¹ G. Williams may fairly be taken as representative, Tradition and Originality in Roman Poetry (Oxford, 1968), pp. 378–83Google Scholar. Cf. Beaujeu, J., ‘Le Mariage d'Énée et de Didon’, Revue du Nord 36 (1954), 115–19CrossRefGoogle Scholar. R. C. Monti goes even further, arguing that there is no ambiguity, and that Aeneas is indubitably married, The Dido Episode and the Aeneid: Roman Social and Political Values in the Epic (Mnemosyne Suppl. 66, Leiden: Brill, 1981), pp. 45–8Google Scholar.

² cf., e.g., Balsdon, J. P. V. D., Roman Women (London, 1962), pp. 181 fGoogle Scholar.

³ Objections of various kinds given by Klingner, F., Virgil (Zürich, 1967), pp. 443 f.Google Scholar; Sparrow, J., Dido ν Aeneas: the case for the defence, The Sixth Jackson Knight Memorial Lecture (Abbey Press, 1973), pp. 5 ffGoogle Scholar.

⁴ On 339; Servius ‘auctus’ has similar comments. On the scale and cost of marriage ceremonies in the upper classes see Friedlaender, L., Darstellungen aus der Sittengeschichte Roms ⁹ (Leipzig, 1919), 1. 274 fGoogle Scholar.

⁵ For the direction of the narrative away from Aeneas here, see Klingner, op. cit. (n. 3), pp. 444–6; Grimm, R. E., ‘Point of view in Virgil's Fourth Aeneid’, CW 63 (1969), 81–5Google Scholar.

⁶ Austin, p. 98.

⁷ Eur. Med. 465–519, A.R. 4. 335–90. For detailed comparisons see, besides the commentators, Heinze, p. 134 n. 1, and Highet, pp. 220 ff.

⁸ p. 98.

⁹ cf. the brief remarks of Clarke, M., ‘Rhetorical influences in the Aeneid’, G & R 18 (1949), 21Google Scholar: ‘the rhetorical system of eliciting all the arguments inherent in a situation lies behind Dido's speech in Virgil as it lies behind Ovid's letter [Her. 7]'; also p. 26.

¹⁰ On 305. If my discussion of Dido's speeches, here and below, appears harsh, let me say that I do not consider we are ever intended to lose our sympathy for the Queen (here I cannot agree with Quinn's idea of a ‘shift’ in attitude as the book goes on, ‘Vergil's Tragic Queen’, Latin Explorations (London, 1963), pp. 48 ff.)Google Scholar. It is not a matter of ‘judging’, still less of deciding which ‘side’ we favour: cf. the sensible remarks of A. E. Douglas, in his excellent address to the Vergilian Society (‘The realism of Vergil’, PVS 1 (1961–1962), 15–24)Google Scholar; ‘What Vergil has done is to present an unpartisan view of all the issues. He has told the truth’ (p. 17).

¹¹ George Eliot catches the sensation in a passage extensively modelled on this scene, when Daniel Deronda attempts to address the shattered Gwendolyn Harleth: ‘he paused a little between his sentences, feeling a weight of anxiety on all his words’, Daniel Deronda (Penguin, 1967), p. 839Google Scholar. Deronda's attempts at self-control remind one irresistibly of Aeneas: ‘Deronda, too, felt a crushing pain; but imminent consequences were visible to him, and urged him to the utmost exertion of conscience’ (ibid.).

¹² CR 7 (1893), 417Google Scholar.

¹³ OLD s.v. ‘res’ lla.

¹⁴ On 337.

¹⁵ Vergil has capitalized upon the antithesis twice in the preceding section: heu quid agat? quo nunc reginam ambire furentem \ audeat adfatu? (283 f.); temptaturum aditus et quae mollissima; fandi / tempora, quis rebus dexter modus (293 f.).

¹¹ Barrett ad loc. compares Arist. Rhet. 1415 b 22, οἱ πονηρ⋯ν τ⋯ ἔχοντεс, ‘those speakers with a bad case to argue’: see LSJ s.v. Ill 4 for this meaning.

¹⁷ See Pease on 333.

