T. P. Wiseman's The Roman Audience, the product of a rich vein of scholarly investigation mined by the author over the course of some fifty years, supports a remarkably consistent and broad hypothesis: that from its very beginnings to imperial times and beyond, ‘all non-technical poetry and prose was composed in the first instance for oral delivery to a large general audience’; written texts, when produced, represented a ‘secondary stage’ in the dissemination of Latin literature (182). While this seems obviously true for certain genres, such as the comedy of Plautus or Cicero's speeches, W. also boldly argues for the performance of works that seem ill-suited to public presentation or whose authors explicitly deny a wish to catch the ear of the crowd.

W.’s attempt to reconstruct a performance history for such a wide range of texts requires him to explore many little considered sources, thereby providing us with a tour de force of historical research in which even the most unpromising scraps of literary and artistic evidence are shown to provide glimpses into the earliest periods (chs 2–3), as well as to support radically new conceptions of the place of literature in Roman social life in better documented times (chs 4–6, on the late third to first centuries b.c.; chs 7–9, on the Augustan period; ch. 10, a short discussion of post-Augustan Rome). As one of countless possible examples, I note here W.’s citation (111) of the ‘good, but neglected, evidence for “Greek games”’ in Republican Rome, reminding us how often we read ancient texts in a way that ignores evidence that does not fit an already well formed image of Roman social life (here, one that might elide indications of ludi Graeci at Rome before Domitian).

When W. deals with the later Republic and earlier Empire, he often demonstrates how even canonical texts can be read in new ways, once we shed our preconceptions. Evidence for performance of Vergil's works is particularly persuasive; and even if the case for many of Vergil's fellow writers is more difficult to make, W.’s picture of the Augustan poets ‘doing their bit’ for the new régime (121) by publicizing, in person or through others, their new works in the great theatres of Rome before appreciative crowds is an exciting and creative reimagining of the period.

Much of this image depends on a narrative of political history that forms a vital thread within the work and with which not everyone will agree. W. draws a stark contrast between supposedly public-spirited leaders (especially Julius Caesar and his adoptive son) who espouse the cause of ‘the Roman People’ and the greedy, arrogant and corrupt aristocracy. No ‘dregs of Romulus’ here, but rather a polity — apparently embodied in the city masses — who were ready both to support their champions and to appreciate an astonishing range of public entertainments, from Greek tragedy to erotic mime to Tibullan elegy. Thus the exhilarating quality of W.’s picture of the Augustan age hinges in no small part on its presentation as the culmination of a centuries long struggle, one in which the public performance of their work by Augustus’ protégés forms part of a grand celebration of victory by what W. would have us see as the only virtuous side in that struggle.

If we wonder why evidence for public performance — not just in recondite sources, but in well-known literary texts — was so long ignored, W. points to the reluctance ‘to let ancient texts mean what they say’, which he finds to be ‘particularly prevalent in the study of Latin literature’ (7). A defence to this charge might refer, first, to the frequent existence of contradictory evidence requiring scholars to interpret a conflicting source in ways that seem to controvert its obvious meaning (as W. himself must sometimes do). A second line of defence resides in the nature of Latin literature itself. Since Latin texts are regularly characterized by a high degree of complexity, subtlety and illusiveness, the demand that they be taken to ‘mean what they say’ in a straightforward way often does violence both to their form and content.

Many of W.’s arguments are persuasive, especially his assertion of the importance of the voice and the auditor in both the private and public communication of literary works. More speculative, and admittedly so, are the author's imagined dramatic enactments of various texts, such as Catullus 64 (109–10) or Ovid, Heroides 7 (158). It is worth remembering here that exercises in enargeia formed a regular part of ancient education, and the fact that a passage stimulates vivid visualization (a constant feature, for instance, of Ciceronian rhetoric) is not evidence that it was actually enacted.

While this challenge to received opinion concerning Roman literary culture might feel like a step backwards — a speculative attempt to undermine a great deal of careful scholarship that has taken place over the last half century or so — W.’s work will surely prove in the end to be a valuable and stimulating reconceptualization of the past. Just as Fergus Millar's assertion of the democratic nature of the Roman constitution and its public assemblies has inspired a spate of indispensable scholarship on Roman political life, often by those who question Millar's point of view, W.’s daring heterodoxy means that henceforth scholars intent on describing various Roman writers and their audiences must, at the least, pose the question of whether their assumption that these works reached only an exclusive and élite audience ought to be revised.