BMCR 2021.10.35

Fragmenta Saturnia heroica

, Fragmenta Saturnia heroica: introduction, traduction et commentaire des fragments de l'Odyssée latine de Livius Andronicus et de la Guerre punique de Cn. Naevius. Schweizerische Beiträge zur Altertumswissenschaft, Band 47. Basel: Schwabe Verlag, 2020. Pp. 473. ISBN 9783796540349 CHF 76.00.

The Saturnian age is returning. In the last decade, new critical texts, commentaries, and monographs have shed light from various angles on the fragmentary remains of Livius Andronicus’ Latin Odyssey and Cn. Naevius’ carmen belli Punici.[1] This new edition of both poems by Antoine Viredaz, based on the author’s Lausanne dissertation, comes therefore at a timely moment.

Viredaz’s work consists of an introduction to both poems, a critical text of the fragments, French translations, a commentary, and appendices (there are no testimonia). The introduction is an easily consultable series of discrete, individually numbered discussions of such specific phenomena as homoeoteleuton (5.1.2) and the orthography of first-declension ablative plurals ( As the author is a linguist by specialization, the section on La langue de l’épopée saturnienne is by far the best; there is much to equip the reader accustomed to classical orthography with the skepticism necessary to approach dubious “restorations” of archaic spelling.

Questions of literary or historical interest are given somewhat perfunctory treatment in the introductions. In a short section on the afterlife of the carmen belli Punici, for example, Naevius’ influence on Vergil is dismissed as “de faible ampleur” (58): Macrobius’ testimony to the contrary is rejected on the grounds that Macrobius also overstated Vergil’s debt to Apollonius’ Argonautica, an imitation which “se borne en réalité à une vague ressemblance de la situation générale” (59)[2] Some readers may disagree with such conclusions, but they are in any case peripheral to Viredaz’s work.

The stated purpose of the commentary is threefold. First, it supports the critical text by justifying Viredaz’s choices among the MS readings and previous conjectures. Second, it discusses whether a fragment should be attributed to one of the two poems, and if so, attempts to attach it to an Odyssean model (for Livius) or a recognizable context within the carmen belli Punici (for Naevius). Here Viredaz is not aiming for originality: “Je m’efforce de discuter toutes les hypothèses dont j’ai eu connaissance…. Je n’en propose que rarement de nouvelles, même lorsqu’aucune de celles qui ont été formulées ne me paraît satisfaisante” (14). The third aim of the commentary is, where applicable, to discuss the linguistic phenomena that occasioned the fragments’ preservation by ancient grammarians and lexicographers.

Viredaz shows rigorous restraint in limiting his commentary to these purposes. Nevertheless, there are moments when some guidance in the interpretation of the fragments’ meaning for the poems themselves and for Latin literature as a whole might have been welcome. At N61, for example (Serv. auct. Aen. 1.273), we learn that Naevius and Ennius both made Romulus the grandson of Aeneas: Naeuius et Ennius Aeneae ex filia nepotem Romulum conditorem urbis tradunt.Nowhere in the introduction or commentary does Viredaz unpack this fragment’s significant implications: that Naevius’ account of early Rome, therefore, omitted the Alban kings, who in Fabius Pictor bridged the gap between Aeneas and Romulus; and that Ennius sided with Naevius’ version against the latest historical research, possibly creating serious chronological problems for his Annales.[3] An explanatory sentence and a reference or two might have helped to orient the interested reader on this and other questions.

Room for such discussions could have been made by omitting the 17 fragments in Viredaz’s collection that have not been printed in any edition of these texts for over a hundred years, because they indisputably do not come from the poems in question. This superfluous material includes sixteenth-century forgeries (L50, N83–85), fragments of drama (N70–74), hendecasyllables by Laevius (N79), hexameters by Ennius (N75–77), and part of line 1 of the Aeneid (N80). Since Viredaz makes clear that none of this comes from Saturnian epic, one wonders why he took the trouble to include it.

The text of the fragments is mostly the same as Blänsdorf’s (2011). Where Viredaz diverges, it is generally towards greater conservatism, sometimes at the cost of intelligibility. Viredaz offers no original modifications to the text of the fragments, apart from superficial orthographic adjustments. Viredaz has not personally consulted any MSS, but handily lists the URLs of those for which scans are available online.

A significant advantage of Viredaz’s text by comparison with Blänsdorf’s is that Viredaz generally presents more of the citing author’s context than the limits of the Teubner series allow. Similarly, in the case of verbatim fragments, Viredaz’s apparatus extends to the citing author’s context, whereas Blänsdorf’s never ventures beyond the fragment itself; Viredaz, however, omits the apparatus when a citation is not verbatim. The apparatus is carefully constructed and silently corrects several of Blänsdorf’s errors.

