The high quality of this work will hardly make headlines: the selective bibliography of the book lists 15 contributions by Edward J. Kenney, and one more is forthcoming (p. 21 n. 78: see infra in this review); starting from the earliest, 1958, Kenney has been indisputably the leader in a new aetas Ovidiana. (I have found out that J. is for John thanks to the new Italian (revised) version of The Classical text: Testo e metodo, ed. by A. Lunelli, Rome 1995). He has not produced the complete Oxford text of the Heroides towards which he has contributed so much, but the offer of a complete commentary, with sketchy apparatus, on the double letters (16-21), poems still unjustly neglected in recent work on Augustan elegy, is a wholly satisfactory Ersatz.
The selectivity of the commentary is no surprise either. The success story of the green-and-yellow series owes much to the cleverness of its Latin editor: Kenney has seen long ago the need for a medium-sized series of commentaries, helpful to both students and scholars. Now he practises what he has been preaching to other contributors. Nobody will complain that his work is too short. Yet there is an awkward problem: the introduction is short on the authenticity debate, and one is forced to refer back to Kenney’s 1979 paper on the authenticity of the two ill-attested passages in 16 and 21, and forward (!) to Kenney’s forthcoming treatment of stylistic and metrical anomalies of the six poems as a whole, to be found in J.N. Adams-R.G. Mayer, The languages of Latin poetry, forthcoming. Now the ‘chorizontes’ have every right to see their charges against Ovidian paternity discussed in full. Their objections, as it happens, have been important precisely in fuelling the analysis of epp. 16-21 in terms of ‘Ovidianness’: quality, wit, programmatic depth, reference to models, have been reappraised because the shadow of spuriousness had been cast—and thoughts of spuriousness almost unavoidably encourage negative evaluation, but also set higher standards for the opposite view.
My very subjective feeling is that Kenney is right on the authenticity of the six epistles in their entirety—except of course that it is fully legitimate to look for occasional interpolation. (Along the same lines, one has the right to protest against the use of Amores 2,18 to claim that some of the single letters are spurious, but not against the search for local interpolations, always possible and insidious in the text of an author like Ovid). Yet I am somewhat uneasy because Kenney dismisses the linguistic arguments against Ovidian authorship with a nonchalance which can appear cavalier—were it not for his forthcoming analysis, inaccessible to me as I write. Important points have been scored on both sides, and Kenney’s appeal to the quality of the double letters as a self-evident criterion for authenticity is bound to appear irritating—although it is more understandable, in its undisguised subjectivity, than recurrent, objectivist invocations of Amores 2,18 as a witness to the authenticity or spuriousness of the single letters.
Anyway, the choice of reading the double letters as first class Augustan poetry, and then seeing what happens, is clearly the right thing to do. The process, once started as a reaction to the authenticity dilemma, can be continued with profit. Take as a small test case a harmless line like 19, 135: flavaque Laodice caeloque recepta Celaeno.
Kenney, perhaps limited by the format of the series, gives only a helpful reference to Fasti 4, 173, a parallel list of Neptune’s women, with a parallel ‘footnote effect’ ( ferunt). Rosati’s new commentary, with more space to fill (see infra), scores the point that the sound effect caelo…Celaeno reinforces the learned effect of the catasterism. Next time, someone else will point out that this is a learned catalogue, the heroine is a reader of learned poetry herself (19, 136 et quarum memini nomina lecta mihi), and the hexameter, framed by the Latin for ‘blonde’ and a touch of Greek blackness (Kelaino related to kelainos, ‘dark, black’), looks forward to a similarly jaded reader, who knows about the stylistic standards of Greek catalogue poetry and its preference for listing women a la Don Juan.
Rosati (referred to above) is a fine Italian commentary on the letters 18-19, that is Hero & Leander: G. Rosati, Heroidum Epistulae XVIII-XIX, Florence (printed July 1996, a few weeks after Kenney’s volume). The coincidence shows that new energies are focusing on the double letters: one looks forward to a commentary on Acontius & Cydippe, although P.A.M. Thompson, Ovid, Heroides 20 and 21: a commentary with introduction, diss. Oxford 1989 offers much of value, to judge from Kenney’s references, and there are several recent interpretations (Kenney in Arion 9, 1970, 388-414 and CQ 29, 1979, 394-431; myself in HCSP 97, 1993, 354-65), but especially to a full commentary on Helen & Paris. The differences between Kenney and the much fuller Rosati on 18 and 19 show that book-length treatments of Ovidian diptych poems are not wasted space, although Kenney’s concentration and selectivity remain admirable.
