This book describes the metaphors the Romans used to envisage their control over their empire. The focus is on Latin authors from the Late Republic to the age of Pliny and Tacitus, although Cassius Dio is not neglected and there are useful excursions into later material. The emphasis is strongly on élite Rome looking out at the provinces, not (as, e.g., in Carlos Noreña's Imperial Ideals in the Roman West) the provinces looking in at Rome. And although this book will be useful to ancient historians, it is strictly a work of Classical Philology rather than Ancient History, because Lavan's eye is firmly fixed on the ancient authors as literary figures, and when he digresses from his catalogue of the metaphors of domination, as he does in chapter 3, on Tacitus, it is to use his analysis to illuminate not historical events, but Tacitus' art and purposes.

Descending from a Cambridge dissertation, Slaves to Rome is a book of many happy insights and surprises. The first, as the book's title implies, is the overwhelming commonness with which metropolitan Roman authors casually described their rule over their empire as that of masters over slaves (chapters 2-3). No less interesting is how long Roman authors continued, more politely, to refer to those they ruled as socii, "allies" (chapter 1), and the expansion of the semantic field of that word beyond the realm of allies stricto sensu to serve first as an everyday term for non-Roman citizens, and soon after for all non-Italian inhabitants of the empire. There were other terms available, such as peregrini, tributarii, stipendiarii, and provinciales, but these tended to be used in technical, legal senses, while as generic terms for imperial subjects they were simply trampled by the overwhelmingly common use of socii. Only from the reign of Diocletian did provinciales decisively replace socii as the word for Rome's subjects. But whatever the words used, there was always the claim of differentness, of exclusivity. As a consequence, a "theme of this book has been the exclusivism implicit in many Roman accounts of empire [...] [T]his is intended as a corrective to conventional accounts which credit the Romans of the Principate with an integrationist ethos" (245).

Romans were also in the habit of emphasizing the loyalty owed them by their subjects by virtue of the boons and services (beneficia) Rome had conferred upon them (chapter 4). But Lavan points out the darker side of this model, because asking for and receiving benefactions was a sign of inferiority in the ancient world, and so by emphasizing the benefactions they bestowed the Romans were also emphasizing their supremacy over their subjects. This makes the Romans parallel to the Persians in Lynette Mitchell's Greeks Bearing Gifts, Persians whose lordly position required that they must always give greater gifts than were given to them. And Lavan is very acute on the mental fancy-footwork the Romans used to conceal from themselves the fact that they were in fact the recipients of bounty from, say, Egypt, rather than Egypt's benefactors.

Next Slaves to Rome leaps into the murky waters of the application of the patron/client metaphor to Roman thinking about relationships with their subjects, and from that scholarly swamp Lavan emerges with a gem (chapter 5). Yes, Latin authors did use the metaphor, but with a tiny number of exceptions, only two of them did so to any great extent, Cicero and Livy. So this famous metaphor for Roman international relations, rather than being the common property of the Roman ruling and literary class, seems to have been a personal idiosyncrasy used for idiosyncratic purposes, in Cicero to sharpen invective, and in Livy to allow comparison to the harsher rule of Hellenistic kings.

In fact, one of the particular strengths of Lavan's book is that his sensitive reading has allowed him not only to identify metaphors that were used, but also to identify surprising absences (like the complete absence of the patronage metaphor applied to interstate relations in Caesar). Another significant absence is the mobilization of familial metaphors to describe Rome's relations with her subjects: Rome is neither mother nor father to the provincials. The Roman emperor may be pater patriae, but he is not pater imperii Romani. And, although provincials may metaphorically be slaves, Lavan notices that they are never metaphorically manumitted. It is also significant that, often, when Lavan has to admit an exception to the rules of conventional usage of the words he studies, that exception is to be found in the neglected pages of the despised second-century-AD historian Florus. In a back-handed way, Lavan is directing the attention of historians of Roman foreign affairs to the unexpectedly rich Florus, an invitation hitherto taken up, so far as I am aware, only by Susan Mattern in her Rome and the Enemy.