¹⁸ ‘quantacumque enumerare potueris in me tuo beneficio conlata, eorum tibi debere gratiam non repugno’ (on 335).

¹⁹ See Martin, J., Antike Rhetorik: Technik und Methode (Munich, 1974), p. 307Google Scholar; for many w examples of the technical use of the verb, see TLL s.v. 618. 44 ff., ‘sensu technico rhet…’.

²⁰ Twice: Met. 1. 215; Ars 1. 254 ff., quid tibifemineos coetus uenatibus aptos / enumerem… ? / quid referam Baias…?

²¹ Seven times: e.g. Ach. 1. 140, sed longum cuncta enumerare; Silv. 3. 1. 102, uix opera enumerem.

²² cf. Nep. Lys. 2. 1, ne de eodemplura enumerando defatigemus lectores; Cic.Plane. 74, omnes (gratias) enumerare nullo modo possent.

²³ p. 75; cf. p. 289. It is surprising that, so far as I discover, Highet is the only writer on the problem to attempt to explain this striking and significant correspondence.

²⁴ cf. 298, omnia tuta timens; 419 f., hunc ego si potui tantum sperare dolorem, / et perferre, soror, potero.

²⁵ Nor, of course, does she know why he is going.

²⁶ Though it does not seem to be recognized that when Aeneas says nee coniugis umquam / praelendi taedas he is in fact admitting a grave moral failing. If a man of Aeneas' (even t compromised) integrity says that he did not use promises of marriage as a pretext, he is not simply telling the woman she has nothing to complain about.

²⁷ See, conveniently, Pease pp. 45 f., Austin p. 105.

²⁸ Such is the preconception behind R. Speaight's bluff appraisal of Aeneas' predicament: ‘There are situations in which there is practically nothing a man can say’, The Vergilian Res (London, King's College, 1958), p. 8Google Scholar.

²⁹ Tolstoy, L., Anna Karenina (Penguin, 1954), p. 776Google Scholar.

³⁰ Further examples in TLL s.v. llb of fingere as the hallmark of oratorical or impassioned speech.

³¹ For the antithesis present here between res and fingere, cf. Cic. Brut. 149, uereor nefingi uideanlur haec, ut dicantur a me quodam modo: res se tamen sic habet.

³² In ‘Dido's Reply to Aeneas (Aen. 4. 362–387)’, in Vergiliana, ed. BardonandR, H. and Verdiere, R. (Leiden, 1971), p. 423Google Scholar.

³³ A convenient collection of the views of the consensus in Monti, op. cit. n. 1, note 11 to chapter 4 (p. 104). Austin is aware of the emotional force of Aeneas' words at some points in this central portion: see his remarks on 351 ff. Similarly Lefèvre, ‘Aeneas’ Antwort an Dido', WSN.F. 8(1974), 111 f.

³⁴ op. cit. n. 1, pp. 42 f. Cf. note 10 to chapter 4 (p. 104): ‘…Aeneas tells Dido not only that his obligations prevent him from remaining with her, but also that he faces the choice between love for her and love for his own people and that he does not decide in her favor. This is not to deny that Aeneas loves Dido; it is a question of choices.’

³⁵ Witt, N. W. De, ‘The Dido Episode in the Aeneid of Virgil’ (diss. Chicago, 1907), p. 78Google Scholar.

³⁶ cf. Part. Or. 121, 137, Brut. 88; in speeches, Lig. 23, Scaur. 39, Flacc. 55, 87.

³⁷ Plaut. Pseud 201, sermone huius ira incendor; Liv. 1. 59. 11, his…memoratis incensam multitudinem perpulit ut…etc.

³⁸ cf. 1. 50, talia flammato secum (tea corde uolutans; 11. 376talibus exarsit dictis uiolentia Tumi.

³⁹ cf. De Orat. 3. 23; Quint. 4. 2. 75, peroratio incendit et plenos irae [iudices] feliquit; ibid. 114.

⁴⁰ 6. 2. 26 ff., esp. 28.

