Bryn Mawr Classical Review 2007.08.47
Nicholas Horsfall, Virgil, Aeneid 3. A Commentary. Mnemosyne Supplement 273. Leiden: Brill, 2006. Pp. liv, 513. ISBN 10: 90-04-14828-0. ISBN 13: 978-90-04-14828-4. €165.00.
Reviewed by Philip Hardie, Trinity College, Cambridge (firstname.lastname@example.org)
Word count: 2691 words
Table of Contents
After his monumental Brill commentaries on Aeneid 7 (2000) and 11 (2003) Nicholas H(orsfall) continues his majestic progress with book 3; book 2, we gather, is now well under way. Into the Italian books 7 and 11 H. poured an expertise based on his lengthy researches into ancient Italian antiquarianism, as well as on his intimate first-hand knowledge of Italian topography. Now a resident of the remote north of Scotland, he has turned to the books that take us into areas far from the Italian heartland. Not that there is any shortage of material in book 3 for areas of Virgilian scholarship which H. has made very much his own over the years: much scope for more ancient antiquarianism, religious lore and aetiology, the importance for the narrative of Aeneid 3 of the traditions of colonisation and ktisis. H. has indeed contributed significantly to the rehabilitation in recent years of Aeneid 3, a book hitherto relatively poorly served by commentaries: R. D. Williams' 1962 edition had remained the best (and will continue to be the first choice for the less advanced Virgilian).
An Introduction briefly surveys the relationship of Aeneid 3 to neighbouring books; language and metre (with particular attention to the range of sources for Virgil's epic idiom: especially striking, here and in the commentary, is the density of Lucretian echoes, often without apparent intertextual point, a further reminder of how important Lucretius continued to be for Virgil as he turned to epic); types of source for the various episodes (of all the books in the poem, 3 has benefited the most from recent attention to Roman poets' use of Hellenistic models; in particular H. makes intensive use in the commentary of Damien Nelis' Vergil's Aeneid and the Argonautica of Apollonius Rhodius (Leeds 2001)). There follows a long section on the old problem, recently revisited by T. Berres and H.-C. Günther, of the 'Growth' of the Aeneid: H. is healthily sceptical of much of the evidence, and of the arguments based thereon, but gives reasons for believing that the third book was one of the earliest to be composed (although not originally as a third-person narrative). For H. the, often criticised, variety of styles and textures in the book is the mark of a poet flexing his muscles in the new medium of epic. Qualified admiration for Virgil's achievement is seen, for example, in H.'s handling of the Etna, and Polyphemus and Achaemenides episodes at the end of the book, once regarded as rhetorical excesses that betrayed lack of maturity and of finish; H. reveals the skill and complexity that went into the crafting of a sequence whose exuberant and inventive grandiosity others have already shown to be anything but otiose within their wider context.
As in earlier volumes, there follows a text without apparatus, but with marginal signs indicating notes in the commentary on points of text, orthography and punctuation, together with a serviceable facing translation.
H. diverges substantially from Mynors' OCT in three places; 127 consita (concita); 618 cruenta (cruentis); 659 manu (manum ).
The virtues of the commentary are familiar: daunting command of the scholarship, careful and revealing attention to the idiosyncrasies of Virgilian language and style, an alert eye for compositional strategy and tactics. The best of the longer notes are very good indeed: one example of many is the densely packed mini-essay on 588-691 Achaemenides and the Cyclopes. One has the sense of a commentator in constant dialogue with scholars and commentators, dead and alive, from Servius and Tiberius Donatus (much despised, and to whom H. refreshingly gives credit where due) to the present day. That conversation can verge on the private, as in the continual noises of approval or disapproval for entries in the Enciclopedia Virgiliana and the Thesaurus. It is unilluminating to read (288 de Danais uictoribus n., on the use of de) 'Gudeman, TLL 18.104.22.168f. unilluminating.' Unexpected abbreviations and ellipses can leave the reader momentarily puzzled: for example, at 82 ueterem . . . amicum n. we do need the reference to the parallel cited (= Hor. Serm. 2.6.81). As in previous volumes, the bibliographical system often leaves the reader scrambling, and there were a number of items that I failed utterly to track down. Progress may be majestic, but there are also signs of haste (not surprising perhaps, given the rate of progress), in typographical errors and repetitions. The printer may be to blame for the irritating tendency to omit a space before an opening bracket.
