2. No one can accuse me of rude haste in making this statement. The term was coined by Julia Kristeva in the mid-1960s. See, for example, her work ‘Word, Dialogue, and Novel’ published in 1967, most easily available to English readers in The Kristeva Reader, ed. by Moi, Toril and translated from the French by Hand, S. and Roudiez, L. S. (New York/Oxford, 1986)Google Scholar. For bibliography on intertextual approaches to Latin Literature, see Appendix 2. The first time that I tentatively tried out the term in print was in 1987. Age and Anglo-Saxon empiricism may make one cautious.

3. Cf. Camps quoted in Appendix 1 below.

4. The object of my attack is, of course, the hoary old ‘Intentional Fallacy’: an ugly and not perfectly apt phrase, but it would waste time and the reader's patience to try to rename it. The body of my text illustrates why I think it wrong to appeal to an author's intentions, and Appendix I contains further reflections on the topic.

5. I am of course self-consciously using the traditional language of Intentionalists.

6. Cf, e.g., Otis, B., Virgil: a Study in Civilized Poetry (Oxford, 1963), pp. 231f.Google Scholar, Clausen, W., ‘An Interpretation of the Aeneid’ in Virgil, a Collection of Critical Essays, ed. by Commager, S. (Englewood Cliffs, 1966), pp. 76ffGoogle Scholar., Lyne, , Further Voices in Vergil's Aeneid (Oxford, 1987), pp. 104ffGoogle Scholar.

7. I do not mean to be dismissive either of Call. fr. 100 Pf. or of Catull. 66 by this phrase. Both pieces are magnificent tours de force. Callimachus' conception is brilliantly original, and Catullus' translation is creative. But I would insist that the context and ethos of both poems is witty and trivial compared with the apparent high tragedy of the context of the Aeneid at this point. On Callimachus and on Catullus' version of Callimachus, see the excellent and succinct comments of Hutchinson, G. O., Hellenistic Poetry (Oxford, 1988), pp. 322–4Google Scholar; also the full discussion of Syndikus, H. P., Catull: eine Interpretation, Teil, Zweiter (Darmstadt, 1990), pp. 199ffGoogle Scholar. In his discussion of the lock's ‘pain at parting’ (lines 39–50), Syndikus, , no frivolous critic, is moved to the word ‘komisch’ (p. 210)Google Scholar. Hutchinson, (p. 322) states ‘In the second half of <Catullus'>; poem the fundamental conception is fantastic and delightful.’;+poem+the+fundamental+conception+is+fantastic+and+delightful.’>Google Scholar Commenting on lines 79–88, which he believes Catullus has added to his Callimachean original, he remarks (p. 323): ‘The strong moral language used of the unchaste makes the interplay with the fantasy the more preposterous.’ Others (cf., eg., Wiseman, T. P., Catullan Questions [Leicester, 1969], p. 21Google Scholar, Tatum, J., AJP 105 [1984], 443Google Scholar; this article is mentioned again below, n. 8) might not share these views. I do. ‘Trivial’ incidentally is the word that comes to E. L. Harrison's mind when discussing Catullus 66.39, and Vergil's, use of it (‘Cleverness in Virgilian Imitation’ [below, n. 8], 241 – p. 445)Google Scholar.

8. Harrison, , ‘Cleverness in Virgilian Imitation’, CPh 65 (1970), 241f.Google Scholar, with notes 2 and 3, reprinted in S.J. Harrison, Oxford Readings in Vergil's Aeneid (Oxford, 1990), pp. 445f. Thornton's, conclusion, AUMLA 17 (1962), 77–9Google Scholar, was independently arrived at by Wigodsky, M., Vergil and Early Latin Poetry (Wiesbaden, 1972; Hermes Einzelschriften Heft 24), p. 127Google Scholar: ‘Thornton is surely right to explain the echo as a deliberate allusion to Aeneas' destiny: he is ultimately to be enskied, like Berenice's lock, and it is to ensure this that the gods ordered him, unwilling and unaware, to leave Dido.’ This gets part of the truth: see below p. 192. Besides this, a sequence of articles in AJP has made some progress: Tatum, J., ‘Allusion and Interpretation in Aeneid 6.440–76’, AJP 105 (1984), 434–52, esp. 444Google Scholar; Skulsky, S., ‘“Inuitus, regina…”: Aeneas and the Love of Rome’, AJP 106 (1985), 447–55, esp. 451Google Scholar; Johnston, P. A., ‘Dido, Berenice, and Arsinoe: Aeneid 6.460’, AJP 108 (1987), 649–54Google Scholar.

9. Austin, ad loc. (who however does not share the view): ‘Modern susceptibilities are pained by Virgil's presumed indifference to the incongruity so produced, and suggest that his line is an unconscious reminiscence …’ Harrison, loc. tit: ‘Most commonly it is assumed that the reminiscence must be unconscious…’; he provides a bibliography of those who state such a view.

