Probably most scholars interested in Ovid's masterpiece already know the characteristics of this Fünfmännerkommentar (thus labelled by Barchiesi himself in the first volume, XII; one is reminded of the Sechsmännerkommentar of the Odyssey published in the same series, which is actually a five-men and one-woman commentary): the Latin text is that of Richard Tarrant's OCT (2004), with few deviations (collected in a short "nota al testo" XXXVI), mostly restorations of lines expelled by Tarrant, and an abridged apparatus taken from the same edition.1 The notes generally provide both laymen and scholars with what they need to understand the mythology, narratology, intertextuality and metapoetry of Ovid's work (for an exception see below the note on 3,721-722). This part seems to me to be successful and often excellent; as it is based on wide bibliographical knowledge and less positivistic, more holistic, sometimes typically postmodern literary criticism, it can be considered as a useful and even indispensable update of the late Franz Bömer's monumental commentary (1969-1986), which however will remain unsuperseded as a repository of all kinds of information, especially those which are insufficiently supplied by Barchiesi and Rosati's commentary. The latter concentrates on higher criticism; it is much less comprehensive and painstaking as far as Latinity is concerned. Bömer's performance in this matter and especially in textual criticism (even though competent verbal scholarship implies excellent mastery of the Latin language) is in my view far from being altogether satisfactory, so that one cannot think that Bömer provides what Barchiesi (books I-III) and Rosati (book IV) do not. Of course no judgement on the whole five-person commentary is possible before we have it all: I guess E.J. Kenney's commentary on books VII-IX will be more satisfactory as far as "lower criticism" is concerned, and I hope that the following volumes can and will be improved in this matter. The example of the six-person commentary on the Odyssey in the same series shows that it is possible to combine lower and higher criticism fairly.
Before entering into the details of Barchiesi and Rosati's handling of lower criticism, I should like to stress again the importance of both scholars' achievement in higher criticism. Each of them has his own style. Barchiesi's commentary is often more intellectually ambitious and has more original, challenging and sometimes farfetched ideas; it is always understandable, but it seems to me to be occasionally "overwritten". See for example the lengthy note on 3,44-45, which contains the following characteristic lines: "Ovidio rovescia la poetica didascalica (il meraviglioso che affascina a distanza, il poeta che "riscalda" la fredezza della scienza astronomica) in una poetica del pericolo terribile e dello spettacolo che schiaccia lo spettatore". When one looks back at the Ovidian lines which inspired this formidable rhetoric, one may think it is much ado about little or nothing. The justification of the choice of the reading satiatae against satiati in 3,140 (about Actaeon's dogs) is a nice piece of ingenious literary criticism. However I do not feel certain that the feminine (the lectio difficilior, if it is not a learned emendation) is right since the following catalogue includes males, a fact which is noticed but perhaps undervalued by Barchiesi. By the way, he surprisingly ignores the relevance of the contrast between crimen and error (3,141-142) in Actaeon's story to Ovid's own case: see e. g. Hugo Magnus in the preface of his indispensable 1914 Weidmann edition, III n. 2. Barchiesi is unsurprisingly at his best when dealing with Narcissus' story, though the comment on 3,348 si se non nouerit doesn't convince me that it was useless to quote and/or discuss Bentley's cj. uiderit. Rosati's less challenging notes show greater coolness and plainness, which doesn't mean that they lack sensitivity. Both scholars' notes display deep and subtle knowledge of Ovid's literary ways. The reader will be grateful for the learned and clever introduction provided whenever a new story is told.
The two scholars' "philological" notes seem to me to be less painstaking, even when no textual problem is at stake. Thus Barchiesi (note on 3,456, et amarunt me quoque nymphae) thinks that it is abnormal for quoque to modify a word which doesn't precede it, and his sophisticated and paradoxical interpretation of the line is based on this assumption. But a careful reader of Latin poetry may notice such a free use of quoque and et = etiam (this is also true of the corresponding word in Greek poetry): see Housman on Lucan 9,463, his addenda to Manilius 1,780, and my forthcoming critical commentary on Statius' Silvae at 5,1,246.2 Addressing his own fleeing image, with which he has fallen in love, Narcissus can plausibly (in my view) observe that even the nymphs have fallen in love with him, not that they have fallen in love even with him (or also with him). I fail to see why Barchiesi's interpretation of quoque renders et unproblematic in the same line. He holds that in 3,530 uulgusque proceresque the first -que is "misurato longo davanti a muta cum liquida, una licenza che si rifà a precedenti alessandrini e virgiliani". Line 4,10, telasque calathosque, but Rosati's note on this passage will show that this is a typical metrical lengthening in arsi (1,193; 5,484; 7,225 and 265; 11,36). Barchiesi's notion (note on 3,690) that the prosody of the long first syllable of Dia "è sicuramente grecizzante" is strange: it is the normal prosody of the word in Greek and Latin; the "non grecizzante" prosody doesn't exist and is unthinkable. In 4,20-21, Oriens tibi uictus adusque | decolor extremo qua tingitur India Gange, Rosati prefers the v. l. tingitur rather than cingitur, rightly in my view, but I wonder how he can argue that "l'idea del Gange che "circonda" l'India è in qualche modo già implicita in extremo".