⁴¹ Speaking of his own prowess in the Orator, Cicero claims nulla me ingeni sed magna uis aninti inflammat, ut me ipse non teneam; nee umquam is qui audiret incenderetur, nisi ardens ad eum perueniret oratio (132).

⁴² Whenever Vergil has dicens (2. 550; 10. 744, 856; 12. 950) or dicente (10. 101), he means ‘even as X spoke’.

⁴³ My description of this second speech is only a precis of the article by Williams (op. cit. n. 32).

⁴⁴ Die Göttin Venus in Vergils Aeneis (Heidelberg, 1967), p. 87Google Scholar.

⁴⁵ Further discussion of the lack of intimacy between Aeneas and Venus in Highet, pp. 37 f.

⁴⁶ And to Orpheus by Eurydice's ghost (G. 4. 497).

⁴⁷ P. 404.

⁴⁸ pp. 22 ff.

⁴⁹ I might dispense with this section if there were an appropriate discussion to which I could refer; but I do not know of one. Indeed, there is surprisingly little written on the subject altogether, as Latacz, J. remarks in the introduction to his useful survey, ‘Zur Forschungsarbeit an den direkten Reden bei Homer (1850–1970): ein kritischer Literatur-Überblick’, Grazer Beiträge 3 (1975), 395Google Scholar.

⁵⁰ The modification of Achilles' position is well traced by Whitman, C., Homer and the Heroic Tradition (Harvard, 1958), p. 190CrossRefGoogle Scholar; cf. Lesky, A., ‘Homeros’, RE Supp. 11, pp. 790 fGoogle Scholar.

⁵¹ cf. Walcot, P., ‘Odysseus and the art of lying’, Anc. Soc. 8 (1977), 1–19Google Scholar; Stanford, W. B., The Ulysses Theme ² (Oxford, 1963), pp. 13 ffGoogle Scholar.

⁵² Griffin, J., Homer on Life and Death (Oxford, 1980), p. 80Google Scholar.

⁵³ op. cit. n. 51, pp. 56 ff.

⁵⁴ cf. Griffin, , op. cit. n. 52, p. 69Google Scholar.

⁵⁵ ibid.

⁵⁶ cf. 18. 330 f.; 19. 334 f.; 23. 144 ff.

⁵⁷ The allusion to Priam's speech must, as Colin Macleod puts it in his commentary, refer to things said while they ate; it cannot refer to the moment, because as they look at each other they are silent. This is plain from the lines immediately following: ‘when they had had their fill of looking at each other, then god-like Priam was the first one to speak’ (632–4).

⁵⁸ Griffin's observation is too bleak: ‘No Egyptian drug can obliterate the sufferings of the Iliad, for which there is no alleviation and the gods can only recommend endurance…” (op. cit. n. 52, p. 69 n. 36). Men can do better than this in the Iliad, even if their sufferings are not ‘obliterated’. Contrast Priam's mad grief before the reconciliation scene (24. 162 ff.) with his self-possession afterwards (713 ff., 777 ff.), a self-possession which is itself in strong contrast with the unrestrained passion of the mourning women.

⁵⁹ In attempting to put Vergil's speeches in perspective, I am afraid I may have put Homer's speeches out of perspective. There is certainly unsuccessful speech in Homer; the first exchanges of words in the Iliad are disasters. But it is Homer's norm which is important (if I may borrow the terminology of Johnson, W. R., Darkness Visible. A Study of Vergil's Aeneid (University of California, 1976), chapter II)CrossRefGoogle Scholar.

⁶⁰ Highet, pp. 23 f. Of these 135, as Highet notes, eight are soliloquies, and there are a few more deductions to be made: see p. 24.

⁶¹ pp. 405 ff.

⁶² cf. Grimm, R. E., ‘Aeneas and Andromache in Aeneid 3, AJPh 88 (1967), 151–62Google Scholar. Grimm analyses well the lack of contact between the two, although I cannot accept his conclusions.

⁶³ No speech between Helenus and Andromache in Book 3. Amata speaks once to Latinus (7.359–72), but he does not answer; at 12.10 ff. Latinus and Amata are both present, with Turnus and Lavinia, but do not address each other.

