The two volumes under review comprise the first instalment of a much-needed effort to provide text, translation and commentary for ‘every reference to Livy made between his own times and the end of classical antiquity’ (vol. 1, p. xvii), loosely defined as 650 ce. Still to come are the volumes covering Periochae 46–142, but even so there is much to be getting on with. L. counts ‘between 225 and 257 references to Livy in all of post-Livian literature’ (vol. 1, p. xx), a range that excludes the periochae, but includes fragments (allusions to the lost books of Livy), citations (references to the surviving books of Livy) and testimonia (references to Livy that are neither fragments nor citations). It is also a surprisingly small number. By contrast, Sallust’s lost Histories alone survive in 550 fragments, and Cato the Elder’s works, including both speeches and histories, in over 400 fragments. Livy, in other words, is under-used, and, as L. notes, it is not hard to guess why that should be: the sprawling size of his Ab urbe condita (AUC) meant that access to any specific section could be tricky, and reading the whole daunting. Still, Livy’s capaciousness endowed him with a certain authority: he became effectively identified with republican history and received frequent credit for historical traditions and details that were not present in the AUC. He was, therefore, under-read but not under-appreciated – the very definition of a classic.

Like all classics, Livy now comes equipped with his own set of ‘SparkNotes’, the periochae, a set of loose summaries whose readership today is even more rarified than Livy’s. Four such epitomes survive: the Periochae; the contents-listing that comes to us attached to the MSS of the Periochae, usually referred to as Periochae A; the Oxyrhynchus Epitome, a third-century papyrus containing summaries of the (surviving) Books 37–40 and the now lost 48–55; and Julius Obsequens’ Liber Prodigiorum, which anthologises omens and prodigies in Livy. The latter two will be treated fully in the future volumes 3 and 4, respectively, as will the rest of the Periochae after 45. The Periochae are a slippery sort of text, with which modern audiences struggle to engage, and in some sense the fact that they are included here is a great service, as they are now possessed of a scholarly apparatus to complement J.D. Chaplin’s 2007 translation for the Oxford World Classics series.

L.’s contribution, of course, goes further than mere acknowledgement, but one of the things he suggests early on is that Livy’s afterlife is fundamentally bound up with the circulation of his epitomes. In addition to the four surviving epitomes, L. is in favour of the existence of a first-century epitome of Livy, possibly the one famously mentioned by Martial in the generation immediately following Livy’s (14.190 pellibus exiguis artatur Livius ingens | quem mea non totum bibliotheca capit), though he does not necessarily see that early epitome as an intermediary source for the Periochae whenever they have a non-Livian version of events. The citation patterns tell a more granular story. By and large, the surviving books of Livy are cited more heavily, with the figures dropping the further one gets into the AUC. Thus Book 1 is cited a lot, then the rest of the First Decade, with each subsequent decade cited less and less until citations drop off almost entirely after Book 45. Thereafter, 35 books lapse with hardly any mention, until there is an uptick with Books 76–83 (the Sullan–Marian wars of the 80s bce), and then a much larger uptick in Books 91–120, the last generation of the Roman republic. Some of this unevenness is the result of the citing authorities’ interest – the Lucan scholia give us many fragments about the civil war, and so on. More interestingly, however, is L.’s suggestion that ‘the pattern of citation … very approximately … correlates with the balance of the narrative of the periochae’ (vol. 1, p. xxviii). Which is to say, the length of the summaries of the periochae follows the fragments’ distributive pattern: the first decade gets fullest treatment, the second less so, and so on, with the civil war books getting slightly fuller treatment than their neighbours. L. does not impute any causation, but his discussion reminds us that Livy circulated at different scales. Some decades were copied deep into antiquity, while others faded away into an epitomised afterlife. More importantly, this kind of observation, whether or not it amounts to anything, is exemplary of the kind of work we can only do and the kind of questions we can only ask by bringing all these disparate materials together.

