¹ Γιὰ μιὰ ποιητικὴ τού έλληνικοῡ δημοτικοῡ τραγουδιοῡ (Herakleion 1988)Google Scholar.

² Parry, Adam (ed.), The making of Homeric verse. The collected papers of Milman Parry (Oxford 1971;Google Scholar henceforth quoted as MHV).

³ The singer of tales (Cambridge, Mass. 1960,Google Scholar henceforth abbreviated as ST); Epic singers and oral tradition (Ithaca and London 1991);Google ScholarThe singer resumes the tale (Ithaca and London 1995)Google Scholar.

⁴ Cf. Lloyd-Jones, H., ‘Remarks on the Homeric question’ in History and imagination. Essays in honour of H.R. Trevor-Roper, Lloyd-Jones, H. (ed.) (New York 1982) 15–29;Google ScholarShive, D., Naming Achilles (Oxford and New York 1986)Google Scholar.

⁵ As Nagy, G. puts it in ‘Homeric questions’, TAPA cxxii (1992) 17–60Google Scholar.

⁶ The dictation theory was first advocated by Lord (‘Homer's originality: Oral dictated texts’, TAPA lxxiv [1953] 124–34,Google Scholar repr. in his Epic singers and oral tradition [Ithaca and New York 1991] 38–48)Google Scholar and then was taken up by others, Cf. Parry, Adam, The language of Achilles and other papers (Oxford 1989) 104–140;Google ScholarJanko, R., The Iliad: A commentary, vol. iv, books 13-16 (Cambridge 1992) 37–8Google Scholar. Powell, B.B. in a recent monograph on Homer and the origin of the Greek alphabet (Cambridge 1991)CrossRefGoogle Scholar pushes the assumption of dictation to an extreme conclusion by putting forward the startling theory that the ‘genius and benefactor of mankind, who invented the Greek alphabet by adaptation from the preexisting Phoenician syllabary’ (12) did so in order to write down what Homer sang, and was thus instrumental in facilitating the production of the Iliad and the Odyssey in a kind of joint venture and cooperative effort with the poet (230); we may even have known his name all along: Palamedes (233-7). However, neither the theories of dictation, nor the idea of a writing poet, nor the ‘evolutionary model’ for the genesis of the epic suggested by Nagy (see now his Poetry as performance. Homer and beyond [Cambridge 1996],Google Scholar especially 74-7, 110) manage to offer satisfactory answers to all the problems involved so as to be fully convincing.

⁷ So M. Curschmann with reference to the Nibelungenlied (‘Nibelungenlied und Nibelungenklage. Über Mündlichkeit und Schriftlichkeit im Prozess der Episierung’ in Cormeau, Ch. (ed.), Deutsche Literatur im Mittelalter. Hugo Kuhn zum Gedenken (Stuttgart 1979) 93–4,Google Scholar quoted by Fenik, B., Homer and the Nibelungenlied. Comparative studies in epic style (Cambridge, Mass. 1986) 173, 202;CrossRefGoogle ScholarCf. Fenik, , Digenis. Epic and popular style in the Escorial version (Herakleion 1991) 17Google Scholar. See also Curschmann, , ‘Oral poetry in mediaeval English, French, and German literature: Some notes on recent research’, Speculum xlii (1967) 36–52CrossRefGoogle Scholar.

⁸ Cf. Sifakis, G.M., ‘Τὸ πρόβλημα τῆς προϕορικότητας στὴ μεσαιωνικὴ δημώδη γραμματεία’ in Panayotakis, N. M. (ed.), 'Αρχὲς τῆς νεοελληνικῆς λογοτεχνίας i (Venice 1993) 267–24Google Scholar.

⁹ Useful in this connexion is van der Werf, H., The oldest extant part music and the origin of Western polyphony i (Rochester NY 1993) 18 ffGoogle Scholar. (on ‘Notation in mediaeval music life’), and G. Amargianakis, An analysis of stichera in the deuteros modes i–ii (Cahiers de I'Institut du Moyen-âge grec et latin xxii, Copenhagen 1977)Google Scholar (offering a breakdown of a group of Byzantine melodies into their constitutive formulas).

¹⁰ See, for instance, Bynum, W.F., Daemon in the woods: A study of oral narrative patterns (Cambridge, Mass. 1978);Google ScholarFoley, J.M., Traditional oral epic: The Odyssey, Beowulf, and the Serbo-Croatian return song (Berkeley and Los Angeles 1990)Google Scholar.

