BMCR 2005.01.26

Seneca. IX Tragedies II, Oedipus, Agamemnon, Thyestes, Hercules on Oeta, Octavia, Loeb Classical Library 78

, , Oedipus ; Agamemnon ; Thyestes ; Hercules on Oeta ; Octavia. Loeb classical library ; 78. Cambridge, MA: Harvard University Press, 2004. viii, 654 pages ; 17 cm.. ISBN 0674996100. $21.50.

The second and final volume of the Loeb tragic Seneca comes out two years after the first (favourably reviewed in this journal by R. Scott Smith, BMCR 2003.02.30), and this time simultaneously with some important textual-critical contributions by the same author.1 Specialists are therefore better qualified now to form an opinion about Fitch’s editorial choices, for this new Seneca is in all respects an independent and original edition, with a Latin text thoroughly revised and a translation which presupposes several years of close engagement with these tragedies.

A Loeb will primarily be assessed on the merits of its translation: study of volume two confirms the favourable judgement expressed by the reviewer of the first. Fine and effective, the translation is remarkable for accuracy, clarity, precision of detail, and concision, the last a remarkable feat, in view of the concentrated nature of Seneca’s Latin, inevitably inviting expansive explanatory renderings.2 An extensive index at the end of the volume has helped to reduce the amount of mythological and learned information in the footnotes. The bulk of the footnotes are therefore free to dwell on the clarification of conceits, obscure allusions, double meanings. All plays are preceded by an introduction, informative and insightful, with some indications for further reading. Finally, an attractive feature of this Loeb is the noting of significant borrowings from Seneca in English sixteenth- and seventeenth-century dramas.

All translations entail an interpretation of sorts of the text at hand, and so does this. Long gone are the days of Leo’s censure of the ‘tragoedia rhetorica’,3 ‘where ethos is nothing and pathos is all’, and even the truth of T.S. Eliot’s acute dictum, that all Seneca’s characters ‘speak with the same voice, and at the top of it’,4 has become less and less evident with the growing of a more nuanced and problematic understanding of these tragedies. Fitch’s translation goes a long way towards consigning to oblivion this early common view, offering a Seneca who is plainer, even conversational at times, though no less subtle, pithy and, in the heated dialogic exchanges, clipped and incisive. His Seneca is rather sombre and pensive than hysterical, or paroxysmal.

Most importantly, the translation brings into sharp relief the psychological tensions, revealing the emotion under the casuistry, the conflict and the paradox under the epigram, without ever overinterpreting the original. This renascence of, in the words of Leo, ethos in Seneca is certainly welcome, and has even affected Fitch’s choices as a textual critic, to advantage. In this translation, the protagonists are less truculent than we expect them to be, and more involved and personal in their expression of feeling. Even secondary characters pronounce more than just cues to let their interlocutors come in with the next thundering salvo of pointed conceit.5

In textual matters, Fitch takes issue with previous editors without antagonism, and Zwierlein’s merits are always acknowledged (in the said volume of ‘adversaria’, Annaeana tragica). Fitch has an easy talent for conjecture, and his excellent knowledge of Latin, sustained by an admirable critical sense, has helped him to heal the text in a significant number of places, either by conjecture or by a different choice of reading.

While generally more cautious than Zwierlein in suggesting corruption or introducing conjectures, Fitch is perhaps bolder than the OCT editor in suspecting interpolation. In fact, I am not sure if some at least of the so-called interpolations should not be considered authorial variants, perhaps present interlinearly or marginally, or subsequently added from a different copy. In any case, a reconsideration of the whole question is certainly in order (as suggested in Annaeana tragica, 4 n. 12). In spite of this persistent doubt on the general issue of interpolations, I find Fitch’s excisions generally acceptable (e.g. at Thy. 58, 388-9, HO 461). I have only one different interpretation to suggest for HO 673-4 nec sibi felix pauper habetur / nisi felices cecidisse uidet (‘and the poor man thinks himself unfortunate, unless he sees the fortunate fallen’, Fitch), which I think could be in place in the section on ‘the poor man is exempt from envy’ at 652-7, after a string of negative clauses. Translate: ‘nor does the poor man thinks himself fortunate unless he sees the fortunate fallen’, that is ‘only when he sees the fortunate fallen’.

