I

These are exciting times for the study of Roman literature. New editions of the fragmentary poets and prose writers either have appeared or are in preparation,Footnote ¹ while renewed interest in the nature and character of the early Roman audience, and in the development of Roman literature itself, has sought to explain the larger context in which these writings arose and were read.Footnote ² Monographs and commentaries on early writers and genres have appeared in profusion, with four volumes on Ennius' Annals alone published within two years, at least one of which has raised serious questions of what we can and cannot know of the structure, orientation and contents of a poem that was (uniquely) influential on both later poetry and history.Footnote ³

Roman historiography has been a beneficiary here as well. In the fifty years since Ernst Badian's mini-masterpiece on the early Roman historians,Footnote ⁴ new approaches and new questions scarcely envisioned by Badian have proliferated. The widening scope brought about by the flourishing of historiographical (as opposed to historical)Footnote ⁵ studies of the Roman historians has led scholars to examine the ways in which the Romans created and utilized their past, just as much as whether this or that historian was a reliable guide to early Roman history. Stephen Oakley's magisterial four-volume commentary on the second pentad of Livy's history is notable not only for its rigorous examination of the historical Realien behind Livy's text (to the extent that these can be determined) but also for its on-going appreciation of Livy as an artist actually shaping and constructing that history for his contemporaries and later generations.Footnote ⁶ Stimulating and illuminating treatments of important aspects of the Roman historical and historiographical tradition as well as nuanced readings of individual historians are now to be found in abundance.Footnote ⁷

Nor is our sense of how the Romans looked at their past bounded any longer by the conventions only of ‘historical’ texts. Though these, of course, loom large, and always will, in any treatment of how the Romans understood and appreciated their past, recent studies have emphasized the many other components of Roman memory. Uwe Walter's superb Memoria und Res Publica gives a comprehensive overview and analysis of how much of the Romans' sense of history was dependent on artistic representations, stage performances, commemorative inscriptions, tombs and monuments, and lieux de mémoire in general, not to mention performative moments such as the Roman funeral and procession, and the pervasive influence of exemplarity.Footnote ⁸ All of these things worked to keep before the eyes of Romans of all orders the great men and, to a lesser (but persistent) extent, women of the past and their achievements, and set a yardstick by which later generations could and did measure themselves.

II

It is into this exciting milieu that a new edition of the fragments of the Roman historians now comes. For the better part of a century, scholars of Roman historiography had to make do with Hermann Peter's problematic and outdated collection of the fragmentary Roman historians.Footnote ⁹ As Felix Jacoby was at work on his Fragmente der griechischen Historiker during the early and middle parts of the twentieth century, it became clear that something similar was needed for the Romans. Jacoby himself called for such a collection already in 1949,Footnote ¹⁰ but while some excellent editions of individual historians appeared in the following decades, no complete collection, done in a modern manner, was forthcoming.Footnote ¹¹ Then suddenly, the dam burst: Martine Chassignet's Budé edition in 1986 of Cato's Origines heralded her comprehensive three-volume collection of the pre-Sallustian Roman historians a decade later,Footnote ¹² and these volumes were quickly followed by a two-volume study edition of the major testimonia and fragments from Hans Beck and Uwe Walter.Footnote ¹³ Shortly thereafter Peter Scholz and Uwe Walter produced a slim volume, similar in orientation to Beck–Walter, which covered the writers of memoirs.Footnote ¹⁴ And now, biggest of all, come these three volumes under the general editorship of Tim Cornell, the fullest treatment yet of the lamentably sparse remains of Roman historiography.Footnote ¹⁵ The collection, proceeding in chronological order,Footnote ¹⁶ begins with the Annales Maximi and Fabius Pictor, and ends in the third century c.e. with Marius Maximus and Asinius Quadratus. It is massive, magisterial and sometimes maddening. It is superbly done, and will be the basis for the study of these historians, and of Roman historiography in general, hereafter.

The physical arrangement of the volumes is unlike any I have encountered before: the first volume contains an extremely valuable general introduction (see below, §III), followed by the introductions to the individual historians and several appendices; the second volume contains the texts of the testimonia and fragments with facing English translation, while the third has the commentary on the fragments. As is now standard practice, testimonia are laid out separately from fragments, and while both are translated, there is no separate commentary on the testimonia: the reader is directed to the individual introductions for discussion of these. In practical terms this requires that a scholar using the collection must have open simultaneously three large volumes; woe betide those with small desks! It also means that it is impossible for several scholars to use the collection simultaneously, since the study of any single historian will demand all three volumes at once. The editors say that they wished to avoid the inevitable flipping back and forth that is so often a feature of commentaries, and while they have done that, their arrangement comes with its own cost.

