Bryn Mawr Classical Review 2004.05.03
Nicholas Horsfall, Virgil. Aeneid 11. A Commentary. Mnemosyne Supplement 244. Leiden: Brill, 2003. Pp. xxvii, 505. ISBN 90-04-12934-0. €110.00.
Reviewed by Alex Hardie, Oriel College, Oxford (email@example.com)
Word count: 4252 words
Nicholas Horsfall's commentary on Aeneid 11 ('H.11') has appeared just three years after his delivery of a similar service for Aeneid 7 ('H.7'), and it is immediately recognisable as (part of) a major contribution on a central area of Latin scholarship. It is also an idiosyncratic work as regards both production and the imprint upon it of the author's personality. As in H.7, Horsfall (H.) is writing for an experienced readership (his Ad lectorem (xxi-xxii) contains more than a hint of a production eruditorum in usum), and H.11 will need specialist review to do justice to its place and likely influence within Vergilian studies. But the non-specialist should not be deterred: H.11 will well repay investment in close reading; and it has riches to offer the Latinist turning to it both for illumination of a relatively unfamiliar book and for what can be learned from H.'s methods.
H.11 comprises an introduction, a text (without apparatus) and translation, the commentary proper and two appendices. At nine pages, the Introduction is exceptionally short in relation to the commentary (416 pages). The intention is patently to shepherd the reader without delay into Vergil's text and thence to the notes, with summary guidance on 'Structure', 'Sources', 'Language' (etc), 'Text' and 'Commentaries'. Readers are thereby encouraged to form their views less from set piece editorial orientation than from continuous engagement with the cumulative detail of the exegesis, above all from immersion in the linguistic analysis. As regards 'Language, grammar, syntax, style, metre' (xv-xvii), the relevant chapter in H.'s Companion to the study of Virgil provides the overview. But in respect to some areas of content, including Camilla, the sparing Introduction carries some risk of loss of cohesion (below).
The Companion (217-48) set out the requirement for systematic study and description, based on full comparanda, of Vergil's language, lexical choices, verbal effects, syntax, grammar and prosody, with sample commentaries, including on 11.378-91; and H.11 represents the continued working out of that agenda. Comparison of sample with commentary reveals three or four points of detail on which H.'s judgment has developed since 1995 (and see below, on 404). That is hardly surprising, and indeed it nicely underlines the elusive complexities of the subject matter.
H.'s methodology typically involves the statistical logging of occurrences as a basis for analysing the tone and resonances of words. The approach is systematic, and system is enhanced (and leavened) by sensitivity, learning and judgment. The assemblage of evidence for distinctively military language is a prominent contribution. H. argues persuasively (xiv-xv) for Vergil's close familiarity with Caesar and with at least the first decade of Livy. The introductory examples of Caesarian, Livian and other loci similes, taken with the Index s.v. 'military prose', tend to substantiate H.'s view (and comparison with citations in their contexts helps). Particularly telling are citations on 460 (cogite consilium), 473 (saxa sudesque), 475 (cinxere), 517 (conlatis signis), 593 (corpus et arma), 609 (substiterat), 616 (tormento), 618 (turbatae acies), 636 (suffoso), 685 (traicit) and 902 (obsessos). H. is also careful to distinguish technical military vocabulary from words that carry military flavour but not distinctively so (notes on, e.g., 599, 608, 618, 666, 684, and 763). As to sources, I did not spot anything new from the documentary material emerging on tablets from Roman frontier posts (other than on 823, citing Dickey's Latin forms of address) and there may be more to be said on that front, elsewhere in Vergil. More generally, CLE is deployed to good effect: for example at 22-23, terrae...mandemus (funerary), citing the Scipionic CLE 9.5-6, mandatus for 'committed' (to earth).
Within book 11, we now have a consistently reliable basis for assessing difficult, ambiguous and controversial phrases. Specific gains include much refined readings of character-development. The notes on 2, 17, 19 and 21 document the language of Caesarian command about, and by, Aeneas; and the key exchange between Drances and Turnus is now richly illuminated by H.'s analyses of style and of Homeric and republican antecedents (succinct summaries at 343-75 and 376-44), thereby accomplishing a piece of self-tasking trailed at Companion 191.