Kenney’s introduction has an important section (pp. 18-20) on 16-21 as a book of poetry, a legitimate Augustan Gedichtbuch. I am glad to have ventured similar suggestions independently ( JRS 85, 1995, 325-27), and will not reiterate the arguments for this approach. It is more important to say a word on how to date this collection. Kenney is on the only possible track when he quotes Louis Purser (p. 25): 16-21 represent a separate volume, written years after the collection of the single letters. Kenney rightly adds that the texts are unfinished—16, with the ill-attested but genuine section, is long and unbalances the book; the final part of 21 has problems which cannot be explained only in terms of transmission. But in what conditions would Ovid circulate an unrevised book? And a book which is not, like most Ovidian works are, advertised by a cross-reference in some other Ovidian publication? Let me come back to the problems of style, meter, and usage normally bracketed under the heading “authenticity”. The double letters feature a series of phenomena which should be set apart (e.g. the only qui = quomodo in the whole (extant) Ovid at 17, 213) and which in my opinion suggest a certain ‘theatrical’ atmosphere: by this I mean that the format of letter-exchange and the teasing realism, the lively dialogue style of some situations, encourage a reference to earlier Roman drama. This can explain a group of anomalies which variously intersect categories like archaic, colloquial, unpoetic, or even ‘Ennian echo’. But the main stretch of anomalies, which always formed the basis of the authenticity dispute, includes material like e.g. the three pentameters ending in polysyllable, the frequent postposition of et (Kenney ad 16, 26), the low frequency of feminine caesura in hexameters (Kenney ad 16, 67). Now all these peculiarities are found in Ovidian exile poetry not in his love elegy. Why should we resist the notion that epp. 16-21 are, so to speak, Epistulae ex Ponto? If we assume that they are late in Ovid’s career, or perhaps drafted in Rome but reworked in exile, the anomalies stop worrying us. The only price to pay would be to assume that Ovid is not very honest when he claims to have given up writing light elegiacs. Not a very high price, with a poet who claimed to have written a Gigantomachy, and to have added a third book to the Ars as an afterthought.
Those who love the idea of a linear evolution in the poet’s mood and style can be interested in the idea that 16-21 have their own dark side, after all (the incineration of Troy, Leander drowning near the Black Sea, Cydippe’s unjust suffering). Those who stick to the traditional idea of the die-hard nequitiae poeta can appreciate the sting in the proemium to Paris’ epistle (“here I am, and I am on fire, I cannot disguise my erotic flame, all too evident now”, plus quam vellem iam meus exstat amor) and the how-to approach to adultery in a Spartan setting, more than a millennium before the Ars and the leges Iuliae: times when a laid-back Trojan can sing veteres…amores to a married lady (16, 257) and the dupe vir can only be a husband, without qualms and ambiguities about the legal status of the Elegiac Woman. Whatever approach one prefers, it is important to try a reading of 16-21 as a product of Ovid’s late career: to quote another Cambridge professor of Latin, “but why should Ovid not have composed Epp. 16-21 in exile?” (M.D. Reeve, CQ 23, 1973, 330 n. 1).
I will append some remarks on problems of detail: of course there is much to add and discuss, but first of all I should probably say more on the pleasures of using this commentary. Here is a handful of samples: “in the epistles the only effective irony is authorial” (p. 96); 17,17 sine crimine vixi“the language of sepulchral eulogy” (without even mentioning in apparatus the variant lusi, which has been so popular. Yet, it is a paradox that, to find out what the paradosis is, I will have to keep consulting Doerrie, a text whose collations need revision, who prints lusi here, and prefers natator to morator at 19, 70 !); 20, 110 ipso nubendi tempore“Lucret. 1, 98 nubendi tempore in ipso, an ironical echo of the sacrifice of Iphigenia, another victim of Artemis”; 21, 31 on te propter as Vergilian echo; 21, 38 proprio vulneror ipsa bono“seemingly an ironic transferral and adaptation to Cydippe’s predicament of Callim. fr. 70 Pf., referring to Acontius”; 21, 215-20 on Cydippe’s symptoms of illness as projections of the love disease which affects Acontius in the Callimachean model: the dynamic of defixiones, I guess; 21, 242 on what Cydippe’s pudor implies. One feels grateful when a commentary is so often able to enhance the literary value of a text.
