Nicholas Trevet (1261-c. 1334) was a high-born English Dominican based in Oxford. In addition to Seneca’s tragedies, he wrote commentaries on Juvenal, Livy, Aristotle, on several books of the Old Testament, and most famously on Boethius; he also composed independent works on astronomy, theology and history.
The editio princeps of Trevet’s commentary on the Senecan tragedies, initiated in 1938 by Ezio Franceschini with Thyestes, is now nearing completion; a group at the University of Bari is working on the two plays still wanting, Oedipus and Phoenissae. We also have useful recent discussions of the commentary by Marchitelli and by Junge, the latter as part of an exemplary edition of Trevet on the Octavia.1
Publication of Trevet’s work takes the interpretive history of the Senecan dramas almost two centuries further back than the first printed editions with commentaries. In fact it takes us back to the very beginning of modern commentary on the plays.2 Hence it is fascinating to find, for example, that the alternative explanations of Phaedra 244 nempe Pirithoi comes given in Mayer’s commentary on the play (1990) are present in essence in Trevet some seven centuries earlier. In my Annaeana Tragica (2004) I traced a particular interpretation of Phaedra 971 (viz. that it summarizes the whole year rather than describing autumn) to a nineteenth-century edition; I now see that it stands already in Trevet (“annus temperet, scilicet calore et frigore, vices suas, id est partes suas”). At Medea 25-26 the unsophisticated and surely wrong reading (see Hine ad loc.), viz. that Medea already intends to kill her children, is there in Trevet, as we might expect. In fact Trevet does frequently get things wrong, and there is much entertainment to be had from the inventiveness with which he ‘explains’ corrupt or puzzling readings.3 But given the limited resources available to him, and the fact that he probably worked from a plain text, it is laudable that he so often gets things right.
It is also fascinating to watch how Trevet is constrained by the epistemology of his age, as all commentators are in their respective ages. He perceives that the phrase flamma iam tectis sonet (Med. 578) could mean either “let fires of sacrifice sound in my house” or “let flames of destruction sound in Creon’s palace.” But he does not raise the possibility — perhaps cannot formulate the possibility — that the phrase means both things. There have been other periods, including much of the twentieth century, in which commentators could not deal with the presence of multiple meaning.4
Much interest has attached to the single manuscript of the tragedies from which Trevet worked, known as τ. Though lost, it can be largely reconstructed from the lemmata embedded in Trevet’s commentary: it was a relatively pure representative of the A family, probably a copy of C. The lemmata contain a substantial number of conjectures,5 some 30 in Hercules alone, and a smaller number of interpolations from E. These were clearly already present in Trevet’s exemplar, and were not inserted by him: every page of his work attests that he saw it as his task to hew some sense out of the text before him, not to improve it.6 Was the manuscript, then, copied in England from C, taken to Italy where conjectures and E readings were inserted into it, then returned to England where Trevet used it? Or could this manuscript have been the source of some of the readings found in Italian MSS? Again, could this MS have been the source of some of the variants inserted in C, rather than vice versa as is usually supposed? Since Chiabò and Roberti do not go into these questions I cannot pursue them here, but it should shortly be feasible to investigate them on the basis of Trevet’s entire commentary, rather than parts of it.
What Chiabò and Roberti provide is a brief introduction, a text of the commentary with the lemmata italicized and the relatively few citations identified, and an apparatus criticus. Since the editors worked collaboratively, their introductions are almost identical, except that each draws examples from her own play: the topics discussed are Trevet’s life and works, the manuscripts, and the nature of Trevet’s commentary.
Chiabò and Roberti make use of just three MSS, Soc. and T representing the α family and V alone representing β. The decision not to use P as a witness to β is probably justified for the practical purpose of constituting Trevet’s text; though it means that singular errors of V cannot be identified by the reader, it does avoid cluttering the app. crit. with the countless singular errors of P. Of the three MSS, I checked only Soc. for the purpose of this review, and that only sporadically.7
In any edition, and particularly in an editio princeps, there are three issues of prime importance: accuracy, editorial judgment, and presentation. So far as my checking goes, Chiabò is sufficiently accurate.8 But Roberti’s text of the Exposicio Medee and of its lemmata is breathtakingly inaccurate.9 The puzzling thing is that many of its errors are not misreadings of the highly abbreviated MSS, which would be understandable, but howlers introduced into the text by Roberti herself: she prints things like rubicundus habent capillos, which a moment’s thought could have prevented or corrected. Even worse, because less easily detected, are the omissions. At 54.19 (on Med. 293) Roberti prints “arguit Medea Creontem, dicit …” I turned to Soc., expecting to find that Trevet wrote dicens not dicit. Instead what I found was this: “arguit Medea Creontem quod in tam modica et rationali peticione se difficilem reddit; unde dicit …” Roberti has simply omitted a whole line. Goodness only knows how many other lacunae are lurking.10 The only possible conclusion is that this edition needs to be redone. Since that is unlikely to happen in the foreseeable future, one must hope that someone will be prepared to collate the MSS against Roberti’s text, and publish a full list of corrections. Until then, Roberti’s inaccuracy will be frustrating for all users of the volume, not least for future editors of Seneca’s tragedies, who can never be certain whether a particular reading printed in the lemmata is indeed that of τ or merely an error by Roberti.
