202 PHOENIX seek out further evidence beyond what this book can offer. This is just what Rives's volume does best: to show us where we stand in a thought-provoking manner that invites further questions about "religion" in the Roman empire. Boston University ZSUZSANNA VÂRHELYI Caesar's Calendar: Ancient Time and the Beginnings of History. Denis Feeney. Berkeley, Los Angeles, and London: University of California Press (Sather Classical Lectures 65). 2007. Pp. xiv, 372 pages, 14 illustrations. Denis Feeney has performed a useful service by compiling evidence concerning time-keeping and chronology on the part of public officials and historians in the Roman world, down through the age of Augustus. While the study offers little that is new, within the boundaries Feeney sets for his inquiry he provides a coherent, lucid, and reasonably comprehensive account. The book is divided into six chapters, corresponding to the six Sather lectures on which it is based. The first two discuss ancient attempts to synchronize chronologies between Rome and Greece (especially Athens) and, to a lesser extent, between the hellenistic world and Eastern, non-Greek kingdoms. The third and fourth consider the boundary between myth and history as represented in foundation stories, myths of a golden age, and accounts of the Trojan War as the commencement of historical time. The final two focus on Rome under Julius and Augustus Caesar, considering the structure and significance of the civic calendar, methods of dating years, and the conflicted contemporary reception of reforms. As the title of the book suggests, the topics covered in the last two chapters determine the structure and content of the preceding four, with Feeney seeking throughout to identify the characteristics of ancient timekeeping and chronology that led to the particular transformations of the late first century b.c. This teleological approach gives the book a certain momentum and coherence, but renders it less than ideal as a survey of Roman (or Greek) notions of time and their cultural import. The book is defined as much by what it excludes, in terms of evidence, approach, and significance, as by what it includes. For all its virtues, it falls victim to what T. J. Cornell has in another context called "the tyranny of the evidence." Feeney assumes that what prevails in the record is what was prevalent on the ground, and fails to consider the extent to which dominant approaches to time were selected and constructed as such or the ways in which they intersect with other structures, beliefs, and practices. Consider for example the issue of chronology. Feeney rightly calls attention to the varying attempts at synchronizing internal and external chronology that characterized the late republic (e.g., Nepos, Atticus, Varro, Cicero's Brutus) and links such efforts, like those of the Greek writer Castor of Rhodes, to the "universalizing atmosphere" of the period, the experience of contact with other peoples and places. But contact with other peoples and places had been the experience of Romans for centuries, indeed from the outset, and it is hard to believe that they were not aware that different people have different ways of structuring the past. Cato grapples with just this problem in Book 2 of the Origines, attempting to sort truth from falsehood; Livy likens the ritual of the praetor's nail to a similar mode of timekeeping at Volsinii; Sallust and Velleius Paterculus refer to African and (perhaps) Capuan chronologies respectively, and the Etruscans almost certainly had a distinctive way of keeping track of cosmic and human time. Is there an ancient society BOOK REVIEWS/COMPTES RENDUS 203 that does not seek to situate itself in relationship to an authorizing past and that does not worry about conflicting stories told by their neighbors and rivals? Feeney's own evidence, which points to the persistent interest in foundations as chronological benchmarks, raises the possibility that it is precisely the process of forming a state, as an abstract, exclusivist, transhistorical entity that generates the need to create clear distinctions between mythical and historical, ours and yours, cyclical time and linear time. To synchronize chronologies is to suggest that there is more than one chronology to be synchronized, a recognition that makes sense precisely in...