A Loeb edition is destined to pass through the hands of specialists as well as general readers, students, and scholars in other fields (for example CompLit and English), all of them, if not necessarily liable to take the left hand side text as littera sacra, naturally less inclined than professional Classicists to query the Latin when the small-print bottom-of-the-page apparatus tells a different story. This consideration may cause one to be perplexed when faced with the considerable number of new conjectures Fitch has printed in the text. Nevertheless, perplexity abates at the quality of these new conjectures, all reasonable, some absolutely persuasive. Even the less successful emendations are always good at advancing the discussion on a specific problem.

Among the most convincing of the conjectures proposed by Fitch I shall mention only HO 318 Argea for the transmitted angor of E, and, on my home ground, Oct. 49 secreta repetit for the transmitted secreta refugit, easy to explain as a kind of polar error. Also beautiful are Thy. 58 stuprator for the pointless proper noun Thyestes, HO 472 nil for non, HO 1079 blanda per inferos for inquirens inferos, HO 1595 sonat ecce uastum (supplement), HO 1885 Nomiaeque for Nemeaeque, HO 1951-2 pervius est Acheron … an remeare licet soli tibi,where an replaces et.6

Fitch has rewritten the transmitted colometry of the anapaestic sections, in accordance with the views set out in his studies on the matter, which I regard as important, especially the case for the more extensive deployment of initial monometers.7 In this edition, however, Fitch has wisely decided to maintain Zwierlein’s line-numbers, going back to Gronovius’ edition. Yet, unlike Zwierlein, he has not adopted the raised dot or other conventional sign to mark extra lines caused by his different colometric arrangement. For this reason, in anapaestic sections, more than five lines are sometimes found between the usual five-unit intervals marked by arabic numbers at the right margins, i.e. in Thy. 805-10, 810-15, 830-35, 850-55, 865-70. Unless I have missed something obvious in the preface, the reasons for this are nowhere clearly set out, nor obvious to see for the non-specialist, and problems may arise when quoting passages in these sections.

The section of the book I have read with the keenest expectation is Hercules Oetaeus. This, perhaps the oddest play in the corpus, has not received the attention of modern commentators: it exhibits a very high number of linguistically questionable features, which have given rise to a great deal of discussion (e.g. the use of genus for ‘the human race’ and certain strange participial constructions; on all these problems cf. Zwierlein, Kritischer Kommentar [Mainz, 1986], 313-20). Fitch has thought out the linguistic problems in depth, and I have found a great deal of illumination in his translation and footnotes. I have therefore chosen to append to my review a list of comments and miscellaneous queries on the text of this play.

At HO 56 quanta nunc fregi mala nudus, ‘what evils … have I now crushed unarmed’, the reading of the MSS, nunc, cannot be right. It is true that a Latin perfect can have that resultative meaning, but the sentence is clearly past: Hercules is not nudus, that is ‘unarmed’, generally; he was so on a number of definite occasions set in the past (for example when he strangled the lion of Nemea).

At HO 185-6 me uel Sipyli flebile saxum / fingite, Sipyli (PT) is a very good choice of reading, though in my view it should be considered a conjectural emendation rather than the reading of the archetype. si syphum of E and sisiphi of CS are unlikely to be independent errors in the two branches.

HO 274 peperi. quid haeres?, in a prayer to Juno (‘I have given birth. Why are you at a loss?’ Fitch); but the MS peperi is not very satisfactory. In the parallel adduced by Fitch in support of this reading, Med. 49-50, Medea lists the murders committed in her flight with Jason: if she shed so much blood as a maid, what worse deeds will she do now she is a woman ( haec virgo feci … maiora iam me scelera post partus decent — where of course post partus is doubly significant). But Deianira cannot use motherhood as the final item of a similar climax (what virginal crimes could she boast like Medea’s?).

HO 309-10 coniugis tacitae fidem / mihi reddis iterum? : tacitae (E: sanctae A) is too weak for ‘long-suffering’, the sense required here.

