¹ The oral tradition of this paper goes back to the Annual General Meeting of the Classical Association in Canterbury in April 1990 (‘Homer and the Mythology Game’) and to the West Midlands Classical Seminar in February 1994. I am grateful for advice and correction given by participants at both and to Professor C.D.N. Costa for his advice and encouragement. In addition I have benefited to a very considerable extent from the advice of editors and referees and, above all, Ahuvia Kahane. The following items of bibliography are referred to frequently or need to be grouped for clarity:

Bethe, E., ‘Homer und die Heldensage: die Sage vom troischen Kriege’, NJ vii (1901) 657–76.Google Scholar

Bethe, E., Homer: Dichtung und Sage, vol. i ‘Ilias’ (Leipzig-Berlin 1914)Google Scholar; vol. ii ‘Odyssee, Kyklos, Zeitbestimmung’ (1922); vol. iii ‘Die Sage vom Troischen Kriege’ (1927).

Clark, M.E., ‘Neoanalysis: a biblographical review’, CW lxxix (1986) 379–94.Google Scholar

Davies, M., The Epic Cycle (Bristol 1989).Google Scholar

Else, G.F., Homer and the Homeric problem [Lectures in memory of Louise Taft Semple] (Cincinnati 1965).Google Scholar

Erbse, H., ‘Nestor und Antilochos bei Homer und Arktinos’, Hermes cxxi (1993) 385–403.Google Scholar

Foley, J.M., Immanent art: from structure to meaning in traditional oral epic (Bloomington-Indianapolis 1991).Google Scholar

Griffin, J., ‘The Epic Cycle and the uniqueness of homer’, JHS xcvii (1977) 39–53.CrossRefGoogle Scholar

Griffin, J., Homer on life and death (Oxford 1980).Google Scholar

Heubeck, A., Die Homerische Frage [Erträge der Forschung xxvii] (Darmstadt 1974).Google Scholar

A. Heubeck, ‘Homeric studies today: results and prospects’, in Fenik 1978: 1–17.

Janko, R., The lliad: a commentary, vol. iv ‘books 13–16’ (Cambridge 1992).Google Scholar

Jensen, M.S., The Homeric Question and the oral-formulaic theory (Copenhagen 1980).Google Scholar

Kakridis, J.T., Homeric researches (Lund 1949).Google Scholar

Kirk, G.S., The songs of Homer (Cambridge 1962).Google Scholar

Kirk, G.S., The lliad: a commentary, vol. i ‘books 1–4’ (Cambridge 1985).CrossRefGoogle Scholar

Kullmann, W., Die Quellen der llias (troischer Sagenkreis) [Hermes Einzelschrift xiv] (Wiesbaden 1960).Google Scholar

Kullmann, W., ‘Zur Methode der Neoanalyse in der Homerforschung’ WS xv (1981) 5–42.Google Scholar

Lord, A.B., The Singer of tales (Cambridge, Mass. 1960).Google Scholar

Lord, A.B., ‘Homer as oral poet’, HSCP lxxii (1967) 1–46.Google Scholar

Lord, A.B., Epic singers and oral tradition (Ithaca 1991) [ch.2, pp. 38–47Google Scholar reprints ‘Homer's originality: oral dictated texts’, TAPA lxxxiv (1953) 124–33].

Merkelbach, R., ‘Die pisistratische Redaktion der homerischen Gedichte’, RhM xcv (1952) 23–47.Google Scholar

Nagy, G., Greek mythology and poetics (Ithaca 1990) 23.Google Scholar

Notopoulos, J.A., ‘Studies in early Greek oral poetry’, HSCP lxviii (1964) 1–77.Google Scholar

Pestalozzi, H., Die Achilleis als Quelle der Ilias (Erlenbach 1945).Google Scholar

Richardson, N.J., The lliad: a commentary, vol. vi ‘books 21–24’ (Cambridge 1993).Google Scholar

Schadewaldt, W., Von Homers Welt und Werk ² (Stuttgart 1951).Google Scholar

von Scheliha, R., Patroklos: Gedanken über Homers Dichtung und Gestalten (Basel 1943).Google Scholar

Schoeck, G., llias und Aithiopis: Kyklische Motive in homerische Brechung (Zürich 1961).Google Scholar

