On first sight, it looks simple. If you wonder what remains of a third-century Roman tragedy by Livius Andronicus called Achilles, you can consult Otto Ribbeck’s Tragicorum Romanorum Fragmenta, widely available in its second edition (1871, rpt. 1972), to find the play’s single surviving line (‘si malos imitabo, tum tu pretium pro noxa dabis’) and, if still curious, to learn from his apparatus that the line survives because in the fourth-century CE the grammarian Nonius Marcellus had it in his stash of useful quotations. Beyond that, a glance at the relevant Loeb volume of E. H. Warmington’s Remains of Old Latin (everyone’s favorite crib) provides a context (‘Achilles refuses to accept the gifts offered by Agamemnon’) and this translation:
If I take woman-cowards for my pattern,
Then you, yes you, will pay the price for wrong.
And at this point things stop looking simple. ‘ Woman -cowards’? Warmington was following Ribbeck’s third edition (1897), where malos was replaced by malas, the reading of Nonius’ MSS, which was also retained by Alfred Klotz in his own volume of Scaenicorum Romanorum Fragmenta (1953). Nor is that all. Closer inspection reveals further uncertainties ( imitabo or imitabis ? dabis or dabit ?), not to mention the fact that ‘Nevius’ and even ‘Lucilius’ have been read for ‘Livius’ and that Warmington’s suggested context has no explicit ancient support: events on Scyros align as well or better with the fragment than the embassy of Iliad 9.1 And these are only some of the problems posed by just a single line quoted by a single source.
Other paths to survival, e.g. when a poet’s words are turned to a new purpose, can create other types of interpretive challenge. So Cicero, in cataloguing the causes of defective argument, cites the case ‘si, cum totum debebit ostendere, de parte aliqua loquatur, hoc modo: mulierum genus avarum est; nam Eriphyla auro viri vitam vendidit’ ( Inv. 1.94). Most editors detect a tragic quotation in Cicero’s words: several Roman tragedies were built on the story of Eriphyla, bribed by Polyneices to persuade her husband Amphiaraus to march on Thebes, and Cicero was fond of spicing expository passages with tragic quotation.2 In reproducing this passage, Ribbeck 2 identified two tragic phrases in uncertain relationship, ‘avarum est mulierum genus’ and ‘auro vendidit vitam viri’ ( Incert. 143-4), while Ribbeck 3, taking up a suggestion by Leo, rewrote the first of these as ‘genus avarumst mulierum’. Warmington, following Cicero more closely, unites the two phrases,
avarum est; … Auro vendidit vitam viri.
Klotz follows Cicero more closely still:
genus avarum est. nam Eriphyla auro vendidit vitam viri.
What, then, did our Incertus actually write? Nobody knows. The one sure thing—assuming Cicero was in fact quoting and not simply alluding to a tragic situation—is that the original lines are impossible to recover with certainty.