Together, but without losing the distinctive features of either style of text or commentary. L. notes that the project is eccentric at the outset, and indeed these volumes do a lot, not least in terms of the different considerations due to fragmentary texts, their corresponding cover-texts, testimonia (the commentary on which is an editorial innovation much to be applauded), and the almost complete and largely independent text of the periochae. This is a lot to juggle, and occasionally things fall through the cracks (e.g. L. gives an introductory note on Sulla on p. 179 [F19], but this is not Sulla’s first appearance in the text, and the note would have been equally useful earlier); on the whole, though, L. does an excellent job of keeping us attuned to the main question of Livy’s afterlife, but without losing focus on the contingent individualities of the specific texts at hand. The division of the volumes likewise adds flexibility; they can currently be purchased both individually and bundled together, and in an electronic edition as well. This will be handy for those who need to work or teach on ‘just’ the periochae, or ‘just’ the fragments, as the volumes are stand-alone enough to be used independently. There are some stylistic quirks, but the only one I found consistent enough to notice was that, while the fragments themselves give a full citation of the cover text (e.g. F9 = Val. Max. 1.8 ext.19), the commentary lists only the fragment number, without the cover text citation. This is only occasionally annoying, and certainly far from insurmountable, but also an easy fix that would prevent some unnecessary toggling.

Reading Livy’s fragments is an exercise in frustration. Only titbits survive for some of the juiciest periods of Roman history. We get a taster of Cicero’s death (F61) but not his life, except that Livy did recount the Catilinarian conspiracy, but no indication as to how (F34, with L. pointing out that Jerome might well be working from the periochae rather than from Livy’s text). Even these meagre traces confirm some of our impressions: Livy was generally fair and clear-eyed in his accounts of the great men of Rome, alive to their faults as well as their virtues, and, in the case of Cicero, able to keep apart his adoration of Cicero the stylist from his more tepid appreciation of Cicero the politician. More tantalisingly, there is F23, the longest Livian fragment, surviving on a single bifolium that was reused in the eighth century to copy out the Book of Judith (p. 185), and containing part of Sertorius’ campaigns in Spain in the winter of 77/6 bce, a time when Rome was trying to steady the ship after a period of turbulence and when Pompey, fresh from putting down the Lepidan revolt, was about to suffer an embarrassing defeat in Spain from Sertorius’ armies. The information in F23 is not recounted by any other source, though Per. 91, Sallust’s Histories and Plutarch’s Life of Sertorius touch on the events. L. does a good job of rehabilitating this bit of Latin from R.M. Ogilvie’s scathing criticism (‘mechanical and careless’, ‘undistinguished’), chiefly by reminding us of one of Livy’s great strength: his ability to ‘achieve powerful effects … by virtue of their contextualization in the wider drama of Livy’s history, as the characters’ plans and behaviour interact to complex effect … and as the dilemmas of maintaining ethical actions in a world of brutal competition are repeatedly explored’ (vol. 1, p. 186). Even fragmentary, we can see the depiction of Sertorius reaches out both to Livy’s account of Scipio Africanus and to Xenophon’s Agesilaus, an intertextual double helix entirely typical of Livy’s surviving work.

Elsewhere, the fragments reassure us that Livy remains Livian even when we cannot read the full original version. F26, from Frontinus Strat. 2.5.34, records that 35,000 armed men fell in Crassus’ battle against Spartacan forces at Mount Cantenna in the winter of 72/1 bce. Those 35,000 men fell along with their commanders, and Crassus recovered five Roman eagles, 26 standards and many spoils, including five fasces cum securibus (1.204–5). As L. points out, Livy plausibly included the detail of the eagles as a matter of irony: Crassus would lose his own eagles in 53 bce, not to be recovered until Augustus. But this kind of numerical reportage is typical of Livy, as any cursory glance at the extant books reveals. It is reassuring, therefore, to see the same habit persist in the same form in Livy’s account of the 70s, when records and memories would have been more reliable. Still, Livy had to have obtained these figures from somewhere, and often in his casualty lists he tells us where: Antias, Claudius, Polybius and so on. The fragment, however, does not preserve this information, either because this was an instance where Livy did not supply it or because the fragment comes from a section of the paragraph before or after names were mentioned or, perhaps, because the process of transmission elided the sources as a matter of course. We might compare here C44, where Orosius cites such figures drawn from Livy, with some crucial differences (on which see L.’s discussion in vol. 1, p. 293). What caught my interest is that Orosius does not report that Livy felt these numbers had been exaggerated by ‘some authors’: 10.30.5 sed superiecere quidam augendo fidem. Whoever these quidam are – and Antias is always a likely bet –, they were important enough for Livy to note, but not for Orosius, nor were they (and my money is once again on Antias) for Frontinus. The fragments, read with and against the testimonia and citations, not to mention extant Livy, deepen our understanding of Livy’s practices, how he maintained them for so very long and what effect, if any, this had on his readership.