Taplin, O., Homeric soundings (Oxford 1992).Google Scholar

Young, D., ‘Never blotted a line? formula and premeditation in Homer and Hesiod’, Arion vi (1967) 279–324.Google Scholar

² Fixed preservation in other traditions, e.g. Finnegan, R., Literacy and orality: studies in the technology of communication (Oxford 1988) 95Google Scholar; ead., Oral poetry: its nature, significance and social context ² (Bloomington 1992) 73–8. Nagy 40–3. Yet the assumption that fixity and writing are the same thing remains prevalent, e.g. Ford, A., Homer: the poetry of the past (Ithaca 1992) 132.Google Scholar

³ Lesky, A., ‘Homeros’, RE Suppl. xi (1968) 687–846, at p. 706Google Scholar views the architecture in particular as ruling out ‘oral improvisation’ though ‘improvisation’ is not quite the right term.

⁴ Kirk 1962: 99. Economics, Jensen 94. Other objections to writing: Kirk 1985: 13.

⁵ It would have to be invented if it did not exist, cf. Merkelbach 40.

⁶ Observed by Bethe 1922: 202, who concludes (241) ‘Ein selbständiges Werk hatten sie niemals sein sollen.’ For the question, ‘why was Chryseis at Thebe?’, obligingly answered by the Kypria, see Taplin 85 & n. 4.

⁷ Davies 4. The opening of the poem appears to have been stylistically late—Davies 3, referring to Wackernagel, J., Sprachliche Untersuchungen zu Homer (Göttingen 1916) 1–159.Google Scholar

⁸ Notopoulos 34 f.; Bethe 1922: 243: ‘Die Aithiopis, d.h. Memnons Aristie und Tod, liegt in einheitlicher Überlieferung vor. Sie bildet ein geschlossenen Ring, ein selbständiger Gedicht.’ Insufficient allowance for differences between Cyclic epics in Davies 4 f. Dating of the Aithiopis: 775 or 760 according to Eusebios Chron. (ed. Helm, R.GCS xlvii, Berlin 1956Google Scholar); born c. 744 and a ‘pupil of Homer’ according to Artemon of Klazomenai (4th century BC?) FGrH 443F2.

⁹ Severyns, A., Le Cycle épique dans l'école d'Aristarque (Liége-Paris 1928) 322.Google Scholar

¹⁰ The date of Eugammon may indeed, as often thought (e.g. Davies 6), simply be based on a decent interval after the foundation of Cyrene (c. 630), but, if so, that is a perfectly reasonable ground for the dating, not a refutation of it. It is fair to assume that Eugammon belongs to the last generation of poets before the Peisistratean recension.

¹¹ It is, however, somewhat disturbing that a Telegony is also ascribed to Kinaithon of Sparta, which would take the story much further back. For supposed influence of the Telegony on our Odyssey, see Davies 87 f.

¹² See Notopoulos, esp. 18–45.

¹³ Foley 8.

¹⁴ One feels the influence of Romanticism in this formulation and it becomes clear that Homer and Vergil are indeed not poles apart. Vergil is rather an oral poet in his closeness to tradition, his focus on recitation, and his tendency to sound patterns.

¹⁵ Lord, unpublished, cited by Foley 11.

¹⁶ Similarly, Young 306.

¹⁷ See Goold, G.P., ‘The nature of Homeric composition’, ICS ii (1977) 1–34Google Scholar, esp. pp. 9, 10–12, 26–30 for the points cited here.

¹⁸ I am realigning Notopoulos's argument (11 f.). The consensus is, however, firmly against the book-division being Homer's own, and gently in favour of its being Alexandrian (Taplin 285; Richardson 20 f.). The ultimate reason is less the apparent absence of book divisions from papyri (in fact there are some signs of recognition of book divisions, West, S., The Ptolemaic Papyri of Homer [Köln-Opladen 1967] 22 f.)CrossRefGoogle Scholar than the absence of any awareness of book divisions in any author before the fourth century BC (West 18). But West is now inclined to push the book division back to the Peisistratean recension; and Jensen (87) is ‘inclined to interpret the arrangement of each poem into twenty-four songs as resulting from the process of dictation’—making the scribe responsible for the assignation of one letter of the ‘Ionic alphabet’ to each book (at the Peisistratean recension, on Jensen's view). Larger units: an anonymous referee of this article comments: ‘The Odyssey falls so neatly into six nearly equal parts (each of four present-day books, except that the third part ends at 13.92), that I find irresistible the inference that it was composed specifically with a view to performance in six instalments.’

