How do you know an institution when you see one? This, in very broad terms, is the question that K. addresses in this rich, engaging and sometimes frustrating book. To be more specific, it deals with the problems involved in devising a visual representation for the new institution of divinisation (that is, the creation of divi) and with the various attempts at solving those problems from the mid 40s b.c.e. down to the 20s c.e. The fact that it is occasionally frustrating is due primarily, I should stress, to the intractability of the problems, so that to some extent the experience of the reader simply reflects that of the contemporary actors who first grappled with them.

As K. points out, the book is ‘organized as a sequence of relatively distinct studies as opposed to a continuous, overall narrative’ (p. 13). The first chapter outlines his main arguments and approaches (with sections entitled ‘Three Claims’ and ‘Four Assumptions’), and introduces a leitmotif that will run throughout the book, the contrast between two potential statue types of Divus Julius that appear on coins of Octavian from 36 b.c.e.: although both depict him holding a lituus, the distinctive cult instrument of an augur, in one he appears as a priest, capite velato, and in the other wears a ‘hipmantle’, leaving the upper part of his body nude. The subsequent chapters parse this difference by exploring the resonances of these two image types across a wide range of precedents and parallels and by connecting them to the larger set of issues and options that swirled around the nascent institution of making men gods. Chapter 2 deals with the divine and semi-divine honours proposed for Caesar both during his lifetime and after his death. In the third chapter K. explores the tradition of augural representation, and in the fourth he demonstrates how it links Divus Julius to the Genius Populi Romani, Romulus and Quirinus. Chapter 5 takes up the problem of Caesar's portraits which, in sharp contrast to those of Augustus, never seem to have been regimented according to fixed types and which surprisingly lack, in all extant examples, the two most distinctive attributes attested in literary sources: the crown voted him while alive and the star that Octavian added after his death. In the sixth chapter K. investigates the roles proper to living emperors and to gods: the former auspicious, guaranteeing divine favour, and the latter propitious, granting divine favour. In Chapters 7 and 8 he traces the spread of the new imagery that had been devised to represent the new institution of the divi: how it was appropriated locally for both public and private cults of the emperor and then, crucially, how it was adopted for the representation of private individuals. As a result, ‘the differing statue types – for men as well as for the divi – crossed the boundaries of the categories for which they were originally devised, in both directions: the new gods continued to look like men, and men began to look like these new gods’ (p. 221).

This brief summary merely suggests the book's range of topics and richness of analysis. I was particularly struck by K.'s skill in teasing out the inconsistencies, ambiguities and uncertainties that arise after a deceased mortal has been made a god, when visual representations that celebrated the living man continued to exist side by side with those that, post-mortem, honoured the deity. Given the Graeco-Roman tradition of representing the gods in human form, how in fact could one distinguish between a mortal and a mortal-made-god? Yet it seems to me that at the heart of the book there is a key problem that remains unresolved and to some extent unaddressed: what in fact constituted the new ‘institution’ of divinisation? The first of K.'s ‘Four Assumptions’ is that ‘the new institution of consecratio profoundly transformed Roman society’ (p. 10); by consecratio he seems to mean ‘the formal senatorial decree of January 42 that proclaimed Caesar a god’ (p. 29), to which he makes repeated reference. In doing so he relies on earlier German scholars, who sketched out in some detail the formal legal procedures (see especially S. Weinstock, Divus Iulius [1971], pp. 386–91, building on the work of Wissowa and Mommsen). Despite their confident reconstructions, however, it is not clear how much we know about what was actually involved (much less when it took place; Weinstock does not cite any clear evidence for January 42). An inscription from Aesernia (CIL IX.2628 = ILS 72) that refers to Caesar as one quem senatus populusque Romanus in deorum numerum rettulit seems to indicate a senatorial decree confirmed by popular vote, but what was its content? Presumably not ‘Caesar is now a god’; the Senate was not in the business of determining whether a particular entity was or was not a god. Their job was instead to determine which entities would receive public cult. Their decree concerning Divus Julius must thus simply have prescribed some or all of the elements of a public cult. In other words, to talk of the Romans ‘making’ a deceased ruler a god, as some ancient as well as modern writers do, is simply a loose way of saying that they treated him as such in cult.

K. is too sophisticated a scholar to be unaware of these complexities, and he rightly insists, in the second of his four assumptions, that ‘institutions not only require representations, in the conventional sense, but they are themselves representations, and must be so regarded’ (p. 11). Nevertheless, there is a recurring assumption throughout that divinisation was primarily a political and religious institution, and that representations in the conventional sense were secondary, attempts to come to terms with a reality already established in another sphere. To take an example more or less at random, he says of a post-Actium coin representing Octavian as a ‘bringer of victory’ that it ‘represents how, as the visual language of the late Republic was transformed under the pressure of changing institutions – institutions both political and religious – the imagery of the monuments required different symbolism that reveals a new significance’ (p. 143). Yet as K. himself has indicated, these political and religious institutions are not qualitatively different from the visual representations that, in his analysis, they generate. It thus seems to me that he concedes to them rather too much authority and that as a result, rather surprisingly for an art historian, he does not give full weight to the crucial part played by visual representations in the process of ‘making men gods’. The visual representations were not simply a response to the institution of divinisation but were themselves an integral part of that institution, just as much as senatorial decrees and cult practices. If institutions are representations, so too representations are institutions; to see an institution is in fact to know one. This has important implications for K.'s overall argument, since it suggests that the instability in the representation of divine status that he so cogently analyses was a characteristic not merely of its representation, but of the status itself.

K.'s book is a crucial contribution to our understanding of that loose amalgam of practice and imagery that we continue, for the sake of convenience, to label ‘imperial cult’. He examines a vitally important element of that amalgam with a degree of subtlety and sophistication that I, at least, have not hitherto encountered. Anyone interested in the problems of divine and human representation in the ancient world or in imperial cult more broadly would be well advised to read this book with care.