The original version of this book was appreciatively reviewed by N. McKeown in CR 60 (2010), 180–1. Because most comparable works treat one or the other, its strength is its coverage of both Greek and Roman slavery, and the opportunity the comprehensive approach presents for understanding the similarities and differences between the two systems. Its guiding theme, the recovery of slave experience, is important, and the topics it addresses are essential: conceptualisation, early forms of slavery and their development over time, demographic and economic aspects, manumission, rebellion, the end of slavery in the Roman west. An understandable national orientation notwithstanding (cf. pp. 3, 5, 106), the English translation will make its findings accessible to a wide constituency, but readers, especially students, should be aware that it contains many inaccuracies, misprints and other slips.Footnote ¹ Also, the bibliography of the French original is not reproduced here, and no account has been taken of new contributions.Footnote ²

The book places an appropriate emphasis on violence as an elemental constituent of slavery (the violence of warfare included), the variety of servile statuses in Greco-Roman antiquity (from which an absence of class consciousness might be inferred), the primacy of social factors in understanding ancient slavery, and the need to distinguish ancient from modern concepts when dealing with slave labour. (In any system economic motives may impel enslavement or lead to exploitation of slave labour, but the relationship between slave and slave-owner is in the first instance a relationship between people and fundamentally therefore social.) Relevant sources are generously quoted. On the other hand there is a tendency to take almost every item of evidence introduced at face value and to generalise freely. On the bedevilling topic of slave numbers, modern minimalist views are rejected chiefly by assertion (pp. 40–52, 60, 65), while detailed countering of rival scholarly views is occasionally excessive (pp. 33–5).

Some particulars. The main source for the slave rebellions in Sicily of the late-second century b.c., the narrative of Diodorus Siculus, was written much later than the events concerned and notoriously contains a number of ‘doublets’. It is also partial, in both senses. In the synopsis of events given here (pp. 145–6), however, there is no indication of the difficulties these factors raise: accuracy of content is assumed without demur, as also for Theopompus on Chios (p. 9, quoted by Athenaeus, how carefully it is difficult to say), for Dionysius and Livy on early Rome (pp. 30–3; cf. 121, where the crucial point should be that neither author could imagine a Rome without slaves), for the Historia Augusta on slave numbers (p. 159, with no indication of the problems of authorship and reliability), and frequently elsewhere. That an improving philosophical response to slavery was ‘by far the most widespread in Roman times’ (p. 134), that fugitive slaves ‘often’ died (p. 137), and that slave-owners' greatest fear was of slaves running away (p. 138) are typical of the broad conclusions offered, all of which may be true but require more documentation than that provided. (Is Seneca's apparently sympathetic attitude towards slaves enough to justify ‘à l’époque romaine'? More tellingly, how can death-rates among fugitive slaves or slave-owners' greatest fears in Greek and Roman history be established?) Perhaps there were no slave revolts in later Roman history (p. 164), but it might be safer to say only that no episodes of revolt are attested.

Issues such as these are compounded by an unexamined assumption that the historical record, although full of lacunae, gives a generally faithful picture of the development of ancient slavery over time. This is perhaps overly optimistic given that the materials on which conclusions about historical change depend diverge in nature, quantity and type. In archaic Greek history for instance the dominance of Homeric epic in one phase gives way to something very different in a later period, so perhaps it is more valuable to emphasise the cultural constant – universal freedom and social equality seem never to have been elements of Greek thought and practice – than to try to specify when a ‘slave society’ first emerged. In other situations little more than impressions are possible, as with the origins of manumission (obscure in both Greek and Roman history), although sometimes there are hints that more might be known from materials not fully investigated, such as the Hippocratic corpus. Certainly the history of slavery and early Christianity can benefit from considering Pauline and comparable evidence.

One topic that seems to stand out across time is the attention slave-owners paid to the practicalities of controlling their slave possessions in order to extract labour and services from them. A considerable body of thought developed in antiquity on how to induce servile compliance, and it displays a remarkable cold-bloodedness of attitude. It is evident especially in Greek and Roman writings on household management (and related texts). In the prescriptions there given slaves merit human recognition only to the extent that owners find it advantageous to do so, their emotions are to be deliberately manipulated, and authority is to be exercised in every mechanistic manner conceivable. Potential rebelliousness is to be defused by ensuring that slaves are procured from different geographical sources, cannot speak the same language and cannot easily communicate with one another. Competitiveness is to be fostered, or else discord and divisiveness, to the same end. A sense of apparent self-worth can be contrived by appearing to take slaves' views into friendly consideration – opinions differed on this point: some thought consultation or friendliness a mistake – and slaves who identify with their owners' interests can be regarded as potential supervisors. Rewards are to be given and incentives held out for obedience and good performance: better clothes and shoes, more food, occasional free time, a few personal possessions. (The slave-owner should save on outlay, however, by limiting the rations of the sick and, to prevent insolence, by forbidding consumption of wine.) Even sexual activity can be regulated: favoured slaves can be allowed partners, and perhaps charged a fee for the privilege of intercourse, and demonstrably fertile women can be relieved from work to allow further reproduction, their children held as hostages meantime to guarantee individual acquiescence. Most effectively of all, the hope of freedom can be offered as a reward for faithful service. If nothing succeeds, however, physical punishment is always available.

These prescriptions, known from Aristotle, Plato, Xenophon, Cato, Varro, Columella, Plutarch and Athenaeus among others, were handed down from generation to generation, reflect real methods of treatment and presumably stretched back to the beginnings of classical history. (When ‘the day of slavery’ came, as Eumaeus said, no man was inclined to work willingly [Hom. Od. 17.320–3.]) The degree of ruthless calculation they embody, the libido dominandi known to Sallust (Cat. 2.2) and Augustine (De civ. D. 1 praef.), is unmistakable. No one will doubt that slavery at large, as stated at the beginning of this book, is ‘morally reprehensible’ (p. 3). In the pages that follow historians will find a compact overview of the principal issues its Greco-Roman forms raise.