Text and apparatus are formatted according to the règles et recommandations prescribed by André (1972), with one notable and unacknowledged departure: where André recommends (where possible) distinguishing corrections by the copyist from those in later hands, Viredaz represents any correction to a MS by the symbols reserved, in André’s system, for one in the copyist’s hand.[4] A note alerting the reader that Viredaz was redefining André’s terms could have helped avoid confusion.

In general, the principle underlying Viredaz’s approach to the texts is extreme diffidence. Sometimes this takes the form of laying at the reader’s feet responsibilities one might have expected from an editor: so, where two variant readings both make good sense and are equally authoritative, Viredaz obelizes rather than decide. He avoids the problems of how to print archaic Latin modernized in transmission by proposing to print the texts not as Livius and Naevius wrote them, but as the citing authorities did (“j’édite le texte de Festus et non celui de Naevius”, 320).

Perhaps most conspicuously, Viredaz refuses to address questions of meter. Not only does he decline to advance an opinion on the nature of the Saturnian or to address the theories so far proposed: there is also no discussion of metrical questions in the commentary, and any previously proposed conjectures are ejected from the apparatus and the discussion if they were motivated by metrical considerations.

…il ne s’agit pas de déterminer si la versification saturnienne obéit à un principe quantitatif, accentuel ou autre; encore moins, de corriger les textes des manuscrits en fonction d’un présupposé de cet ordre. Il me semble qu’une telle entreprise est prématurée, tant que la composition du corpus des saturniens n’est pas établie avec la plus grande rigueur. (14)

One might have hoped that, since the rigorous establishment of the text is precisely what this edition aspires to achieve, its preparation would have enabled the editor to offer new insight on the “Saturnian question”—particularly as this is the first edition of Saturnian epic to appear since Mercado’s (2012) groundbreaking study. But the question will have to wait for the next editor.

Typographically, the most radical consequence of Viredaz’s fundamentalist agnosticism is that he prints all the fragments as prose. He reasons that, if it is impossible to scan a Saturnian, it is also impossible, in the absence of the interlinear marks found on inscriptions, to tell where one starts and ends. In longer fragments, the conventional line-divisions as printed by Blänsdorf are indicated “pour la commodité de la lecture” by superscript numerals within the text. If the aim here was to show that alternate lineations are possible, this might have been better achieved by discussing possible line divisions in the commentary, or at least noting the cases where editors have lineated the fragments differently. As it is, the ultimate effect is to follow Blänsdorf’s line divisions anyway, only less legibly and less accurately—for in two cases Viredaz misrepresents Blänsdorf’s lineation (N1 and N55).

In contrast to Viredaz’s scrupulous caution elsewhere, his placement of the fragments of Livius’ Odyssey is sometimes questionably confident. Many of Viredaz’s Fragments dont l’emplacement est certain are in fact highly controversial and can by no means be certainly placed. Among the Livian fragments of certain position:

  • • L2 (neque enim te oblitus sum, Laertie noster) is interpreted as a translation of 1.65. If so, then Livius recast the rhetorical question and made Jupiter, while addressing Minerva, speak in the vocative to Ulysses, who is not there.[5] Certainly possible; hardly certain.
  • • Viredaz connects L5 (sancta puer Saturni filia regina) with πότνια Ἥρη at 4.513. The Latin, however, is a translation of a phrase from the Iliad, πρέσβα θεὰ θυγάτηρ μεγάλοιο Κρόνοιο (Il. 5.721 = 8.383), and therefore either a literary “contamination” or an Iliadic interpolation in Livius’ text of the Odyssey. Another reference to Hera in the nominative could have served for a context (e.g. Od. 12.72), and it may have come from a passage that was in Livius’ text of the Homeric Odyssey but not in ours, for there is no telling what Livius’ copy of Homer looked like in the wild West of textual multiformity before Aristarchus.
  • • L12: nam diua Monetas filia (filiam) docuit. The MSS’ filiam must be emended to filia to make the fragment resemble Μοῦσ’ ἐδίδαξε (8.481 or 8.488), of which Viredaz regards it as a certain translation. This is satisfactory, but not much more so than emending to filia me and hearing the words of Phemius at 22.347–8 (αὐτοδίδακτος δ᾽ εἰμί, θεὸς δέ μοι ἐν φρεσὶν οἴμας | παντοίας ἐνέφυσεν).
  • • L14 (quando dies adueniet quem profata Morta est) could plausibly correspond to any of five Odyssean passages (2.99–100 = 19.144–5 = 24.134–5, 3.237–8, 10.175).

