¹ I have previously discussed this problem in ‘Defensor Vergilii: considerazioni su Richard Heinze’, my introduction to Heinze, R., La tecnica epica di Virgilio (Italian translation, Bologna 1996) 9 ffGoogle Scholar.

² The idea is now increasingly widespread among interpreters of Homer that ‘despite the lack of direct intervention by the primary narrator-focalizer, the Iliad is far from objective or impersonal: it is full of implicit colouring or focalizers’ (Taplin, O., Homeric soundings: the shaping of the Iliad (Oxford 1992) 52Google Scholar). Among the many analyses of the problems of the point of view and of narrative focalization in Homer, see e.g. Delrieu, A., Hilt, D., Letoublon, F., ‘Homère à plusieurs voix: les techniques narratives dans l'épopée grecque archaïque’ in Lalies: actes des sessions de linguistique et de littérature 4 (1984) 177–94Google Scholar; Scully, S., ‘Studies of narrative and speech in the Iliad’, Arethusa 19 (1986) 135–53Google Scholar; de Jong, Irene J.F., Narrators and focalizers. The presentation of the story in the Iliad (Amsterdam 1987)Google Scholar; Bremer, J. M., de Jong, I.J.F., Kalff, J. (edd.) Homer: beyond oral poetry, recent trends in Homeric interpretation (Amsterdam 1987)Google Scholar; Martin, R.P., The language of heroes. Speech and performance in the Iliad (Ithaca and London 1994) 235–7Google Scholar; Segal, C.P., Singers, heroes and gods in the Odyssey (Ithaca and London 1994) 125–6Google Scholar; Felson-Rubin, Nancy, Regarding Penelope: from character to poetics (Princeton 1994) passimGoogle Scholar. On the objectivity of Homer's narrative style see Bassett, S.E., The poetry of Homer (Berkeley 1938) chs. 2 and 3Google Scholar; Griffin, J., Homer on life and death (Oxford 1980) ch.4Google Scholar; Effe, B., ‘Epische Objectivität und authoriales Erzählen’, Gymnasium 90 (1983) 171–86Google Scholar.

³ The variant spes ampla (Pr) might find support from Prop. 3.22.41–2 hic ampla nepotum | spes et uenturae coniugis aptus amor (see Austin ad loc.).

⁴ On this and other examples of ‘deviant focalization’ see Fowler, D.P., ‘Deviant focalization in Virgil's Aeneid’ PCPS 36 (1990) 42–63Google Scholar.

⁵ Sometimes the pathetic amplification can occur even in a supplementary relative clause, like 2. 248–9 nos delubra deum miseri quibus ultimus esset | ille dies, festa uelumus fronde per urbem, or at 2.426–8 cadit et Rhipeus, iustissimus unus | qui fuit in Teucris et servantissimus aequi | (dis aliter uisum): the added thought integrates the utterance and intrudes as a bitter reflection on events. At other times it is the similes – like expanded narrative epithets – which become the source of pathetic intensification; they illustrate the narrative by glossing it with subjective and sentimental connotations. On this aspect of Virgilian similes, see e.g. West, D.A., ‘Multiple correspondence similes in the Aeneid’, JRS 59 (1969) 40–9Google Scholar (= Harrison, S.J. (ed.), Oxford readings in Vergil's Aeneid (Oxford and New York 1990) 429–44)Google Scholar; Williams, G., Technique and ideas in the Aeneid (New Haven and London 1983) 60–7Google Scholar; Lyne, R.O.A.M., Further voices in Vergil's Aeneid (Oxford 1987) 119–20Google Scholar; id., Words and the poet (Oxford 1989) 63–99.