Detailed notes follow (by line number); the accumulation of small criticisms in no way detracts from my admiration for the overall achievement.
2-3 superbum | Ilium 'Again a run-over adj.': in fact a noun.
13-68, end of n.: Servius' reference to the story that Romulus' spear took root where it fell does not seem 'unnecessary': the allusion would point up the Polydorus episode as a perverted foundation story.
27 quae prima . . . arbos: it is the noun rather than the adj. that is taken over into the relative clause, as at 94-5 (so Williams).
32 temptare may be 'an unsurprisingly Lucretian application', but it coheres with the Lucretian colouring of causas penitus . . . latentis, whether or not one wants to develop a reading in terms of forbidden knowledge.
46 telorum seges: here the artefacts, the spears, turn into vegetable growth, whereas at G. 2.142 and A.7.526 spears are, literally or figuratively, the product of a kind of vegetable growth. seges anticipates and helps to explain the following increuit.
51 Curious notion that the suppression of the name Polymestor may partly be so as not to confuse Dido with two names starting with Poly-.
55 Polydorum obtruncat: is there perhaps a pun on truncus 'tree-trunk', as there may be in the case of the boy's father at 2.557 truncus, taken with the felled-tree simile at 2.626-31? Polydorus in some sense undergoes a literal metamorphosis into arboreal matter.
62-3 ingens . . . tellus: H. points out an embarrassing error in my note on Aen. 9.709 dat tellus gemitum et clipeum super intonat ingens, but if clipeum is there governed by super it is more natural to take ingens as qualifying Bitias rather than tellus ('the huge earth thundered on top of him' makes no sense).
69 prima fides pelago: does this have to mean 'first moment of safe sailing' in the sense 'beginning of the sailing season', rather than the first moment that it seemed safe to sail, without reference to the season?
72 recedunt: not so much 'exchange of subjects', as registering the illusion that the land, not the ship, is moving, as at Lucr. 4.388-90; see 205 n., and M. Murrin The allegorical epic (Chicago and London 1980) 36-7.
85 propriam . . . domum: propriam is more naturally taken 'our own, for ourselves', rather than 'for your very own' (or is 'your' a misprint for 'our') or 'of your (Apollo) very own gift'. On Thymbraeus see J. D. Reed, 'The death of Osiris in Aeneid 12.458', AJP 119 (1998), 399-418.
91 Earthquake does seem the easiest way of understanding moueri (either way there is some paradox in the contrast with 77 immotam); the matter is further complicated by the fact that, as often, uisa could be taken either as 'was seen' or 'seemed'. summissi certainly looks like terror, especially in the light of Lucr. 1.92, cited ad 93.
104 Od. 19.172-3 begins one of Odysseus' false Cretan tales; the allusion is perhaps a hint that we should not believe in Anchises' interpretation of the oracle.
126-7 sparsasque per aequor | Cycladas, et crebris legimus freta consita terris: Ovid may implicitly correct Virgil when he uses sparsas in another sense at Met. 2.264 exsistunt montes et sparsas Cycladas augent (i.e. not crebras).
129 If Servius is correct to see Cretam proauosque petamus as in the metre of the celeuma, then why is it a 'clear error' (128 n.) to invoke the celeuma?
132 'heroes who build walls are turning into citizens': but Aeneas never stopped being a citizen, unlike, arguably, the Greek heroes of the Iliad.