But I query what Barchiesi and Rosati's notes say less than what they don't. Faced with 4,54, talibus orsa modis lana sua fila sequente), the reader is left with the translator's remote rendering "scorrendo il suo docile filo di lana" and Rosati's note, which dicusses the textile/textual metaphor without saying how he literally understands the difficult phrase. The reader busy with the Latin will too often experience such helplessness either in cases where the transmitted text may be correct but is difficult to understand or in cases where a textual corruption may lurk. The reader may well wonder how ubi conditus antro (3,31), where ubi refers to the same thing as antrum, can mean "nascosto in quell'antro", which translates hoc conditus antro or quo conditus antro. Neither Tarrant's apparatus nor Barchiesi's note help the reader, who will find help in Magnus' 1914 edition, which reports the readings quo and hoc. The same experience awaits the reader when he tackles 3,598-599, applicor et dextris adducor litora remis | doque leues saltus udaeque immittor harenae: the translator's rendering "mi trovo sull'umida sabbia" seems to betray some meritorious uneasiness with the text she had to translate. But how can immittor harenae mean "mi trovo sulla sabbia"? I immediately thought of innitor, a reading which I also found afterwards in Magnus' edition. But if the reader goes as far as to wonder how in 3,77-78, ipse (sc. anguis) modo immensum spiris facientibus orbem | cingitur, interdum longa trabe rectior adstat, the word cingitur can mean "s'attorce", as the Italian translation has it, he will then find no help even in Magnus' edition: some sort of help can be found in Revue de Philologie 78, 2004, 63, where I suggested that flectitur might give the required meaning. I do not mention this case in order to boast but only with a view to indicating how much exegetical and critical work must still be done on Ovid's masterpiece. Indeed, Tarrant's edition is an improvement on any modern preceding text (not apparatus), but if any of the five men working on the new Mondadori edition and commentary thought that this improvement was great enough for him to dispense with careful consideration of the textual issues and to content himself with as few deviations from Tarrant's text as Barchiesi and Rosati have allowed themselves in books I-IV, he may have been wrong.3 My query will not I hope be considered to be mere pedantry. Or should one be happy with a critical edition in which the translation too often renders another text than the one it faces and in which the commentary is silent or vague? I think a commentary where there is room for a note on the wavering prosody of Sidonius (3,1294) or the orthography tinguit versus tingit (4,127) should say something about the above-mentioned issues and at least some of the issues I deal with at the end of this review.
On textual issues, both Barchiesi and Rosati may be questioned. At 4.243-4 Rosati disagrees with the conjecture of one of the foremost 20th century critics of Latin literature, the late William Watt, nec tu iam poteras enectum pondere terrae | tollere, nympha caput corpusque exsangue iacebas. Watt (Classica et Mediaeualia 50, 1999, 168) observed that enectum, from a verb exceedingly rare in high poetry, is, as an epithet of caput, both odd in itself and an unpleasant anticipation of corpus exsangue iacebat. He suggested the excellent oneratum, "which differs from enectum by little more than one additional letter; for pondere onerare see TLL 9.2.632.19ff". Now, according to Rosati, Watt "trova "impoetico" il verbo, ma vorebbe sostituirlo con l'ancor più prosastico oneratum". But Watt's essential query was the oddity of enectum in its context, and it cannot be objected to oneratum that it is "even more prosaic" than enectum: contrary to the latter, the former gives excellent sense in its context and onero is so far from being prosaic that Ovid uses it more than a dozen times. I am confident that such a note is due to haste and doesn't reflect its author's capacity in the relevant field. Here and elsewhere, Barchiesi and Rosati's Latin text might do well to agree less often with Tarrant's and more often with the intuitive Italian translation. Then the Latin (4,416) facing "per tutta Tebe si celebra Bacco" might be uenerabile, which now stands in the apparatus, and not memorabile. Word repetitions could be also dealt with more thoroughly and more acutely, e. g. in the note on 3,55-58.