Kenney agrees with G. Wentzel’s (1890) thesis that Ovid is indebted to the Cycle: Wentzel’s source analysis had persuaded Pasquali ( Storia della tradizione e critica del testo, Florence 1962, II ed., 97-98) that the Paris letter is in its totality the work of a well-read Augustan poet (note, ibid., Pasquali’s even more confident claim, in 1952!, that nobody has any further doubts that Sappho’s Herois is authentic…). Given the importance of the Cypria and the epic tradition (see also S. Rocca, in Ead (ed.), Latina Didaxis X, Genova 1995, 159-67; and note Naev. (?), Cypria Ilias (?), fr. 2 Courtney penetrat penitus thalamoque potitur, presumably Paris vs Helen, a macho-epic version of what Ovid now describes as a sly infiltration) for this letter, I am tempted to claim that the initial emphasis on the protection of Aphrodite and the fateful building of the fleet (vv. 16, 15-38; 105-26) results in constructing Paris as an anti-Aeneas avant de la lettre. Readers of the late Augustan period could have been sensitized to the story of a Trojan hero protected by Venus starting a divinely sanctioned mission: to sail westwards and find himself a new bride ( hoc mihi quae suasit mater Amoris iter…divino monitu…coepto non leve numen adest…Aegeis ire iubemur aquis: iubemur, cf. the Itali’s iubebar, is considered then discarded by Kenney ad 16, 118 and at CQ 29, 1979, 401 with n. 34, but it must be right because of the recurring theme of Venus’ leadership. Cf. Aen. 3,5-9 auguriis agimur divom…pater Anchises dare fatis vela iubebat.). This is a man who can say, as Aeneas did: … dea monstrante viam data fata secutus ( Aen. 1, 382): not a mother goddess, true, but the same goddess, and a mater Amoris, at least. Both cousins start their respective mission with spectacular tree-felling on Phrygian Ida (16, 107 ff. Troica caeduntur Phrygia pineta securi…innumerasque mihi longa dat Ida trabes; Aen. 3,5-6 classemque…molimur montibus Idae; Aen. 9,80 ff. Phrygia formabat in Ida / Aeneas classem…Pinea silva…trabibusque; Ov. fast. 4, 273-74 protinus innumerae caedunt pineta secures / illa, quibus fugiens Phryx pius (i.e. not the naughty cousin) usus erat).
The link with the Cypria could be the surprising testimony that Aphrodite had convinced his son Aeneas to join the expedition (figurative evidence in LIMC I 1 382-83 (F. Canciani); prose summary, Cypria p. 39. 10-11 Bernabé). the Ovidian Paris, to be sure, does not mention this early voyage of Aeneas, nor does Vergil, but one could argue that this undesirable witness gives edge to polemical references to Aeneas as a raider and a new Paris in the Aeneid : he had had the occasion to learn the tricks, after all. So part of Paris’ pride in retelling the expedition would be that he is forestalling, with a tradition of erotic elegy backed by time-honored Greek epic, the unique trip from the Troad of the arch-epic Roman hero Aeneas. Helen, a cunning heroine and ace reader, gets the point right at the start of her letter, 17,5 scilicet idcirco ventosa per aequora vectos…, where the irony of the two initial words clashes with a half-line imported from Virgil, Aen. 6,335: Trojans are always on the move, always with a divine apparatus and a mission to hold on to…
(16, 37-38) For the idea of longing for Helen even before seeing her, Kenney rightly refers to Hesiod, Eoeae fr. 199, 2-3 M-W; in fact the old Greek poem appears to have had a large section on the Suitors of Helen. It is amusing that Paris sees no problem in joining the long list, now that the lady is married, and that the two epistles totally ignore the dangerous Suitors’ Oath. I take this occasion to ask whether somebody is interested in a link between the Eoeae and the Heroides in general. The two poems share an important feature, the angle on famous women and their lovers, and the Ovidian work frequently refers back to a tradition on divine amours as a kind of previous stage, now that the single Heroides deal with the loves of half-gods and heroes, and the double Heroides move on to famous boy-meets-girl stories like Leander or Acontios. Perhaps the influence of Catalogue of Women deserves more attention.