10. I confess that my slightly cautious phrasing in Further Voices in Vergil's Aeneid, p. 103 was leaving room for some sort of escape clause like this. I am slightly suspicious too when I read Conte, G. B. distinguishing between ‘modelo-antigrafo’ and ‘modello-codice’ (Il genere e i suoi confini p. 147Google Scholar; see Appendix 2), but I am not sure that I exactly understand C. at this point, so my suspicions may be unfounded. Cf. too Wigodsky, , op. cit, pp. 80–1Google Scholar on Vergil's borrowings from Pacuvius and Accius.

11. Honourable exceptions: Tatum, Skulsky, and Johnston in a 8 above.

12. Nothing too remarkable in this, however: fair hair conventionally belonged to heroes and heroines of legend (Fordyce on Catull. 64.62), nor was ‘the blond type in antiquity’ that uncommon according to the massive note of Pease on Aen. 4.590. But ‘flauus’ is not the only way to say ‘blonde’.

13. Lyne, , Further Voices in Vergil's Aeneid, pp. 18–27, 194–7Google Scholar.

14. Postgate's correction of the corrupt hi dii uen ibi.

15. On such ‘contrast similes’, see Porter, D. H., ‘Violent Juxtaposition in the Similes of the Iliad, CJ 68 (1972), 11–21Google Scholar; Macleod, C. W. (ed.), Homer, Iliad Book XXIV (Cambridge, 1982), pp. 48f.Google Scholar; Moulton, C., Similes in the Homeric Poems (Göttingen, 1977), p. 31Google Scholar; Lyne, , Words and the Poet: Characteristic Techniques of Style in Vergil's Aeneid (Oxford, 1989), pp. 135–40 and ffGoogle Scholar.

16. And there are further echoes. Cf, e.g., the two ‘whither shall I turn?’ passages, 4.534ff, 64.177ff, and the two imprecation passages, 4.608ff, 64.188ff.

17. Cf., e.g., Griffin, J. in Griffin, J., Boardman, J., Murray, O. (eds.), The Oxford History of the Classical World (Oxford, 1986), pp. 631–2Google Scholar; Lyne, , Further Voices in Vergil's Aeneid, p. 200Google Scholar (and the preceding pages).

18. Lyne, , Further Voices in Vergil's Aeneid, pp. 173–6Google Scholar.

19. Dido too cites it (4.492f, above p. 190), offering a different – a pathetic – irony.

20. Further Voices in Virgil's Aeneid, pp. 40–1. This was back in the days when I was using the term ‘allusion’.

21. There are vital pages on this topic in Hardie, P. R., Virgil's Aeneid: Cosmos and Imperium (Oxford, 1986), pp. 33–50, 158ffGoogle Scholar.

22. I was put on the track of this example by D. P. Fowler.

23. Characteristically Norden on Aen. 6.851 posits an Ennian source for both. This may be right. But the text whose ideology most recently and vividly attaches to the phrase must be the DRN.

24. For an appropriate emphasis on the fact that the Lucretian text has significance for the (then) present day see Fowler, D. P. in Griffin, M. and Barnes, J. (eds.), Philosophia Togata: Essays on Philosophy and Roman Society (Oxford, 1989), pp. 143–4Google Scholar.

25. See Lyne, , Further Voices in Vergil's Aeneid, pp. 214–15Google Scholar; but this quotation – and 6.847–50 as a whole – raise further and complex questions: cf. Hine, H., ‘Aeneas and the Arts (Vergil, Aeneid 6.847– 50)’ in Whitby, M. et al. (eds.), Homo Viator: Classical Essays for John Bramble (Bristol, 1987), pp. 173–83, esp. 179, 181–3Google Scholar.

26. Iliad 16.470, 22.362f., etc.

27. Cf. Eur, . Alc. 74ffGoogle Scholar. and Austin on Aen. 4.698. Cf. too Hor. Ode 1.28.30 with Nisbet-Hubbard, ad loc. Eur. Alc. is already an intertext (we are already – or contrasting – Alcestis with Dido) at Aen. 4.648ff.: cf. Alc. 175ff.

28. Note the elaboration of her curse at 4.607ff. with its promise of an ultor (Hannibal, ?) in 625f.Google Scholar; note too 8.18ff. with Lyne, , Further Voices in Vergil's Aeneid, pp. 125–32Google Scholar. We might observe, however, that at 4.379f. Dido produced this opinion: scilicet is superis labor est, ea cura quietos / sollicat; and her ironic point there was, as Servius saw, almost exactly the Epicurean first Kuria Doxa: ‘The Blessed and Immortal Nature knows no trouble itself nor causes trouble to any other…’ One might wonder how an Epicurean Dido could promise to pursue Aeneas as a ghost; one might even reckon that an Epicurean Dido should (as it were) endorse the Epicurean description of her death at 4.704f. But inconsistency in a character's views are less disturbing than subversions built into the narrative. There is no psychological reason, for example, why Dido should not genuinely believe she can persecute Aeneas after death and yet, for ironic rhetorical purposes, proffer her sneering version of the Epicurean Kuria Doxa at 4.379f. Cf. her sister at 4.34, ‘id cinerem aut manis credis curare sepultos?’: Epicurean stuff on the insensibility of the shades, for a rhetorical purpose.