Others are similarly debatable. It is dangerous to present these possibilities as certainties, since they form the basis of one of Viredaz’s more significant arguments. Against the view that Livius’ translation featured only excerpts from the Odyssey, Viredaz argues it was complete. (He is right to point out that a frequently misread passage of Gellius (NA 18.9.5) does not, in fact, imply that the entire Livian Odyssey fit on one papyrus roll.) One of Viredaz’s arguments for completeness is that “autant qu’on puisse en juger, la plupart des épisodes du poème homérique trouvent une contrepartie dans les fragments de l’Odyssée latine” (35). This argument is based, of course, on the distribution of episodes represented in the Fragments dont l’emplacement est certain, as well as on those episodes only attested in the puzzling hexameter fragments quoted by Priscian (L31–34). We simply do not know who wrote those hexameters or when, and the assumption that anything in this Latin hexameter Odyssey was also in Livius’ Saturnians (35) seems to me unjustified. Given the controversial nature of the material, it might have been preferable to be more sparing with the fragments of certain placement.

Notable omissions from the bibliography include Most (1997), indispensable for editors of fragments; Goldberg (1995); Elliott (2013), an invaluable study of Republican epic and a sobering warning against the dangers of overbold arrangement of fragments; and anything by G. Nagy, who has worked for decades on the “multiformity” of Homeric texts in this period, a question with serious implications for the approach to Livius’ translation.

The Latin text of the fragments themselves is mostly free of typographical errors, but note N14 (inclitus in the text, inclutus in the commentary) and L3 (“ut rusus”, where the r is in the MS reading utrius and so should not have been italicized). The translations are clear and faithful, and offer a significant aid to interpretation.

To conclude, it is obvious on almost every page that great effort and meticulousness have gone into the preparation of this work. It is not always as obvious what sort of audience the editor hopes will benefit most from it. From an authoritative critical edition we might have expected, ideally, personal collation of the MSS, or at least a bolder and surer editorial hand where the text requires it. From an interpretative commentary, greater attention to the poems’ literary qualities or to their cultural and historical contexts could have been wished. The most useful aspects of the edition are the good translations, the learned linguistic discussions, and the commentary’s handy résumé of previous editors’ arguments on textual problems. If Viredaz’s edition does not exactly put Saturnian epic on a new footing, it may be of considerable help to those who will.


J. André (1972). Règles et recommandations pour les éditions critiques (série latine). Paris.

J. Blänsdorf (2011). Fragmenta poetarum Latinorum epicorum et lyricorum praeter Enni Annales et Ciceronis Germanicique Aratea. Berlin–New York.

T.J. Cornell (1986). “The Annals of Quintus Ennius”. JRS 76: 244–250.

W.D.C. De Melo (2014). “The Latin Saturnian revisited: A critical look at Mercado’s Italic Verse, followed by a fresh analysis of the metre”. Kratylos 59: 53–81.

J. Elliott (2013). Ennius and the architecture of the Annales. Cambridge.

D. Feeney (2016). Beyond Greek: The beginnings of Latin literature. Cambridge, MA.

E. Flores (2011a). Cn. Naeui Bellum Poenicum : introduzione, edizione critica e versione italiana. Naples.
id. (2011b). Liui Andronici Odusia : introduzione, edizione critica e versione italiana. Naples.
id. (2014). Commentario a Cn. Naevi Bellum Poenicum. Naples.

S. Goldberg (1995). Epic in Republican Rome. Oxford.

S.M. Manzella and M. Paladini (2014). Livio Andronico : Odissea. Commentario. Naples.

S. Mariotti (1986). Livio Andronico e la traduzione artistica : saggio critico ed edizione dei frammenti dell’Odyssea. 2nd ed. Urbino.

A. Mercado (2012). Italic verse: A study of the poetic remains of old Latin, Faliscan, and Sabellic. Innsbruck.

G. Most, ed. (1997). Collecting fragments = Fragmente sammeln. Göttingen.

D. Nelis (2001). Vergil’s Aeneid and the Argonautica of Apollonius Rhodius. Leeds.


[1] Heartfelt thanks are due to Caitlin Hines and Jarrett Welsh for generously sharing with me their thoughts on some of the questions treated here.
Editions: Blänsdorf (2011), a revision of his strongly criticized Teubner of 1995; Flores (2011a); Flores (2011b). Commentaries: Manzella and Paladini (2014); Flores (2014). Monographs: Feeney (2016); Mercado (2012). See also De Melo’s (2014) critical response to Mercado.

[2] On Apollonius in Vergil see Nelis (2001).

[3] See Cornell (1986: 247).

[4] André (1972: 22).

[5] Cf. Mariotti (1986: 36 n. 54) on other possibilites.