III

The first volume opens with a massive general introduction (I, 3–137), which contains: (i) a brief review of earlier editions; (ii) the scope, structure and contents of the present edition; (iii) a study of the prose styles of the authors;Footnote ¹⁷ and (iv) a summary of each of the citing authors (Appian to Velleius Paterculus), with observations on how each used historical works and the consequences that this has for our understanding of the fragments themselves.

The first section, after pointing out that this new edition is the most comprehensive produced,Footnote ¹⁸ turns to the criteria for inclusion in the volume. The last two decades have seen a number of important studies on the rationale and presuppositions of collections of fragments and commentaries, and it seems clear that the editors are familiar with them, given the caution with which they proceed in their collection.Footnote ¹⁹ Scholars have pointed out, for example, that the ‘cover text’, the text in which the ‘fragment’ is embedded, is crucial in evaluating the fragment itself, since the quoting author often employs such citation or quotation in disagreement or polemic: we cannot, therefore, take at face value the remarks made by those who cite or quote. It is often very difficult, moreover, to know for certain the extent of a ‘quotation’, since the citing authors often add remarks of their own, especially by way of explanation, not to mention that the ‘quotation’ itself may be largely paraphrase, often unbeknown to modern readers.Footnote ²⁰

The lessons imparted by this recent scholarship have been learnt by the editors, and caution may be said to be almost a defining motif of this collection. At the very beginning they alert the reader to the limits of the collection (I, 3, 7 with my emphases):

Pointing out that their arrangement of the fragments is ‘more conservative’ than that of their predecessors, they emphasize that their ‘arrangement and methodology [are] … designed to make clear the distance that separates the lost originals from the surviving remnants’, such that conjectures and theoretical possibilities are ‘confined … to the commentary’ (I, 7). Even within the individual entries, reminders are given of just how fragile our knowledge of the contents and arrangements of fragmentary histories is.Footnote ²¹ Indeed, one can hardly read a few pages without encountering peremptory dismissals of this or that scholar's reconstruction and/or interpretation, often without much discussion. It may seem that we are moving backwards, but in fact we are not, since we are reminded again and again of how little we know about these historians and their works. Given the fantasies that have too often been spun from the slender filaments of fragmentary authors (not least the Roman historians), I think this edition's approach is appropriate and admirable.

Sometimes, however, the editors' caution might go too far. In the section on the Annales Maximi, for example, only explicit citations of the Annales are deemed sufficient to be considered a fragment. Chassignet, by contrast, in her Budé edition included a section ‘Libri Annales sine nomine’, which collected all the references in later authors to ‘annales’; here Chassignet seems to me to have understood better what was helpful to scholars. The editors of FRHist could have made clear that some of these were doubtful fragments (as the editors in fact do with many of the historians elsewhere in the collection), but they ought nevertheless to have been included for two reasons: first, where the testimonia and fragments are so sparse as with the Annales Maximi, and where scholarly uncertainty is so great, it is best, I think, to err on the side of comprehensiveness, provided only that the reader is warned that such ‘fragments’ are not necessarily to be considered certain. Second, although it seems pretty clear that a number of Chassignet's fragments are not from the Annales Maximi, there are several that I think are likely to be.Footnote ²²

The criteria for inclusion in FRHist are clearly spelt out, although not entirely consistent (as the editors themselves admit). Roman historians are defined as all Romans who wrote chronologically-ordered prose narratives of primarily political and military events (I, 7). Hellenophone historians are included, of course, as are freedmen, but not non-Romans (so no Theophanes or Juba II), not poets (regrettable but understandable), and not authors whose works survive in substantial part and can be consulted elsewhere (for example, Sallust's Histories, Granius Licinianus or Julius Exuperantius). Biographies of famous men, however, as well as memoirs and autobiographies are included, but antiquarian writers are not. One can, of course, justify the inclusion of biographies because of the important content which they contain. Yet having transgressed their generic boundaries so as to accommodate biography, the editors raise questions as to why this historical literature is included but not other types.

On the issue of ‘antiquarian’ writers, the editors, it seems to me, tie themselves in knots by trying to separate out historians and antiquarians, even though (as they well recognize) the latter term is not ancient and would not have been employed by Roman writers.Footnote ²³ Yet even so, they include some works, such as Varro's on Trojan families and Hyginus' on Italian cities, noting that ‘they would probably not have been regarded by the Romans as “history”’ and concluding, ‘we have had to make some more-or-less arbitrary choices in marginal areas’ (I, 9). But why, then, is Verrius Flaccus not included? If the dividing line is narrative, that will not do, since it is clear from Gellius 4.5.1–7 that Flaccus' Book of Memorable Events did contain narratives, including the famous one about Horatius Cocles and the treacherous Etruscan haruspices — which the editors duly print as F6 under the Annales Maximi since Gellius says that the story was in both. Here again, the issue is more than academic, for it goes to the heart of the reason for such a collection's existence and its putative audience. As it happens, Verrius Flaccus, along with a number of other writers, is assigned to an (excellent) appendix of thirty-nine ‘also-rans’, authors who were not included in the collection. Here the reader may find the relevant testimonia and bibliography, together with the reasons in each case why the author was excluded.

