The shadow of Fergus Millar falls obliquely across Olivier Hekster's latest imperial venture. Caesar Rules is dedicated to Millar's memory. Its subtitle knowingly riffs on Millar's influential The Emperor in the Roman World (31 BC–AD 337) (1977; 2nd edn 1992), reviewed brilliantly and with careless cruelty in this journal by Keith Hopkins (‘Rules of Evidence’, JRS 68 (1978)). Aside from an excoriating critique of Millar's methodology, Hopkins objected to the limits of the project and its emphasis on the emperor as administrator and judge. Millar's credo that ‘the emperor was what the emperor did’ took no account of ‘what people thought about the emperor, what they believed him to be doing or to have done’ (186).

Hopkins’ review is not cited by H. But it is worth a skim read by those interested in H.'s construction of his own intellectual genealogy. It also underscores his diplomatic delicacy: Caesar Rules boasts a genuine reverence for Millar's work (17, 160), while offering sotto voce a productive response to Hopkins: ‘The emperor may well have spent most of his time on administrative affairs (and in that sense the emperor was what he did), but most of his subjects expected him to be doing something different’ (163).

For H., these expectations are key. Ch. 1 (‘Portraying the Roman emperor’) looks in detail at imperial titulature and at how emperors were represented. It offers a fine discussion of imperial regalia, especially the crown and sceptre (81–101). Ch. 2 (‘Playing imperial roles’) examines the expectations of emperors in religious, civic and military contexts, the last a welcome corrective to Millar's desk-bound monarch. H.'s use of numismatic ‘big data’, in part based on the soon to be published work of Corey Ellithorpe, to capture some sense of mass image production is particularly interesting (170–7). Ch. 3 (‘Being around the emperor’) explores the relationship between emperors and courtiers, senators, bishops and the imperial family, especially wives and heirs. Ch. 4 (‘The emperor in the capital and provinces’) offers a set of rewarding observations on the expectations of the emperor in Rome, Constantinople and the provinces as a civic, military and religious ruler.

While emphasising the constraints imposed by the ever-accumulating weight of tradition, H. also insists on the ‘kaleidoscopic image of emperorship’ (330) that resulted from varied — often conflicting and incompatible — expectations of imperial rule. Emperors were ‘men for all seasons’ (331), yet ‘there was no endless flexibility’ (331–2); ‘emperors could not simply present themselves as they saw fit’ (182 and 332). Rather, ‘context created emperorship’ (326). ‘Playing the right roles for the relevant people was the best way to become the perfect emperor’ (182). Expectation, variation and limitation are fused. H. allows the near-paradox to stand: ‘Roman emperors continued to be different figures to different people. They differed in similar ways for a very long time’ (329).

H. is a confirmed gradualist. His account foregrounds the continuities across six centuries of Roman imperial rule. The contrast between the bare-headed Augustus and his bejewelled late-antique successor is ‘only part of the story’; apparently later elements ‘were already part of the (visual) vocabulary of the early empire’, sometimes criticised at their inception, but ‘less contested in the course of time, mainly through repeated exposure’ (103; see too 14, 242–3, 257–8, 326–9). Similar arguments are traced, for example, on Diocletian's ceremonial reforms (74–7, 105, 168); on the rise of court eunuchs (216, 257–8); on Constantine's diadem (‘a substantial innovation … that developed from established practice’ (90)) and his favouring of Christianity (‘perfectly in line with centuries of precedent’ (141; see too 145, 182, 279)); on the relationship between late-antique emperors and bishops (‘in many ways analogous to that of emperors and senators in the earlier empire’ (202)); and, perhaps somewhat less convincingly (even on H.'s own terms), on the shifting composition of the emperor's entourage (259) or Justinian's abolition of the consulship (194–5).

In Caesar Rules there are no Roman revolutions. Some will prefer to strike a different balance between continuity and change. Others may wish to propose something rather more analytically hard-edged than H.'s artful irresolutions: ‘everything changes, nothing perishes’ (104, channelling Ovid, Met. 15.165). However this is to be argued out, H.'s signal achievement is to offer — generously, intelligently, magisterially — a set of rich and detailed discussions of remarkable range and scope. Importantly, too, H. is as attentive to the visual as the textual. Millar's Emperor was unillustrated and uninterested in images. Caesar Rules is beautifully presented with over sixty high-quality plates. In short, beginning with Julius Caesar and pushing two centuries further than Millar, H. has put thinking about the emperor in the Roman world on an impressive new footing.

But there is still room to dream. There are very few fantasies in Caesar Rules (105, 221). Yet, it might gently be suggested, the construction of an imperial image is as much about imagination as it is about expectation. In thinking about power, legitimacy or authority, make-believe matters. Once upon a time the body of a centaur was sent from Arabia to Rome. The creature — as it turned out, disappointingly diminutive — was put on display in the palace and then carefully conserved and stored in an imperial warehouse (Phlegon of Tralles, On Marvels, 34–5). Of course that is not true. There are no such things as centaurs. Only a Roman emperor could plausibly be imagined to be sufficiently powerful to acquire a specimen and order it to be stuffed, mounted and added to the royal collections.