Bryn Mawr Classical Review

Bryn Mawr Classical Review 2000.06.06

Rebekka Junge, Nicholas Trevet und die Octavia praetexta. Editio princeps des mittelalterlichen Kommentars und Untersuchungen zum pseudo-senecanischen Drama.   Paderborn:  Schoeningh, 1999.  ISBN 3-506-79064-1.  DM 88.00.  

Reviewed by Rolando Ferri, University of Pavia (
Word count: 3662 words

Nicholas Trevet, a Dominican of Blackfriars, Oxford, started writing acommentary on the ten tragedies of the Senecan corpus at the invitation of Cardinal Nicholas of Prato in 1311. By 1317 at the latest his task was accomplished, and the fruit of his labours had taken the form of a continuous paraphrasis of Seneca's text accompanied by sparse historical and mythographic annotations. As declared in the preface to Thyestes, Trevet's objective was that of making Seneca accessible to a larger reading public, hitherto deterred by the linguistic and mythological difficulties for which Seneca's text is noted.

With the present edition of Trevet's Expositio Octauiae, Junge (henceforth J.) has given us one of the best-edited parts of this commentary (still missing are Phoen. Phaedr. Med. Oed.). J. has a full-scale apparatus criticus, and her textual choices are explained in a further critical commentary on selected problems. The text has been established on a sound basis, with thorough and direct collations of the four most important MSS (Soc. T V P) and with accurate reports of other crucial witnesses. Trevet's commentary has been transmitted independently as well as in the form of scholia surrounding the text of the tragedies. Seneca's text, rewritten in prose word order, is underlined, with greater or lesser accuracy depending on the scribe's attention.

There is a great deal to be learned here about Trevet, about Seneca, and about Medieval Latin in general. J. has produced a number of insightful and convincing emendations (e.g. 30.14 ff. 'suspiria'; 33. 11 'maiorum'; 48. 5 'protulit'), but she has also shown good judgement in the weighing of divergent MS readings. At the same time, she has been careful not to do away with seemingly odd linguistic features, when they could be paralleled with the help of lexica and database-assisted word-search (for Medieval texts she mentions the Cetedoc Library of Christian Latin Texts): cf. e.g. 42. 17 where the rare adjectival use of 'maledicta' is defended on the strength of Aug. Serm. 56 and Hildeg. Bing. 18R. The Kommentar also offers a number of lexicographical contributions to Medieval Latin (e.g. 5.9 retrodating the first occurrence of 'substantiuare'; 13. 35 on the meaning of 'pro tanto').

On consideration, however, Trevet's commentary is of only limited interest to specialists of Senecan drama. The lost MS available to him, t, was a copy of C which is of little stemmatic importance: J. lists divergencies from Zwierlein's OCT at p. 155, as well as Trevet's relation to the recc. (98) and t's position in A (99). By a striking coincidence, 1999 saw the publication of another important addition to the field of Senecan scholarship, namely S. Marchitelli, 'Nicholas Trevet und die Renaissance der Seneca-Tragödien', Museum Helveticum 56 (1999), 36-63, 87-104, which deals both with Trevet in the transmission of Senecan tragedy and with the tradition of Seneca-commentaries up to the first printed editions, accompanied by substantial sets of paraphrastic marginalia in the style and lay-out of MS scholia (cf. the notes appended to his Seneca by Gellius Bernardinus Marmita [Lyon, 1491] and those of Daniel Caietanus of Cremona in the 1493 Venice edition).

A steady flow of conjectures, often successful, seeps into the Italian, lower branches of the Seneca tradition from the mid-1300s onwards, haunting modern critical editions with the tantalizing siglum recc. In this respect, however, Trevet lives in a different age: his Octauia is still full of hardly intelligible readings and metrically impossible lines. Marchitelli (59), who bases her argument on collations of Hf, ascribes to Trevet a number of tacitly made conjectural emendations (to be precise, coincidences with E in right readings): cf. e.g. 408 'sed nunc pereat omnis memoria', a reading in which E and t coincide, against the rest of A with 'pergat', nonsensical in this case. The question is whether the (relatively few) right readings setting t against A were arrived at independently or through contamination with E, and secondly whether it was Trevet who produced them or the scribe of the MS he used. Contamination, even at such a relatively early stage, is probably the answer (cf. R.J. Tarrant, in his edition of Sen. Agam. [Cambridge, 1976], 82-3). The attitude exhibited by Trevet in the Oct. makes one disinclined to attribute to him a great talent in textual criticism: he accepts unquestioningly whatever reading he finds in his MS unquestioningly, despite his declared awareness of its limitations (e.g. the deminutio of which he complains in ad Thy. p. 30 Franceschini), even when the paradosis yields a very dubious sense. In this, he is presumably acting more out of a kind of scruple than because of limited linguistic abilities.

