BMCR 2001.06.09


, L. Annaeus Seneca Troades : introduction, text, and commentary. Mnemosyne, bibliotheca classica Batava. Supplementum, 212. Leiden/Boston/Köln: Brill, 2001. 573 pages : illustrations ; 25 cm.. ISBN 9004120041 $196.00.

Troades is the most popular of Seneca’s plays, and readers looking for an English commentary have a range of choices. Boyle’s edition with translation (1994) is suited to the needs of undergraduates, providing businesslike help on linguistic points together with stimulating literary comment. Fantham’s editio maior (1982), also with translation, offers a fuller commentary with a substantial “literary introduction” of 122 pages. Now Keulen’s editio maxima provides a commentary on the fullest scale, over two-and-a-half times the length of Fantham’s. In addition, K’s focus is rather different from those of Boyle and Fantham, as will become clear.

K’s edition is the late and happy issue of a longstanding love-affair with this play. He first “felt attracted to these tragedies” in the 1960’s and began work on a Troades commentary, the preliminary result of which was a Dutch school edition (1971). The pressures of a schoolteaching career and the appearance of Fantham’s edition led him to shelve the project, but after retirement c. 1993 he rewrote and enhanced the commentary, presented it for a Groningen Ph.D., and published a Frisian translation separately in 1995. In part because of this history, one gains a stronger than usual sense of the person behind the commentary—still in love with his subject, hospitable to a wide range of opinions, offering his own well-founded views with modesty.1

The Introduction is relatively brief, at 36 pages; it discusses the play’s date, sources, structure, themes, metres, staging and influence. Its longest section, on “the characters of the play,” is instructive on the diverse ways in which the dramatis personae have been read and assessed. This is in keeping with K’s interest in the varied reception of the Senecan plays. Six fascinating pages of the Introduction are devoted to “the play(s) through the ages.” (Here K. justly emphasises the role of the Low Countries is publishing commentaries on the plays, notably those of Farnaby and Gronovius, in the seventeenth century.2) Similarly the Commentary is rich in instances of imitation of Seneca, notably in the neo-Latin tragedies of Grotius and in quotations reflecting critical views of the play at various periods.

The substance of this book lies in its Commentary. It reflects immense diligence on K’s part; over the years he seems to have read most of Latin verse before and after Sen., together with commentaries on that verse, all relevant Senecan scholarship, and much scholarship on the Latin language. The result will be an indispensable resource for future scholars. K. is less interested than Fantham and Boyle in large issues of interpretation; his focus is on the details of literary texture, such as expressive metre, alliteration, intertextuality, word-choice, textual variants. Opening the Commentary at random at Ode 4, I found perceptive comments, some from W. Marx and some from K. himself, on the importance of parallelism in the ode’s structure, and a particularly fine characterization of the shift towards the end of the ode, as comparable to the turn between the octave and the sestet in a sonnet.

Word-choice is a major concern of K. In the comments on 83-85, for example, he discusses the relative frequency in the dramas of fidus against fidelis, maestus against tristis, capilli against coma and crinis, and sheds light on the lexical registers of these words. I should have liked to see some general themes emerge from this mass of material, e.g. the influence on Sen.’s diction of his iambic (as opposed to dactylic) medium, and the relationship of his diction to republican drama. It must be admitted that K. sometimes indulges in hard-core philology with no redeeming relevance to the text (e.g. on tenus 89). I sometimes felt, too, particularly in the choral odes, that more attention might have been paid to the resonance of words and phrases in their poetic contexts. At 71 K. notes that secuit means “cleaved” (the sea), and illustrates this usage amply; is there no interplay between this sense and the cutting of trees for ships (72, cf. Eur. Hec. 631-34, Hel. 229-35) or the cutting of the harvest (76)? Similarly on 72 “the pine sacred to mother Cybebe” K. devotes six lines to the form Cybebe, five lines to her title “mother” and 21 lines to the Cybele cult, but is silent on the suggestion of sacrilege in the felling of the goddess’ trees. Fortunately he redeems himself on 76 by commenting, “the simple picture of the summer-activity [of reaping] might have a grim association for an audience remembering Catul. 64.353-5” (where Achilles slaughtering Trojans is compared to a reaper: add Iliad 11.67-71).

Now to the text. K’s discussions of text problems are detailed, giving the history of the issues and taking account of recent work. His own orientation is conservative, showing a healthy scepticism about alleged grounds for emendation, deletion etc., and a definite tendency to retain the transmitted text (viz. the reading of EA or of E plus one branch of A). He sides with the paradosis against Zwierlein at some 27 places.3 Now conservatism is a strong position from which to start in text criticism, and at 11 of these places I believe K. is right (100 dimissa, 120 iacent, 176f. retained, 197 divisit, 246 placita, 273 brevis, 296 gregis, 338 exsolvit, 586 timens, 922 Paridi, 1080 cacumine); Boyle also concurs at 10 of these 11 places, and Fantham at 8 of them. It seems clear, then, that Zwierlein was somewhat too ready to alter the transmitted text.4 I myself would not go as far as K. in the opposite direction; yet even at places where I could not agree with him, I generally felt that he made a substantial case for retaining the paradosis. In view of his conservatism K. naturally offers no new conjectures (though he did consider alto editus at 1149). He also rarely offers a conclusive new argument, but he does strengthen or extend existing arguments (notably at 100, 296, 586), and he advances the discussion of 197 by adducing the imitatio at Val. Fl. 5.89-95.