Good sense and care are evident in the layout of the testimonia and fragments. I was troubled only by the typographical conventions which are designed to identify more clearly what exactly is part of the fragment and what is not. Bold type is used for any material attributed to the lost source by the citing author, while bold italic indicates what purports to be a verbatim quotation. In the case of uncertainty, bold type is used only for what can ‘with reasonable confidence be taken as attributed to the source’ (I, 15). So, for example, an entry will look like this (Cato, FRHist 5 F141):

Quite apart from the value of such a convention, I confess that I find it to be somewhat at odds with the caution displayed in the volumes as a whole, and I worry that material so marked by the editors will come to be taken by scholars as being an accurate or a verbatim quotation by the citing author, even though we know so well how careless (to use a more benign term) such citing authors can be.

IV

The editors are very clear as to what constitutes a fragment as distinct from a testimonium (I, 14 with my emphases):

The editors in a footnote make reference to Fabius Pictor and the Tauromenium inscription which both Chassignet and Beck–Walter printed as F1 but which FRHist prints as a testimonium (1 T7), and this particular case raises interesting issues. The passage is by now a familiar one and first characterizes Pictor himself (‘Quintus Fabius, surnamed Pictorinus, a Roman, son of Gaius’), and then gives a brief (lacunose, alas) summary of his work, which includes the arrival of Heracles, Lavinia, Aeneas and the ‘later foundation’ of Romulus and Remus.Footnote ²⁴ The first four lines are clearly a testimonium, but what of the remaining ones? Given that what is summarized here are the contents of Pictor's work, it seems to me an over-fine distinction to say that because it does not refer to a specific passage it is not a fragment. But it does refer to specific passages, does it not, those places where Fabius treated Heracles and Aeneas and Romulus and Remus? And when one considers that in none of the fragments of Fabius that follow as printed in FRHist is there any mention of Heracles, the reader who relied on the fragments alone would not know that Fabius had in fact treated the hero. In addition, if, as the editors say, Pictor proceeded chronologically, Heracles must have been treated before Aeneas, such that the appearance given by the fragments — namely, that the work began with Aeneas — is misleading. We can be pretty sure, thanks to the Tauromenium inscription, that it did not. So here it seems that it would have been preferable to split this up into a testimonium and a fragment (or to repeat the latter part of the inscription in the fragments).

Sometimes the failure to place a remark into the fragments results in possible distortion. It is fine to print in the testimonia for Pollio's history the remark (FRHist 56 T5, from Tacitus' speech ascribed to Cremutius Cordus) that ‘the writings of Asinius Pollio preserve a highly complimentary record of these same men [sc. Cassius and Brutus]’. But it ought to have been in the fragments as well since these two men appear nowhere else in the fragments, which are dominated by ‘particular passages’ on Caesar; and it would have been helpful to be reminded amongst such references that Pollio's work also contained (favourable) accounts of the deeds of Cassius and Brutus. I realize that in both these cases one can argue either way; but if it is the aim of a collection such as this to put before the reader all the relevant information, then anything that has to do with the contents of the work (be it specific or general) should be in the fragments.

The absence of a commentary on the individual testimonia is said to be made up for by taking full account of them in the introductions to the individual authors, but this is not, alas, always the case. Indeed, of all the editorial decisions made, this was the most regrettable. There is, as might be expected, much in the testimonia that cries out for comment but is not addressed in the introductions. I leave aside here the General Testimonia on Roman historiography (GT1–7, found at II, 2–9) where commentary would have been extremely helpful in making links with the other material found in these volumes (and in preserved authors, of course). To mention only Fabius Pictor again: since the authors do not comment directly on the testimonia, there is no discussion of why Pictor might have treated Heracles; the hero is mentioned only in passing in the Introduction, where he is included as forming the ‘legendary prehistory’ of the city along with Aeneas and Evander. Likewise, the remark concerning Cremutius Cordus' ‘spirit in which he deplored the civil wars’ (FRHist 71 T3) elicits no comment, yet if it is an accurate characterization of Cremutius' history (and there is no reason to think that it is not), it should have been linked with Velleius' remark that ‘no one has lamented the fortune of this entire era [sc. that of the civil wars] in a sufficiently worthy manner’.Footnote ²⁵ Perhaps even more importantly, amidst all the questions of Cremutius' affiliations, this valuable indication of his disposition towards the events he narrated has been passed over. These are but a couple of examples; lost opportunities in the testimonia abound.