If the fragments titillate, the periochae are a funhouse mirror, using Livy to create an experience of Livy that is, if not entirely new, then at least creatively engaged with the AUC. The volume has its own independent introduction, where L. lays out some of the main issues. Did the author of the periochae work directly from Livy, or was there an intermediary text? And did he aim only at miniaturising Livy, or rather at doing something more creative? Consequently, can we rely on the periochae as complete or faithful summaries of their respective AUC books? The answers, naturally, are all related. L.’s view is that ‘the Periochae is not merely summarizing Livy, or even reproducing Livy’s words … but appears to be engaging with his text in a creative way, the sort of creativity which only make sense if the author has the actual text of Livy in front of him’ (vol 2, p. xxiii). So, for instance, the author of the periochae regularly uses Livian phraseology, but draws it not from the relevant bit of text he happens to be epitomising, but rather from callbacks later in the text. The periochae’s account of Servius Tullius’ birth takes its working not from Livy’s Book 1, but from a reference back to the story in Book 4 (4.3.12). The same is true for Marcius’ saving of the Roman armies in Spain in 211 bce, where the periochae look to a callback in 26.41.5 rather than the original moment in Book 25. Nor are the periochae without agenda. The Trials of the Scipios get a far more pro-Scipionic treatment than Livy’s complex account, with the accusations of the Petilii stated as outright false and the atrocities of Pleminius at Locri entirely unmentioned. Finally, the periochae are no stranger to the general identification of Livy and republican history, adding in details of general knowledge that are not in Livy, but are compatible with his history.

From these and other authorial choices we can start assembling a sense of the periochae’s sensibilities. The author was a careful reader, ‘looking synoptically at Livy and relating different parts of his text to one another’ (vol. 2, p. xxiv), which is not to say that he shared all of Livy’s interests. Like Livy, he was interested in the growth of the Roman state, with what we might term a more bureaucratic slant. He records almost every instance of the expansion of the tribes, all colonisation, every census and is very focused on firsts. Unlike Livy, the periochae have no interest in religion, no consular dates and only a few AUC dates (this is distinctive, since other epitomators preserve both chronological markers and religious material), and by and large remove alternative versions. The periochae also lean towards ethical simplicity, and while there is a clear interest in individuals, the treatment is uneven: Tiberius Gracchus occupies the whole of Periochae B58, but his brother Gaius is treated more diffusely, over several books. Finally, it mattered to the author that his reading of Livy be recognised: we have seen already that the periochae are sensitive to the way in which Livy disposed of narrative, hearkening back to important events in later books. The reverse also applied: ‘sometimes the Periochae alludes to the wider context in Livy not so much by direct use of Livy’s language, but rather by the inclusion of a detail which is apparently superfluous to the story as the Periochae narrates it’ (vol. 2, p. xxv).

A good edition leaves readers not only with questions and threads to pull at, but with the means to do so profitably. It provides a solid launching pad for future inquiry. L.’s lost Livy achieves all that and more; it will be a useful tool in the hands of students and teachers, encourage readers to reach beyond extant Livy and remind us all how fundamental Livy was to the project of Roman history. All this in sturdy, readable and accessible format. Who could ask for more?