⁶ In this respect the book by Irene de Jong (n. 2) is highly debatable and essentially inexact. The author, perhaps with the simple intention of provoking debate (which nonetheless stands in contrast to the attitude of cautious moderation that she maintains in other parts of the book) actually claims that the presentation of the narrative in the Iliad should be considered ‘subjective, engaged and emotional’ (227). Behind this claim lie a misunderstanding and a distortion: she confuses the narrative features of the contents (sorrowful events and feelings) with the narrative features of the form (the way of representing this sorrow). The fact that the external narrator (Homer) inclines to a narrative largely composed of dialogues should not be confused with a ‘subjective style’, which is quite another matter. I suspect the author is confusing the ‘primitivism’ of Homer as narrator (an illusion of late-Romantic criticism, now rightly rejected) with the detachment of the narrator from the reported events (a quite different thing, and the foundation of Homeric objective narrative). Even the emotions aroused in the reader by the narrative cannot be invoked as proof of an affective narrative; indeed it is possible that the poet narrates in order to arouse emotions, without indicating his own emotional participation and his own presence in the form of the text. That is just what Homer does. Nevertheless, one should not underestimate the closeness of narrator and audience: see Edwards, M.W., Commentary on Iliad 17–20 (Cambridge 1991) 2–7CrossRefGoogle Scholar.

⁷ Here I recapitulate some of the ideas and arguments that I expounded in The rhetoric of imitation. Genre and poetic memory in Virgil and other Latin poets, transl. from the Italian, edited by Segal, C. (Ithaca and London 1986) 152–61Google Scholar.

⁸ Cf. Enn., Ann. 291Google Scholar Vahl. Romanis Iuno coepit placata fauere (the reconstruction of the hexameter proposed by Hug, based on Serv. ad Aen. 1.281): compare also Vahlen's discussion pp. cxix ff.; Buchheit, V., Vergil über die Sendung Roms (Heidelberg 1963) 54 f., 144 fGoogle Scholar. The hostility of Juno had its first beginnings in the goddess's hostility towards Troy. It is likely that the motif of the goddess's appeasement did not have the same prominence in the Annals as in the Aeneid; but it is difficult to say how relevant it was to Naevius: cf. Skutsch, O.The Annals of Quintus Ennius (Oxford 1985) 46Google Scholar. An important contribution is Feeney, D.C., ‘The reconciliation of Juno’, CQ 34 (1984) 179–94CrossRefGoogle Scholar (= Harrison, S.J. (ed.) Oxford readings 339–62Google Scholar) where the abundant specific bibliography is discussed. See also the useful observations of Johnson, W.R., Darkness visible (Berkeley and Los Angeles, 1976) 124–7CrossRefGoogle Scholar.

⁹ Twenty years ago, with somewhat different critical interests, I defined the point of view as ‘the form of relationship that passes between the semantic system of the text and the represented universe’ and spoke about a ‘relationship of truth which binds the manner of representing to the object of the representation itself’. With these presuppositions I was able to define the point of view of individual characters as ‘the semantic position which each character occupies in the text’: cf. Conte, G.B., Virgilio: il genere e i suoi confini (Milan 1985 ²) 68 n. 10Google Scholar (= The rhetoric of imitation 155 n. 10).

¹⁰ Fundamental here is Ziegler, K., L'epos ellenistico. Un capitolo dimenticato della poesia greca (Italian transl., ed. de Martino, F., Bari 1988)Google Scholar. The original German edition was 1934. The Italian version has an instructive preface by M. Fantuzzi, who also offers a valuable catalogue raisonné of the epicists who cultivated the tradition of the (non-Callimachean) large-scale poem in the various forms of mythological, local and encomiastic historical epic from the age of Alexander to the last century before Christ (pp. lv–lxxxviii). One should also read Kroll, W., ‘Das historische Epos’, Sokrates 4 (1916) 1 ffGoogle Scholar., though it is less reliable on Naevius' historical epic. Most important is the recent study of Cameron, A., Callimachus and his critics (Princeton 1995) especially 263–302Google Scholar. The provocative stance of this learned and brilliant book (easy to guard against) is better at opening and reopening problems than at indicating peaceable solutions. On the influence of the Aitia on the Aeneid and Augustan poetry see George, E. V., Aeneid VIII and the Aitia of Callimachus (Leiden, 1974)Google Scholar and the useful compilation of Miller, J.F., Callimachus and the Augustan aitiological elegy, ANRW II 30.1 (Berlin 1981) 371–417Google Scholar, and Briggs, W. W., Virgil and the Hellenistic epic, ANRW II 31.2 (Berlin 1981) 948–84Google Scholar. On the Alexandrianism of Virgil in general see Clausen, W., Virgil's Aeneid and the tradition of Hellenistic poetry (Berkeley and Los Angeles 1987)Google Scholar.