Let us consider for instance Trevet's note ad Oct. 761 'maneat praesens metus', where the sense required is 'may my fears prove unfounded'; modern editors alter either maneat or metus, but Trevet clings to the incomprehensible paradosis: [40.14] 'ut maneat metus presens' which he glosses with a ludicrous 'scilicet hostibus meis'. In fact he is at pains to justify even the most obviously corrupt readings, such as 82 [7. 38-39] 'non uota tua, o nutrix, regunt casus meos, sed uota regunt, scilicet aliorum', and [6. 10-13], infra. Possibly his contaminated MS was in a worse way for Oct., for which E was not available.

Trevet's metrical competence also appears to have been primitive, as he systematically ignores all clues to interpretation disclosed by prosody. Whatever learning was available in his times on the metrics and prosody of classical Latin (both Junge, 162 and Marchitelli 46 cite Servius' De centum metris, Keil IV 456-467 as Trevet's metrical source) was not at Trevet's fingertips. How little conversant he was with this topic, we grasp every time he mistakes ablatives for nominatives or vice versa when the metre shows that it cannot not be so, because an obligatory short is required in an even foot: to the list of examples given by J. at p. 161, n. 340, e.g. 21. 37 'neglecta terra', add 48. 31 'non funesta scilicet numquid non funesta, id est funeri addicta, ante uiolenta manu remigis ... iacuit uictima?'. Examples are easily multiplied. True, he found the right solution, confirmed by E, at Agam. 57, '[o] regnorum magnis fallax ...', where the uneven number of syllables must have put him on his guard. It is however only a drop in the ocean of MS howlers which he swallows at a gulp without even so much as the shadow of a doubt (picking up at random: 49 'quam secreta'; 379 'grauis'; 762 'loquar'; 887 'fortuna leuis').

In the light of these flaws, one cannot agree with J. when she affirms that Trevet's metrics could equal or rival Lovato's (as known from the nota domini Louati iudicis et poetae Patauini, in Vat. Lat. 1769, 246v). Trevet is below even mere syllable count at times, e.g. 379 'grauis' which was emended by someone later in the century (it is correctly grauius in the recc.). One misses a reference to J. Leonhardt, Dimensio syllabarum. Studien zur lateinischen Prosodie- und Verslehre von der Spätantike bis zur frühen Renaissance (Göttingen, 1989), 144, and n. 63-66, listing passages of ancient grammarians relevant to metrics, and commenting on Lovato's own metrical analyses. Also important is G. Billanovich, "Il preumanesimo padovano", in G. Arnaldi-M. Pastore Stocchi (edd.), Storia della cultura veneta. II (Vicenza, 1976), esp. 58.

The main preoccupation of Trevet was to construe Seneca's Latin (on the metalanguage of medieval paraphrases see S. Reynolds, Medieval Reading. Grammar, Rhetoric and the Classical Text (Cambridge, 1996), 110-20), a task made simpler in the Oct. by the anonymous poet's more contained stylistic ambitions in verse writing: his hyperbata are less difficult, and the number of trimeters filled by two noun-adjective pairs is low (36), compared with 114 in Oed. and 95 in Agam. Yet even in this play there are syntactic ambiguities which have often led to textual corruptions, sometimes simply because scribes tended to introduce improvements of their own. One of these cases is, I believe, 797-8 'membra per partes trahunt diducta laqueis', where Trevet either ignored or disregarded the reading deducta, found in recc. and accepted by Leo and Peiper-Richter. Trevet understands, in my view correctly, 'membra diducta, id est separata, trahunt laqueis, scilicet per plateas'; deducta, 'pulled down', goes more naturally with laqueis but yields an unsatisfactory sense after 796-7 'afflicta uulgi manibus et saeuo iacet / euersa ferro'; it must have found its way into the text because somebody was worried by the isolated diducta, which gives an imbalance to the sentence. Trevet is also right in his interpretation of 'per partes' (if J.'s supplement is right, 42. 15), which is directional ('in all directions'), and should be tied to trahunt rather than to diducta: cf. Thy. 60 'membra per partes eant / discerpta' where per partes is an expressionistic detail, almost visualizing the dismemberment of Thyestes' sons as an explosion of disintegrated bodies.

