Evidence for discussing grand strategy in the Roman Empire is extremely limited. In reaction to Luttwak’s thought-provoking book, ancient historians have tended to focus on limitations in the information available to emperors and hence their capacity to determine priorities. Kagan’s reformulation of Luttwak’s thesis to use troop movements as a proxy for strategic decision making does not adequately take into account the personal and cultural considerations that often influenced rulers, and in particular that suppression of internal rivals took priority over external threats. That said, grand strategy remains a useful tool for investigating imperial decision making, especially for emperors such as Augustus and Diocletian, both of whom had to stabilize the empire after protracted bouts of civil war and enjoyed long reigns. It is also relevant to the later empire (after AD 300) when rulers regularly had to balance threats and opportunities on different frontiers against the more limited resources that were available.

My title quotes the opening sentence of the chapter on ‘The civil service’ in A. H. M. Jones's magisterial survey of the social, economic and administrative history of the late Roman empire published in 1964 – a work that still defines the field even after half a century. One might quibble about the emphasis in ‘before all things’, but it is difficult to deny that the late empire demonstrated many aspects of a developed bureaucracy as defined by Max Weber – for example, full-time salaried officials organized into hierarchies of authority, both centrally and regionally, with particular spheres of responsibility and effective communications to ensure cohesion. Members of the administration might not always be exactly aligned with the policies of their imperial masters, as can be seen in the ruse used by Emperor Anastasius (r. 491–518) to abolish the chrysargyron tax, or in the opposition of the praetorian prefect, John the Cappadocian, to Justinian's attack on Vandal Africa in 533. This capacity to resist emperors, however, also meant that administrative structures could survive through regime change, as in the transition from Roman to Ostrogothic Italy in the late fifth century, or the political chaos of the early years of Zeno's reign (late 470s) or the usurpation of Phocas (r. 602–10). The late empire was a territorial state whose frontiers were, in the main, fixed, though this did not prevent the Romans from attempting to exercise authority outside their formal territory, in line with their centuries-old approach to giving orders. Jones's perceptions of the late empire were undoubtedly influenced by the nature of imperial control and organization in the earlier stages of the Roman empire. He, after all, embarked on his study of late Rome after producing authoritative volumes on The Greek city and The cities of the eastern Roman empire, as well as working on the construction of the Roman empire under the first two holders of the title of princeps, Augustus (d. 14 CE) and Tiberius (d. 37 CE).

The early Roman empire has been characterized as ‘undergoverned’ by comparison, for example, with the Chinese empire. This assessment appears in the chapter entitled ‘Government without bureaucracy’ in the authoritative introduction to the empire's economy and society by Garnsey and Saller.

The acts of church councils offer us an exceptionally rich source of information about the utterances and behaviour of large numbers of bishops while engaged in one of their most important duties, namely the collective establishment of orthodoxy and the identification of heresy. They provide examples of Christian leaders in action, not only of individual bishops who in their own cities would be regarded as leaders but in the context of an ecumenical council were overshadowed by their metropolitans or the patriarchs or other key figures prominent in a particular debate, but also of a small number of international leaders. Although the absolute accuracy of council records is open to challenge, with the impossibility of verbatim precision, especially during heated moments, being acknowledged by those responsible for attempting to produce the records, the general impression of the tone and conduct of debates is beyond challenge, while the subscriptions by individual bishops to various decisions also offer insights into how the participants wished their involvement to be registered and remembered.

Episcopal behaviour has recently been the subject of two illuminating studies. The first, by Claudia Rapp, identifies different strands of episcopal authority – spiritual, ascetic and pragmatic –, probes how these were combined to legitimate that authority, and considers how bishops as men of power operated within the evolving economic and social structures of their cities.