Reviewing Arthur Palmer's edition of the Heroides,1 Housman commented that 'for the elucidation of Ovid there is not much to be done: both he and his imitators are very straightforward writers, and their words are seldom obscure unless they are corrupt'.2 Times have changed. The last few years have seen a plethora of substantial commentaries on the Heroides: in addition to the two on 12 here under review, we now have that of Alessandro Barchiesi (Florence, 1992) on 1-3, Sergio Casali (Florence, 1995) on 9, Marcus Beck (Paderborn, 1996) and Gianpiero Rosati (Florence, 1996) on 18-19, as well as the admirable smaller commentaries by Peter Knox (Cambridge, 1995) and E.J. Kenney (Cambridge, 1996), on selected single Epistles and the double Epistles respectively. Palmer's commentary on Medea's letter to Jason runs to 14 pages, Heinze's (hereafter H.) is ten times, Bessone's (B.) fifteen times as long. The weight of a century of scholarship has done much to cause, indeed necessitate, this expansion, and two factors may be singled out as contributing to it more than any others. One is the much greater sympathy which Ovid now enjoys with critics, who are willing as never before to search out, evaluate and applaud his skill in manipulating the conventions and content of the literary tradition in which he is writing. The second, much less pleasant, but demanded by the nature of the textual tradition and by perceived anomalies, whether of meter, of diction or of thought, is the Echtheitsfrage. Most modern scholarship on the Heroides, not only on the double letters and the Epistula Sapphus, which have long been controversial, but also on many of the other single letters, is dominated by this problem.
Both B. and H. devote very considerable space to elucidating allusions and to a vigorous and unhesitating defence of Ovidian authorship. Many of the Heroides concern stories otherwise little known to us, and certainly none provides a more fertile ground for the study of literary allusion than does the Medea-letter, given its strong links with Euripides, Apollonius of Rhodes, Vergil, Seneca and Valerius Flaccus, and with Ovid's own handling of the story elsewhere, in Hypsipyle's Epistle, in Met. 7 and in Trist. 3.9, to say nothing of his lost Medea, or that of Ennius or the version of Apollonius' poem by Varro of Atax. The radical change from Palmer's approach is evident right from the start of both commentaries, in the notes on at ... regina (1), as an echo of Verg. Aen. 4.1, which refers to another exotic and abandoned royal lady, herself modelled on Medea, and on memini (2), which B. discusses (cf. also pp. 287f.) with particular effectiveness as a means of signalling a poetic allusion.
In his introduction to Palmer's commentary, L. C. Purser dismissed Lachmann's attempts to establish that 12 and 13 are spurious by remarking that 'it is only on aesthetic grounds that [he] separates them from Ovid's works ... On such questions of taste it is well not to dispute' (pp. xxx f.). Such a dismissal would not do nowadays, especially since Peter Knox's forceful arguments3 have given the Medea-letter the dubious distinction of having its authenticity scrutinised more closely than that of any other of the mythical single epistles. H. discusses this problem at length in his introduction (pp. 51ff.) and in his notes, while B. presents the issues succinctly in a substantial introductory footnote (p. 18, n. 17), and directs us to discussion of specific aspects of it in her commentary.
H. devotes rather more space in his introduction than does B. to passing in review the literary history of the Argonautic saga. He focuses particularly on the tantalising question of the relationship of this Epistle to Ovid's Medea.4 The comparative dating of most of Ovid's works before exile is notoriously problematic. H.'s view that the Medea was written after the first edition of the Amores but before Epist. 6 and 12 is reasonable, given the paucity of our evidence. Paucity of evidence, however, along, of course, with the loss of the Medea, also means that our restricted ability to assess the relation of the poems is little affected by consideration of the relative dating. H.'s sequence depends on the assumption that Ovid wrote the Medea soon after he announced in Am. 3.1 his intention to turn to tragedy on completion of the Amores. Whereas, however, it is likely enough that Ovid's boast in Am. 2.18 of his success as a tragedian was an addition written first for the second edition, and refers to the favorable reception accorded to the Medea (cf. Quint. Inst. 10.1.98, Tac. Dial. 12.6), it is not certain that we should link Am. 3.1 specifically with the Medea. As H. himself very valuably points out (p. 223), Tragedy's injunction to Ovid at Am. 3.1.25 cane facta uirorum does not suit the Medea-story (especially if the play is set, as H. very plausibly maintains, in Corinth, rather than at Colchis [which seems incompatible with frg. 1] or on the Argonauts' return journey), and his speculation that 'die Ähnlichkeit der Formulierung in am. 3.1.60 munus habes, quod te iam petit ista, meum mit [Epist.] 12.197 te peto und 206 hoc ipsum, ingratus quod potes esse, meum est [cf. also 110 munus, in exilio quod licet (v.l. quodlibet) esse, tuli] lässt sich ... vielleicht als Anspielung auf den Medeastoff verstehen' (p. 223, n. 269) will not perhaps command much assent. In Am. 3.1, Ovid speaks as if he intends to devote a substantial period to writing tragedies (cf. esp. 68 tu labor aeternus), and we simply do not know whether it is the case that he turned straight to writing the Medea, and then, for whatever utterly unfathomable reason, abandoned the genre after one successful attempt. If the Medea was indeed his only tragedy, perhaps he simply missed the opportunity to indulge his penchant for lasciuia, the fault for which Quintilian criticises his other poetry, though praising the Medea in such unexpectedly strong terms.