More generally, H.11 offers new resources for assessing the sources, inventiveness and resonances of Vergilian diction, including:
- the interface between prose and poetry in high Augustan poetry. See especially the Sallustian component in Drances' character sketch, noted at 336-42. H. does not seem to claim a requirement for reader recollection of the original context of any verbal borrowings, so that they may not carry the allusive resonances of full-fledged Homeric or Callimachean intertexts; but at 525, the Sallustian background to furta...belli (Turnus' tactics), at Hist. 1 fr. 112 on the Trojan horse, comes close. On a separate but related matter, where H. speaks of 'prosy' language, terminological cross reference (whether or not in agreement) to the new discussions of 'prosaic', 'formal speech or prose', 'neutral' and 'colloquialism' in Adams-Mayer, Aspects of the language of Latin poetry 3-10 would be helpful. The Index gives two entries (10, 144), omitting H.'s judgment (xii; cf. on 302-55) that Latinus' speech is 'couched in calculatedly flat, prosy, language'.
- the colloquial element in high epic, as 'part of his campaign to keep the narrative in touch with the realities of every day life' (Oliver Lyne, Words and the Poet 59, cited on 281); to the indexed 408, add notes on, esp., 227 (the blunt nihil...actum of Diomedes' reported response to the Latin embassy), 333 (connected with 281); cf. 389.
- Vergil's use and variation of technical language: especially important for literary representation of the sacral are H.'s comments on religious terms, e.g. at 217 (dirum... bellum and the technically ill-omened character of the war), 271 (horribili visu as variant on prodigy-style), 739-40 sacra... nuntiet: the verb technical, the combination unique); and 789 (da pater as 'elements of prayer-language rearranged...by a poet'). Also, by contrast, the atypical ritual technicality of decurrere (189). A surprising omission is the allusion in pestis (792) to Soracte Apollo's ritual interest in pestilentia, available in Gransden's note ad loc.
- the Greek component in the language of Latin poetry. Consistently informative here, including on Homeric lexical antecedents (see for example on cara ... pectora [215-16] and its affective extension in sense of Homeric φίλα). The Index s.v. 'Grecism' gives seven entries under four categories, which might have been ordered more clearly, again perhaps cross referenced for categorisation to Roland Mayer's discussion in ALLP (cited by H. for individual features). Thus, no sub-entries 'lexical', 'morphological' (but cf. notes on 775, cassida, and esp. 657, Italides) or 'word order' (the postponed et  is indexed without sub-entry, but postponed nec  is not, and it is not clear whether H. would take anastrophe of this sort more generally as a Grecising feature): so some slippage here in relation to H.'s general commitment to systematic categorisation in matters of grammar and figures of speech (below). H.'s qualification of Servius' view of 416, fortunatusque laborum as a Grecism (unindexed) is important (cf. Mayer, ALLP 161-70 for modern and ancient approaches in this field); and Servius' comment on timoris...argue (383-4, Turnus challenges Drances to accuse him of fear), est de Graeco: nam ita dicunt κατηγορῶ σε φόνου, might have been cited, for its relevance to him as criminosus (and possibly [see below, on 126] mimicking his capacity for Grecisms?). Less readily categorised (but cf. Francis Cairns on 'comparative philology' in Horace at PLLS 8  92-96) as an element in Graeco-Latin linguistic usage are bilingual homophones, with or without semantic equivalence: in book 11, a Homeric context (Il. 16.667-75, Sarpedon's grave) yields τύμβῳ/tumulo (594); and Turnus' philosophising variique labor mutabilis aevi (425) may suggest the Greek word for cyclical change, metabole.
- Vergil's powers of suggestion, or implication (including reader-visualisation): a more elusive virtue, where H.'s sensitivity is at a premium, and where references to autopsy are rewarding: on, for example, spectant (200, the funeral as spectacle for the survivors), deficit (231, Latinus' emotional and political collapse) or, most delicately, the metrical effect of spumeus (626): 'run on here to suggest that last line of foam at the very top of the beach'.