¹⁹ The comparison of an anonymous referee.

²⁰ Nagy 38.

²¹ This is particularly well shown by Kirk 1962: 280 f., who finds himself driven by an entirely reasonable argument to the desperate solution of a genius-Homer defying normal performance conditions. Kirk 1985: 12 talks of its having ‘been performed in a special way at which we can only guess’ rejecting as unlikely that it ‘was never intended to be heard as a whole’. The suggestion of assistants (or sons/apprentices) comes from an anonymous referee and would somewhat recall the recitals of Vergil. Ford (n. 2) 133 also arrives at this position, speaking of ‘a still largely illiterate age in which they would have been rarely read and nearly impossible to perform in toto’.

²² Lloyd-Jones, H., ‘Remarks on the Homeric question’ in: Greek epic, lyric and tragedy: the academic papers of Sir Hugh Lloyd-Jones (Oxford 1990)Google Scholar ch. I [reprinting Lloyd-Jones, H., Pearl, V., and Worden, B. (eds.), History and imagination: essays in honours of H.R. Trevor-Roper (London 1981) 15–29], at p. 3.Google Scholar

²³ Foley 12 on the Pabuji epic in Rajasthan and on the Mwindo tradition, citing this particular passage from Biebuyck, D., The Mwindo epic from the Banyanga (Berkeley 1969) 14.Google Scholar

²⁴ Dowden, K., The uses of Greek mythology (London 1992) 7–9.CrossRefGoogle Scholar

²⁵ Hatto, A.T., ‘Towards an anatomy of heroic/epic poetry’, in: Hainsworth, J.B. (ed.), Traditions of heroic and epic poetry, vol. ii ‘Characteristics and techniques’, (London 1989) 145–306, at pp. 147 f.Google Scholar

²⁶ Kullmann 1960: 12 f.

²⁷ Smith, J.D., ‘How to sing a tale: Epic performance in the Pabuji tradition’ in: Hainsworth, J.B. (ed.), Traditions of heroic and epic poetry, vol. 2 ‘Characteristics and techniques’ (London 1989) 29–41, at p. 36Google Scholar; Lord 1987: 67; Lord 1960: 99. Cf. Parry, Milman, The making of Homeric verse: the collected papers of Milman Parry, ed. Parry, A. (Oxford 1971) 446Google Scholar on ‘stable or essential’ themes.

²⁸ Finnegan 1992 (n. 2) 174. Lord 1960: 151.

²⁹ Lord 1960: 159, Young 305.

³⁰ Slatkin, L.M., The power of Thetis: allusion and interpretation in the Iliad (Berkeley 1991) 108.Google Scholar

³¹ Cf. also Slatkin (n. 30) 15 on the ‘exclusion of such traditional mythological material, or its displacement into more or less oblique references’.

³² Homeric ignorance: Dowden, K., Death and the maiden (London 1989) 11 f.Google Scholar

³³ Gaisser, J.H., ‘Adaptation of traditional material in the Glaucus-Diomedes episode’ TAPA c (1969) 165–76, at p. 170.Google Scholar

³⁴ No foreseeing: Griffin 1977: 48, Kullmann 1960: 221–4.

³⁵ Griffin 1977: 40 f. The qualitative distinction is an important point in Schadewaldt's work too, see Heubeck 1974: 43.

³⁶ Patroklos invented by Homer: Scheliha 236–51, 391; Schadewaldt 178–81; raised from obscurity by Homer: Kullmann 1960: 44 f., 193 f.; Janko 313. Hektor: Scott, J.W., The Unity of Homer (Berkeley 1921) ch.vii esp. 226Google Scholar; Scheliha 221 f.; Kullmann 1960: 182–8; Janko 312. According to Bethe 1901: 674, H. Usener tried to reconstruct an lliad with Paris replacing Hektor.

