UNLIKE the Western Mediterranean, the Eastern Mediterranean saw a continuous survival and succession of state systems throughout the Middle Ages. At no time was the contrast between the two sharper than in the later sixth and seventh centuries, when in the West - outside the Byzantine lands - the fragments of Roman imperial government and authority were being reconstructed in an ad hoc way by all manner of ethnic leaders (Roman, German, Berber) whereas in the East, despite huge conflicts between the principal states of western Eurasia, the Roman/Byzantine and the Sasanian empires were succeeded by the Arab-Islamic Caliphate without a break. At both ends of the Mediterranean, nonetheless, this century and a half saw major discontinuities in nearly every set of traditions that had been inherited from the Roman empire, and the establishment of the ancestors of the major polities that dominated the Mediterranean for the rest of the Middle Ages. Given the relatively small amount of evidence for any of the major historical sequences, and given that in every region much of it is late or otherwise problematic, it is also not surprising that almost everything about these discontinuities remains under debate. Indeed, almost nothing is agreed in this period, outside the narrowest political outline, either in the West or in the East. Overall interpretations of political, social, economic, culturaVreligious change are particularly resistant to consensus.

I make such comments not because they are new or controversial - about these, at least, everyone will probably be agreed - but in order to frame a brief concluding commentary on the articles presented in this book. It is in fact daunting to realise that these articles, none of them particularly polemical, and all of them firmly grounded in a wide range of data, disagree on almost every issue that affects any form of wider understanding of Byzantines, Sasanians and Arabs. Where they do agree, other equally authoritative experts could in every case readily be found to contest them. Was the Sasanian state highly centralised (Howard-Johnston) or potentially undermined by the old aristocracy (Rubin)? Was there a strategic and organisational centralisation of the early Islamic polity (Donner) or a largely autonomous system of local armies (Kennedy)? (This difference in emphasis between the two latter writers remains, despite their substantial accord in empirical detail.) The Byzantinists were more in agreement, both with respect to the essential stability and strength of the political system up to 600 (Carrié, Whitby, Isaac), and its very slow and ad hoc reorganisation after the Persian wars and the catastrophe of Islamic conquest (Lilie, Haldon), but even here a gallery of active historians had to be invoked at every stage who thought the opposite. It may be noted that some of these disagreements are linked: people who think, for example, that the sixth-century Byzantines were well-organised and active are more likely to think that their enemies the Sasanians were highly organised as well. Fortunately, however, these logical chains did not turn the contributors simply into members of rival historiographical teams; rather, their disagreements resulted in a truly stimulating and wide-ranging debate in the Workshop itself, which was for many the most exciting part of the proceedings, and which has in many cases continued informally since.