In his last substantive chapter (chapter 6), Lavan turns from Latin authors to the official language of the empire, the language of imperial communications to cities and individuals. A scholar less intellectually honest might well have excluded this chapter from his book, because, as he must admit and as any epigraphist knows, the rhetoric of these inscribed documents from emperors and governors does not accord well with the major roles - master of slaves, commander of allies, patron of clients - that Lavan identifies in metropolitan literature as the way Romans thought about their relations with their subjects. Such documents display, in fact, "the total absence of the categories that are the subjects of this book" (224). Or, to put it more kindly, such documents emphasize one of those pairs of roles - benefactor and recipient - nearly to the exclusion of the others (218-220). And rather than, in one way or another, emphasizing a symbolic opposition between the blessed race of Romans and all outsiders, as metropolitan Roman literature had done, the dichotomy in official documents is between a lofty individual - the emperor (or governor) - and his undifferentiated subjects.

Forthright and true to his literary purposes, here, in the muddle of conflicting discourses, Lavan stops. "Roman literature presents us with a range of different visions of empire. This polyphony reflects the coexistence of divergent values and ambitions throughout the period covered by this book" (244). He has done his job well, and so ends his tale. He has also done his duty by the modern scholarship (even that in French and German, not to be taken for granted in an English dissertation these days). And for the incompetent printing of the book, which scatters bold-faced letters (or fragments of them) through the text ("wider [...] of [...] Romans" [108, cf. 15, 175]), Cambridge University Press' editor Michael Sharp is justly punished, because it afflicts his own name where he is thanked in the preface (ix).

But the historian cannot halt where Lavan does. He wants somehow to order, to rank, to make sense of this collection of metaphors of domination. To the historian it is perhaps not all that surprising that the Romans called their subjects slaves when talking to each other, but not when addressing those subjects: one feels that a greater sense of audience would have reduced the confusion of Lavan's shouting polyphony. And the historian wants to go where Lavan did not: into the writings of Rome's subjects (almost all the evidence is in Greek). Did they feel themselves slaves, or allies, or recipients of benefactions, or clients? My instinct is that they did not consider themselves slaves (despite a passage in Plutarch urging them not to act slavishly so often quoted as to suggest it must be nearly unique [Praec. Ger. Reip. 814f]), and that Greek-speakers at least did not consider themselves, in a Roman sense, clients (a peculiarly Roman relationship that the proud Greeks show few signs of having grasped). But Rome's subjects likely did consider themselves allies of Rome: and perhaps, unknowingly, Lavan has hinted at a solution to one of the puzzles of Roman imperial history, which is why the embassies from cities to the emperor and his messages in return (the topic of Fergus Millar's The Emperor in the Roman World) were dressed up as "diplomacy" between independent states.

Finally, the provincials certainly did regard themselves as recipients of imperial benefactions, and when they speak of their own loyalty to the emperor, it is (a little surprisingly to us) chiefly in terms of gratitude for benefactions that they speak. But I suspect that Rome's Greek-speaking subjects would not agree with Lavan that the deeper meaning of begging the emperor for benefactions was to stress their humility and his superiority. Rather, they begged the emperor for boons - money, buildings, civic titles - to beautify their cities and advance those cities in their relentless competition with their neighboring cities. In their eyes, the emperor was not an arrogant grandee, to whom their begging emphasized their lowliness: no, what they got from the emperor they merited because of the ancient glory, nobility, and standing of their cities. And rather than the cities of the empire being drawn into a position of subordination to Rome by the emperor's boons, the emperor was drawn into their contests, older than the empire itself, and therefore fiercely fought, because as a distant and relatively disinterested giver of benefactions and honorific signs of approval he played a valued role in their parochial contests - that of referee.

J. E. Lendon