The volume brings together a number of papers presented at a conference in Trier in 2006 on the same subject. The editor, Altay Coşkun, provides an introductory essay and an ‘Anhang’ that provide structure for the eight much more detailed discussions. The contributions have been arranged in a roughly chronological order. It should be noted up-front that the title is somewhat misleading as the cases discussed here in detail relate solely to communities and individuals in the Greek-speaking East. Only in the Anhang and the chapter by Craig Williams does one find any references to the practice of amicitia and clientelae elsewhere in the Empire, and even this discussion is on a much more general level than is to be found in the other studies.

Friendship is one of those concepts that we all claim to recognize, but also find difficult to characterize in a consistent and categorical way. Much as one would like to establish universal criteria for understanding such relations and for placing them in a contractual context, the prospects for success are limited. Coşkun and his colleagues recognize the difficulties. The first two and the very last chapters, including the paper by Craig Williams, ‘Friends of the Roman People. Some remarks on the language of Amicitia’, make it abundantly clear that the language of friendship is elastic at several levels. Much as one would like to believe that there could be, indeed should be, a genuinely altruistic basis to friendship, the overwhelming number of cases discussed here involves behaviour that may reasonably be described as ‘opportunistic’. In fact, all the authors in this volume assume that individuals acted as ‘friends’, and/or promoted themselves as friends, in order to achieve some goal for themselves or for their community. Furthermore, the actors believed that amicitia was the appropriate vehicle for such promotion. Amicitia then involves the exchange of benefactions, actual or expected.

The chapters by Luis Ballestros Pistor (‘Cappadocia und Pontus’) and Axel Niebergall (‘Rom und die griecheschen Eliten’) look primarily to the cases of friendship between the élites of the Anatolian provinces and Rome/Romans during the late second and early first centuries b.c. That the local and Hellenized élites employed the language of ‘friendship’ to characterize their relationship with Rome and Romans is manifest in the literary and epigraphical sources. Again and again the authors point to the ‘utilitarian’ nature of the exchanges (e.g., 50 ff.). Friendship served the interests of all parties, primarily by creating an atmosphere of conciliation and cooperation. Niebergall argues effectively that the divisions among local élites are reflected in the changing patterns of friendship between natives and Romans (71 ff.) as the former sought advantages in local conflicts.

These ambiguities become clear in the case of Artavasdes II, an Armenian king who might, as Henrik Prantl observes, be described as a ‘Freund oder Feind der Römer’. In this case, Artavasdes used the language of friendship to navigate between the Parthians and the Romans, and also between factions of the latter. His behaviour is comprehensible (friendly or hostile) only when perceived as a strategy to maximize his own and Armenian independence. The language of ‘friendship’ allowed the actors considerable latitude, but did so at a cost.

The chapters by Johannes Engels and Julia Wilker deal with the evolution of political amicitia from the Late Republic into the Principate. The connection between Roman politicians and the leading intellectuals of the Hellenized cities of the eastern Mediterranean served the interests of both parties: Romans acquired access to information about how best to administer their provinces and enhanced their reputations as defenders of Greek learning by receiving embassies and encomia. Wilker's article, relying extensively on the narrative of Josephus, documents the transition whereby local dynasts, like the Herodes, reconstituted themselves as fully integrated members of the Roman élite. The more integrated they became, the less successful they were about mediating between their subjects and the Roman authorities.

Heinz Heinen adds a useful discussion of several very fragmentary inscriptions from the Cimmerian Bosphorus area. His contribution is a reminder that the process by which historians develop an historical argument depends on such survivals, and contrasts strikingly with the account of Wilker. The latter builds her argument around the more coherent narrative of Josephus. Nonetheless these articles work together to explicate the various rôles played by the philorhomaios and by the philokaisar in more informal connections between rulers and subjects.

In the last chapter, labelled an ‘Anhang’, Coşkun addresses the contractual character of amicitia. The issue is an important one and extends the discussion well beyond the practice of amicitia between individuals. How much can be read into the Roman tendency to link overlapping concepts like amicitia, hospitium, foedus and societas? May one conclude that some kind of permanent relationship was so constructed because the terms are used together in some constellation? Coşkun's discussion of this issue is thorough and balanced, and rejects the overly legalistic interpretation. It is a discussion that is marred by his reliance on an unpublished manuscript of Andreas Zack.

To paraphrase an aphorism, friendship meant different things to different people at different times. Here we one can only conclude that amicitia could mean different things to the same people at the same time. Indeed the very elasticity of the concept allowed each party in the relationship to interpret the arrangements in such a way as to maximize benefits for himself. Despite the ambiguity, the very fact that amicitia and its cognates continued to be used in literary, legal and epigraphical texts confirms the power of the words to disguise manifestly opportunistic and self-serving behaviour in an attractive and altruistic light. One cannot help but believe that Cicero would understand and agree.