(16, 45-46) Hecuba dreams that she has given birth to an ingentem…flammiferam…facem; the rare double epithet construction is presumably intended to sound Ennian, and in fact Alexander 50-51 (Jocelyn) provides the model for the content, Alcmeo 25 J. the precedent for the use of flammifer (the only attested pre-Ovidian example, again in the context of a vision). Ingentem“should not be emended away”: Kenney defends the force of the epithet as meaning “larger than life” and suitable to such a big omen (the conflagration of Troy). Perhaps it is also relevant that ingens can be perceived by Augustan poets as related to gens (see A. Keith, “Etymological play on ingens in Ovid, Vergil, and Octavia“, AJP 112, 1991, 73-76): the torch in fact portends a huge family crisis for the house of Priam, with its fifty-one bedrooms
(16, 53-56). The commentary should make clearer that the idyllic setting is traditionally linked to Paris’ status as a herdsman, especially in the rich figurative tradition on the Judgment; so when Paris specifies “no sheep no goat no cow” was grazing the locus, this must be part of his strategy to emend away his former status as a Phrygian cowboy (contrast the hyper-bucolic Oenone epistle) and to concentrate on his aristocratic nature. See A. Cucchiarelli,’ “Ma il giudice delle dee non era un pastore?” Reticenze e arte retorica di Paride (Ov. her. 16)’, MD 34, 1995, 135-52 (his argument is stronger if taken in conjunction with the bucolic colouring of the Judgment in the visual arts). Kenney is regularly alert to Paris’ rhetoric and agenda, but not in this connection, perhaps because he has a special theory on the Ovidian chronology of Paris’ early life (cf. his Introduction, 5-6).
(16, 59) K. finds pedum pulsu visa est mihi terra moveri more suitable to a military onslaught than to the start of a divine beauty contest, yet cf. Enn. ann. 1 Musae quae pedibus magnum pulsatis Olympum, inspired by a Hesiodic tradition of theophany. There would be more to say on echoes from Republican poetry, especially drama, in 16 and 17.
(16, 83) Dulce Venus risit : I agree on the reference to the Homeric “laughter-loving” goddess (see especially Il. 3, 424, in context with Paris & Helen), but in the poetics of Ovid it is surely also relevant that dulce ridentem is an Horatian coinage (c. 1,22,23) responding to a tradition of Catullan (51,5) and Sapphic erotic poetry. The perfect style for a goddess unde movetur amor (16, 78). Yet of course the goddess’ smile is mysterious and dangerous, more so in the epic tradition: previous examples of sweet laughter and smile by Aphrodite include situations as disastrous as hymn. Hom. Aphr. 5,49 and Theocr. 1,95. Considering the ever present authorial irony in the Heroides, one might even say that Paris gets the erotic side of the cliché, but the reader is exposed to the less comfortable aspect of Venus’ smile, and both sides are implicated by intertextuality.
(16, 259) The excellent note on Clymene and Aethra transformed from epic amphipoloi into elegiac go-betweens should be repeated at 17, 268: otherwise the ending of 17, “let’s proceed through the intermediary of Clymene and Aethra” sounds a bit weak: the reader should be informed that the two elegiac maids (one of them Theseus’ mother…) are going to resurface, quantum mutatae!, in the Iliadic teichoscopia.
In general, Kenney’s grasp of Helen as an Ovidian and intertextual character is so consistently good that when, for once, he explains her behavior at 17,197-98 with “this is all very natural and feminine” he is probably trying to simplify some more sophisticated argument.
