The historian Livy is known to us through two distinct bodies of evidence. The first is the transmitted text of the surviving portions of his great work. These portions comprise the first ten books and the almost complete text of books 21 to 45 inclusive―that is, a total of thirty-five books from a work that originally ran to 142 books in all. [1] The information that the original work contained 142 books is taken from the second body of data mentioned above, which is dubbed ‘para-Livian’ material by David Levene in the preface to his new edition. [2] The first two volumes of it are reviewed here. Levene defines para-Livian material as summaries, fragments, and all other places in ancient texts where Livy is explicitly cited or mentioned by name. The new edition aims to set these out in full in the original languages, with facing English translation and full commentary. It needs to be said at once that Levene has done a superb job, and that the two published volumes are a triumph of scholarship.
The first volume contains the fragments, and three subsequent volumes are set aside for the summaries that were compiled in antiquity. The most important of these, the so-called Periochae, occupy volume 2 and the first part of the promised volume 3, the later part of which will contain the so-called Oxyrhynchus Epitome, while the final volume will include the Liber Prodigiorum of Julius Obsequens, the list of portents assembled from Livy. The present volume 2 deals with the Periochae of books 1–45, but the long and instructive introduction deals with the Periochae as a whole, and readers will need a full text to hand if they wish to follow the argument. [3] The detailed introduction to Periochae 1–142 leads one to hope that the work of editing, translating and commentating on Per. 46–142 is nearing completion and will be published soon.
Volume 1 begins with the “fragments,” a misleading term used by scholars to describe references in other authors to passages of lost ancient texts. These can take the form of quotations, paraphrases or otherwise vague statements about what the author originally wrote. It is an unfortunate misnomer, because only rarely does the term refer to fragments in the literal sense. Only three fragments of Livy fall into this category: a scrap of a fifth-century parchment codex preserves two severely mutilated passages of book 11 (one on each side of the parchment), printed here as F1 and F2, and a few pages of a palimpsest contains part of book 91, here included as F23. This last, the longest fragment by some distance, which describes Sertorius’ activities in Spain in the winter of 77–76 BC, receives exemplary treatment from Levene, who has been able to take advantage of the digitized images recently produced by the Vatican Library, which have finally made it legible once again; at the time of its last edition, by Robert Ogilvie in the late 1970s, the text was completely invisible to the naked eye, or at least it was to mine, when Professor Ogilvie asked me to examine it on his behalf (see his posthumous publication in PCPhS 30, 1984, 119). Levene’s edition offers several new readings and the commentary illuminates Livy’s narrative technique: for instance, his description of Sertorius’ military preparations recalls those of Scipio Africanus, and also draws upon Xenophon’s account of Agesilaus’ actions before his campaign against Persia (I, 191–192).
To the “fragments” of the lost books,[4] Levene adds two further categories of material: first, what he calls “citations”―places where secondary authors cite the surviving books―and, secondly, a complete collection of testimonia, statements by ancient writers about Livy’s life or work but without reference to particular passages of his text (i.e. every mention of Livy that is not either a fragment or a citation). It is surprising that testimonia have only recently become a standard feature of editions of fragments of lost texts (one would have expected them to be included as a matter of course once Jacoby had shown the way over a century ago).[5] To my knowledge this is the first time they have ever been assembled for Livy. The citations are another welcome novelty, because they allow us to check the secondary quotations against the full text, and to examine the accuracy, as well as the aims and methods, of the citing authors. The opportunity to compare what is transmitted by an indirect tradition with the original text preserved in manuscript(s) arises again with the Periochae, and is one of the most revealing features of Levene’s edition.
A further innovation (this one explicitly acknowledged: I p. ix) is the inclusion of commentary on most of the testimonia, as well as over 50 percent of the citations―a revealing figure, this, given that accurate quotations of surviving passages of Livy scarcely need, and do not here receive, any commentary. As for the testimonia, the utility of Levene’s decision to provide commentary is obvious, and a likely cause of embarrassment for the editors of comparable editions (such as The Fragments of the Roman Historians) that do not possess this valuable feature. By commentating directly on the texts that furnish the evidence Levene is able to treat important matters that would otherwise have had to be dealt with in passing in the introduction, if at all. These include, for example, the question of Livy’s dates of birth and death (T1, T11), his relations with Augustus, who reportedly called him “a Pompeian” (Pompeianus) (T7), and the much-debated opinion of Asinius Pollio, who is said by Quintilian, twice (T9a–b), to have found fault with Livy’s language on the grounds of its patavinitas (“Paduanity”). The commentary allows Levene to discuss this famous judgement at length, and he does so with a balanced assessment of the various possibilities. The context shows that Pollio was referring to Livy’s use of language, not to his social standing or outlook, as Syme famously thought. It is however perfectly possible, as Levene rightly notes, that Pollio was criticising Livy’s spoken Latin. But if so―if Pollio was sneering at Livy’s provincial accent―we might also infer a jibe at his bourgeois origins and his small-town outlook; in other words, Syme’s theory might still have legs.
