This book looks at fundamental problems of our academic activity and of dealing with texts in general. The preface sets out the basic assumption: the focus is not on genre, but on the ‘textual dimension’. The introduction develops this: the ‘loss of the original’, or rather the absence of the original, means that the text we have today is a remnant of an earlier wholeness. The ‘formal phenomena’ for identification of epitome include fragmentation, re-composition and selection. The questions of ‘when, why, for whom and how’ are interdependent and interact with each other. Sacchi is right to state that ‘creation, presentation and transmission of information’ are the general cultural mechanisms. He warns against the functionalist approach of redemption of what is lost or missing. But what distinguishes the act of appropriating epitomised texts from any other appropriation of texts does not become clear. The criteria of absence and interruption are defined as inherent qualities of epitomai; one wonders if this does not apply to any text, only in different dimensions.

The scope of the fourteen essays is very broad, extending from antique and late antique texts to Pascal Quingnard, Antonin Artaud and Roland Barthes. In the following I will concentrate on some essays which are closer to my own expertise. Sowers on Ausonius assembles a range of examples, ‘literary miniatures from fragmented pieces of the cultural past’. Digests, catalogues, assemblage, commemorator all have something to do with each other, they build up the ‘deprecating trap’. Ultimately, the existence of the texts is explained as a literary parlour game. Hudson tries to get closer to Nepos’ Vita Catonis. He interprets Nepos’ concise statements as voluntary briefness. But is it possible for today's interpreter to think past the information transmitted by Cicero, Plutarch, Lactantius? What safeguards us from over-interpreting the text? Hudson is over-confident about that when he uses phrases like ‘(Nepos) may distil’, ‘may have chopped off’, ‘won't say’, ‘tacitly … …’: I have doubts about whether we can actually look into Nepos’ workshop.

Liverani and Elsner offer valuable observations on image and text. The comparison of topoi and spolia is thought-provoking. But spolia are firmly fixed in their new place and context, while topoi are reproducible and movable. Liverani choses late antique sarcophagi with repeated images in a ‘columnar arrangement’ and without chronological order as a plausible comparison with topoi in texts. Yet he insists that the frescoes in Old St Peter's are original creations whose genesis and intent cannot be compared to epitomised texts. Drawing from disparate sources alone cannot be a criterion for defining epitome. We return to images with Jas Elsner. Images, in his view, become independent of a narrative because they have to make a choice. The image – not the illustration – in the codex inevitably becomes a paratext. Elsner's insightful descriptions of the Vergilius Romanus lead us to the fundamental difference between images and texts. Images must select the highlights, the ‘key events’. Elsner stresses the difference between summarising and selecting. Do Elsner and Liverani contradict each other or is there an implicit understanding? Works of art like those described here are luxuries, unique objects, and not intended for wide proliferation. This applies to the Vergilius Romanus, the Bible manuscripts and also the Tabulae Capitolinae. Therefore, we must endeavour to differentiate between the image as a (perhaps even widespread) cipher, and the epitomic quality of a single artifact.

In Payne's contribution on Nonius one recognises his effort to look at this text as a whole. But I miss the indication that it is precisely Nonius who so wonderfully depicts the different layers of what we are used to carrying with us as literary history. The author could have profited from the contributions in G. Most (ed.), Collecting Fragments (1997). More convincing is Burrus’ article about Dionysius Areopagita. This text, Burrus suggests, exposes itself as incomplete. Is it a characteristic of theological texts only to omit what is omitted because it is found to be inexpressible? Or is it not the nature of the epitome that it makes itself known as such? Hardie on Symphosius stresses the contrast of epitome and epitomic quality, and convincingly argues that certain qualities of texts to which we too readily attach the label ‘late antiquity’ are in fact always present. McGill on the verse arguments to the Aeneid insists on their entertainment value. He arrives at a contextualisation, at a social situation (‘keyword’ of the Deipnosophists). We are again dealing with the question of choice. McGill shows how new ‘forms of expressions’ are wrested or extracted from the established canon. The final essay on Nabokov and Pliny by Noens is problematic. Noens’ reading of Pliny enables him to cast a thematic net with the modern text. But the difference between the letters’ addressing fictional or real contemporaries and the autobiographical approach of Nabokov's work must not be neglected. Here a discussion of Plin., ep. 6. 'aliud est enim epistulam aliud historiam, aliud amico aliud omnibus scribere', as well as Plutarch's (Pelopidas 1.1) remark about historia and bios, might have been helpful.

The volume is carefully edited, with few misprints. The layout is not really reader-friendly, the pages being very full. The quality of the illustrations is high. On the whole, the editors have assembled stimulating, while not always convincing, essays, and, by connecting a wide range of topics, texts, authors and eras, encourage us to think about our fragmented knowledge.