¹¹ Att. l.16.15.

¹² Apollonius' poem is in four books, perhaps not accidentally, if we recall that Aristotle (Poetics 24) called the poems of the ancients too long and wanted to allow only those comparable in length to a tragic tetralogy: cf. Vahlen, , Beiträge 287Google Scholar.

¹³ Cf. Courtney, E., The fragmentary Latin poets (Oxford 1993) 192–200Google Scholar.

¹⁴ The crucial point where myth joins history in a vertiginous exchange of inverted perspectives is in book 6. There the mythical past contains and narrates the national history of Rome as a succession of future events, when Anchises in the Underworld reveals to Aeneas forthcoming history: quae postquam Anchises natum per singula duxit | incenditque animum famae uenientis amore | exim bella uiro memorat quae deinde gerenda (6.888–90). Fama ueniens – History – itself becomes child of myth, and inversely myth finds its validation by assimilating itself to past historical time. The tradition founded by Naevius and continued by Ennius renews itself here; the timing of events loses its linearity of development (the ‘beforehand’ of myth which in the sequence of the narrative should precede the ‘afterwards’ of history) to become a complex narrative plot.

¹⁵ The classic article of Williams, R.D., ‘The purpose of the Aeneid’. Antichthon 1 (1967) 27–41CrossRefGoogle Scholar (= S.J. Harrison, op.cit. 21–36) is still very helpful.

¹⁶ Cf. Barchiesi, A, La traccia del modello. Effetti omerici nella narrazione virgiliana (Pisa 1984) 93–5Google Scholar.

¹⁷ Here I repeat the substance of what I wrote in ‘“Rhetoric of imitation” as rhetoric of culture’, Vergilius 38 (1992), 45–55Google Scholar.

¹⁸ In this connection we cannot fail to recall the famous remarks of that great romance philologist Contini, Gianfranco, ‘Dante come personaggio–poeta della “Commedia”’, in Contini, G., Varianti e altra linguistica: una raccolta di saggi (1938–1968) (Turin 1970) 335–61Google Scholar.

¹⁹ See on this whole question the excellent work of Fantuzzi, M., Ricerche su Apollonio Rodio. Diacronie della dizione epica (Rome 1988)Google Scholar.

²⁰ ‘A proposito dei modelli in letteratura”, MD 6 (1981) 148 fGoogle Scholar.

²¹ On Virgilian formularity Moskalew, W., Formular language and poetic design in the Aeneid (Leiden 1982)Google Scholar is accurate and often acute in interpretation.