Another testimonium that allows Levene to dilate on a major issue is Martial’s famous reference to a shortened version of Livy in the form of a parchment codex (14.190 = T23). The commentary offers a detailed examination of the issues which are taken up again in the introduction to volume 2. Levene concludes by rejecting the prevailing scepticism and reverting to the traditional view that Martial is referring to an epitome. He also argues that there is independent proof of an abridged version of Livy existing at this time (around AD 100). The key text is a passage of Plutarch (Marc. 11.4 = Livy C53) which cites Livy as stating that Marcellus’ victory at Nola, while not a major event in terms of casualty figures, was nevertheless important because it encouraged the Romans to think that Hannibal could be beaten. A similar point is made in Periocha 23.4, which says that Marcellus was the first to give the Romans hope in the war. This is interesting because Livy himself (23.16) does not say anything about Marcellus’ enhanced reputation or about the Romans’ improved morale; he merely states that the battle was important because at the time (just after Cannae) it was an achievement not to be defeated by Hannibal. Levene thinks that the difference between Livy and the sources that purport to depend on him is significant, and can be explained by their common use of an intermediate source that had read more into Livy than was really there; and that the said intermediary was most probably a first-century epitome.
Readers may or may not find this argument compelling. For my part I am not entirely persuaded that two authors could not have independently made the jump from Livy’s report of a minor but significant victory to the inference that it was significant because it gave the Romans cause to hope. The important point, however, is that Levene’s lucid discussion, spread across four separate passages (I, xxv–xxvi; 294–95; II, xiii; 486–87), provides all the information necessary to help readers make up their own minds. The same applies to his wider discussion of the long-held belief, going back to Mommsen, that the Periochae depended entirely on a lost epitome, which was used also not only by Plutarch but by other historians such as Florus, Eutropius, Orosius and others, who used it as a substitute for Livy. Levene steers a careful course between this extreme and its opposite, the currently fashionable view that rejects the idea of a lost epitome and maintains instead that the author of the Periochae always worked directly from the full text of Livy.[6] Levene offers a compromise, arguing that the author sometimes worked with Livy’s full text but also made occasional use of a pre-existing epitome. This conclusion is cogently argued and generally persuasive, even if decisive proof in either case is strictly unobtainable.[7] Once again the key point is that Levene sets out the arguments in full and makes it possible for readers to decide for themselves.
In common with much of recent scholarship Levene treats the author of the Periochae as a distinct literary personality who set out to do something more complicated than simply to abridge Livy.[8] His aim was to produce a self-contained, coherent narrative history―based on Livy, admittedly, and adopting the structure of his work, but designed to be read independently, revealing its own aims and prejudices, and frequently including historical details, versions of events, and interpretations different from Livy’s own. These are set out clearly in the introduction and the commentary, and allow Levene to delineate a portrait of the author and to identify his distinctive historical interests. These include his emphasis on the development of institutions and the growth of Roman power (with a particular emphasis on “firsts”), and his concern with the influence of speeches and letters on events; his lack of interest in religion (in marked contrast to Livy), his focus on individuals and their achievements, his failure to address chronological issues, his oversimplified ethics, and his monocausal explanations. There are exceptions to all of these features, however, and Levene’s discussion of them (xlviii–xlix) is carefully qualified and nuanced. Another notable feature of the text, which serves further to identify it as the work of a single author with distinct individual characteristics, is its literary style. Levene analyses this in a brief but illuminating section (lii–lv) which is actually the first proper study ever published of the Latinity of the Periochae.
As for the author himself, Levene speculates that he was writing probably in the third quarter of the fourth century (certainly before 380, as he was used by Eutropius), that he was “manifestly a Christian” (lvi: the adverb strikes me as unwarranted), that he may have known Greek (suggested by his use of Greek placenames), and that he was based in northwest Italy (where he knew the geography better than Livy), possibly in Milan (where there were good libraries). These conjectures are followed by an essay (“Reading the Periochae,” lx–lxxiv) on the difficult question of intertextual allusion in ancient historiography, a process constrained by the fact that the intertexts also served as historical sources, and that the intertextual references reflect the attitude of the later author not only to the texts of the authors they are using but also to the underlying historical reality. The essay is a modified and updated version of previously published material,[9] but with the focus now firmly on the Periochae and the author’s dual purpose: to produce both a summary of Livy and an independent narrative that incorporated much non-Livian material and often strayed away from Livy by substituting alternative versions drawn from his own general knowledge and from sources such as Cicero and Valerius Maximus.
That the Periochae achieved its aim is evident from the fact that it was widely used―by Eutropius, Augustine, Orosius and the anonymous de uiris illustribus among others (as demonstrated on pp. xxviii–xxxii). It obviously contributed significantly to the process that had given rise to its existence in the first place, namely the gradual disappearance of Livy’s full text and the fact that, despite his reputation, Livy was actually little read. This is the insistent message that will impress itself on anyone who studies these texts: the total number of fragments and citations of Livy is small by comparison with Sallust, not to speak of Cicero, Virgil, Horace and Ovid. Large parts of Livy’s text yield almost no fragments at all: the central section from book 51 to book 75 (almost a fifth of the whole work) is cited just once (F16, from book 56; notice, incidentally, that in this fragment the Numantines are translated as Numidians―one of the very few errors that I have been able to detect in these two volumes). Admittedly the number of fragments picks up as the narrative reaches the period of the civil wars, a pattern that is broadly matched by the amount of space devoted to each book by the Periochae (cf. I p. xxix). But the general impression remains: the first five decades were the most often quoted, are given the most space by the Periochae, and have provided the books that still survive today. The citations of the surviving books (one quarter of the total) are almost as numerous as the fragments of the three quarters that are lost.