In B.'s introduction, her discussion of 'parole grandi' (pp. 28ff.) is perhaps particularly valuable, highlighting the challenge faced by Ovid in presenting Medea, an archetypal tragic heroine, in the humbler genre of love-elegy. Her acute observations and speculations on what amounts practically to a Kreuzung der Gattungen bring sharply into focus the frustrations which we inevitably feel at the loss of the Medea. For the text, both B. and H. have relied, faute de mieux, primarily on Dörrie's rather uneven edition (Berlin, 1971), though B. has also made a fresh inspection of the primary mss and of a selection of the others. On many points, not surprisingly, the divergences from Dörrie match. It may, however, be worth remarking on the very considerable number of places in which B. and H. differ in punctuation, not only from Dörrie, but also from one another. This type of variation is symptomatic of the desperate need for an authoritative text of the Heroides, by far the most corrupt and controversial of all Ovid's works. A Teubner edition is keenly awaited from J. B. Hall, which, let us hope, will soon fill this need.5 Since Knox has cast such a dark shadow of suspicion over the authenticity of this Epistle , many notes on the text in both commentaries are in reaction to his view that the tradition, even when the text itself is not in dispute, betrays signs of spuriousness. Two of the passages which attract particular suspicion from Knox are 121 compressos utinam Symplegades elisissent and 161f. deseror amissis regno patriaque domoque / coniuge, qui nobis omnia solus erat. Both passages elicit persuasive defence from both B. and H. One might perhaps add that, even if the meter of 121 lacked sufficient elegiac parallels, which it does not, it is much assisted by Catull. 64.15 aequoreae monstrum Nereides admirantes, where a Latin verb-form is found as a spondeiazon, with the Argo(nauts) as its object, immediately following a Greek feminine nominative proper noun of the same prosody. One might also add that Medea was the most celebrated exponent of sigmatism, a feature which enhances the magnificent vehemence of the line so splendidly; cf. Eur. Med. 476f., with the parody at Plato Com. frg. 29.2 Kassel-Austin. On 161f., both B. and H. supply firm parallels and justification for the expression deseror .../ coniuge, with B. making the fine point that it depersonalises the hated Jason: 'l'omissione di a rende più distanziante l'espressione: Giasone é allineato alle altre cose perdute'. It is probable that the truth is generally preserved somewhere in the testimony of the mss, and that an editor's best approach in such an open tradition is to be broadmindedly eclectic. Both commentaries adopt this principle, and are admirably thorough in their discussion of the merits of the various readings and the conjectures of earlier critics. B. never prints a conjecture of her own, H. only once. At 71, B.'s lemma (p. 137) noscis? an exiderunt mecum loca?, anticipated by Merkel, Palmer and others, is an improvement on noscis an exciderunt m. l.?, printed by most editors. 'L'interrogativa diretta disgiuntiva, adatta all'aggressività di Medea' may be supported by the sarcastic agnoscis ...? at Sen. Med. 1020. Heinze tentatively proposes nosti, an exciderunt m. l.?,, an improvement on the too daring nostin', an exciderunt m. l.?, which Heinsius advocated, wrongly supposing that to be the original reading in P. For the elision of a long i in elegy before a short a, however, Platnauer6 cites only Am. 2.17.29 noui aliquam and Fast. 1.147 sumpsi animum, and such an elision would perhaps be particularly remarkable here, before a sense-pause.