²² Certain sequences can be repeated even in Virgil as quasi fixed structures, especially in the repetition of typical scenes (speeches, battles, deaths of heroes, conventionalized similes, etc.). I noted in Memoria dei poeti e sistema lelterario (Turin 1985 ²) 42 n. 26Google Scholar, that at times it is a matter of short cola of one or two words, at other times of hemistichs and even whole lines or pairs of lines (e.g. Aen. 10. 745 f. = 12.309 f.). In a way distinct from that of the Homeric poems, their function is not connected with modular requirements of the oral composition, but is a procedure aiming to restore a mark of style. Indeed it sometimes seems that Virgil is afraid of the presence in his modern poetry of an unmotivated formular composition and seeks somehow to give even this procedure a function in the new dramatic purpose of his poetic art. Thus certain details are repeated, and motivated, each reclaiming the other as internal to the work, as samples of a strong consciousness of unity, as structural articulations that set the action in motion again or signal a significant reversal of the narrative situation. The case of 1.100–1 is exemplary: … ubi tot Simois correpta sub undis | scuta uirum galeasque et fortia corpora uoluit (‘where the waters of Simois will turn over so many bodies of heroes and shields and helmets’) compared with 8.538–40) quas poenas mihi Turne dabis! quam multa sub undas | scuta virum galeasque et fortia corpora uolues | Thybri pater! (‘how you will have to pay for your savagery at my hands. Turnus! What brave bodies of heroes and shields and helmets your waters will turn over, O father Tiber!’) Cf. Iliad 12.22 ff. … (‘The Simois, where many shields and crests will fall in the dust and the race of heroes half-divine’). In the two Virgilian passages the verse is not repeated with a stereotypical formulaic function: it is enough to see how the second context is constructed as a mirror image and reversal of the first. Here is the first context: in 1.90 a threatening thunderclap explodes in the sky during the storm with which Aeolus is trying to destroy the Trojan fleet; in 94–9 follow the desperate laments of Aeneas, who calls upon the bravest of his Greek opponents (o Danaum fortissime gentis | Tydide) as the man who should have killed him; thus Aeneas would have died as a warrior, in battle on the plains of Troy, where (100 ff.) ‘the Simois overturns the corpses of so many heroes’. In the second case, in perfect parallel, the same triadic structure returns, but reversed as a positive sign. Now Aeneas – such is the will of Fate – is not a defeated man as he was then, but prepares to become victor. In 523–9 the thunder explodes in a clear sky (this is the signum faustum of divine favour); then there follow his words of triumphant defiance, calling on Turnus, chief of the new enemies (Turne dabis) as responsible for the many deaths which the Tiber, a new Simois, (but as a benevolent father) ‘will overturn in his waters’ (538–9). The mechanical formula of the inner ear has become an intratextual allusion. And the new Aeneas is to be found entirely in the suggested opposition.

²³ Cf. Conte, G.B, ‘Fra ripetizione e imitazione. Virgilio Eneide 10.24’, RFIC 111 (1983) 150–57Google Scholar.

²⁴ The Aeneid even includes a sort of ‘repetition’ of Homer. For example, the war in Latium between Trojans and Italians is often seen as a repetition of the war of Troy (see e.g. 6.89, 538 ff., compared with 1.100 ff. on which see note 22 above). But this is not a passive mirroring: at the beginning the Trojans are besieged and near death, as if they were condemned to their unfortunate destiny; at the end, however the Trojans are victors, and Aeneas kills Turnus, as Achilles killed Hector in the Iliad. In the new ‘Iliad’ the Trojans are destined to win. One can see that the repetition is also a surpassing of Homer; war, with all its struggles and suffering, will not lead to destruction, but to the construction of a new unity. In the end Aeneas reclaims in himself the image of the victorious Achilles and above all that of Odysseus who conquers his native land after so many ordeals, and restores peace there. For an acute and innovative consideration of the repetition of Homer in the Aeneid see Quint, D., ‘Repetition and ideology in the Aeneid’, MD 23 (1989) 9–54Google Scholar (= Epic and empire (Princeton 1993) 50–96Google ScholarPubMed).

²⁵ The discovery of a tragic dimension converges with the most vital thread of Virgilian criticism in this century. In his famous study Virgils epische Technik (Leipzig and Berlin 1915 ³) 321ff., 467 ffGoogle ScholarPubMed. (= Virgil's epic technique, transl. Harvey, H. and Harvey, D. and Robertson, F. (Bristol, 1993) 252 ff., 370 ffGoogle Scholar.), Richard Heinze identified the presence of the tragic element with the concrete influence of the great Athenian dramatists, but also introduced the fruitful idea that the entire narrative technique of the Aeneid was in fact a dramatization of Homeric material, not only in the compression with which Virgil conducted the action, impressing upon it a strong progressive dynamism, not only in that the narration proceeds articulated by scenes, but also because from the beginning of the poem strong tensions build themselves into the poem which demand solutions and thus instal into the poem a teleological movement dramatic in type. Other more historically motivated critics have identified the tragic feeling of the poem with the sympathy turned on the defeated and the irresolute (Aeneas) which would be a sort of tearing of the Augustan ideological veil. Anyone who like Pöschl, V., Die Dichtkunst Virgils (Berlin and New York 1977 ³) 39CrossRefGoogle Scholar (= The art of Vergil: image and symbol in the Aeneid (Ann Arbor 1962))Google Scholar, sees the unifying clement of the poem in its tragic component, makes it consist in the opposition between the sense of history (Römertum) and the eternal toil of the individual, a price paid in the historical and political dimension of the Aeneid. The war in Latium is tragic because it is virtually a ‘civil war’ (128 f., a very happy formulation); this tendency finds itself in agreement with Augustan policy, concerned with re-evaluating the Italian (as we see there is much compromise in this solution). In German studies of the post-war period the concept of the tragic is analysed from a discouraging generic aspect (bibliography in Wlosok, A.. ‘Vergils Dido-Tragödie. Ein Beitrag zum Problem des Tragischen in der Aeneis’ in Görgemanns, H. and Schmidt, E.A. (eds.), Studien zum Antiken Epos (Meisenheim am Glan 1976) 228–50Google Scholar). There is little gained, I think, by refining the concept through a pedantic application of the categories, themselves quite disputed, as articulated in Aristotle's Poetics. On these tendencies see the above-mentioned work of Wlosok, and also von Albrecht, M., ‘Zur Tragik von Virgils Turnusgestalt’ in von Albrecht, M. and Heck, E. (eds.), Silvae: Festschrift E. Zinn (Tübingen 1970) 1–5Google Scholar: he argues that the Aeneid would pass through all the phases of the Aristotelian definition to dissipate itself in an unspecified Metatragisches that would be peculiarly Roman.