Surprisingly, H. suggests (xxii) that Robert Maltby's Lexicon of ancient Latin etymologies is superseded for Vergil by James O'Hara's True names: no more than OLD is superseded for 11 by H.11! A modest haul of three new etymologies (92, 531 and 557-8) is advanced, out of eleven noted (eight indexed). Other possibilities include 26-27, Euandri ... / non virtutis egentem, of Pallas: cf. O'Hara, 226, and Ov. Fast. 1.479 (exhortation of Evander to act viriliter), so suggesting etymologically that Pallas is his father's son; 571, armentalis (of wild herds of mares suckling (?) Camilla), where, in etymological coherence with character and nurture, variant derivations (Maltby, s.v. armentum) quasi apta armis and ab arando respectively connect with the arming of Camilla at 572 (armavit), and contrast with her wild, untilled, environment at 570; note also 721 (ales ab alto).
Central to H.'s methodology is system and precision in analysing and categorising grammatical and syntactical phenomena. This is painstaking work which inter alia will refine readers' ability to recognise genuinely special effects. Thus (out of so much that is new), H.'s analytical sensitivity to syntactical tone is well exemplified in the full discussion of nec veni... (112) in an unreal conditional clause (in Aeneas' diplomatic reassurances to the Latins), yielding the suspicion 'that V. here concentrates grammatical singularities as an expression of intense and confused thinking'. Of course, not everything can be pinned down, and it is a strength of this commentary that it recognises ambiguity and tends not to press for single, exclusive readings. The Index, s.v. 'ambiguity, syntactical', cites donum Triviae (566), 'a gift from/to Diana' (Triviae genitive in both, the latter with archaising flavour). To which could be added the preceding tua (558): nom. pl. with tela, in the sense of the spear as 'your [Diana's] weapons', but then coloured by tuam (560), suggesting also tua nom. sing., i.e. Camilla as 'yours', 'belonging to Diana', with hymnal repetition (cf. H. on 560). There seems no way of choosing, and H. is surely right to judge that no choice is needed. Still more important, because of what it says about Aeneas' capacity for disingenuous rhetorical ambiguity, is the note on nostra ... hospitia (113-14): Latin guest friendship offered by or to Aeneas? The hero implies the former, but the commentator's reference to book 7 reveals it had been the latter.
The emphasis post-Knauer is on refinement of understanding rather than the discovery of new Homeric intertexts. But gains there are: on the structure of 445-6 (citing Margaret Hubbard), for example. One or two Knauer parallels are dropped, for example at 378 (Il. 2.796). The introduction eschews a general summary (xiii) and lists Vergil's debt under five headings: phrases taken from Homer (xvi); Homer as a source of 'facts' and myths; structural reworkings; motifs; and the scholia. A useful start for the newcomer; but a longer overview might have helped, covering, for example, the importance of Iliad 7, reference to which turns up only at 203-224. Among the Homeric set pieces, the treatment of the Turnus 'free horse' simile (492-97) is masterly, noting the gender of the armenta equarum (494) towards which the beast gambols, and the significance for Turnus' character development of Homer's application of the simile to Paris. H. as often cites Donatus (on 497) to good effect, on Turnus'/the horse's mental state; worth adding are the Homeric scholia on the horse's 'irrational' character (Robin Schlunk, The Homeric scholia and the Aeneid 28). Proper weight is given to P.F. Burke's demonstration that complexity of character is underpinned by a heterogeneous range of Iliadic sources that leave the reader to ponder the balance of intertextual implications. Against this insight into Vergil's methods, the debate on Camilla's address to the doomed Ornytus (686-9), whether nomen tamen haud leve patrum / manibus hoc referes, telo cecidisse Camillae (688-9) is truly consolatory (so H., citing Iliadic parallels) or ironically so (Henry), might be further illuminated from Il. 7.89-91, where Hector prophesies his naming as victor/killer in the putative epitaph of a (nameless) Greek champion: similarly, the 'name' that Ornytus will carry to his ancestors, is Camilla's own (instead of, or in addition to, nomen as fame accruing to Ornytus from death at her hands [thus, H.]): further Iliad-based complexity of character, therefore, and sustained sarcasm from the heroine, not just 'soldierly pride' (H.); and just as Hector's speech is echoed at his killing by Achilles (79-80/22.342-3), so is Camilla's after her fall (898).