The second criterion is editorial judgment. The text of Trevet’s commentary itself is relatively uncorrupt (except for the usual headaches over proper names), and there are few points of substance where I would dissent from the choices made by these editors.11 In the lemmata the situation is more complex. When the text differs from that of C, is this the result of error or conjecture by τ, or of error by Trevet or his copyists? Often Trevet’s own explanatory comments tell us what he found in τ. So at Pha. 662 his gloss “id est filasset” shows that he read what Seneca wrote, viz. nevisset (rightly restored by Chiabò), though Soc.TV have venisset. Chiabò and Roberti generally make the right inferences from Trevet’s glosses, though not always.12 In the process of rearranging Seneca’s words into prose order and simultaneously interposing his own comments, Trevet quite often omits a Senecan word or phrase, e.g. gazas Med. 485, dote 489; consequently I would not be inclined to reinsert Seneca’s iussit with Roberti 71.18 (Med. 460). Conversely Trevet sometimes clarifies the meaning by repeating a word not repeated by Seneca; hence on Pha. 538-39 I see no reason to delete the second dederant with Chiabò 70.19. In my view, then, the judgment of these editors is middling, but by no means dependable.
Under my third criterion, presentation, there are two issues of importance: identification of the lemmata, and punctuation. The MSS use underlining to distinguish the lemmata from Trevet’s comments; inevitably individual scribes sometimes extend the underlining too far, i.e. into Trevet’s comment, and sometimes not far enough. In a modern edition the only sensible policy is to ignore these vagaries and italicize Seneca’s words. But these editors sometimes italicize too many words, sometimes too few; some of these errors come from the MSS, but others are their own, so far as I can tell from Soc. The result is confusion.
Punctuation is worse. Junge adopted the only sensible policy, which is standard in editions of ancient and medieval texts, viz. to use modern punctuation throughout for the guidance and convenience of readers. But Chiabò and Roberti offer a strange mishmash. First, they place a semi-colon after each unit of exposition, i.e. lemma plus comment. This is both superfluous (since lemmata are already distinguished by being italicized) and confusing (since semi-colons are also employed elsewhere in their modern grammatical function). The editors also scatter modern punctuation about in a way that does not reflect either the sense or the MSS. Take Chiabò 53.25 (Pha. 278): ” iaculatur tela quam certo, id est ualde certo; arcu, plaga data!, id est uulnus illatum ab ipso. Non habet latam frontem…” Here the punctuation obscures the fact that Trevet is taking plaga data rightly as the subject of non habet. Or take Roberti 98.6 (Med. 734): ” scelerum artifex, id est Medea; hec discreta, id est diuisa et distincta ab inuicem; ponit hiis, scilicet medicamentis uel herbis; inest uis rapax ignium…” Here the omission of punctuation after ponit obscures the fact that Trevet understands hiis correctly with the following words, not the preceding ones. This is punctuation fit for the Mad Hatter, not for academics who will want to use these volumes.
While nothing can excuse Roberti’s carelessness, some responsibility must rest on Luigi Piacente, who wrote the preface to both volumes, and on the Dipartimento di Studi Classici e Cristiani at Bari, which published them, for not checking the work adequately before publication. One can only hope that much more care will be taken in the promised editions of the remaining two plays, particularly over accuracy and punctuation, so that they will at least be serviceable and reliable tools for the scholarly community.
1. S. Marchitelli, “Nicholas Trevet und die Renaissance der Seneca-Tragödien,” MH 56 (1999) 36-63, 87-104; R. Junge, Nicholas Trevet und die Octavia Praetexta, Paderborn 1999. Particularly noteworthy among older studies is C.E. Stuart, “The MSS of the interpolated (A) tradition of the tragedies of Seneca,” CQ 6 (1912) 1-20.