I shall now append a few notes on selected passages where I disagree with J. (the numbers refer to page and line numbers in her edition of the text).

6.10-13 Hoc dicit, quia illius maris insulas subiecit, in quod influit Tanays. Cuius ingressum in mare uocat 'ora Tanays'. Non enim legitur quod triumphauerit regiones Scithie, sed tantum Britanniam et Orchades > J. prefers to write tamen for the unanimous enim of w. The change is palaeographically minimal, but I think that the text must stand. Trevet is making the best of the preposterous ora Tanais [sc. qui ... primus imposuit iugo, 41], which Scaliger corrected to Britannis. 'Tanays' (i.e. the Don) is a river flowing through Scythia, too far away from Britannia to be of much use. What Tanais and Britannia have in common, however, is their position with regard to the (North?) Sea, into which the long river, for Trevet at least, exhausts its course ('illius maris insulas subiecit ...'1). 'enim' is meant to clarify the use of ora Tanays, as opposed to Tanays alone: 'the poet said the estuary, not the river, because we are told that Claudius conquered only Britain, not Scythia', and the North Sea, by the figure synecdoche, may be thought to include Britain as well. The word tantum is also significant: Britain stands to Scythia as ora to Tanays.

14.29-30 (= l. 234) rigens Arctoo frigido, id est polo aquilonari, qui ab ursa arctos [a: artoos b] dicitur eo quod 'arctos' Grece, Latine 'ursa' dicitur > J.'s choice of reading, favouring a, is questionable, because the text as it stands is tautologous and fails to account for the lemma (actually corrupt: the right reading is 'frigore Arctoo rigens'), where the word Arctoo was in fact the one requiring an explanation. Trevet cannot have interpreted Arctoo as an inflected form of arctos.2 Compound nouns and derived forms, especially from Greek, are an object of special interest for Medieval commentators and glossators: cf. Reynolds, cit. 80-86. In Seneca, Greek derivatives in -(i)os are not uncommon, e.g. Eoos at Agam. 483, Hf 25, and Arctous occurs in Hf 1326 'Arctoum ... mare', where Trevet understands it correctly (= 'aquilonare'). Similarly, in Trevet's note ad Hf 6 Ussani, the equivalence posited is ab ursa = arcticus: 'unde polus ille [sc. aquilonaris] ab ursa, que grece artos dicitur, polus arcticus appellatur'. Since the Oct. 234 is virtually identical to Med. 683, it would be useful to have Trevet's note on that passage to settle the question.

15.32-4 Romanum enim imperium ... et deformauit et minuit inusitate luxurie sumptuumque > an explanatory note would have been welcome: the sense suggests that this sequence of genitives inusitate luxurie sumptuumque indicates a cause, but is this possible? Alternatively, this might be an instance of independent genitive of quality, unheard of in Classical Latin but quite normal later (in Trevet cf. 41. 25-6 'et efferata id est mentis ferine effecta', a genitive of quality depending on a verb).

35.12 I am not sure if Trevet, as J. is inclined to believe, really takes sides with anti-Seneca scholars in the question of authorship. His merely adducing Eutropius bk. vii to illustrate Agrippina's prophecy is not a conclusive argument. Like some modern commentators, he fails to tell us whether he considers Agrippina's words to be a prophecy fulfilled or simply written post factum, but he may well have believed the former to be the case. Compare, by way of contrast, the more explicit anti-Seneca stance taken by the author of the marginalia in BM Harl. 2484, who transcribes excerpts of Trevet's commentary, often expressing his disagreement (cf. A.P. MacGregor, 'The MS tradition of Seneca's tragedies: ante renatas in Italia litteras', TAPA 102 (1971), 347 n. 29; Tarrant, Agam. cit., Introd. 39-40). At l. 620-1 'dignum parat letum tyranno', Trevet's cross-reference to Eutropius is transcribed, but an X marks the scribe's own comment: 'describit mortem Neronis futuram et ex hoc tu potes scire quod Seneca non composuit hoc opus quia Nero necauit Senecam ut dicit Boetius de consolatione et alii'.

37.18 o miseranda memor socerum ... diri uiri, scilicet Neronis soceros dicit Claudium patrem suum, qui uere socer fuit, et affines ratione soceri, scilicet Claudii > we receive the information that Claudius was Nero's father-in-law twice within the same sentence. Trevet is not averse to repetition, but perhaps the sense gains something if we read 'Claudios'. With Claudios, the affines notorious for their cruelty are the members of the famous Roman gens, whose haughtiness and ruthlessness Trevet could at least evince from the story of Appius Claudius, who figures as the defiler of Verginia at 17. 35 ff. A similar case of confusion between nominative and accusative is discussed by J. in 28.1-2. Alternatively, as L. Prauscello suggests to me, Claudii could be understood as an instance of unconstrued nominative.

