The ‘Italic tradition’ of F.’s title refers to a cultural koinê that putatively existed in the Archaic and early Classical periods among the peoples of central Italy. Elements of a common material culture we know there were, in some areas at least; and this evidence suggests the existence of analogous social practices, perhaps extending beyond imported items like the symposium to include shared or similar political and religious institutions. F.’s hypothesis is that Ennius’ Annales contains linguistic evidence that refers to this cultural koinê. But his aim is not so much to strengthen the case that this shared culture existed: he believes that the material record proves as much. Building on this belief, he argues that Ennius refers to this culture often enough and in a sufficiently distinctive way as to support a general interpretation of his epic. Instead of a one-way translation and mediation of Greek literary culture into Latin by an outsider who was an adept of both, F. suggests that the Annales represents a more complex triangulation of the cultural vectors representing ‘Hellenisation’, the hegemonic ambitions of the Roman ruling class and the understandings that Romans shared with their Italian friends, relatives, allies and subjects. Accordingly, the poet, being himself of a people belonging to this cultural koinê, takes on a role different from that of the culturally marginal arriviste or poeta cliens. This is a new and provocative conception of the cultural work performed by Ennius’ masterpiece, and one that might also apply to other aspects of Roman literary production in the third and second centuries b.c.

The potential appeal of such a book is obvious. Up to now, linguistic evidence has enabled editors to invent unattested forms of archaic Latin to try making sense of passages garbled in transmission, but has not contributed much else. F., working almost entirely with the text as printed by O. Skutsch (1985), identifies collocations of words that he believes belong to identifiable discursive systems based on religious ritual, military concepts, family relations and so on. Chapter 1, ‘Ennius and the Italic Tradition’, lays out the general framework of assumptions on which he bases his approach. Chapter 2, ‘The Annals and the Greek Tradition’, refocuses Ennius’ act of representing the Roman experience in terms borrowed from Greek culture, as one of finding common ground in concepts shared by the Greeks, the Romans and the other peoples of Italy. The remaining three chapters concentrate on specific aspects of ritual as represented in three particular contexts, two of which belong to Book 1 and the other to Book 6; these are ‘Ritual and Myth in the Augurium Romuli (Annales 72–91)’, Chapter 3; ‘Ritual, Militia, and History in Book 6 of the Annales’, Chapter 4; and ‘Ritual, Kinship, and Myth in Book 1 of the Annales’, Chapter 5. It is not surprising that the main examples come from two of the best attested books in a very fragmentary record, but this distribution must limit somewhat the material's explanatory potential for the poem as a whole.

There is a more basic issue, as well. The idea that there existed, across the languages of central Italy, certain verbal formulae relating to such discursive systems seems hardly unlikely, but the nature of any relationship among them is difficult to assess, and its existence impossible to prove. The problem is that the evidence is so noisy. First, a number of the comparanda F. uses to identify these formulae involve similar collocations in the Annales and later texts in Latin, which may be alluding to Ennius himself, or else to concepts germane to their own times, rather than to any archaic inheritance. Second, the independent comparanda that F. finds in Italic languages – mainly inscriptions that either antedate Ennius or are extremely unlikely to refer to him – often seem to derive from a common Indo-European source, or even to be the kind of phrase that might arise spontaneously in any language. That being the case, the existence of a central Italian cultural koinê that expressed itself by using similar phrases across language barriers is not the only, or the most economical, interpretation of the evidence.

An example: one of the simplest forms of ritual collocation that F. adduces is that of a verb plus a noun complement, such as donum dare (to which he believes vv. 183–91 Sk. alludes, though the phrase does not actually appear there: see pp. 92–7). Often, as in this example, both words are etymologically related, and the noun is a direct object, but variations on both these general rules are found (cf. pede(m) pulsare, v. 1 Sk., pp. 29, 41–4; voce vocare, v. 49 Sk., pp. 146–50). Cognate phrases in Oscan or Umbrian, particularly in what seem to be ritual contexts, suggest to F. that the Latin forms of these phrases in Ennius allude to similar ritual contexts. His interpretations of different individual collocations seem to me more or less plausible, and I imagine that other readers will keep their own counsel in each case. But the most basic problem in all cases is that every language I know possesses a number of similar, if not identical, fixed phrases. A few of the more obvious ones in English are ‘do a deed’, ‘give a gift’, ‘sing a song’ and ‘tell a tale’. Some of these exist in other languages, too: ‘give a gift’ is exactly equivalent to donum dare, and all these phrases share certain properties with that phrase (alliteration, isosyllable, etymology, verb plus cognate substantive etc.). Admittedly, even if these are very common phrases, it may be that a special significance can be activated in certain situations, but it might equally not. For instance, the collocation ‘sing a song’ makes one of its signal appearances in the King James rendering of Ps 96:1 as, ‘Oh, sing unto the Lord a new song’. It had been equally at home in biblical Hebrew ( shiyrû lahwäh shiyr chädäsh), and the translators of the King James Version follow those of the Septuagint (ᾄσατε τῷ κυρίῳ ᾆσμα καινόν) and of the Roman psalter ( canite Domino canticum novum, which was approved and adopted by St Jerome) in finding a close English equivalent. But they were hardly using a ritual phrase familiar from earlier English to render a devotional text from the Hebrew. And even if this were their intent, when such common collocations are involved (the phrase is plentifully attested in Old, Middle and Early Modern English according to OED s.v. song), is it really possible to tell? F. attempts to find his way through this problem in his treatment of voce vocare (cited above), for which he finds partial parallels in later Latin and also in Umbrian and the Homeric collocation εἰπεῖν ἔπος (Il. 7.394; cf. εἶπας ἔπος 1.108, εἶπῃσθα ἔπος 20.250). As he admits, however, Homer never uses this collocation in a ritual context; and he might have added that it forms the basis of the most common clichés in the Greek language (ὡς ἔπος εἰπεῖν).

In general, F. is commendably alert to the slippery nature of his evidence, which means that he uses adverbs like ‘possibly’ and ‘perhaps’, and modals like ‘could’, ‘would’ and ‘may’, to a very unusual extent. As a result, I believe that this book contains not a single dispositive, conclusive or irrefutable argument. It does contain a wealth of learning, and it puts forward a number of ideas and interpretations that, individually and collectively, merit further discussion. But I would caution anyone, particularly any less experienced readers who are, happily, being attracted in numbers to the study of Ennius by the bounty of innovative and exciting work that has recently been appearing, to bear in mind that this contribution is more valuable for the possibilities that it raises than for what it actually proves.