⁶⁴ Aeneas and Anchises only converse in the Underworld. Evander speaks once to his son alive (…dum te, care puer, mea sola et sera uoluptas, / complexu teneo...etc., 8. 581 ff.), and once to his corpse (11. 152 ff.). Pallas never speaks to his father.

⁶⁵ Apart from the ‘unreal’ conversation between Aeneas and the disguised goddess in Book 1 (321 ff). At 8. 612 ff. she speaks briefly to Aeneas, but he makes no reply.

⁶⁶ op. cit. n. 59, p. 179. See also Lieberg, G., ‘Vergils Aeneis als Dichtung der Einsamkeit’, in Vergiliana, ed. Bardon, H. and Verdiere, R. (Leiden, 1971), pp. 175–91Google Scholar.

⁶⁷ Earlier in Book 2, when Aeneas and Creusa are together in Anchises' house, she speaks once to him, without receiving a reply (657 f.).

⁶⁸ ‘[Aeneas] is never heard saying any special words of love to Dido, as Paris does to Helen (Il. 3. 438–46); nor is he ever seen embracing her, like Odysseus with Penelope (Od. 23. 231–40)’, Highet, p. 35.

⁶⁹ ‘Nicht Rede und Gegenrede’, says Heinze, (p. 410)Google Scholar of the bursts of speech in Anchises' house.

⁷⁰ One line of reported speech, pared and bald: Anchisen facio cerium remque ordine pando (179).

⁷¹ cf. Austin's introductory remarks on 679–702 (ed. A. 6); Highet, p. 34. Lieberg, art. cit. n. 66, 189, has some good comments on the emptiness of contact between Aeneas and Anchises at this point.

⁷² See esp. 1. 646; 4. 234, 274–6, 354 f.; 12. 436 f.

⁷³ Except for Books 8 and 9, when Aeneas is at Pallanteum without Ascanius.

⁷⁴ Il. 6. 476–81.

⁷⁵ Soph. Aj. 550 f., or Accius, Armorum ludicium fr. 10 Ribbeck.

⁷⁶ This is so even if we accept the relocation of Ajax's speech in Accius, as suggested by Jocelyn, H. D., CQ n.s. 15 (1965), 128CrossRefGoogle Scholar.

⁷⁷ Highet, pp. 41 f.

⁷⁸ On Aeneas' solitude, see Lieberg, art. cit. n. 66, 176 ff.; Elftmann, G., ‘Aeneas in his prime: distinctions in age and the loneliness of adulthood in Vergil's Aeneid’, Arethusa 12 (1979), 175–202Google Scholar; especially Liebing, H., Die Aeneasgestalt bei Vergil (Diss. Kiel, 1953), pp. 22, 95, 110, 152Google Scholar.

⁷⁹ Highet, p. 39.

⁸⁰ The second longest he makes.

⁸¹ cf. Quinn, , op. cit. n. 10, p. 48Google Scholar: ‘One of the functions of Book V is to heal this alienation of commander from his men.’

⁸² His third longest speech in the poem.

⁸³ A similar perfervid atmosphere prevails at the beginning of Book 12(10–80), when we hear speeches from Turnus, Latinus, Turnus, Amata, Turnus. The appeals of Latinus and Amata, so far from inducing Turnus to give up the war, drive him into an even greater frenzy.

⁸⁴ talibus insidiis periurique arte Sinonis / credita res, captique dolis lacrimisque coactis / quos neque Tydides nee Larisaeus Achilles, / non anni domuere decent, non mille carinae (195–98).

⁸⁵ See Heinze, pp. 421–4; Highet, pp. 285–90. On Anna's speech in Book 4 (31–53), see West, G. S., ‘Virgil's helpful sisters. Anna and Juturna in the Aeneid’, Vergilius 25 (1979), 10–19Google Scholar.

⁸⁶ p. 289.

⁸⁷ pp. 277–90.

⁸⁸ See Kennedy, G., The Art of Rhetoric in the Roman World (Princeton, 1972), pp. 390 ff.Google Scholar; Russell, D. A., Criticism in Antiquity (London, 1981), pp. 126–8Google Scholar.