The Apollonian loci listed by Damien Nelis (The Aeneid and the Argonautica of Apollonius Rhodius) are relatively few in number (but see below) and Apollonius is not indexed in H.11. In Appendix I 'Camilla and the Epic Cycle', H. argues, as he did in 1988, against Vergil's access to and use of the Cycle as a source for the Amazons (i.e. for Camilla), and it is not easy to imagine that archaic corpus being familiar even to the most learned among his ancient readers (for awareness, however, cf. H. on 257, per orbem; and Vergil seems to have used an equally archaic epic, now lost to ancient testimony, on Heracles in book 6). On the allied matter of Vergil and Quintus Smyrnaeus, it is worth noting the correspondences (presumably available in the specialist treatments) between Camilla and Achilles/Penthesilea in QS 1: at 705-6, QS 1.575 ('woman' taunt); at 721-2, QS 1.569-72 (simile on dove as easy prey for hawk); and at 803, QS 1.594 (spear penetration around breast).
The Introduction delineates (xi-xiii) the book's tripartite structure (in line count, a quarter each to funerals and the Latin debate, and half to battle/Camilla). Here and in introductions to narrative episodes, H. economically helps the reader navigate his/her own path through the complexities of the plot, its transitions and structural linkages. A 12-line summary of Camilla's aristeia (664-835) provides succinct encouragement to look carefully at neglected pace and structure. Some key developments receive new emphases (Aeneas' calculatedly generous speech to the Latins at 108-119, for example; and the 'lines in which [the Latins'] grief is transmuted into political opposition (213-24)': xi). An exception to this exemplary plot-exegesis is Jupiter's incitement of Tarchon (725-8), a 'brief and discreet intervention on the Trojans' behalf', comments H., laconically, but one that leads by a direct chain of events to Camilla's death, hence Turnus' abandonment of plans to ambush Aeneas, and the final unravelling of the Latins' cause; a decisive intervention, then, that needed to be signposted.
Book 11 gives us Aeneas as orator, with four separate utterances in the first 120 lines. H. is excellent here, cumulatively establishing Aeneas, consistent with the typology of the good king, as emerging commander-statesman, still tight-lipped, but human and capable of strong emotion, most expressively in his failure to protect Pallas (for his tears at 41, cf. especially Hercules' weeping at his inability to respond to Pallas' appeal at 10.464-5); capable too of calculated benevolence and rhetorical ambiguity as he plays (108-19) on latent/emerging discord among the Latins. The Companion chapter on book 11 is on 'Rhetoric', which, with the Introduction and prefatory notes on speeches, keeps the reader well supplied with pointers to structure, speaker-motivation, stylistic traits and impact. S/he may however wonder about H.'s view on the issues of rhetoric and truth as set out in Philip Hardie's 1998 paper 'Fame and Defamation in the Aeneid'. He touches on this in a sensitive introduction to Diomedes' rewriting of the (Iliadic) 'historical' facts about Aeneas as warrior (243-95); but the juxtaposition of (reported) forensic vituperation and (quoted) tactical flattery (of Aeneas) at our first encounter with Drances, the criminosus (H. on 122), invites further comment. The extravagant artifice of his opening o fama ingens, ingentior armis (124), a comparative form unparalleled before C.3, the suspect validity of the jingling antithesis of famaand arma (fama surely cannot signify 'peaceful virtues' [H.]), and the startling Grecism at 126 (mirer plus genitive) all inspire distrust and suggest deeper Vergilian interests in the rhetoric of praise and blame, and its ethnic affinities. Only the Latins react (132); tacet Aeneas.
On a related point, the reported parting exchanges between Evander and Aeneas (45-8) and between Evander and Pallas (paired, at 152-5) would appear to be susceptible of rhetorical analysis as summary examples of speeches of departure and sending off (syntactica and propemptica). H. (on 55) rightly stresses the need to read Aeneas' bitter self criticism at Pallas' death in its rhetorical context, and, as we reconstruct the drama of these remembered words, the standard epideictic topics of promised return and faithfulness might bring closer definition to his regrets (54-5) about his rhetoric on that occasion.