³⁷ Kullmann 1960: 122 f. Kullmann challenged, e.g. by Heubeck 1974: 45; for such criticism and its validity, see Clark 382.

³⁸ On the Trojan side, surely Pandaros is of this type: he is a Paris-avatar, who exists to break the truce and be killed (on Pandaros as Paris, now see Taplin 104 f.); Euphorbos is a similar figure, who in killing Patroklos foreshadows Paris killing Achilles (16.812), cf. Janko 410, 414 and Clark 385, referring to Mühlestein, H., ‘Euphorbos und der Tod Patroklos’ SMEA xv (1972) 79–90.Google Scholar For this avatar technique, cf. Phoinix who is a Nestor-avatar who can be left at Achilles' tent (Phoinix as Nestor, cf. Erbse 387).

³⁹ Just as points forward to the return of Odysseus from Troy, Nagy 23.

⁴⁰ Kullmann 1981: 23–5, now accepted by Richardson 202, 246.

⁴¹ Proklos' summary of Hagias of Troizen, Nostoi; Eur. Tro. 65 ff., 70, 90.

⁴² Cf. Richardson 78.

⁴³ Iliad 5.627–98: Bethe 1901: 668 f.; Robert, C., Studien zur Ilias (Berlin 1901) 402Google Scholar; Cauer, P.Grundfragen der Homerkritik ³ (Leipzig 1923) 242Google Scholar; Bethe 1927: 65. ‘Bethe's fundamental elements, the duels, are very meagre and somewhat uninteresting myths’, Nilsson, M.P., Homer and Mycenae (London 1933) 48.Google Scholar

⁴⁴ For a list of possible allusions to Antihomerica and Posthomerica see Kullmann 1960: 6–11.

⁴⁵ E.g. Kullmann 1960: 365 f. (‘In der ganzen Ilias kann man die Beobachtung machen, dass dies Epos in seinem Aufbau den ganzen Krieg zu repräsentieren scheint’ plus detailed table); Griffin 1980: 1 (’The wrath of Achilles and its consequences are made to represent the whole story’).

⁴⁶ Fenik, B.C., Iliad X and the Rhesus: the myth [Collection Latomus 73] (Bruxelles 1964).Google Scholar

⁴⁷ Notopoulos 34 f.

⁴⁸ Pestalozzi was in effect the first to adumbrate the full picture of the influence of the Aithiopis on the Iliad, though the way was pointed by Kakridis 1949: 93–5 (1944 in Greek). This picture has subsequently been developed, notably by Schadewaldt (155–202, ‘Einblick in die Erfindung der Ilias: Ilias und Memnonis’), Kullmann 1960 (from his Habilitationsschrift of 1957) and Schoeck. In my opinion, the most thoroughgoing and dependable of these texts is Kullmann's. For a full account of neoanalysm, see Clark.

⁴⁹ Else 39.

⁵⁰ 24.77 and 24.79 look artificial.

⁵¹ Connection of the scenes: Bethe 1914: 109–12, Pestalozzi 10, Schoeck 20–2. Accepted and, revealingly, muddled by Davies 4—it is hard to hold the scenes apart. Typical or meaningfully borrowed?—fair, if cautious, discussion in Fenik, B.C., Typical battle scenes in the Iliad: studies in the narrative techniques of Homeric battle description, [Hermes Einzelschrift xxi] (Wiesbaden 1968) 231–40.Google Scholar For a different view, that Arktinos is here developing Homer, see Erbse.

⁵² Kullmann 1960: 32. The replay in the Iliad is taken rather lightly by Kullmann 1981: 25, who sees it as the Aithiopis minus the tragedy. In contrast, Erbse 394–7, though I disagree with his ultimate conclusion, shows that the episode has a proper function in the Homeric text.

⁵³ Pestalozzi 11 f., Schadewaldt 164, Schoeck 29 f.

⁵⁴ Schadewaldt 165 f.; Clark, M.E. & Coulson, W.D.E., ‘Memnon and Sarpedon’ Museum Helveticum xxxv (1978) 65–73.Google Scholar There is the problem of why Death should remove someone granted immortality (Davies 57), but perhaps it is no more a problem than why a dead Sarpedon should be anointed with ambrosia, given immortal clothing and transported if permanently, and Homerically, dead (in implicit contrast not only to Memnon, but also to Achilles on Leuke).