40. 25-27 cuius dum tergo Europa delectata eius pulchritudine insidere uellet, transportabat eam ultra mare et ibi concubuit cum ea > the occurrence of a verb in the imperfect where the context requires a perfect is striking, though examples of anomalous occurrences are documented in Medieval Latin (cf. P. Stotz, HLSMA 4. 320, 58): cf. Arnold. Lub. Greg. 2. 10. 19-22 'forte casu ... nolens ledebat filium'. From a sophisticated Latinist such as Trevet, this sentence is decidedly odd, even if one bears in mind that he was a native speaker of a language not possessing the imperfect.

42. 38-9 ille [sc. amor] ferocem Achillem iussit pulsare lyram quia captus amore Polixene filie Priami noluit cum ceteris pugnare, sed remansit domi modulans carmina amoris in lyra > The Oct. poet has in mind a famous scene described in Hom. Il. 9. 186 ff.: Polixene is still out of the picture, Achilles sings out of boredom rather than to console himself for his lost love, and the theme of these songs is that of brave feats of war. The romanticizing of Achilles' character is a post-Homeric feature, of which we have evidence in Ov. Trist. 4. 15-6 'fertur et abducta Lyrneside tristis Achilles Haemonia curas attenuasse lyra'; Sen. Troad. 320-1 'segnis iacebat belli et armorum immemor, leui canoram uerberans plectro chelyn'. Trevet's note on the carmina amoris must be an extempore inference founded on the words 'amor iussit', since even the mythographical handbooks to which he may have had access know nothing of love songs: cf. Hyginus, Fab. CVI Rose 'quam ob iram Achilles in proelium non prodibat, sed cithara in tabernaculo se exercebat'.

The second part of J.'s monograph is a different book altogether, and one may have reservations about the wisdom of combining it with Trevet. As the author declares in her preface, the Untersuchungen are essentially a survey (accurate and balanced, let me add) of research done on Oct. in the last hundred years or so. Nevertheless, some interesting discoveries do emerge, especially in the chapter on Language, thanks to the extensive and diligent analyses conducted by J. on the vocabulary and style of the anonymous Oct. poet.

J. discusses various topics relating to Oct. and traditionally the focus of scholarly interest, namely the questions of the historical sources, of the play's structure and characters and of the tragedy's language and style, especially in relation to the Senecan corpus.

I have little to add to J.'s discussion of the historical sources. She sides with the mainstream opinion that the anonymous poet pre-dates Tacitus, and that he lived at some time close to the events. In her discussion of divergences from Tacitus, however (193-196), J. fails to mention Poppaea's pregnancy (cf. Oct. 591) as one of the decisive points of difference between the Oct. and Tacitus, as first observed by J.P.V.D. Balsdon [in C.J. Herington, "Octavia Praetexta: A Survey", CQ n.s. 11 (1961), 29 n.3. I have myself added some elements to the discussion in HSCPh 100 (1998), 342-3, n. 7].

J. does not contest the poet's use of written accounts, but argues for the author's first-hand knowledge of events (197). This, as things stand, is in my view impossible to prove. Let us consider by way of example the mention of Acte's monumenta in 193-7 ('uiolare prima quae toros ausa est tuos ... iam metuit eadem ... subiecta et humilis atque monumenta exstruit quibus timorem fassa testatur suum'). These monumenta are not on record in extant historical accounts of this period, and F. Ladek aroused a wave of excitement (in 1891!) when he called attention to CIL XI 1414, CE]RERI SACRVM AVG. LIB. ACTE, possibly the personage in question. It strikes me as high time the excitement accompanying this find subsided. These words are inscribed on a fragment of limestone, 0.52 high, 1.77 long, probably an epistyle, now in the Cimitero Monumentale in Pisa (35A est. [= parete esterna] in Cristiani-Arias-Gabba, Il Camposanto monumentale di Pisa. Le antichitá [Pisa, 1977], 77), thought to have come in medieval times from the region of Olbia, where the rich liberta was the owner of large estates. At any rate the identification of CIL XI 1414 with the monumenta of Oct. 196 presents a number of problems: is Ceres the most appropriate goddess to turn to for assistance in matters of love? Is Acte likely to have sought such publicity, on such a grand scale (if the inscription was part of a temple), for her private dealings with the emperor? More to the point, what prevents us from ascribing the story to a historian rather than to the poet himself? Indeed, the cryptic, allusive character of this remark can equally well be explained either as an innuendo for the court cognoscenti or as a reference to written accounts well-known at the time. Plinius, to mention one possible candidate, is known to have used epigraphic evidence when other sources were deficient or not available (Suet. Cal. 8 'Plinius Secundus ... addit etiam pro argumento aras ibi ostendi inscriptas 'ob Agrippinae puerperium').