²⁶ Aen. 12.930–1 ille humilis supplex oculos dextramque precantem | protendens. See the subtle note of Traina, A. in his commentary on Aen. 12Google Scholar (Virgilio: L'utopia e la storia (Turin, 1997) 185Google Scholar): ‘humilis “from the ground”, takes up with a reversed direction from low to high, ad terram of 927, but the etymological denotation (humilis … tractus est ab humo, Serv. Dan. ad Aen 4.255) doubles itself with typical Virgilian polysemy, by a psychic connotation, activated by supplex; the proud one “humiliates” himself to supplicate. Thus the reversal is achieved, as is the redemption of Turnus' ethos, which will make Aeneas' resolution more dramatic and conflictual.’

²⁷ Cf. Cicero, , De invent. 2. 65–6Google Scholarnaturae quidem ius esse, quod nobis non opinio, sed quaedam innata uis adferat, ut religionem, pietatem, gratiam, uindicationem, obseruantiam, ueritatem; [ … ] uindicationem per quam uim et contumeliam defendendo aut ulciscendo propulsamus a nobis et a nostris, qui nobis cari esse debent, et per quam peccata punimus.

²⁸ See the fine interpretation of Barchiesi, A., ‘Il lamento di Giuturna’, MD 1 (1978) 99–121Google Scholar. See also Lyne, R.O.A.M., Further voices in Vergil's Aeneid (Oxford 1987) 86–7, 139–44Google Scholar; O'Hara, J.J., Death and the optimistic prophecy in Vergil's Aeneid (Princeton 1990) 114–16Google Scholar; Perkell, C., ‘The lament of Juturna: pathos and interpretation in the Aeneid’, TAPA 127 (1997) 257–86Google Scholar.

²⁹ Note the brilliant observations on the killing of Turnus as a sacrificial act in Hardie, P., The epic successors of Virgil: a study in the dynamics of a tradition (Cambridge 1993) 32–5Google Scholar. On the theme of sacrifice in the Aeneid see also Bandera, C., ‘Sacrificial levels in Virgil's Aeneid’, Arethusa 14 (1981) 217–39Google Scholar; J.J. O'Hara, op. cit. 19–24, 28–35, 82–4, 106–10.

³⁰ Heinze defined the first as ‘Subjektivität’, the latter as ‘Empfindung’. Klingner, F., Virgil: Bucolica Georgica Aeneis (Zurich and Stuttgart, 1967) 434, 451Google Scholar and passim, talks of ‘lyric insertions’ to indicate ‘Subjektivität’, a rather inexact definition, which reveals its idealistic origins, dominated by the generic definition of all poetry of emotion (indeed of all poetry) as ‘lyric’. Otis, B. followed this path, in Virgil: a study in civilized poetry (Oxford 1963) 49–52, 61–92Google Scholar and passim, and rendered ‘Subjektivität’ by ‘sympathy’, and ‘Empfindung’ as ‘empathy’. The term empathy draws on the idea of identification with the thing represented which has had so much importance in studies of the psychology of art (see e.g. Wörringer, W., Astrazione e empatia (Italian transl., Turin 1975)Google Scholar. The problem is discussed more fully in my Virgilio: il genere e i suoi confini, 80–6.