HO 465 F. adopts the reading of A, and his text is better than Zwierlein’s: Quas pontus herbas generat aut quas thessala / sub rupe pindus aut ubi inveniam malum / cui cedat ille?, but I find the initial indicative odd for an interrogative of this kind: better to write perhaps generet ?

HO 821-2 truncus in pontum cadit,/ in saxa uertex: unus ambobus iacet : one expects unus to be opposed to duo, not to ambo : cf. Ov. Met. 8.461-2 pugnat materque sororque, / et diuersa trahunt unum duo nomina pectus. Perhaps editors have been too rash to adopt Grotius’ only apparently attractive antithesis: ‘only one lies in both’. The transmitted funus, also meaning ‘corpse’, should perhaps be retained: Lichas’ corpse lies in both elements, land and water.

HO 1124-7 quis tantum capiet nefas / fati, quis superus locus / pontum Tartara sidera? / regna unus capiet tria? (‘what place in heaven will hold such fate-wrought havoc — what single place will hold three realms, Tartarus, sea, and stars?’, Fitch). I view this apocalyptic vision as an implosion, with the sky, aether, collapsing onto the earth and underworld. But if the sky falls, what is the locus superus left to contain the final indistinct mass?

HO 1272 fletum abstulisti‘have drawn tears from me’, but aufero is not found with the meaning exprimo. I first thought of excitasti but HO 1711 si uoces dolor abstulerit ullas supports the MS reading. Is the meaning ‘to take, steal away’, as if speaking of a prize or a prey?

HO 1560 parcite, o dites, inhibete dextras‘forbear, you magnates, stay your hands’: I have found no parallels for this use of diues. Cf. however, Quint. Decl. min. 345.16.2 scimus tyrannidem praecipue ad diuites pertinere, which may point to a declamatorial origin for the play. TLL V, 1589.9-10 gives an example of diues. translating δυνάστης in Christian Latin (Itala), but that is too late to be relevant.

HO 1678-9 I was at first surprised by the phrase laniare uterum as sign of mourning but cf. Quint. Decl. maiores 10.3 modo super ora pallentis infelices lacerabat oculos, nunc siccata frustra ubera querebatur, nunc superstitem caedebat uterum.

More than a discussion of select passages would be required to do full justice to these 654 pages of Latin, backed by another 277 of ‘adversaria’, but I hope I have succeeded to convey something of the appreciation I feel for this book, and of the stimulus for further research I have found in it. This new Loeb Seneca is a fine, well thought-out and original piece of scholarship, which will advance considerably the debate on and understanding of these tragedies.


1. J. G. Fitch, ‘Textual Notes on Hercules Oetaeus and on Seneca’s Agamemnon and Thyestes, CQ 54.1 (2004), 240-54, and id. Annaeana Tragica. Notes on the text of Seneca’s Tragedies (Leiden, 2004)

2. Only in a handful of passages did I feel that a different interpretation of the Latin was perhaps preferable. HO 415 quisquis alius orbe consaepto iacet does not seem to mean ‘whoever else lies at the world’s edge’ but ‘on dry land’, that is, the inhabited world in general, which is surrounded by the Ocean (other editors believe the passage to be corrupt). HO 1568-69 unde commisso resonare ponto / audies Calpen does not mean ‘(listen from there as Calpe resounds) to its warring seas’, but ‘while sea is joined to sea’, that is the Ocean to the Mediterranean: for this use of committo‘to join two previously separated entities’ cf. Sen. NQ sic momento se redundantia pluribus locis maria committent; Med. 35-6 gemino Corinthos litore opponens moras / cremata flammis maria committat duo. Too colourful or too low-register perhaps is Thy. 283-4 ingesta orbitas in ora patris, ‘childlessness stuffed down the father’s throat’; the alternative proposed in the note is the only one possible in the context, ‘thrust in the father’s face’, (after tota iam ante oculos meos / imago caedis errat‘the whole picture of the carnage hovers before my eyes’, Fitch, with the same insistence on seeing). A note was perhaps needed at 1101-2 THY. natos parenti … ATR. fateor, et quod me iuuat / certos, ‘you gave sons to their father … I admit it — and definitely your own sons, I am delighted to say’, in Fitch’s translation. ‘Definitely your own sons’ is not entirely clear to me: is it opposed to ‘not mine’? Perhaps the translation should be rather something like ‘and what is more, your legitimate sons’. Atreus, obsessed with adultery and fatherhood, turns against Thyestes what he perceives as his brother’s greatest advantage over himself, the fact that no doubts exist about his being the father of his sons. The conversational phatic rendering of nempe with ‘you see’ is slightly undertranslated. The particle seems to me to possess a more strongly adversative force. ‘Surely’ or ‘indeed’ would suit much better the tone of the sarcastic passages where it occurs, as at Oct. 195, which Fitch prints as one sentence said by the nurse, iam metuit eadem nempe praelatam sibi (‘[the very woman who first dared to dishonour your marriage bed] is now fearful, you see, of the woman favoured above her’). Other editors usually give the words after nempe to Octavia, still preferable in my view.