2. Albertino Mussato’s commentary was probably written at about the same time as Trevet’s but was never as widely circulated or as influential; it survives only in fragments. For awareness of Trevet’s work in fourteenth- and fifteenth-century commentators on the tragedies, see Marchitelli 87-104. A commentary apparently uninfluenced by Trevet is that of Giovanni Segarelli, probably written before 1400: K. Hafemann, Der Kommentar de Iohannes de Segarellis zu Senecas ‘Hercules Furens’. Erstedition und Analyse (Berlin 2003) 243. Of older interpretive material the only traces left are those that have intruded into the actual text of the tragedies: see ‘glosses’ in the index of my Annaeana Tragica.
3. At Med. 622 he interprets his text auris amisse (a corruption of Aulis amissi) as an allusion to Midas, whose human ears were “lost” when transformed into asinine ones. At Pha. 780 he identifies Liceus (a corruption of licens) as “deus Archadie in cuius templo fiebat abusus masculorum,” and says that the turba consisted of “immundorum puerorum.”
4. On this issue see the index of my Annaeana under ‘multivalent language.’
5. At Pha. 1099 the reading erecto (probably a conjecture rather than an error) has been attributed tralaticiously to τ by many editors, including alas myself. But Chiabò reports the reading as eiecto, the standard A reading, and this is confirmed by my check of Soc. (Is erecto perhaps found in other MSS of Trevet than Soc.TV?) In Annaeana I argue that eiecto (‘thrust out’) is what Seneca wrote.
6. Where his MS offers variants he explains them both rather than choosing between them (Marchitelli 45 fn. 46). The very fact that he cleaves to one MS, rather than comparing it with others, suggests that he eschews textual criticism. Does he adopt the same policy in his Boethius?
7. My thanks to Adrian James of the Society of Antiquaries of London for loan of a microfilm.
8. The only errors I noticed were iuuentum for inuentum at 61.14, quid for quod at 74.21, prius for patris (cf. 77.1) at 77.15.
9. The following are corrections to Roberti’s text, with her error following in parentheses; in each case my check of Soc. confirms that the error is Roberti’s. 28.5 parientibus (-etibus), 28.8 ordinantur (-atur), 31.19 que (qui), 33.8 inducitur (inducit), uirgines (-eas), 39.10 scelera (-re), 41.15 sic (hic), 49.10 rubicundos (-us), 55.23 mutabilibus (-bilis), auream (-rem), 60.8 nauis (-ui), 61.13 quos (quod), 62.7 inolitam (molitam), 62.12 regum (regnum), 63.17 seris (series), 63.17 laxet (latet), 64.10 festinas (-ans), 64.13 talis (-li), 65.8 cessabit (cass-), 66.4 morari (mi-), 68.23 celebrantur (-atur), 68.25 permittes (-ens), 69.2 proposito (-sit), 71.1 transfretandum (tranf-), 71.3 pontici (-cis), 75.1 septimo (-ma), 78.18 potestatem (-um), 80.18 profundam (-fudam), 90.11 interfecit (intef-), 91.25 face (-cie), 93.13 furiosos (-sus), 93.25 tristicie (-ia), 94.16 nunc (tunc), 96.3 illius (-is), 98.1 incantatis (-ti), 107.8 celaret (celerat), 108.7 agitatur (-tus), 109.17 miror (minor), 110.3 unda (unde), 113.16 punientur (-etur).
10. Those I happened to catch are: 52.4 refundere Jasonem, 60.6 nauigacio timorem, 73.19 opibus quas procul querit, 95.4 homo cinctus <serpente; soluat manus pressas, scilicet quibus stringit> serpentem, 100.19 modicum, 111.15 quedam .
11. At Roberti 50.9, “que uolet fortuna — scilicet aduersa michi — premat” (Soc.T) seems more likely to be what Trevet wrote than “que uolet fortuna — scilicet aduersa — imprimat” (V); at 71.12, “feci” (V) is more probable in this first-person context than “fecit” (Soc.T) At Chiabò 41.25, “casto, id est a casto” (TV) is surely correct since Trevet regularly glosses ablatives of separation by adding prepositions.
12. At Pha. 33 Trevet’s comment shows that he read collo forcia (V), not colla forcio (Soc.T) which Chiabò 33.16 prints. At 1063 the gloss “id est sonantes” shows that Trevet read crepitantes, not trepitantes (Chiabò 110.17). Of course the lemmata represent what Trevet read in τ, not necessarily what Seneca actually wrote; hence it is wrong-headed of Chiabò 67.24 to print sata (a conjecture of Heinsius on the Senecan text) at Pha. 494, where EA and Soc.TV have sita.