How does a modern reader cope with such complexities, especially now that Roman tragedy has begun attracting attention beyond its traditional cadre of eruditi ?3 Nineteenth-century tools can no longer meet the demands of twenty-first century work. Although some subjects acquire a bibliography over time that may over time be consigned with a clear conscience to the scrapheap of superseded and superfluous efforts, this is not one of them. The last century and a half of scholarly labor has produced a substantial body of evidence, conjecture, and discussion with an enduring claim to attention; doing serious work on Roman tragedy requires a firm grip on our predecessors’ legacy. Thus the need for, and the mission of, this new series of Tragicorum Romanorum Fragmenta developed under the guidance of Widu-Wolfgang Ehlers and Bernd Seidensticker, which strives to put between hard covers as complete as possible a record of evidence and analysis.4
So formidable a task has produced a formidable tool. This first volume begins with an immensely useful compilation of material that, though headed ‘Testimonia Tragoediae Romanae’, actually covers the full gamut of play-production in Republican Rome, from Plautus’ metatheatrical asides to Cicero on translating Greek plays to Livy’s notices of ludici scaenici to Cassiodorus’ allusion to the mysterious genre of the ludus talarius. And more. Then come sections arranged by author: Andronicus, Naevius, and a bevy of Minores from Pacuvius’ protégé Pompilius to Lampridius of Bordeaux, who earns a place in the list because Sidonius Apollinaris ( Epist. 8.11 car. 26) said he was inspired by the epos tragoediarum. An extensive and variously problematic set of Adespota follows. For each of these we are presented, as appropriate, with a full collection of testimonia (numbered T 1 etc., following the current convention), a bibliography of editions, and then the fragments (F 1 etc.), introduced by what is known or conjectured about each play. It is a lavish display. Fragments are reproduced in each of their original contexts and then in the versions our editor prefers (with meters identified whenever possible), the whole complex supported by a four-part apparatus: the record of MS variants (CC = Conspectus codicum), the various readings in different editions (CE = Conspectus editionum), readings and conjectures found in the secondary literature (CS = Conspectus studiorum), and a set of useful parallels and echoes in Greek and Latin sources (CP = Conspectus locorum parallelorum). All this is followed by a full array of supplemental material including bibliographies, concordances, and indices. The approach is so exhaustive and so complex that each volumes includes an extensive explanatory introduction, German and English versions provided in tandem, followed by a two-page, coded sample entry. There is surely no more emphatic way to say, ‘User, you really need to read this!’ Which you do. 5
This presentation, designed to suit the varying needs of all four volumes in the series, reflects a firm set of editorial principles and an unwavering commitment to their consequences. Among the most straightforward of these is the decision in printing fragments to capitalize only personal names, thus refusing to commit to syntactic judgments that must ultimately be based only on conjecture, and to make no effort to maintain (or restore) archaic orthography. Fragments not specifically identified by title (or author) in an ancient source or otherwise securely linked to one or the other are relegated to the Incerta (or Adespota), with references provided to conjectural identifications proposed in the secondary literature. Still more significant is the decision to present the fragments of each play in a chronological sequence established by the transmitting sources and not a sequence based on the editor’s conjectural reconstruction of the plot. This decision not only necessitates a new numbering system for each author in TrRF, but demonstrates most clearly the series’ fundamental emphasis on the process of transmission over the work transmitted, which is a less perverse decision than may at first appear since the latter is inevitably contingent on the former.
How do these decisions shape the presentation of our two examples? For Andronicus’ Achilles, it means we get extensive bibliography, the text of Hyginus 96 (Achilles on Scyros), the two citations in Nonius quoting the line and Schauer’s preferred version (all three in fact identical), and then our four Conspectus. Ribbeck’s single verse and six-line apparatus thus grows to two pages and a bit more in TrRF.6 The entry for our Adespotum (F 12) provides the source passage in Cicero, the virtually identical (because derivative) passage in Julius Victor, Schauer’s version (basically that of Warmington), followed by three Conspectus (no CP because no parallels). Once you get the hang of it, the exposition is clear enough, but while its focus on textual minutiae produces meticulously drawn boxes, it provides little incentive to think outside them. So Schauer reports that the word he prints as pretium in Andronicus may appear in the MSS of Nonius as praecium and praetium, but he does not point out that the choice between malas and malos might be relevant to conjectures about the dramatic context. The presentation of Adespota 12 makes it possible to trace with precision the various metrical schemes proposed for these phrases but offers no clue that the fragment itself may be only a phantom.