⁵⁵ Slatkin (n. 30) 28–33. On Uṣás,- see also Puhvel, J., Comparative Mythology (Baltimore 1987) 60.Google Scholar

⁵⁶ 16.651–5. Clark-Coulson (n. 54) 66 f. Euphorbos as Paris-avatar, above n. 38.

⁵⁷ A long shot, this, but see Hommel, H., ‘Der Gott Achilleus’ SB Heidelberg Abh. i (1980).Google Scholar

⁵⁸ Willcock, M.M., ‘The final scenes of Iliad XVII’ in: Bremer, J.M., de Jong, I.J.F. & Kalff, J., Homer: beyond oral poetry: recent trends in Homeric interpretation (Amsterdam 1987) 191Google Scholar (‘as if to prepare for what will happen after the end of the Iliad’); Kullmann 316 (‘bereiten offenbar bewusst den Aithiopisstoff vor’).

⁵⁹ Lohmann, D., Die Kompositio der Reden in der Ilias (Berlin 1970) 169–73CrossRefGoogle Scholar; Richardson 5 f., 17.

⁶⁰ Briseis and Helen functionally compared already by Bethe 1901: 667.

⁶¹ Willcock, M.M., A companion to the Iliad (Chicago 1976) 287.Google Scholar Even Kullmann, in a concessive mood, allows the concept ‘zumindest assoziativ von ihr beeinflusst’ 1981: 20.

⁶² Thus I think Page's sarcasm recoils on itself, when he derides the neoanalysts for treating Homer's use of his predecessors like Vergil's use of Homer and contrasts them with ‘those of us who have long understood the process of growth of the traditional oral epic’ (Page, D.L., ‘Homer and the Neoanalytiker’ CR xiii [1963] 21–4).Google Scholar

⁶³ Cf. Schadewaldt 168; Kullmann 1960: 38 f. There is of course a problem here with whether ‘forgetful of your horsemanship’, is appropriate, something which it is in the one other use at Iliad 16.776 of Kebriones; but equally one may query whether ‘huge hugely’, a rare ‘formula’, is justly deployed on such a minor and expendable figure and it is this that is the issue, not the tagging on of the second half-line. On the other hand, Kebriones' death is sited somewhere near the Skaian Gate (16.712) in an area of text where Apollo is very active, and may confront Patroklos-Achilles with a vision of the death and fight for the body (16.765–80) awaiting Achilles in person.

⁶⁴ References: Kakridis 84, Gaisser (n. 33) 176. Catalogue of Ships: Kullmann 1960: 157–68, 1981: 23, 38 (‘offenbar der Katalog zum grössten Teil wörtlich von anderswoher übernommen wurde’).

⁶⁵ On verbatim repetitions scarcely attributable to formulaic composition, see Young 311 f.

⁶⁶ or the accusative, are very Iliadic. The phrase occurs 10 times and the word occurs only another three times with different substantives. (In the Odyssey it occurs only seven times, once of an in Apollonius only four times, of which only one occurrence is associated with in Quintus only once, of .) The first six references in the Iliad are at 4.19, 134, 217, 5.99, 110, 278. The first three references are to the wounding of Menelaos; the second three to the wounding of Diomedes. The two scenes are linked in our minds, but perhaps more by the rarity of arrows than by the word

⁶⁷ A different link between these scenes is found by Taplin 107, unaware, I think, of Else's argument.

⁶⁸ Foley 8.

⁶⁹ Kakridis, J.T., Homer revisited (Lund 1971) 23.Google Scholar

⁷⁰ Foley 247 addresses this problem by deriving ‘extratextual meanings’ from the text itself, but it is plain—cf. 247 n. 6—that, even if effective, this method can only lead to a poor-quality understanding of the text. The real problem is the legitimation, and disqualification, of critical language: Foley's excellent chapter on Iliad 24 acquires legitimacy through conforming to his oral- ‘inherent’ discourse, but it does not say anything which is specially surprising or categorically different, i.e. which reveals an unperceived Homer.