For the time being, I must content myself with praising the two men's accomplishments in the higher criticism of books III-IV and with offering the following selection of supplementary remarks mostly on book III, in which I abstain from commending readings which already stand in their abridged version of Tarrant's apparatus. My remarks aim at proving and illustrating my contention that more attention has to be paid to "micro-exegesis" and textual problems. They are also a kind of tribute to the pleasant Italian translation of Ludovica Koch, who may more than once have guessed the truth lurking under the suspect Latin text and may then be said to have successfully mistranslated it.
-- 3,33 corpus tumet omne uenenis. I suspect corpus (anticipation of l. 40?) has taken the place of guttur. guttura and corpora are alternative readings at 15,464. In 4,352-353, ille cauis uelox applauso corpore (v. l. pectore) palmis | desilit in latices, I conversely suggest gurgite.
-- 3,73 causa recens. From l. 260? Here the apposite word would be plaga, a reading reported by Magnus 1914.
-- 3,99 ille diu pauidus. Diu is otiose. I suggest metu, the word ille having absorbed the first syllable of metu. Compare 4,228 pauet illa metuque; 6,706 pauidamque metu; 9,249 metu paueant.
-- 3,169-170 sparsos per colla capillos | colligit in nodum, quamuis erat ipsa solutis. Though Barchiesi does not see anything wrong here, the second hemistich of line 170 seems strange as far as meaning is concerned and its Latinity is at least awkward. I see in Magnus 1914 that the 16th century scholar Jean Crespin suppressed the hemistich. quos pexerat ipsa solutos would give plausible sense. It is perhaps impossible to find a certain solution; cruces seem to me to be needed here.
-- 3,222 hirsutaque corpore Lachne: I suggest tergore.
-- 3,241 circumfert tacitos tamquam sua bracchia uultus. There is no note either in the apparatus or in Barchiesi's commentary, though the comparison, as it stands in the transmitted text, is puzzling and the preceding lines show that Actaeon is by no means silent. I now suggest circumfert, tendat tamquam sua bracchia, uultus.
-- 3,271-272 fallat eam faxo, nec sum Saturnia si non | ab Ioue mersa suo Stygias penetrabit in undas. The translator must have felt uneasy with fallat, which she inaccurately renders "saprò rivoltarglielo contro". perdat would do (confusion of P/F, anticipation of faxo?). I query mersa and suggest tosta, which perfectly suits Semele's fate, and provides an antithesis with undas, a word which I think is bizarrely anticipated by mersa.
-- 3,314-315 inde datum nymphae Nyseides antris | occuluere suis lactisque alimenta dedere. Wasn't Ovid too resourceful a poet not to avoid the repetition datum/dedere when it was so easy? I would consider tulere. It might be a mixture of "Perseverationsfehler" and well-attested confusion between tuli and dedi.
-- 3,378. I remark that the translator's "raccogliere" renders my cj. exceptare (Revue de Philologie 78, 2004, 64), not the printed reading exspectare.
-- 3,388 et uerbis fauet ipsa suis. The translator's loose rendering "esaltata lei stessa da quello che ha detto" suggests to me calet.
-- 3,403 luserat hic nymphas, sic coetus ante uiriles. I have little doubt that coeptus, a reading mentioned by Magnus 1914, is right. The translator renders "gli approcci dei maschi", which suggests rather coeptus than coetus. -- 3,538 uosne, senes, mirer, qui longa per aequora uecti | hac Tyron, hac profugos posuistis sede Penates, | nunc sinitis sine Marte capi?. Can this mean "E potrei non stupirmi di voi"? If it cannot, then mirer is problematic. Lactantius Placidus on Statius Theb. 7,164-165 has miseri (from 7,4?). Did mirer, a stop-gap, take the place of such a word as clari, "you glorious old men? Then, with uosne, acrior aetas, | o iuuenes etc., one has to supply sinitis sine Marte capi.
-- 3,601-602 exsurgo laticesque inferre recentes | admoneo. The right word is adferre, a reading reported by Magnus 1914. Persequor still stands in the text at 4,151 and 551 where Latin seems to me to require prosequor. Confusion of prefixes leads to corruption.
-- 3,618-619 qui requiemque modumque | uoce dabat remis. Here also the translator may have found something wrong, since she seems to render requiem with "tempi". Does requiem come from 10,377, nec modus aut requies, nisi mors, reperitur amoris, where it is unquestionably apposite? J. H. Voss's regimen (twice in Ovid, only in this poem) should I think be mentioned.