V

In this review article I have tended to focus on those areas that I have found wanting or problematic, so let me assert again that these volumes have a wealth of important and useful information, and all who work in the fields of Roman history and historiography are greatly in the editors' debt. Excellent treatments of important issues abound, and here I limit myself to but a few examples.

John Rich's overview of the issues surrounding the Annales Maximi could not be better (I, 141–59): he gives an extremely helpful summary of the testimonia, separates out carefully what we do (or do not) know about the tabula apud pontificem, explains what, if any, is the relation between the tabula and the eighty-book edition of the Annales, and what the contents and nature of each entity might have been. There is a clear summation of scholarship on the matter, some rejection of unnecessarily hypothetical relationships (for example, between the tabula and the calendar, going back to Mommsen), but at the same time a recognition that no one hypothesis has yet been able to account for all of the testimonia. Likewise, Tim Cornell's masterful discussion of Cato's treatment of the Aeneas story seems to me pretty much perfect for a work of this sort. In just two-and-a-half pages he manages to sum up decades of scholarly discussion in a succinct and clear way, point out where the contradictions appear and expose the shortcomings of the traditional way of reconciling the fragments (III, 65–7). Even though no solution to the dilemma is proposed, the reader well understands what the issues at stake are and from where, if anywhere, a solution might be reached.Footnote ²⁶ (Indeed, his commentary on all of Cato's fragments will now be the standard discussion.) John Briscoe offers a ‘fresh assessment’ of the distribution of the material in Coelius Antipater's history of the Second Punic War, and of how the individual years fit into Coelius' seven books, making a compelling case for a more evenly distributed number of years per book (I, 257–60). The reconstruction has, of course, a certain speculative quality to it, but in keeping with the policies of the collection, the arrangement proposed by Briscoe does not affect the way the fragments are arranged in the collection and therefore the fragments themselves do not prejudice the reader into accepting Briscoe's hypotheses. Mark Pobjoy's long discussion of the textual problems with Sempronius Asellio's remark on the distinction between history and annals (20 F2) is invaluable for offering a comprehensive overview of the ways in which scholars have tried to emend the text (III, 278–81).Footnote ²⁷ Andrew Drummond's introduction to Pollio's career and history is exemplary: it avoids making him a zealot of either side, and is careful to note that his legendary independence did not necessarily indicate hostility towards or criticism of the new Augustan régime (indeed, there is evidence that he fitted quite comfortably into that régime). He warns that Pollio's letters to Cicero should not be read as statements of his political convictions, but rather as mechanisms to safeguard his own interests,Footnote ²⁸ and that we should not reconstruct Pollio's history based on inferences from his career (I, 430–45).

These and many other discussions of a similar high quality can be found throughout the collection. It is to be regretted, therefore, that the collection as a whole is not more unified. I do not mean simply in the matter of formatting or the like, where there is a general, though not uniform, consistency.Footnote ²⁹ I mean in the more substantial matter of the material itself. The entries, except where there are clear indications of cross-referencing in the testimonia or fragments themselves (such as when several historians are cited for a particular matter), are mostly self-contained, and there is very little cross-referencing to issues or discussions elsewhere in the collection. This is partly ameliorated by the fine index of topics (III, 759–72), but it would have been good to have such references (and discussion) in the commentaries themselves.

If I had to give a general characterization of the collection, I would say that it is exceptionally strong on historical matters, and somewhat less strong on literary and historiographical questions. This is not to say that there are no important discussions of the latter, only that in general the work will, I think, satisfy historians more than students of literature and historiography.Footnote ³⁰ For those interested in linguistic and historiographical issues, both Chassignet's and Beck–Walter's collections will continue to be of great use and should be employed alongside FRHist. Of course, the danger in a review of this sort is that with such an enormous text to choose from, any examples run the risk of appearing arbitrary; but I have worked through the whole collection, and I think that I have represented the contents fairly if not fully.

VI

Finally, a word about the work's form. Even the most devoted lover of books cannot fail to note that this collection in its physical form is something of a throwback to an earlier era. In an age when collections of fragments are increasingly found online — with all the conveniences of access and use that such a format brings — the appearance of these volumes solely as three print volumes must raise questions. What can possibly be the justification for the fact that there is no electronic version, no e-book, no online access and, so far as I can tell, not even the promise or preparation of such? An electronic version can more easily accommodate corrections (inevitably, I noticed quite a number of typographical and other errors) and additions. The admirably full bibliographies that precede each author will, in this paper version, quickly become outdated; online they could have been consistently updated.Footnote ³¹ Since the collection is currently searchable on Google Books, an electronic edition exists somewhere. It is to be hoped that the editors and the publisher will address the issue of an electronic form of FRHist immediately. The magnificence of this collection and the scholarship on display in it demand that these volumes reach the widest possible audience.