The Siberian language called Ket is sometimes reckoned to have fewer native speakers than there are linguists studying it. It is impossible to say whether fewer people ever composed verse in the Latin Saturnian metre than there have been modern scholars who have written on it, but the published work on the Saturnian is out of all proportion to the number of surviving verses. M. reckons that there are 114 textually secure Saturnian verses, 77 literary and 37 epigraphic (p. 56). Rather worryingly, he states early on in this work that he is in agreement with earlier investigators who have constructed this corpus from verses that cannot be scanned in any other known metre (p. 27). However, it is clear that M. here sells himself short, since he has taken great care to follow ancient indications about the literary figures known to have written in the metre, principally Livius Andronicus and Naevius, and he is careful in his assessment of putative inscriptional examples.

M. seeks to crack the long-standing puzzle of the Saturnian with all the armoury in the scholarly arsenal. He displays a command of metrical theory, modern theoretical phonology (including optimality theory) and metrics, statistics (which go far beyond the χ ² test), comparative philology, textual criticism and epigraphy. Furthermore, the second half of this book is devoted to a consideration of metre and phonology in the sister languages to Latin: Faliscan, Umbrian, South Picene and several varieties of Oscan. The reader is given an early indication of what will be required to follow M. through this terrain on pp. xxiv–xxvi, where three pages are taken up with the description of the different sigla employed in the text (including ten different types of brackets and ten different arrows). M. guides us along the path through clear signposting, numbered sections and frequent pauses for recapitulation or to explain what is coming next, but there might still be some weaker individuals who turn back when confronted by sentences such as the following, even before the metrical foothills have been reached (p. 22, in a discussion of the difference between two separate uses of the square brackets used in transcriptions of two examples, labelled (5a) and (5b)): ‘Here in (5b) is potential for confounding editorial and linguistic conventions, though the x grapheme should break the tie, and the unlikelihood that a phonetic representation would interrupt a sequence in orthography in such a way as in (5a) should decrease the chances of misapprehension’. By the ‘x grapheme’ he means nothing more than the letter x which occurs inside a section in square brackets in one of the examples.

So what is M.'s theory of the Saturnian after all this? In his view, the basis of the metre is accentual, rather than quantitative, and the surviving lines can be analysed in terms of a quadripartite structure, with four metra, two in each half of the line. Underlying trochaic or amphibrachic metres can be disturbed by processes such as the suppression of colon initial syllables, and anaclasis, the inversion of metrical positions at the colon, metron or even foot level (hence an iamb can count as a trochee, for example). The resulting schema on p. 89 looks at first glance a bit like the analysis of the Saturnian made by J. Parsons in Transactions of the American Philological Association 129 (1999). However, M. argues through statistics that the Parsons formulation will allow far too many potential verses, whereas his analysis better predicts the forms which actually occur, and even has a good fit with the frequency of different types of lines attested in the corpus.

Once M. has settled the structure of the metre, he is able to go through the less secure textual examples suggesting how they will or will not work in his scheme, and offering possible emendations or explanations to make them fit. He also examines questionable epigraphic examples concluding, for example, that the monument to the baker, M. Vergilius Eurysaces, still standing outside the Porta Maggiore in Rome, does in fact contain a Saturnian, as Bücheler and Ritschl suggested (pp. 222–3).

In the second half of the book M. turns his attention to the verse-patterns of the other Italic peoples. All the examples are taken from inscriptions, and none has any explicit indication that it is metrical, such as a special marker employed at each verse-end. M. has some good things to say here; I particularly liked his comparison with the legend on the back of his iPod (‘Designed by Apple in California Assembled in China’, which he is able to scan as an ‘anapestic hexameter’) with the makers' marks on ceramics from Teanum Sidicinum (such as ‘At Minis Beris's workshop, (these) were made in Teanum’), which Poccetti claimed were written in Oscan Saturnians (pp. 311–14). M. considers both the iPod and the ceramics to display ‘sub-literary jingles’. (It is worth noting here that he also finds a surprisingly high number of possible Saturnian sequences in a test on samples of archaic Latin prose, pp. 203–5).

Much space is given in this second section to the discussion of the possible scansions of some well-known Italic texts, including four South Picene inscriptions, dating from the fifth or fourth century b.c.e., and two late second- / early first-century Paelignian epitaphs. For those who like such things, this section will be a delight, but I was left rather with the impression that every scholar who has set out to prove that these texts contain complex metres has been able to do so to his own satisfaction. Those who are sceptical about the results of earlier attempts may not find themselves convinced by M.'s conclusions. For me, the final section concerned with comparative metrics and the reconstruction of Proto-Italic / Proto-Italo-Celtic / Proto-Indo-European metre was equally unpersuasive. Despite M.'s efforts to show that it is possible to trace the history of metrical patterns through the Italic and Indo-European language family, the evidence does not seem compelling enough to rule out the possibility of borrowing of rhythmical or metrical patterns from other cultures, or indeed the independent creation of verse types.

The strength of M.'s book lies is his use of statistical analyses on the Latin material in order to assess how closely his predictions of the Saturnian correspond to the attested lines, and the results do indeed look impressive. Such tests are not possible on the tiny number of examples from other Italic languages. However, once one looks at the very helpful table of scanned lines included in an appendix, some doubts begin to emerge. It is striking, for instance, that the metrical analysis requires there to be a stress on mihi in both the initial line of Livius Andronicus' Odysseia: uirum mihi Camena insece uersutum, and in Odysseia 7: tuque mihi narrato omnia disertim. In both cases the dative mihi lies in a position where Classical Latin, Indo-European philology and the Greek originals would suggest an enclitic pronoun. It would have been helpful if such examples could have been fitted into M.'s discussion of ‘function words’ on pp. 112–13, where the contrasting need not to have any accentuation on tibi in CIL I² 10.4 is explained.