A major grumble: the lack of an apparatus criticus is inexcusable in an edition of this scale and nature. Doing text-criticism without an app. crit. is like doing carpentry without a saw. It is not always possible, and certainly not easy, to infer the MSS readings from the commentary. And where K. chooses not to comment on variant readings, e.g. at 352, 1021, the reader has no clue that they exist. A minor grumble: some carelessness has crept into the printed text in the last third of the play. It does not reflect K’s own text choices at 913, 932 and 1031 (he often points out such inconsistencies in Fantham’s edition); an obelus is omitted at 844, the word hic at 811, and the last e of querelae at 802.

In a month’s acquaintance with this edition I have scarcely begun to absorb the wealth of material and thought that it provides. For that reason, and because of space limitations, I shall not offer a more detailed critique; in any case, many of my views on Tro. were stated in a review article on Fantham’s edition at EMC 29 (1985) 435-53. Instead I shall append below two comments on specific points, viz. Ode 2 and the anapaests. Let me simply say that this is a book to which readers including myself are likely to return regularly, and with pleasure, as to a learned and genial friend. * * * * *

1. K’s treatment of Ode 2 is a good example of his interest in reception. He reports various critics’ views as far back as Gronovius, and interweaves comments and arguments in a helpful way. The ode is a meditation on the question whether the human spirit survives post mortem or dissolves at death; it concludes by adopting the latter view. In itself the ode is of a magnificent and sombre beauty, yet critics have made heavy weather of its relationship, or lack of relationship, to three elements in the play: (a) the chorus’ own earlier picture in Ode 1 of Priam’s shade in the Elysian fields; (b) the report in Act 2 of the very lively apparition of Achilles’ ghost; (c) Andromache’s description of the Trojan women, immediately after this ode, as weeping, tearing their hair etc., which does not square with the mood of the ode. K’s own solution is to assign the ode to a secondary chorus of Greek soldiers. This takes care of problems (a) and (c), but it exacerbates (b), and introduces a new puzzle: why should the Greeks, fresh from their victory over Troy and eager to return home, muse about death at all, much less welcome it as a release from sufferings?

Some thoughts on these problems, stimulated by K’s discussion: (a) Choral lyrics in Greek tragedy sometimes respond to an immediate dramatic situation but sometimes are more universal and separate from specific plot events (these functions are formulated, and the tension between them noted, in Taplin’s Greek Tragedy in Action 13, 183.) Much the same is true of Sen., and Ode 1 of Tro. has the first role, Ode 2 the second. It is easy to see how lines 158-60 of Ode 1 develop out of Hecuba’s makarismos of Priam (142-55): they represent a brief, emotional response concerning one particular person, and should not be exaggerated into a settled and generalised eschatology. (b) The orientation of Ode 2 towards death-as-cessation of course reflects the desirability of that view for the Trojan women in their hopeless plight. Yet its thinking is highly generalised: the initial picture of the spouse closing the dead one’s eyes (373) is normative, not related to the women’s situation (since their spouses are dead), much less to Achilles or Polyxena. The ode show no awareness of the apparition of Achilles; it no more comments on the events of Act 2 than Ode 3 comments on Act 3 or Ode 4 on Act 4. K, like other critics, worries about the question whether the chorus was “present” or “absent” during Act 2: on the unreality of this issue see my HF commentary p. 255. (c) For this problem cf. HF 827-29, where the external characterisation of the chorus’ mood does not fit the content of its song (at least for the first 45 lines). A possible hypothesis is that Sen. composed some choral odes separately from the actions, and did not subsequently integrate the two components fully.

2. K. has a useful introductory section (23f.), as does Fantham, on the structure and composition of the iambic trimeter. But his treatment of the anapaests is less satisfactory. The colometry of the anapaests in Tro. and the rest of Senecan drama is now settled for perhaps 75% of lines but remains under debate in the remainder. For the 75% Richter’s criterion of sense-correspondence offers a reliable guide, rightly followed by Zwierlein in the OCT and simultaneously confirmed by the metrical evidence published in my Seneca’s Anapaests (1987). Unfortunately K. follows Zwierlein not only in the 75% but also in the debatable 25%. Any other editor inclined to follow suit should first read pp. 182-202 of Prolegomena, where Zwierlein struggles with unimpressive ad hoc criteria, and acknowledges that at times he relies on nothing more than personal taste.5 This was perhaps the best that could be done in the early 1980’s, but the statistical evidence provided in Seneca’s Anapaests placed (or should have placed) the issue on a different footing. K’s characterisation of the issue as still “subjective” (128) shows that he has simply not engaged that evidence. Prospective editors need to come to grips with it, and either accept its results or attempt a reasoned refutation.


1. The agreeable flavour of K’s work is enhanced—I say this in a friendly way—by some delightful learned modifications of English, e.g. “a graffity,” “variated,” “the rapture of Europa” for her rape, “deducted” for deduced.

2. That tradition is continued by Brill, who have now published three full-scale commentaries on Senecan drama, the other two being Marica Frank’s Phoenissae and Margarethe Billerbeck’s HF.

3. By contrast K. diverges from the paradosis against Zwierlein only once, at 1031-33, where he uncharacteristically and unfortunately revives Leo’s substantial rewriting of the text.

4. Similarly Billerbeck and I retain the paradosis against Zwierlein at HF 113, 299, 436, 456, 490, 505, 625, 660, 703, 866, 928, 1075, 1091, 1135f., 1223, 1285, 1317. In a recent review in this journal of Hine’s edition of Medea, I agreed with all his decisions in favour of EA.

5. Naturally a colometry with such a basis was unstable. At Prolegomena p.196 lines 5-9, Zwierlein listed eight “plausible” monometers, but seven of them had been dropped three years later in the OCT.