Kenney well discusses the problems of the hypothetical lost Greek model, cf. now Rosati, pp. 11-22; Kenney argues that the recent papyrus findings show the existence of a pre-Ovidian Greek poetic tradition, but probably do not represent a main influence; Rosati observes that even if Musaeus knew Ovid, a possible hypothesis, this does not exclude the presence of a Greek model active on both Ovid and Musaeus. Kenney notes that the new fragments are not likely to be by an influential author like Parthenius. Where are we supposed to look for a model? A poem on Hero and Leander does not emerge as a very influential text in the Alexandrian period, to judge from silence. On the other hand, as both Kenney and Rosati duly acknowledge, the brief allusion to the story in Vergil, Ge. 3, 258-63 already presupposes familiarity with at least one poetic treatment of the story (cf. Thomas ad loc.); and Propertius was also struck by such a treatment (so T.D. Papanghelis, Propertius: a Hellenistic poet on love and death, Cambridge 1987). The story, as Rosati puts it (1996, 12) becomes epistolary because it is a well known text. Otherwise, 18-19 would be the only moment in the Heroides where no intertextual foil enhances Ovidian innovation. Perhaps the picture is difficult to guess because it is complicated. A Greek poem, as Kenney suggests in a footnote (11 n. 40), could have been collecting dust for a century or so before the story became a hot theme in the Triumviral period. The Ovidian poems 18-19 show an interest in various obscure corners of literary history. Leander lists women loved by the gods who became stars, but also steps back from this tradition with a funny move: publica non curat sidera noster amor (18, 150). We think of Alexandrian catalogue poetry: catasterisms, divine amours, pseudo- and post-Hesiodic texts. Hero caps this learned recusatio with a list of Neptune’s amours, names she remembers having read somewhere:
et quarum memini nomina lecta mihi.
Has certe pluresque canunt, Neptune, poetae (19, 136-37)
The two young lovers are a new step in the tradition of love poetry: like the next couple in the collection, Acontius & Cydippe, they represent non-mythological narrative about humans, but they are aware of a previous tradition of mythological love poetry: the Eoeae and Alexandrian catalogues (a tradition of learned but sometimes dry lists of women: nomina lecta…has certe pluresque, indeed). But now, in the age of Augustus, Hero & Leander too have achieved the status of exemplary tales. When was their consecration accomplished? The next double letters are based on Callimachus. Letters 18 and 19 feature allusions to Cicero’s Alexandrian juvenilia (cf. R. Lamacchia, in Poesia latina in frammenti, Genova 1974, 350-56, a stimulating paper). It is also impressive that the two epistles share an allusion to the only fragment by Cornelius Gallus known to us before Qasr Ibrim (fr. 1 Morel and Courtney), uno tellures dividit amne duas, echoed at 19,142 seducit terras haec brevis unda duas, but also at 18, 125f.
ei mihi, cur animis iuncti secernimur undis
una que mens, tellus non habet una duos ?
as noticed long ago by Scevola Mariotti, ‘Intorno a Domizio Marso’, in Miscellanea Rostagni, Turin 1963, 613 n. 88 (the two parallels in Rosati, the first only in Kenney, none in Courtney ad Corn. Gall. fr. 1). The fragment has nothing to do with the Hellespont and the two lovers, although it could have a link with the theme of lovers and separation in Gallan elegy. It is notoriously dangerous to use Gallus as a peg for literary genealogies, but the interest in Liebestod and death by water shared by Vergil in the Georgics and Propertius in books II and III would make sense if the two poets had a recent model in Gallus’Amores, a model which would explain the revival of an older Hellenistic text about Hero & Leander. If Gallus had an interest in Hero & Leander, perhaps as an erotic exemplum, one finds tantalizing that Leander’s non curat…noster amor (18, 150) sounds like Verg. ecl. 10, 28 amor non talia curat, and other themes of separation and longing could have a source in Gallus, before Propertius and Vergil. The theme ‘I will go to the edges of the world for your love’ is probably too commonplace, but the idea of “I will go as far as Colchis and the farthest Pontus” (18, 157) resembles modern speculations on why Gallus mentioned, in fr. 1, the Caucasic river Hypanis-Kuban.
It is enough to keep in mind that the two letters are rich in literary reflexivity. In their notes on publica non curat sidera noster amor both Kenney and Rosati, remarkably, point to Call. epigr. 28 Pf. I would be curious to know their opinion on 18, 133-34,
iam patet attritus solitarum limes aquarum
non aliter multa quam via pressa rota
which looks like one of the closest Latin translations from the Aitia prologue: Leander complains that he is on a beaten track (itself, by the time of Ovid, a trite programmatic topos). Perhaps he is aware of more predecessors than we are able to spot.
