays out the available evidence for late antique and medieval North Africa and a methodological framework for analysing spatial change. It emphasises the chronological and regional limitations of the evidence, and suggests that an over-reliance on ceramic finewares as dating tools has affected our ability to identify seventh- and eighth-century activity.;Chapter 4 examines the discursive nature of Byzantine imperialism. It argues that the shift towards a Christian spatial and ideological conception of the Roman empire, Romanitas and the role of the emperor has important implications for the study of Byzantine imperialism in the sixth century.;Chapter 5 moves from the rhetoric of urban renewal to explore how urban topographies were affected on the ground. Detailed analysis of urban plans from Tunisia, Algeria and Libya exposes major changes in built space and a mass investment in new religious, military and economic networks. I suggest that this reflects the Byzantine state's attempt to use churches and forts, as both ideological and economic institutions, to legitimise the conquest, control trade and taxation and create social consensus through religious orthodoxy.;Chapter 6 is in many ways more a methodological critique of current assumptions about the rural landscape in the Byzantine period than a study of the relationships between the Byzantine state, rural peoples and exploitation. It suggests that the dating tools currently used by archaeologists to identify late antique occupation---finewares, especially African Red Slip ware---underestimate the real level of rural settlement. The chapter concludes that until we have a better understanding of coarsewares, scholars should move away from attempting to address demographic issues from survey evidence and instead focus on the more measurable questions of interregional connectivity and settlement morphology.;Chapter 7 provides for the first time, a systematic overview and analysis of the archaeological evidence for early medieval North Africa. Despite difficulties in dating early medieval occupation, synthesis of the available evidence reveals that the Arab conquest was not catastrophic for settled life, as is often assumed. In the Umayyad and 'Abbasid periods, the Arabs were only a small minority, who maintained their dominance through military means. Most major towns seem to have been garrisoned by a djund of Arab troops who were housed in the Byzantine fortress, or less frequently in a newly constructed fortress. Mosques were built inside Arab-garrisoned forts or in separate extra-mural settlements. Islam, then, was used to reinforce divides between colonising Arabs and colonised Christian North Africans. Motives of group solidarity, political cohesion and a desire to establish separateness from their new African subjects contribute to this spatial segregation between djund and North Africans.;(Abstract shortened by UMI.).