Ovid was forever altering himself – never failing to ‘advertise the revised status of all his major works’ (p. 1) and their own capacity to effect revision. He invites the reader to tease out the ‘commentary’ inherent in these revisions and to construct from it further revisions of the poetic career and authorial identity that his works create. His ‘revisory practices’ interact with other types of revision in the Augustan milieu, and the persistence of revision throughout his œuvre ‘offers a set of important insights into a particular aspect of the self-construction of the writing subject – and, therefore, of subjectivity more generally’ (p. 2). Such are the basic contentions of M.'s monograph. It is a novel, ingenious and ambitious first book. But it is also full of problems, with significant shortcomings in (a) conception and structure, (b) methodology and (c) readability (of the last already an inkling in the above-quoted statement from p. 2).

(a) M.'s book is not one study, but at least four studies, which loosely constitute a ‘set’ only in that each addresses ‘revision’ in one or more works of the same poet: Amores; Ars and Remedia; Fasti; and the exilic corpus (plus the epilogue of the Metamorphoses as a key intertext for the Tristia). These studies are not (inter)linked by any coherent thread(s) of argument, and the alleged common factor of ‘revision’ itself often slips out of sight (even in Ovid's own text it is hardly always ‘advertised’ and sometimes visible only to readers with an oddly focused microscope). True enough, literary revision comes in many forms and may encompass both an author's (or editor's) physical altering of an earlier version of a text and a text's potential altering of an earlier perspective on almost anything, including other texts, events and images. True also that both types feature in Ovid's works in authorial first-person voice. But the two are not the same, and they will not in tandem necessarily support plotting of any œuvre-wide rolling revision of Ovid's career-profile and identity such as M. posits. Sustained pursuit of this form of Ovidian revision emerges only in the book's second half, in the two chapters (5 and 6) on the exilic corpus, which do provide some tenuous cross-chapter coherence. Ovid's naming/not naming of himself alongside others in his Tristia, particularly when recapitulating the (ostensibly) earlier epilogue of the Metamorphoses, marks his revision (M. argues) of his once confident self-image as a poet, and further revision of his authorial identity results from his naming of the addressees in his letters Ex Ponto as those in whose hands his reputation now lies. The awkward questions of relative compositional chronology introduced by possible physical revision of parts of the Metamorphoses (and Fasti) in the exilic period M. dismisses with the claim that it does not matter whether or not Ovid actually revised this or that passage or work, only that he writes in such a way that the reader thinks he may have done – in a kind of revising editorial mode.

This position on physical revision is the only point of contact with the chapter on the Amores (2), the first of three effectively freestanding studies of Ovidian ‘revision’. Taking a cue from Genette's theorising on paratextual influence upon a reader's perception of the main text, M. prises out of the extant three-book edition of the Amores (with some straining of Latin, e.g. p. 40 on praetulit, epigr. 2) a nod-and-a-wink ‘commentary’ on the earlier five-book edition mentioned in the prefatory epigram. This ‘commentary’, she argues, makes the earlier edition seem to have existed, whether it actually did or not, and the Ovid now writing therefore seem already to be revising an earlier literary self. She misses a trick here, though, in not attempting to connect this with those later chapters specifically exploring revision of career-profile and identity.

In the truly bizarre Chapter 3 virtually no connection with the rest of the book suggests itself. Inspired by the Freudian theory of Peter Brooks, M. construes the erotodidactic works as a single narrative of sexual desire, with two restarts involving thematic repetition-with-inversion (Ars 3 and Remedia). These restarts allegedly prevent premature ending of (the narrative of) desire before its proper moment of extinction by the male's alleged ‘castration’ when confronted with the ‘disgusting’ sight of female genitals (pp. 89–92). The editorial author Ovid M. finds reconfiguring middles as pause-points and endings as middles, and so progressively overwriting his entire structure. But she has to assume that ‘a sequel (or two) is at least anticipated [by Ovid himself] from the very first’ (p. 69 n. 1), and that a reader will naturally reboot the reading process several times to keep abreast of the rolling programme of revision. Neither assumption is safe.

The strongest chapter is the central one (4) on the Fasti, which M. successfully demonstrates is both revised (in conception), by dint of its premature closure after only six books (one on each of the first six months of the year), and itself revises earlier revisions of the Roman calendar, including the latest by Augustus, whose authority Ovid must therefore seem to challenge. This is also the most accessible chapter, not least because it is largely free of the theoretical methodology and language that dominate the rest of the book.

(b) I have no quarrel with M.'s recourse to theory per se, though I raise an eyebrow at the positivist tone that frequently surfaces in conflict with the non-definitive approach that she personally claims (p. 29) and that theorists (theoretically) embrace. A ‘best’ or ‘right’ way is repeatedly pointed; the promise of ‘insights’ itself suggests revelations of hidden truth; and everywhere is the complicit ‘we’/‘us’ (of M. and her readers). My greater reservations, however, despite M.'s attempt to forestall this criticism (pp. 33–4), are about the number of different theoretical methodologies used – or, at any rate, invoked, along with their proponents – within this one book: too many for any but the most specialised theoretical readers. Collaborative revision, distributed authorship, performativity, difference, etc. etc.: they are all there, but how many are really helpful? In general I find too much packing in of multifarious ideas (I could comment on only a tiny selection) and far too little clarity.

(c) Opacity of expression is the book's most serious weakness. Parts of it are virtually unreadable, including, unfortunately, much of the introduction. Just one example (p. 9):

It is easy to blame theoretical idiolect for this alienating jargon, but there are other indications of imperfect control of English: for example ‘concision’ for ‘abridgement’ (p. 36); ‘travesty’ for ‘transgression’ (p. 88); ‘dilemma’ for ‘problem’ (p. 188). The modish ‘she’ in generic reference to an author is unnecessarily jarring cheek by jowl with specific reference to the indisputably male Ovid (e.g. p. 4). A firmer editorial hand could have eliminated such unwise choices, as well as steering M. towards a generally more lucid and tightly focused book. Misprints are few, the most disconcerting being ‘invested’ for ‘interested’ on p. 3.

M.'s monograph is a tour de force in its demonstration of how a particular modern mind-set can impose itself on Ovid's poetry, but what in essence it has to say is that Ovid at the end of his career was different from Ovid at the start. And does he not say that himself: non sum ego quod fueram (Tr. 3.11.25)?