133 Here and elsewhere (e.g. 302 n.) H. is critical of David Quint's and Maurizio Bettini's influential reading of the Trojans' regressive attachment to the past, unfairly I think. It is true that there are plentiful historical parallels for attaching the name of an old city to a new foundation, but the desire to remain in the past at Buthrotum goes far beyond this.
139 Mynors on G. 1.444 is misreported: he says 'V. adds the subject-matter of G. 3', with no hint at the Noric plague.
147-91, 9 lines down: Ov. Met. 15.10 ff. is the story of Myscelus, not Numa (who is the audience).
178 As support for the claim that intemerata means unmixed cf. Stat. Silv. 1.2.205 (Alpheus) flumina demerso trahit intemerata canali (and cf. Ecl. 10.5). Is the usage perhaps influenced by the sound of tempero?
180 That prolem ambiguam imports a hint of the Cretan labyrinth may be supported by 181 nouo ueterum deceptum errore (cf. Aen. 6.27) locorum: a new kind of straying, in contrast to the old story of the labyrinth?
190 paucisque relictis: 9.482 is not a parallel for the sense of someone being left behind in death.
207 The contrast with remis insurgimus works better if uela cadunt means 'sag'; the men rise expansively on to their oars.
215 ira deum suggests the etymology of Dira from dei ira (see Maltby Lexicon of ancient Latin etymologies (Leeds 1991) s.v. dirus).
237 'eptic' for 'proleptic' in the last line of the note is a more puzzling typographical error than most.
240 noua proelia in fact marks the first time that the Trojans attempt physical violence against the Harpies.
271 -que in Dulichiumque does not 'look both ways' (and Wills does not say that it does).
272 On the location of ancient Ithaca see now R. Bittlestone Odysseus unbound: the search for Homer's Ithaca (Cambridge 2005).
281 H. perhaps underestimates the paradox in patrias . . . palaestras. Mynors scents oxymoron at G. 2.531 agresti . . . palaestra. oleo shows that we are not just dealing with the athletic contests shared between Greeks and Trojans in Homeric epic.
296 If we hear Enn. Ann. 167 in Aeacidae, we may reflect on a reversal of expectation of a different kind.
303 I am reluctant to exclude the possibility that manisque uocabat hints at necromancy; 312 shows that Andromache is in the state of mind where she might not be surprised to see Hector. That state of mind is not easy to pin down, and is comparable to the sleeping confusion of Aeneas as to what he sees in the dream-vision of Hector in book 2.
329 transmisit habendam: transmitto is used in a legal sense (OLD 9): cf. 333 reddita cessit n.
342 animosque uirilis: this is not so much an example of 'virtues beyond years' (as at 9.311), as the wish that his father Aeneas and uncle Hector will stimulate Ascanius to grow into the manly virtues proper to his family.
415 aeui longinqua uetustas: the Lucretian echo is more extensive, including Lucr. 2.70 (69 is quoted) ex oculisque uetustatem subducere nostris, with the difference that Lucretius is talking of gradual, not sudden, change.
422 rursus is 'back again' (OLD 1b) rather than 'again, for a second time': no logic is infringed.
438-9 dominamque potentem | supplicibus supera donis: the conceit of humility conquering the proud deserves note (here stronger than in the parallel 8.60-1).
453 It's not clear why genitives of value are 'somehow conceptually alien to the subject-matter of Aen.': Virgil is not shy of commercial or financial imagery, as at 10.532 commercia belli: see R. O. A. M. Lyne Further voices in Vergil's Aeneid (Oxford 1987) 52.
460 cursusque . . . secundos: here the repetitions (454 cursus, 455 secundos) seem to have point: visiting the Sibyl will be worth the wait, since she will vouchsafe a fair voyage and more. But in general H. does well in drawing attention to the incidence of non-significant verbal repetition.
