Despite the value of the pseudo-Caesarian Bella as evidence for the development of Latin prose in the mid-first century b.c., these texts have received little recent critical attention. This book constitutes a welcome exception to the trend. The approach taken is analytic, and the principal theses defended are as follows: the Bellum Alexandrinum is a heterogeneous work, different parts of which were written by different authors; the individual responsible for stitching together and supplementing its various parts was Aulus Hirtius; the diction, style, narrative technique and historiographical character of paragraphs 1–21 enable those paragraphs both to be clearly distinguished from the rest of the work and to be attributed to the hand of Caesar himself (as the latter part of the novissimus imperfectus commentarius mentioned by Hirtius in the Epistula ad Balbum).

As the authors acknowledge, neither the heterogeneity thesis nor the attribution of paragraphs 1–21 to Caesar are new; both were first put forward as early as 1888 (by G. Landgraf), and both have since found further supporters. The novelty of G. and H.'s contribution to the debate is twofold: first, the linguistic and stylistic evidence is expanded (see especially pp. 55–72), and presented with a clarity and a level of detail that has not previously been matched (see also Appendices E–J); second, the authors have succeeded in incorporating new historiographical and narratological considerations into the analysis (see pp. 74–154). As a result, this book constitutes the most convincing demonstration to date of the Bellum Alexandrinum's heterogeneity: in the wake of G. and H., any thesis according to which the entire text was written from scratch by a single author will not deserve credence; this is perhaps the book's principal achievement.

After a brief introduction, Chapter 2 deals with the ancient evidence for the publication of the corpus Caesarianum. The testimonia quoted here do not offer a precise account, and it is accordingly no surprise that the chapter's arguments are often grounded in quite weak assertions about the relative likelihood of different interpretative positions (the series of arguments at pp. 15–21 in favour of the posthumous publication of the Bellum Civile are a case in point). It may have made more sense simply to admit that any more-or-less precise claim concerning the publication history of the Caesarian corpus is likely to be under-determined by the available evidence, before moving on to the more solid ground of Chapters 3 and 4.

Chapter 3 contains linguistic and stylistic arguments that attempt to prove the book's central theses: (a) that the text is heterogeneous (and, in particular, that paragraphs 1–21 have distinct origins from the remainder of the work), and (b) that paragraphs 1–21 may, in some form or other, be attributed to Caesar himself. These two theses, which the authors have a tendency to run together, are quite distinct: it is easy to imagine circumstances in which (a) may be true even if (b) is false and, indeed, the authors demonstrate (a) rather more convincingly than they demonstrate (b). The authors present compelling new evidence in favour of the text's heterogeneity, successfully demonstrating that the lexicon, diction and syntax of 1–21 mark those paragraphs out as distinct from 22–78 (which make up the remainder of the text). The attempt to prove Caesarian authorship of 1–21, however, is made difficult by the presence in those chapters of at least eighteen expressions that contradict Caesar's stylistic preferences as established in Bellum Gallicum and Bellum Civile. The authors offer two explanations for this: the continuator Hirtius may have made minor changes to a Caesarian original; and the Caesarian original in question may have been no more than a draft. Both explanations are plausible, and both speculative. Elsewhere, the authors rightly express a healthy scepticism in the face of speculative arguments made by other scholars; this is one of a few cases in which just such a sceptical attitude may profitably be brought to bear on their own conclusions. While the eighteen exceptions by no means falsify the hypothesis of a Caesarian core to 1–21, the case in its favour remains unproven.

Chapter 4 extends the analytic approach of Chapter 3 into the domains of literary technique and historiographical method. Again, the authors set out to demonstrate the existence of variation (or, as they put it, ‘inconsistency’) between different parts of the text (and especially between 1–21 and 22–78), and again, some of their results are truly compelling: the analysis of political and ethical value terms (which, it is shown, occur with much greater frequency in 22–78 than in 1–21; see also Appendix L); the demonstration that quite distinct attitudes to historical and divine causation are exhibited in different parts of the text; and, above all, the differentiation between historiographical influences (Thucydides behind 1–21, Hellenistic tragic history behind parts of 22–78). This opening up of the text to historiographical and literary analysis is another of the book's major achievements. It is also in Chapter 4 that one of the more significant problems with the authors’ overall approach is most apparent. The problem in question has to do with the different degrees of attention paid to different parts of the text. The heterogeneity thesis defended in this book is not limited to a distinction between 1–21 and 22–78; it is also claimed that different parts of 22–78 have discrete origins, probably as drafts prepared by different Caesarians and stitched together by Hirtius (see pp. 53–4, 154, 164–5). In Chapter 3, and despite this supposed heterogeneity, the passage 22–78 is mostly treated as a monolithic unit with whose language and style that of 1–21 may be compared (an approach defended at pp. 53–5), with the result that the existence of parts with discrete origins within the passage 22–78 is not fully demonstrated in Chapter 3. In Chapter 4, however, when the unevenness in the density and quality of information on offer in 34–78 is taken to demonstrate that different parts of that passage are based on sources of differing quality, the authors treat this as corroboration of the thesis that ‘the transmitted text is little more than a patchwork of separate reports’ (p. 93). This move is unwarranted: the internal heterogeneity of 22–78 has not yet been demonstrated, meaning that there is no good reason to slide from a claim that the text is based on different sources to a claim that the text is composed of different reports. This example also speaks to a more general concern: the relative lack of attention paid to delineating the supposedly different parts of 22–78 is a real shortcoming in what is quite a long book about rather a short text. The authors should be thanked for re-opening and reinvigorating the debate surrounding the origins and authorship of the Bellum Alexandrinum. Much remains to be done, especially on paragraphs 22–78.