A.'s monograph offers a detailed study of the concept of concordia in the Late Republic. The methodology, which he explains and justifies in the first pages, is primarily philological: an examination of all occurrences of the term, its cognates and their Greek equivalents (primarily ὁμόνοια, but also ὁμοφροσύνη) in texts dating from or referring to the Republican period. The seven parts (bookended by introduction and conclusion) into which the volume is divided move chronologically from the Second Punic War to the death of Cicero, and form a convincing narrative of the nature and evolution of the term. Concordia was a concept to conjure with, not a reality: our sources locate it definitively in an ideal future or an idealised past, implying that their own times were characterised by its absence, and centring their narratives of recent history on the question of when and at whose hands it was ruptured. A. demonstrates that concordia referred essentially to a horizontal relationship between equals, and particularly between magistrates of the same rank. In a system which offered supreme power to not one but two chief magistrates, both of whom had reached the apex of their careers by competition rather than co-operation, it had a vital role to play. From this basis it could be expanded to include agreement among the governing class as a whole, especially since a lack of concordia between the consuls or censors of the year threatened elite control of politics more broadly. Only at the very end of the Republic, after Cicero's consulship and in particular during the Civil War, was concordia expanded to include all citizens. At all times different interpretations of how concordia should be achieved jostled for prominence: in particular, we see arguments between those members of the elite who used it to argue for the necessity of large-scale compromise, and others who believed that the need for concordia justified excluding certain individuals from the body politic.

Especially given the importance of speech in Republican politics, it is right to see language as encoding and determining the structures of political thought. Yet this sample of occurrences of concordia cannot capture in full how Romans understood the concept–not because we should imagine that it exists in some supposed realm of meaning which is separate from language, but because of the limited state of our sources. A. does indeed move beyond instances where the word itself is used to more general episodes which concern themes of concord, largely by taking each occurrence in a very broad context. This is a fine line to walk, however, and sometimes the incisiveness of A.'s philological method (well demonstrated in, for example, his reading of the De lege agraria speeches at pp. 253–7 or the correspondence between Cicero and Caesar in early 49 at pp. 353–6) is lost amid overly speculative interpretations of episodes on a larger scale.

Any analysis of Republican politics–and especially one focused on language–runs into the Cicero problem. In this case A. shows that previous discussions of concordia have been hampered not only by the predominance of Ciceronian evidence, but also by a modern tendency to concentrate on concordia ordinum at the expense of other flavours of concordia which are in fact predominant even in Cicero's own writings. The analysis of the varied and changing use of concordia in Ciceronian texts, going well beyond concordia ordinum, is a great strength of this volume. A. also makes valiant attempts to look beyond Cicero, searching carefully for material in later texts which can eventually be traced back to a non-Ciceronian Republican source. His decisions about what can be included in this category are often somewhat optimistic: a more pessimistic reader might conclude that Caesar's words over Pompey's corpse as reported by Lucan, or the speech of Calenus in Dio, say more about imperial than Republican concepts of concordia. A similar objection could be raised to A.'s treatment of Sallust: the length of time involved is much shorter, but given the close attention A. pays to the evolution of concordia during the Late Republic, can we really assume that the speeches given to Lepidus or Philippus really reflect the conditions of the 70s rather than the 40s? In other places, though, A. is at his best when showing how each generation read its history anew according to its own developing concepts of concordia. He takes Livy and Dionysius out of chronological order, often returning to a particular episode more than once as he demonstrates how narratives and exempla mirrored and commented on contemporary events.

There is plenty more to be said about this book's treatment of individual episodes and questions, but I want to conclude by focusing on a single aspect of concordia which does not get its due in A.'s work. Concordia was not just a word or a concept to the Romans of Cicero's era: it (she?) was a goddess. Latin speakers did not face the problem of whether or not to capitalise her name or gender her pronouns, and it is hard to believe that her divine aspect was ever absent when her name was invoked. Of the evidence A. cites, the earliest (a name painted onto a third-century vase) and the only non-Ciceronian items clearly datable to the 50s or 60s (coins bearing the head of Concordia) both refer definitively to the divinity. The temple of Concordia, whose use as a meeting place for the Senate A. persuasively and fruitfully argues was indeed symbolic, drew its power from the fact that the political idea could not be separated from the divine force. A. devotes parts of his introduction to the debate over the date of the arrival of the deity in Rome, her links with divinised ὁμόνοια and the question of her various temples (he holds that there was only one), but after this the divine nature of concordia is missing from his analysis, and A.J. Clark's Divine Qualities: Cult and Community in Republican Rome (2007) is not in the bibliography. On p. 13, the lack of attention to religious aspects is specifically noted as a failing in previous studies of the subject, and in a footnote we do get a reference to F. Pina Polo, ‘Concordia y libertas como polos de referencios religiosa en la lucha política de la república tarda’, Gerion 18 (2000), 261–92, along with a number of other works on the cult of Concordia, but the note finishes with a remark about the paucity of source materials on the subject. That granted, it is a great pity that A. does not integrate more consistently what little we do know about the cult, and the rather larger amount we know about divine qualities and the inseparability of Roman religion and political life, into his study. The goddess's absence does not vitiate this book's arguments, particularly not in the interpretation of individual episodes, but it does mean that it falls short as an overall exploration of the nature of Republican concordia.

Copy-editing is generally good, though on p. 264 the second Catilinarian is dated to 1963. More importantly, the graphs on p. 15 are misplaced: the caption given for number 1 actually belongs to number 2 and vice versa. As printed, they are nonsense, and since the axes are not labelled it is not easy to recognise what has gone wrong.