Before reading the commentaries in detail, I noted some 60 passages in which I hoped for guidance on a variety of questions, both large and small, on textual, metrical, linguistic, literary historical and literary critical matters. Commentaries are notoriously difficult to review, since one tends to notice most readily points on which one disagrees. I should therefore wish to emphasise that I found help from both B. and H. in a gratifyingly large number of these passages, with honors just about even, though B., with her somewhat fuller commentary, often finds slightly more space for discussion of the particular point and tone. For example, on 10, turbaque Phasiacam Graia bibistis aquam, she notes the very different use of the same phrasing in making a curse on the Argo at Am. 2.11.6 Argo funestas pressa bibisset aquas!; the reference to Jason in 48 as an agricola when he sowed the dragon's teeth might seem inappropriately facetious without her citation (on 18) of the parallel at Stat. Theb. 1.8 (of Cadmus); she is probably correct in seeing, by contrast, 'un effetto giocoso' in the sleepless Medea's description of the serpent in 60 as peruigil. (Occasional points might merit fuller notice: for example, on 2, petere ut ... seems to be largely a prosaic construction; on 11, both commentators note the elegiac color of flaui ...capilli, but it might be worth adding that Ovid is here reversing the norm, in that it is usually the elegiac puella whose blond hair is attractive; the word-order in 20 quam mala multa deserves comment because it is unusual, that in 39f. dura ferorum / insolito premeres uomere colla boum because it is so impressively chiastic, especially in contrast with the prosaic tone of the first hemistich of the hexameter, dicitur interea tibi lex.) It is a testimony to their excellent understanding of Ovidian poetics and their full grasp of the scholarly literature that so often the notes by B. and H. are so closely comparable. At times, this consensus can be eloquent: in particular, it greatly strengthens the case that the difficult phrase mensa ... deserit ... toros in 52 means 'the guests leave the couches' to find that this meaning is not only proposed by H., but also endorsed by B. as the interpretation offered to her independently per litteras by Adrian Hollis. At other times, we are presented with the privilege of choice by the two commentators' differing interpretations: the similarity of 116 sic ego, sed tecum, dilaceranda fui to Am. 3.14.40 tunc ego, sed tecum, mortuus esse uelim is taken by B. as an indication that tecum refers to Jason, not Absyrtus, and provides a foreshadowing of her curse on Jason, that he should perish along with Creusa; H., however, may well be correct to suggest that 'die Stelle ist offensichtlich gezielt zweideutig'. Nevertheless, although I should not wish to detract in any way from the solid worth of the contribution made by both commentaries, it is rather frustrating to find so many couplets on which the notes, while excellent, are so similar.7 Most Classical texts are still without adequate commentaries, and Classical scholars are few. Nowadays, when communication between scholars is so easy, reduplication of effort, at the cost of neglecting other texts, is an unfortunate luxury and one which could be avoided. The writing of commentaries and the editing of texts are precisely defined fields of study, and they are tasks which take time. It should be possible to ensure that long-term research projects are more fully advertised in the small world of Classical scholarship than they are at present.
1. P. Ovidi Nasonis Heroides. With the Greek translation of Planudes. Ed. A. Palmer. Oxford 1898.
2. CR13 (1899) 176 (= Classical Papers, edd. J. Diggle and F.R.D. Goodyear, Cambridge  477).
3. Ovid's Medea and the authenticity of Heroides 12', TAPhA90 (1986) 207-223. Contra, S.E. Hinds, 'Medea in Ovid; Scenes from the Life of an Intertextual Heroine', MD9 (1993) 9-47, H. himself, 'The authenticity of Ovid, Heroides 12 reconsidered', BICS38 (1991-93) 94-97.
4. H. has a useful Appendix on the play, in which he discusses its probable setting and the question of performance, provides testimonia to Ovid's career as a tragedian, and gives a thorough commentary on the two surviving fragments, along with a sceptical notice of a possible third, detected by P. Faider in 'Sénèque De ira 1.1.4 et la Médeée d'Ovide', Musée Belge 27 (1922) 131-3.
5. Apart from anything else, an authoritative text would remove the bothersome lack of consensus in line-numbering. In this review, I adopt the numeration of B. and H., rather than that of Dörrie.
6. Latin Elegiac Verse, Cambridge (1951) 74.
7. The same might be said to some extent about the commentaries on 18 and 19 by Rosati and Beck, but the fact that they argue respectively for and against Ovidian authorship ensures a constantly different perspective.