By far the most interesting chapter of the Untersuchungen is devoted to the play's vocabulary and style. J.'s diligent word-search, using the Packard PHI CD-Rom #5.3 (1991), and the Poesis CD-Rom (Bologna, 1995), has brought to light a goldmine of loci similes, and I shall acknowledge my debt more fully when my commentary on Oct. is finally published. J.'s conclusions are always sound: the play's author was a fairly unsophisticated rhetorician; the frequency, greater than in Seneca, of high-flown, typically poetic expressions is a feeble attempt to compensate for the author's comparative weakness in the composition of more elaborate syntactical units; the poet's vocabulary is firmly rooted in post-Ovidian Latin, but there are no clear-cut linguistic elements allowing us to date the play (239). Following in the footsteps of O. Zwierlein (Gnomon 64, 1992, 502-6), J. has a good section against Billerbeck's contention that the play's vocabulary reveals Tacitus' influence (260-264) and she must be right to interpret obsequium, at 82, as a reminiscence of Ov. Ars 2. 179 (248-9). The recurrence of expressions like post fata (230) and attonitus (224) is seen against the background of Silver Latin poetry, where, most notably in the work of Statius and Silius, many such iuncturae popularized by the Augustan poets undergo a devaluation process: this reveals important insights into the poet's literary and cultural referents.

Against the reasons possibly supporting a much later date for the play J. makes no conclusive remarks, even if I say this partly to agitate the spectre of pyrrhonian doubt: indeed, the events of 62 would fail to have an impact on an audience chronologically distant from the events. J. rounds off her monograph with a chapter on imitation, where she (281, n. 904) dismisses Helm's argument that there is a link between the Siluae and the praetexta. I hope to prove in a forthcoming paper that there is in fact a much stronger case to be made. The Octauia reveals an acquaintance with Statius' encomiastic poetry, the Siluae, and, more specifically, with the epithalamium of Stella and Violentilla (1. 2). This, if accepted, is important in several respects and gives us a terminus post quem for the play, since the marriage can be assigned with a certain precision to 89-90, by means of Mart. 6. 21 which celebrates the same event.

Just a few points of detail:

234-5, on 435 'sed ecce gressu fertur attonito Nero' J. says 'vergleichbar in der lateinischen Literatur ist nur Med. 675f. 'ut attonito gradu euasit''. Not true: cf. Val. Fl. Arg. 8. 21 'attonito qualis pede prosilit Ino'.

259, n. 785: J. states that the nexus anim-(am) trist-(em) in Oct. 375-6 'animam tandem per fera tristem uulnera reddit' has no parallels in extant Latin literature, meaning that 'anima-/trist-' in whatever sequence is not attested in Classical Latin. The intrinsic interest of this find is hard to assess, and perhaps it was better to couple it with the information that anima ... tristis, in this context, is an adaptation of such classic death-scenes as Verg. Aen. 12. 952 (= 11. 831) 'uitaque cum gemitu fugit indignata sub umbras' (imitating in turn Homer, Il. 16. 857 [= 22. 363] ὃν πότμον γοόωσα, wherein the soul escapes from the body with the last sigh of the dying).


1.  . Elsewhere Trevet is of a different opinion about the location of Tanais. In ad HF 1323 (174.21 Ussani), for instance, he copies out Isid. Etym. xiii.21.24, where this river is said to divide Europe from Asia and to flow into the Black Sea ('... in Pontum fluens'). My colleague Lucia Prauscello rightly observes that Isidorus' entry resurfaces again in ad Troad. 8 (4. 29-30 Palma) 'Tanais ... fluit in pontum', which she suggests correcting into Pontum.
2.   A rare word, but known to medieval writers from Lucan and Statius: cf. MLW 904-5 Paul. Diac. Lang. 1. 1. 'arctoo sub axe' and, as a noun, Anon. Astrolab. p. 371, 31 'inter summos arctoum ambitus'; also in Dictionary of Medieval Latin from British Sources I.121.

Read Latest
Index for 2000
Change Greek Display
Books Available for Review

HTML generated at 13:26:50, Friday, 03 April 2009