476 cura deum: the parallels are not for cura applied to a person or thing, for which OLD s.v. cura 8 gives as the first instance Ecl. 1.57, very likely alluding to Gallus. In the present context an amatory overtone readily presents itself. This passage is included in TLL 4.1466.57ff. deliciae, is a quo aliquis delectatur uel cui cura impenditur, a sense which overlaps with ibid. 4.1475.42ff. metonymice de persona amata.
493 uiuite felices: worth citing (as Williams does) the makarismos at 1.437 o fortunati, quorum iam moenia surgunt! With the contrast in 494-5 nos . . . uobis cp. Ecl. 1.1-4 tu . . . nos . . . nos . . . tu. 495 parta quies resonates with 1.205 sedes . . . quietas, the promised land for Aeneas and his men. A passage dense with Virgilian thoughts on home and exile.
510-11 There seems to be a playful contrast between in litore sicco and sopor inrigat artus, of a kind that Virgil indulges in elsewhere.
514 H., with other commentators, compares G. 1.376 captauit naribus auras. The Georgics passage on weather signs seems to be more widely present to Virgil here: with 510 in litore sicco cf. G. 1.363 in sicco ludunt fulicae; with 515 sidera . . . labentia caelo cp. G. 1.365-6 stellas . . . caelo labi; and note that G. 1.332 alta Ceraunia (Aen. 3.506 Ceraunia) occurs in the description of the storm that shortly precedes the weather signs.
516-17 Od. 5.271-5 (Odysseus watching the stars at sea) is a more important model than H. allows, and may ironically anticipate Palinurus' fate at the end of book 5: Homer emphasizes that sleep did not fall on Odysseus' eyelids.
556 Good note on the personification in fractas . . . uoces: add that it also anticipates the personified landscape of Aetna. For a related equivocation between human and natural sounds cp. 5.866 tum rauca adsiduo longe sala saxa sonabant (no more sweet-voiced Sirens, but a hoarse sound). Both epithets occur at Hor. Odes 2.14.14 fractisque rauci fluctibus Hadriae, where Nisbet and Hubbard comment 'the participle hints at a hoarse voice, like rauci'.
568 fessos is most naturally taken of the effects of their exertions in escaping Charybdis than of the effort needed for 'a final pull in to land' once the wind drops; adlabimur does not suggest effort.
571 horrificis . . . ruinis: Williams' 'falling material' is more helpful than the examples of ruina in different senses given by H. tonat imports a hint of Gigantomachy, I suggested in Virgil's Aeneid. Cosmos and imperium (Oxford 1986) 264 (elsewhere cited generously by H.).
588-691 Stephen Hinds made the metapoetical point about 591 noua forma uiri, in Allusion and intertext (Cambridge 1998) 114, published the year before Papanghelis.
592 Achaemenides isn't strictly a shipwrecked sailor.
600 G. 2.340 cum primae lucem pecudes hausere deserves to be quoted (in Williams).
607 genibusque uolutans: Suet. Cal. 42 does not describe 'wallowing on the knees'. 'Wills 374' in this note should read '272'.
613 infelicis Vlixi: contra H.'s interpretation it should be noted that elsewhere Ulysses does not share in the progressive rehabilitation of Greeks in the poem.
614 mansissetque utinam fortuna: this is different from Hor. Odes 3.29.53 laudo manentem, which refers to the good fortune of prosperity, etc., notoriously fickle; Achaemenides chose to swap his lowly station for something a bit more ambitious. There is then perhaps a hint of paradox in the way he puts it.
668 certantibus . . . remis: another possibility is that the contest is with Polyphemus, who will soon be in pursuit.
679 concilium horrendum: perhaps glancing at an etymology of concilium a concalando, id est uocando (Paul. Fest. 38); the Cyclopes have been summoned by Polyphemus' shout.
690 relegens: relectos is Heinsius' conjecture at Hor. Odes 1.34.6, possibly strengthened by other echoes of the Horatian lines in the vicinity: H. adduces 3-4 nunc retrorsum | uela dare at Aen. 3.686 dare lintea retro, and cp. 3 erro with Aen. 3.690 errata.