29. Times Literary Supplement Jan. 19–25 (1990), 71.

30. jRS 80 (1990), 211Google Scholar.

31. I now use the more familiar, if not very pleasing term, as a shorthand. See my comments (above, pp. 187–9) on unjustified preconceptions about an author's intentions, and Appendix 1 below.

32. Actually, I would dispute that we know this sort of thing – always or even often – but I let it stand for the moment. Cf. the thoughtful comments of Taplin in n. 42 below.

33. Readers of Philip Larkin might like to call it the ‘Fallacy of the Comparability of the Sale and Movement of Poultry (Domestic) Act, 1943’. Cf. his interesting remarks in ‘Master's Voices’, Required Writing, Miscellaneous Pieces 1955–1982 (London, 1983), p. 137Google Scholar, very interesting since they issue from someone who appears mainly to be victim to the Intentional Fallacy. What I call the ‘Fallacy of Audience Limitation’ is distantly related to the ‘Affective Fallacy’ (see Wimsatt, W. K., The Verbal Icon [London, 1970], pp. 21ff.Google Scholar, and the entry in Preminger, A. et al. (eds.), The Princeton Encyclopedia of Poetry and Poetics [full ref. in Appendix 1], p. 8Google Scholar), but only distantly related.

34. I know that this sounds a bit pretentious. It also sounds more universally applicable than I perhaps intend; scholars will no doubt be able to think of instances where an audience's incapacities should be brought into the reckoning. I wish simply to get at a general (rather than, necessarily, an absolute) fact of literature, and to provoke thought A portentous and catchy label may therefore be in order.

35. Lebeck, , The Oresteia: a Study in Language and Structure (Washington D.C., 1971Google Scholar).

36. E.g., The Stagecraft of Aeschylus: the Dramatic Use of Exits and Entrances in Greek Tragedy (Oxford, 1977)Google Scholar; note, e.g., pp. 311–15.

37. Spurgeon, , Shakespeare's Imagery and What it Tells us (Cambridge, 1935)CrossRefGoogle Scholar.

38. Taplin, , The Stagecraft of Aeschylus, pp. 12–18Google Scholar.

39. See William Shakespeare: the Complete Works, ed. by Wells, S. and Taylor, G. (Oxford, 1986), pp. xxx–xxxiiiGoogle Scholar. The only works that Shakespeare himself seems to have cared about putting into print are the narrative poems Venus and Adonis and The Rape of Lucrece. The first efficient publication of his dramatic works was the First Folio, produced by John Heminges and Henry Condell in 1623, after Shakespeare's death. It is a sobering and revealing fact that, had it not been for their work, eighteen of Shakespeare's plays, including Tempest, Macbeth, and Antony and Cleopatra, might, for all Shakespeare knew or apparently cared, have vanished. Wells and Taylor conclude on p. xxxii: ‘John Marston, introducing the printed text of his play The Malcontent in 1604, wrote: “Only one thing afflicts me, to think that scenes invented merely to be spoken, should be enforcively published to be read.” Perhaps Shakespeare was similarly afflicted.’ Of the inaccurate Quartos produced in Shakespeare's lifetime, some at least were produced from the ‘foul papers’, another sign of Shakespeare's lack of interest in accurate textual survival: see Campbell, O. J. and Quinn, E. G., A Shakespeare Encyclopedia (London, 1967)Google Scholar, s.vv. ‘foul papers’, ‘Quartos’.

40. David, R. on Shakespeare in PBA 47 (1961), 158Google Scholar, quoted by Taplin, , op. cit, p. 18Google Scholar with approval. But Taplin in fact shows much greater subtlety of critical position later. See below.

41. The very title of Goldhill's, S.Reading Greek Tragedy (Cambridge, 1986CrossRefGoogle Scholar), is of course indicative.

42. Cf. my remarks below on ‘intention’ in Appendix 1. Cf. too the perceptive comments of Taplin, , op. cit., pp. 18fGoogle Scholar. After citing David with approval, he writes: ‘But these are negative cautions, and one should take care in pressing them. Especially one should be careful not to dogmatize too confidently about what an author could or could not put into his work, for there are many levels of creative consciousness besides clearly formulated deliberation; and [Taplin continues, interestingly] similarly with what an audience could or could not register during a work in performance, for there are many degrees of apprehension beside the full and conscious recognition which it is the critic's task to formulate.’ I, of course, would not accept that limitation on the critic's task in the last clause. But those who do not like to by-pass the presumed capacities of the immediate audience (as I do) and yet wish to argue for, say, patterns of symbolism in Aeschylus or intertextuality in Vergil could exploit Taplin's (true) statement that ‘there are many degrees of apprehension besides … full and conscious recognition’ in an audience's response to a play or poem.

43. In the Oxford Honour School of Literae Humaniores Latin Literature paper for 1990, there was the following question: ‘Can we ever know that a Roman writer wished his readers to notice an allusion to earlier literature?’ Radiating, as it does, the Intentional Fallacy and the Fallacy of Audience Limitation, it was presumably inviting the candidate to dispose of both.