3. De Senecae tragoediis observationes criticae (Berlin, 1878), 148.

4. Seneca in Elizabethan Translation (1927) in Essays on Elizabethan drama (New York, 1960), 4.

5. Sometimes, a mere change in punctuation can produce the effect, as at HO 897, where a question mark in uirum sequeris? (‘You will follow your husband?’, Fitch) may seem a slight change on earlier editors’ plain uirum sequeris. (a phrase often belaboured to extract a rhetorical point), but is enough to make the nurse’s reaction a half-stifled cry of despair, in the service of characterization rather than of rhetorical climax: ‘is that so, are you really determined to die?’ Likewise, a different speaker-attribution at Thy. 308-09 not only makes the progression of argument more plausible but also succeeds in making the attendant a character with stronger individual features, who will finally succeed in making Atreus recede from the idea of involving his sons in the plot.At HO 930-1 DEIANIRA: interim poena est mori, / sed saepe donum. pluribus ueniae fuit, the transmitted version, preferred by Fitch, transforms Deianira into a more positive-thinking, hopeful figure ( ‘sometimes death is a punishment, but often a gift: for many it means forgiveness’, Fitch) than Grotius’ determined suicidal character pluribus uenia obfuit (‘for most, pardon led to disgrace’, my translation). At HO 1340-1, membra complecti ultimum, / o nate, liceat (‘let me embrace your limbs for the last time, my son’, Fitch, printing Bothe’s ultimum), yields a less gruesome text than membra complecti ultima of the MSS, apparently meaning ‘what is left of your limbs’, consumed by the poison.

6. Among the less convincing emendations I should perhaps mention: HO 388 pariter soror / materque multum rapuit ex illo mihi (‘sisterhood and motherhood both stole much of it [sc. beauty] from me’, Fitch), where Fitch writes pariter soror in place of pariter (E; partu A) labat in the MSS. However, Deianira’s grief for the death of her brother Meleager does not seem to be mentioned elsewhere in the play. In addition, failure to understand the allusion would not be enough to explain the change to labat. — HO 1099 quod natum est iterum mori, ‘what is born, dies once more’ (Fitch), but why ‘once more’? ‘Back to the original state’, argues Fitch in CQ, cit., 245 (see note 1), with various parallels, but this meaning is usually evident in the context, for example when two contrary movements are being described. I still find Ackermann’s properat a more satisfactory solution to replace poterit. — HO 1245-6 ubi uires, pater, / in me sepultae?, with Fitch’s pater replacing the MS prius : ‘where is my strength, father, now buried in me?’ — but ‘now buried’ for the past participle sepultae is not appealing. Gronovius’ prius / memet, is still best, in my view, despite the absence of this form in the play. memet is commonly found in the other plays, although admittedly not as a second term of comparison. — Oct. 858 PRAEFECTVS tua temperetur ira in place of tua temperet nos ira : too impudent a remark for the prefect; Nero’s response would be way too sedate if the prefect had really spoken this line. I have tried to defend the MS reading in my edition of the play (Cambridge, 2003, ad l.).

7. Notably Seneca’s Anapaests. Metre, Colometry, Text and Artistry in the Anapaests of Seneca’s Tragedies (Atlanta, 1987), and, with renewed arguments, Annaeana Tragica, 263-77.