H.'s view of Camilla as a 'great, learned, literary invention' was set out in his distinguished 1988 paper 'Camilla o i limiti dell'invenzione', was restated in H.7 and is taken as read in the principal note on the heroine (on 535-96; cf. on 546), which also summarises her various analogues in legend and literature. The treatment in H.11 unfolds with Vergil's own: her courtesy at the meeting with Turnus (good on etiquette) and urgency (defence of the 'commander's quick instincts and good eye'; qualified reservations on incaution firmed up on 586 'undervalue(s)... the Trojan threat'); imitatio Dianae as leader of her followers (on 533; 586); impact of wild childhood surroundings, nurture and training as huntress (on 569, and especially 571 [Scythian ethnography and Amazons]; cf. 652) and contrast with her adult dress (576, not developed: see below); the rationale for her going to war (586; 589: no guilt incurred as a beloved of Diana; rather, 'her will (to fight) and doom (to die young) coincide'); Amazonian dress and weapons in battle (649, 651, 654; cf. 659-663; add, on arma relinquunt (830), the delicacy with which in iconography, Penthesilea's weaponry slips from her relaxing fingers in death: e.g. LIMC VII.2.52a and d); her luxury accoutrements and alleged 'vanity' (652), yet (on 664, aspera virgo) 'rough of character or nature, in a harsh sense'; 'soldierly pride' on killing Ornytus (688-9; but see above); 'enraged by disparaging use of her gender' (705); on Camilla's fatal, 'feminine' attraction to Chloreus' apparel (782), 'as a warrior, above gender' (yet vulnerable to gender insult); 'as a woman she was born with a love for pretty things' (yet nurtured in rough clothes); and, finally, on the proximate cause of her fall: 'her blindness is human' (suggestive note on memor, 802: failure to heed the familiar sound of the spear and associate it with danger to herself).
Perceptive touches and suggestive connections are not lacking, but I missed an extended comment on Camilla's integration into the wider scheme of things (that is, a companion to the note on 7.803-17, esp. ad fin.). What are we to make of the piquant juxtaposition of two asexual warriors, each devoted (H. on 591) to a divinity of the mountains, eunuch Phrygian Chloreus and Amazonian femina Camilla, each interested in artificially gorgeous clothing, especially against the (sexually charged) Apollonian background in Medea/Jason's cloak suggested by Damien Nelis? And more might have been made of the long vein of gender interplay, worked out through Turnus (above gender in book 11: contrast 7.444 for his 'war is for men' assumptions), Ornytus (above), the son of Aunus (705-6: cf. Achilles/Penthesilea, cited above), and especially Tarchon upbraiding the Etruscans (734; also, for criticism of unbecoming behaviour that echoes Phrygian stereotyping , cf. 9.615 and 618). Others might formulate other questions; and the commentator might respond that his task is to supply resources to inform critical explorations of this sort, not to write one himself. Yet the key words femina (x2) and femineus (x3) need more, and the impression lingers that H. is not yet fully on terms with Camilla, whose book this is, or else has consciously held back from fuller exegesis.
The text is relatively uncontroversial. Noting (xvii-xviii) just two divergences from Mynors 1969 OCT, H. advances the suspicion that some parts of the poem, book 11 included, have been better transmitted than others, a phenomenon that H. tentatively attributes to the earliest period of the history of the text. Where the Capital MSS MPR are all extant, and in agreement, there is doubt or debate about their text in four places only: 230, pacem petendam, where archaising petendum is rightly preferred; 338, lingua melior (Drances), retained in preference to linguae, the lectio difficilior, with a fair account of the evidence for experimentation with a rare genitive, to which might be added the points on Drances' capacity for Grecism, noted (on 126 and 383) above; 691, aversum, of Butes, defended against the Carolingian adversum: yes, but perhaps to be translated 'when his head was turned away', rather than H's 'from behind', given the placing of Camilla's spear between breastplate and helmet, which presumably prompted the late correction; and 839, mulcatam ('lacerated', of Camilla's corpse), where mul(c)tatam (ω), 'punished', is defended in the commentary and is incorporated in the translation but not in the text. MSS variants are selectively recorded in the notes: among the absentees are R's dura for archaism dia (657) and P's designavimus (367), neither of which represent what Vergil wrote, but they might be significant for the distribution of scribal errors/conjectures and for such patterns of MSS (dis)agreement as there may be. H. notes R's preservation of sound readings 'contra mundum' at 487 and 728 (cf. on 466); for similar virtues in P, there is 126, 151, 168 and 381; and for PR agreement on apparently sound readings against M, there is 145, 236, 304, 410, 595, and 844. 'T' (indicating a textual issue) is marked at two points in the text (87, 266) where there is no corresponding note. And a T might have been entered at 269, given Edward Courtney's suggestion of a lacuna in all MSS and the importance of such phenomena, if accepted, for his thesis of a late archetype lying behind the Capital MSS.