A general bonus in Kenney’s interpretation is that Cydippe’s perspective and suffering are valorized, against recurring attempts to read the plot as a bland narrative about happy endings and mutual love. (I also recommend F. Spoth, Ovids Heroides als Elegien, Munich 1992, 161 f.). Kenney also improves on previous Ovidian scholarship with precise remarks based on Callimachean scholarship. Some more guesswork could be useful. Debitus…coniunx (20,8) resonates against οὔνομα κουρίδιον (Call. fr. 67,4 Pf.). In 20,27 the conjecture a se (Nisbet apud Thompson) is deemed unavoidable because arte cannot be defended, yet one wonders about the Callimachean emphasis on τέχνη (67,3). ( Carmina in 21, 235-36 is a problem, but strictly speaking the Callimachean Acontius is a poet—he composed the formula on the quince, so he is labelled magne poeta at 21, 110). The striking enumeration of religious Sehenswürdigkeiten in Delos at 21, 77-100 must be fully vindicated to Callimachean influence: Cydippe, a character from Aitia Book III, is especially attracted by features and memorabilia immortalized by Callimachus (cf. my HCSP paper, quoted above). For example, one should not read Cydippe’s quid me fugis, insula? (21, 83) without glancing at Callimachus’Hymn to Delos, the basic narrative on how Delos stopped floating (has it? 84 laberis in magno numquid ut ante mari?). Kenney rightly connects me fugis with the Vergilian idea of ‘fleeing Italy’ (an idea in turn indebted to the Delian Hymn, as I suggest in CQ 44, 1994, 443). (21, 243) cetera cura tua est supports Housman’s supplement at Call. fr. 75, 40 Pf. (G.B. D’Alessio, Callimaco, vol. II, Milano 1996, 485 n. 74).
Even with this fine commentary at hand, the text of the double epistles remains a difficult and rewarding subject. Of course Kenney’s decisions can suggest further discussion. I remain unconvinced that ave at 17, 114: nec spolium nostri turpe pudoris ave, is Ovidian ( habe MS.); Kenney accepts from Palmer this buzzword of textual conjectures, but he gives us ways to ponder alternative solutions: the apparatus mentions Reeve’s attractive nec…habe, and the note informs that the only Ovidian parallel for aveo is not a good one, Met. 2, 503 recedere aventi. On the other hand, at 17, 102 nec tibi plus cordis sed nimis oris adest it is good to see Riese’s nimis in the text, yet from apparatus and notes one could not guess that the manuscript tradition has minus or magis. As for the issue of interpolations, the search is very welcome, and could be pushed further, but the athetesis of 18, 181-82
velle quid est aliud fugientia prendere poma
spemque suo refugi fluminis ore sequi?
deserves second thoughts. I cannot see big problems on the stylistic level: quid est aliud ( sc. nisi) is, we are told, colloquial and unique in poetry, but Kenney repeatedly warns us against taking hapax as an incriminating feature; and besides that, the full construct, with nisi, is indeed found in the late Ovid; fugientia / refugi suits the rhetoric well; flumen is a nice designation for the receding water. Conceptually, Tantalus is an inexhaustible exemplum for ancient representations of desire (see Rosati ad loc.), and the thirst imagery is weirdly apt for one who is going to drown because of his devouring ‘thirst’ for love.
I must stubbornly insist that this is a difficult text: under a smooth and easy surface, most of the variants make sense and demand insidious choices. One is readily convinced that Kenney’s text of 19, 208, tum placidas tuto pectore finde vias, is without problems. The Latin is so transparent, Hero’s intention so easy to figure out: she wants Leander to wait for good weather then swim safely to Thrace. But it is good to spare a moment for toto in the apparatus: she also wants to see Leander coming her way soon, her letter is torn between passion and the dangers of the sea, and we know that Leander will not let another stormy night to pass by. So toto compensates the unspoken thought that Leander will have to show his enthusiasm after this long delay, and swim safely but passionately…the weather is not his fault, but, as a lover… The choice between the two variants is teasing because it recapitulates Hero’s emotional wavering in the whole letter. Now one feels that toto pectore is right, and tuto simply a fallout of the previous word, placidas. Whatever one prefers (Reeve and Rosati recommend toto) the basic thing is that the double Heroides are a difficult text because they are poetry from the major league of Augustan verse.