These priorities reflect in part the difference between an edition and a commentary, but they also reveal a conservative tilt with further consequences. The system wobbles noticeably, for example, when presenting fragments of dubious legitimacy. Thus unwary readers may assume, because TrRF slips a passage of Terentianus Maurus (1931-8 GLK) into the regular sequence of plays between Hermiona (F 15) and Tereus (F 17), that Livius Andronicus wrote a play called Ino, which, however improbably, contained hexameter verses (F 16.5-8). In fact, scholars since the sixteenth century have doubted Terentianus’ attribution to ‘Livius ille vetus Graio cognomine.’ The consensus is put concisely by Courtney (not cited by Schauer): ‘It is inconceivable that these verses could be by Livius Andronicus. We may suppose that Terentianus wrote Laevius and by Graio cognomine meant Melissus…but it is more likely that the source of Terentianus just named Livius and that this was corrupt for Laevius…’7 This is why Ribbeck, Klotz, and Warmington all distinguish this passage from clearly authentic fragments. The truth is to be found in TrRF, too, but buried in the bibliographic note of Schauer’s Conspectus studiorum rather than manifest in the presentation of the sources.
Readers experienced enough to extract that crucial information from the apparatus are clearly the target audience for TrRF, and so it may be unfair—but remains nonetheless inevitable—to wonder whether so meticulous a presentation of all these scholarly thickets is an act of great generosity or arrant pedantry. Much here may well intimidate the neophyte and discourage the faint of heart. It must be said, however, that the editors of TrRF had little choice. Roman tragedy comes to us tightly wrapped in the history of its reception, which itself becomes an important part of tragedy’s story. Cicero, to offer just one simple example, famously declared that the plays of Andronicus were not worth a second reading ( Brut. 71 = Livius T 4), and it is surely no coincidence that barely forty lines of his tragedies survive, not one of them quoted by a literary source or cited for its content. That sort of fact, readily gleaned from Schauer’s Index fontium, is itself of some importance for understanding this problematic genre and is precisely the sort of evidence made most readily accessible by an edition conceived and executed on this scale and with these values.8 When faced with a scholarly record on this order of complexity, there is much to be said for the thoroughness we sometimes call pedantry, and while this volume may on occasion make us gasp or sigh, it will continuously and rightfully earn our respect and our thanks.
1. Warmington’s suggestion follows the lead of O. Ribbeck, Römische Tragödie im Zeitalter der Republik (1875) 25, who had quoted Achilles’ words at Iliad 9.384-7—and still had malos in his text.
2. For the story in drama, cf. Cic. Ver. 2.4.39: ‘Eriphylam accepimus in fabulis ea cupiditate ut, cum vidisset monile, ut opinor, ex auro et gemmis, pulchritudine eius incensa salutem viri proderet.’ Though no Eriphyla is specifically attested, her story figured at least in the Alcmeo plays of Ennius and Accius and in Accius’ Epigoni.
3. For recent work, see G. Manuwald, ‘Römische Tragödien und Praetexten republikanischer Zeit: 1964-2002,’ Lustrum 43 (2001) 11-237, S. M. Goldberg, ‘Reading Roman Tragedy,’ IJCT 13 (2007) 571-84.
4. Four volumes of TrRF, as it styles itself, are planned. The first two, this one by Markus Schauer and Gesine Manuwald’s edition of Ennius, appeared in 2012. Volumes dedicated to Pacuvius and Accius are in preparation by other editors.
5. Though an older generation mutters in my ear ‘Why not write the preface in Latin and have done?’, these editors know the world in which we live. Scholarly explication within the work, however, is in Latin.
6. Just to continue the count: all of Andronicus occupies six pages in Ribbeck 2, eleven in Klotz, twenty-two in Warmington (including translations), and forty-four in TrRF (without translations).
7. E. Courtney, The Fragmentary Latin Poets (Oxford 1993) 129, s.v. Laevius (Melissus?). A second source (Marius Victorinus) is derivative. Further discussion in Ribbeck (above, n. 1) 33-5.
8. The situation is essentially the same for Naevius. Cicero’s passion for tragedy begins with Ennius, as the comparable index in TrRF, vol. II reveals. Other sorts of useful information are doubtless waiting to be gleaned from the corresponding indices of meters and vocabulary.