Levene’s two volumes comprise the most sustained and comprehensive study ever published of the ancient reception of Livy. The story he reveals is paradoxical. The reputation of Livy’s work at the time of writing was considerable (see T4–T6 and comm.). With its lengthy coverage of the whole of Roman history it dwarfed most of its predecessors, and as a literary masterpiece it surpassed them all. The result was that his competitors rapidly fell into obscurity, so that only Livy remained; and by then his work had become synonymous with the history of the Republic. The consequence was that his name was attached to familiar stories even when they were different from what was actually recorded in his text. The sheer bulk of the work meant that, over time, fewer and fewer people had access to copies, and even those who did rarely made the effort to read them.
A final word about the commentary. In the second volume, which deals with the Periochae of books 1–45, there is a marked difference between the commentary on the second decade (books 11–20), for which the corresponding books of Livy are lost, and that on the other decades (books 1–10 and 21–45) for which Livy’s text is preserved. In the latter case the focus is entirely on the relationship between the Periochae and the original Livian text, how faithfully the author summarises the original, and what his selections, additions, and distortions tell us about his historical and literary aims, methods, interests and prejudices. The historical events themselves require no comment, except when the Periochae offer a version that differs from Livy’s. The situation is completely reversed when the corresponding books of Livy are lost. The Periochae of the lost books provide the only evidence of Livy’s narrative of events; in doing so they also provide vital historical evidence for the events themselves, especially for those periods for which there are few other sources, if any. Levene therefore provides a detailed historical commentary on the Periochae of books 11–20, and fully immerses himself in the events there recorded: the final stages of the Roman conquest of Italy, the plebeian secession that led to the lex Hortensia between 289 and 283 BC (his comment on the date, pp. 220–21, is instructive), the invasion of Pyrrhus, the First Punic War, and the difficult period of the 230s and 220s, including the conquest of Sardinia and Corsica, the Illyrian Wars and the Gallic campaigns of 225–222 BC. On these and related matters the historical commentary is both comprehensive and excellent. Just as these two volumes will provide the definitive account of Livy’s reception to the end of antiquity, so too the commentary on Periochae 11–20 will serve as the essential starting point for all serious historical research on the obscure period they cover. Historians will be impatient for the publication of the remaining volumes, and the commentary on the destruction of Carthage and Corinth, the Numantine War, the tribunates of the Gracchi, the career of Marius, and the Social War―to name just some of the events for which the Periochae are a major source.
Notes
[1] The text of the final surviving pentad, books 41 to 45, is known from a single manuscript (the fifth-century Vienna 15), which at time of its discovery in the sixteenth century had a number of substantial gaps. For details see the introduction to J. Briscoe’s 1986 Teubner edition (summarised in English in his A Commentary on Livy 41–45, Oxford 2012, 3–4).
[2] The number is not without problems and has been challenged; it is also uncertain whether Livy intended to end at book 142, or to continue until he had reached, say, 150 in all, but was unable to complete the task. See Levene’s discussion in vol. II, lxxviii-lxxix.
[3] Most conveniently Arthur Schlesinger’s 1959 Loeb (with English translation), or Paul Jal’s 1984 Budé (with French translation and notes: see further n.6, below).
[4] They number 89 in total, to which Levene adds four further Livian fragments (F90–93) on the art of rhetoric.
[5] F. Jacoby, Die Fragmente der griechischen Historiker I (Berlin 1923).
[6] Thus C. M. Begbie, CQ 17 (1967), 332–8; P. Jal, Abrégés de l’histoire romaine de Tite-Live, XXXIV.1 (Paris 1984), xxvi–lv.
[7] This point was already made by John Briscoe, Gnomon 57 (1985), 419–20.
[8] Thus Jal (cit. n. 6, above), lxvii–xc; W. J. Bingham, A Study of the Livian ‘Periochae’ and their Relation to Livy’s ‘Ab Urbe Condita’, Unpublished diss., University of Illinois at Urbana-Champaign, 1978; J. D. Chaplin, in M. Horster, C. Reitz (eds), Condensing Texts―Condensed Texts (Stuttgart 2010), 451–67; R. L. Love, Writing after Livy: Historical Epitomes in the Livian Tradition (Diss. Yale University, 2019.
[9] Levene asks readers to read the current version in place of the electronic publication Histos Working Papers 2011.01. Both depend heavily on his chapter ‘Sources and Intertexts’ in his Livy and the Hannibalic War, (Oxford 2010), 82–163, where the emphasis is mainly on the relationship between Livy and Polybius.