⁸⁹ These words should be read with Aeneid 7, 11 and 12 in mind.

⁹⁰ See Kennedy, G., The Art ofPersuasion in Greece (Princeton, 1963), pp. 14 ff.Google Scholar; vonArnim, H., Leben und Werke des Dio von Prusa (Berlin, 1898), chap. 1, pp. 4–114Google Scholar, ‘Sophistik, Rhetorik, Philosophie in ihrem Kampf um die Jugendbildung’; Vicaire, P., Platon: Critique Litter aire (Paris, 1960), pp. 276 ff.Google Scholar; Hellwig, A., Untersuchungen zur Theorie der Rhetorik bei Platon und Aristoteles (Hypomnemata 38, Göttingen, 1973), pp. 304 ffGoogle Scholar.

⁹¹ Despite Highet's claims (pp. 287 ff.) that Aeneas distorts the truth like other speakers. His examples are ill founded. At, e.g., 1. 200 f., Aeneas tells his men, uos et Scyllaeam rabiem Tpenitusque.sonantis / accestis scopulos. ‘Aeneas speaks as though he and his men had actually braved the dangers of Scylla and Charybdis, whereas the narrative shows that they avoided them by sailing southward along the Sicilian coast…’ (Highet, p. 288). Aeneas in fact tells his men that they went close to Scylla and Charybdis (accestis, 1.201), and this is precisely what happened: laeuam cuncta cohors retnis uentisquepetiuit. / tollimur in caelum curuato gurgite, et idem / subdueta ad Manis imos desedimus unda…etc. (3. 563 ff.).

⁹² e.g. 5. 45–71, 8. 127–51, 11. 108–19.

⁹³ See Highet, pp. 39 f.

⁹⁴ He concludes, ‘For Vergil both Augustus and his prototype Aeneas were more godlike than human; and a god, as we know from Aristotle, cannot have human friends’ (p. 43). But his consequent remarks are more valuable: ‘Did he not also wish to show him as one who, after almost unendurable losses and sufferings, had grown into the melancholy of middle age and the grave contemplation of approaching death?’ (ibid.).

⁹⁵ ‘Mortality, morality and the public life. Aeneas the politician’, Ant. R. 32 (1972–1973), 649–64Google Scholar.

⁹⁶ 650.

⁹⁷ 659 f. A fine example is the worthy but flat oration with which Aeneas inaugurates the funeral games (5. 45—71); this is the ancient equivalent of the Cabinet Minister's speech at the opening of the bridge or factory.

⁹⁸ 660.

⁹⁹ p. 42. One suspects that it was the early part of Augustus' principate, when Vergil knew him, that provided the material for the conventional assessment of his comitas (Suet. Aug. 53.2 f., 66. 1–3, 74). Vergil did not see the Augustus who lived through the deaths of all his early friends, and of Gaius and Lucius, who saw the disgrace of daughter and granddaughter, and survived into ‘the atmosphere of gloom and repression that clouded the last decade of the reign’ (Syme, R., History in Ovid (Oxford, 1978), p. 205)Google Scholar.

¹⁰⁰ art. cit. n. 10, 21.

¹⁰¹ Following Ihm's Teubner text for [t]umentibus.

¹⁰² cf. Douglas, loc. cit. n. 100, ‘Vergil was dead before these events, but he knew all about dynastic marriages of convenience’.

¹⁰³ ‘Hero and theme in the Aeneid’, TAP A 94 (1963), 157–66Google Scholar.

¹⁰⁴ cf. Stewart, art. cit. n. 95, 651: ‘The Aeneid is a study of the preternatural strains and anxieties a political vocation brings to mere natural man.’

¹⁰⁵ But I cannot resist quoting the following passage from Suetonius, to which Prof. Nisbet referred me as being illustrative of Augustus' constrained and inhibited use of speech: ‘sermones quoque cum singulis atque etiam cum Liuia sua grauiores non nisi scriptos et e libello habebat, ne plus minusue loqueretur ex tempore’, Aug. 84. 2.

¹⁰⁶ loc. cit. n. 98.