³¹ Hunter, R.L., The Argonautica of Apollonius Rhodius: literary studies (Cambridge 1993) 172CrossRefGoogle Scholar.

³² Cf. Faedo, L., ‘L'inversione del rapporto poeta–Musa nella cultura ellenistica’, Annali Scuola Normale Sup. di Pisa 39 (1970) 377–86Google Scholar.

³³ On point of view in Apollonius see e.g. Fusillo, M., Il tempo delle Argonautiche: un' analisi del racconto in Apollonio Radio (Rome 1985) 347–59Google Scholar; Knight, V., The renewal of epic: responses to Homer in the Argonautica of Apollonius (Leiden 1995)Google Scholar index s.v. ‘focalization’. On the relative incidence of direct speech and of poetic intervention in Apollonius see Farrell, J., ‘Towards a rhetoric of (Roman?) epic’ in Dominik, W. J. (ed.) Roman eloquence (London 1997)Google Scholar.

³⁴ On authorial interventions in Virgilian narration see Williams, G., Technique and ideas in Virgil's Aeneid (New Haven and London 1983)Google Scholar ch. 7 ‘The poet's voice’.

³⁵ Cf. Lefèvre, E., ‘Dido und Aias. Ein Beitrag zur römischen Tragödie’, Abhandlungen der Akad. derWissensch. in Mainz (1978)Google Scholar no. 2.

³⁶ Enn., Scaen. 246–54Google Scholar Vahlen (= 208–18 Jocelyn) utinam ne in nemore Pelio securibus | caesa accidisset abiegna ad terram trabes … (but see Jocelyn's comments ad loc).

³⁷ Cat. 64.171–2 utinam ne tempore primo | Gnosia Cecropiae tetigissent litora puppes.

³⁸ See Bonfanti, M., Punto di vista e modi della narrazione nell'Eneide, Biblioteca di MD 3 (Pisa 1985) 46–51Google Scholar.

³⁹ Cf. Conte, G.B., ‘Verso una nuova esegesi virgiliana: revisioni e propositi’, in Virgilio: il genere e i suoi confini, 143Google Scholar.

⁴⁰ As I tried to show in Virgilio: il genere e i suoi confini 87–96 (= The rhetoric of imitation 173–82) the figure of Aeneas has a double literary status in the text. Like every other character Aeneas has a personal and limited point of view, but he is also the representative of the will of Fate, executor of that cosmic will which suffocates the rights of individuals (and so also the subjective rights of Aeneas). In short Aeneas fills two functions in the text: he is a character, possessor of a personal voice and a relative truth, but also the protagonist, bearer of an absolute and objective truth. Thus the vehicle of Fate's constructive will sometimes shows his personal features (as a character) marked by uncertainty and doubt, but he is also forced to shed his identity and repress his own subjectivity to take on his mission instead. As a character Aeneas can recover his subjective personality only by seeking a different dimension from that of Fate, but to do this he has to withdraw and turn back to the past. If he attempted to look in the future for paths other than those imposed by Fate, he would betray his own elect status, and merely live a life like his own memories, something only possible before he was called from on high. See Aen. 4.340–44 me si fata meis paterentur ducere uitam | auspiciis et sponte mea componere curas, | urbem Troianam primumque dulcisque meorum | reliquias colerem, Priami tecta alta manerent, | et recidiua manu posuissem Pergama uictis. Unlike the other characters of the poem in the mission that has fallen upon him (and in this respect close to Fate and the omniscient narrator), Aeneas is not unlike them in the defeat which his personal ‘I’ must undergo.

My warmest thanks go to Elaine Fantham for once again translating my original and to Philip Hardie and the referee for their helpful comments.