H. is rarely prescriptive, and he resists standardization or uniformity in orthography. Ironically, where he is more dogmatic on a textual matter, there is room for doubt. H. is emphatic that 404 (Turnus to Drances), nunc et Tydides et Larisaeus Achilles (sc. 'would fear Trojan arms'), is interpolated from 2.197, perhaps via a reader adding a marginal expansion on Myrmidonum proceres (403). Really? The speech echoes earlier passages elsewhere (e.g. with 393-4, cf. 6.87); his subject is defeated Trojans, as is Aeneas' at 2.195-8; in both contexts, rhetorical artifice (artificis scelus; 2.125; arte, 2.195; artificis scelus, 407, of Drances) is central to the drama; and the objection of unreality (the dead Achilles) falls away in the face inter aliaof the adynaton in 405. If genuine, 404 underlines Turnus' wrongheadedness, for in expressing respect for Trojan/Phrygian arma (293), Diomedes has just come close to doing what Turnus suggests is impossible for him; and Turnus' rhetorical artifice is arguably leading to the ruination of his own cause: in short, a telling comment on specious rhetoric in this most rhetorical of books. If not genuine, 404 is an uncommonly appropriate interpolation, skilfully inserted (positives replacing the 2.195 negatives, yielding et... et.../405...et). This paragraph, it might be added, was written quite without recollection of Companion 191, where the line is assumed genuine and integrated into context on other grounds. The Latin Index (with absences: look in vain, as H. might say, for spectant) and the Index of Proper Names are helpful aids to navigation. As already indicated, however, the English Index does not do justice to the contents. The selection of entries seems at times to take on a whimsical quality (the doubling of 'periphrastic synecdoche' and 'synecdochic periphrasis', for example). Sadly, this will detract from the book's reference usability, for students coming to H.11 from, e.g., ALLP, to see what H. has to say about (say) 'colloquialism' or 'archaism'. Enough has been said in reviews of H.7 about multiple parentheses; but what does demand comment is the number of misprints, missing brackets, and so on (I counted 40) and omissions/slippages between text and translation, or translation and commentary (10 noted, some serious). The publishers are doing a service in taking on commentaries on the scale of H.7 and H.11; but speed of production (or cost containment) seems to be taking precedence over professional copy-editing.
In substance, however, H.11 is to be welcomed with gratitude as a quite new commentary that will fully open up the book for extended study. Its claim to be comprehensive is justified in relation both to its range and to H.'s encyclopedic embrace of Vergilian scholarship, distilled in the notes and also in bibliographical lists for which researchers will be properly grateful. The positions taken in it will invariably command attention and respect, if not always critical assent. Where future study of Aeneid 11 moves on from H.11, it will likely enough be on the basis of material H. has made available; and informed debate will be in line with H.'s encouragement, indeed challenge, to readers to think things through for themselves, not least in his summation of the evidence for haste on the poet's part, and his concluding evaluation of quality, in Appendix II, Dormitatne Maro quoque?).
This is a work with which anyone with a serious interest in Latin poetry will need to be familiar. It is also a work that graduate supervisors might profitably place in the hands of their research students, with encouragement to immerse themselves in H.'s methodology and its philological disciplines. Moreover, we may now look forward (ix ad fin.) to 'H.3'.