-- 3,626-627 is mihi, dum resto, iuuenali guttura pugno / raptat et excussum misisset in aequora. raptat is Heinsius' cj. for rupit, which presumably comes from 15,464-465 uituli qui guttura ferro | rumpit. Such a phrase as pectora/tempora... pulsat would more plausibly describe the blow which made the speaker lose consciousness and might have thrown him into the sea.
-- 3,646 totumque immurmurat agmen. The context and the translation "mi mormora contro" suggest obmurmurat, a verb found at Her. 18,47.
-- 3,665 grauidis distinguunt uela corymbis. This miracle is expected to have a result parallel to that expressed by impediunt l. 664. This would require a verb formed with -stringunt, perhaps neither the variant distringunt nor Heinsius' destringunt but constringunt.
-- 3,676 at Libys, obstantes dum uult obuertere remos, | in spatium resilire manus breue uidit. The oars were impediti (see l. 664) and that may be what obstantes should mean. The translator's rendering "dar volta ai remi incagliati" suggests she may have felt the presumed difficulty. Furthermore obuertere seems to point to a movement which differs from the one implied by the transmitted verb (cf. 11,475 obuertit lateri pendentes nauita remos), so I suggest adstrictos... conuertere, "shake the entangled oars" (note that Cicero Rep. 1,56 uses conuerto to render the verb which means "shake" in the corresponding line of the Iliad, 1,530, a testimonium absent from West's Teubner edition of the Homeric poem). I suppose the first step which led to obstantes was obstrictos, a word which Ovid doesn't use and which is sometimes a variant of adstrictus.
-- 3,688-689 pauidum gelidumque trementi | corpore uixque meum firmat deus. This is implausibly awkward phraseology, as Lejay's 1894 school edition (p.69) is aware. The best attested reading is pauidus gelidusque, which is impossible. The corruption may be old and the reading adopted by Tarrant and Barchiesi may be only a poor attempt at emendation. Lejay quotes and adopts Havet's pauidum gelidoque trementem.
-- 3,721-722 illa quis Actaeon nescit dextramque precantis | abstulit; Inoo lacerata est altera raptu. The flat abstulit may have been substituted for what possibly is the right word, abscidit (compare the similar word used by Euripides, Bacch. 1127). Lacerata and raptu are both remarkable, though there is not a single word about the line in Barchiesi's commentary. raptu seems to be explained by Euripides, Bacch. 1127-1128 (see J. E. Sandys' note in his 1885 Oxford second edition): through frenzy the god supplies the woman with enough strength to sever the part of the body without the normally required tool. But instead of lacerata "was rent" I suggest resoluta "was detached". Lacerata might be a correction of something like *luserota: for such mistakes see Housman in his edition of Manilius' book I, lviii-lix, from whose well-known collection I point to M's mistake lacertis for relatis in Met. 13,122.
1. Restored lines: 3,200 (maybe rightly), 230 (maybe rightly), 415 and 417 (I disagree), 576 (I disagree). Choice of another transmitted reading: 4,69 uoci versus uocis (granted); 121 et versus ut (granted); 417 numen versus nomen (I disagree). Barchiesi (wrongly in my view) attributes to Narcissus, not to the narrator, the second hemistich of 3,447 and misrepresents Tarrant's punctuation in the "nota al testo". Neither Barchiesi nor Rosati seem to contribute any personal emendation to the text of books III-IV. The reader must be aware that the abridged apparatus may deprive him of salutary warnings such as the doubts entertained by Tarrant about 4,704 et orant (read opemque, cf. 737 auxilium?).
2. In 3,291 et is to be connected with deorum, not with the immediately following deus. The translator gets it right.
3. See Georg Luck's detailed and learned review of Tarrant's edition in Exemplaria Classica 9, 2005, 249-271. I have abstained from discussing passages rightly (in my view) criticised by Luck. Tarrant's apparatus should have reported more conjectures, as Luck shows. I should like to advertise here Felix Gaertner's revival (Rheinisches Museum für Philologie 150, 2007, 93-95) of a most admirable emendation in 6,343, Verheykius' Melitensis for mediocris, mentioned by Magnus 1914 but not by Tarrant. The commentary on books V-VI will by Rosati.
4. It's misleading to say that poets (including "Homer") who use only one scansion "adottono una soluzione coerente" and that those who use both are incoherent. "Homer" always uses the word with short second syllable because in most cases if not all in which he needs the word the other scansion would not fit the hexameter.