¹¹ Adam Parry in his Introduction to MHV xli.

¹² Cf. ST 100, and Lord, Epic singers and oral tradition 76, 102, 130, 209; The singer resumes the tale 23 ff., 95, 108.

¹³ ST 68.

¹⁴ ST 50-3. Bowra, C.M. (Heroic poetry (London 1952) 234 ffGoogle Scholar.) had already denied the existence of thrift in other traditions outside Homer.

¹⁵ The flexibity of the Homeric formula (Oxford 1968)Google Scholar. See also now his concise analysis of Homer's formulaic style in the introduction to his volume of The Iliad: A commentary iii: books 9-12 (Cambridge 1993) 1–31Google Scholar.

¹⁶ ‘Towards a generative view of the Homeric formula’, TAPA xcviii (1968) 269–311,Google Scholar and Spontaneity and tradition: a study in the oral art of Homer (Berkeley and Los Angeles 1974) 13–15Google Scholar. Nagler's ‘Gestalt’ or ‘sphota’, though, is a concept too vague ‘and frankly mystical in tone’ (Spontaneity 14) to be really useful as an analytical tool, so it has not been used by others. Also, his use of the term ‘allomorph’ is arbitrary because it refers to very dissimilar forms supposedly originating from the same template; but it has the advantage of dispensing with an Ur-formula from which others are derived (see p. 146 below).

¹⁷ Cf. The Iliad: A commentary iii 26.

¹⁸ The phrase τεύχεα καλά is indeed a formula that occurs many times elsewhere. But does this mean that the two words should be taken as a formula under any circumstances? It all depends, of course, on how the formula is defined, so even an uncommon case like this is consistent with Hainsworth's definition (which ignores metre and versification patterns).

¹⁹ Cf. Ingalls, W., ‘Another dimension of the formula’, Phoenix xxvi (1972) 115Google Scholar.

²⁰ All this literature, up to the mid-eighties, is reviewed by Edwards, M.W. in an invaluable critical essay: ‘Homer and oral tradition: The formula’ I–II, Oral Tradition i/ii (1986) 171–230Google Scholar, and iii. 1-2 (1988) 11-60. See also Shive (n.4 above).

²¹ To the earliest major study of this kind, Arend, W., Die typische Scenen bei Homer (Berlin 1933),Google Scholar many more and very good ones can now be added, e.g., Armstrong, J.I., ‘The arming motif in the Iliad’, AJP lxxix (1958) 337–54;Google ScholarAustin, N., ‘The function of digressions in the Iliad’, GRBS vii (1966) 295–312;Google ScholarEdwards, M.W., ‘Type-scenes and Homeric hospitality’, TAPA cv (1975) 51–72,Google Scholar‘The structure of Homeric catalogues’, TAPA cx (1980) 81–105,Google Scholar and Homer, poet of the Iliad (Baltimore 1987);Google ScholarFenik, B.C., Typical battle scenes in the Iliad, Hermes Einzelschriften xxi (Wiesbaden 1968),Google ScholarStudies in the Odyssey, Hermes Einzelschriften xxx (Wiesbaden 1974);Google ScholarKirk, G.S., ‘The formal duels in books 3 and 7 of the Iliad’ in Fenik, B.C. (ed.), Homer: Tradition and invention (Leiden 1978) 18–40;Google ScholarTsagarakis, O., ‘Oral composition, type-scenes and narrative inconsistencies in Homer’, Grazer Beiträge viii (1979) 23 ffGoogle Scholar.; Form and content in Homer, Hermes Einzelschriften xlvii (Wiesbaden 1982)Google Scholar. For a survey of this literature on type scenes, see Edwards, M.W., Oral Tradition vii.2 (1992) 284–330Google Scholar. See also the recent monograph of Reece, S. on The stranger's welcome. Oral theory and the aesthetics of the Homeric hospitality scene (Ann Arbor 1993),CrossRefGoogle Scholar and Lord on arming scenes in The singer resumes the tale 75-95 (and passim on themes in other traditions).

²² Only one major work, the epic/romance of Digenes Akrites, whose origins go back to medieval epic tradition, has survived in fifteenth century mss., as has the shorter lay of Armoures, alongside the acritic songs. Critical edition: Alexiou, S., Βασίλειόσ Διγενὴσ Άκρίτησ καὶ τὸ ἀσμα τόῦ Άρμόύρη (Athens 1985);Google Scholar traditional style: B. Fenik, Digenis (n. 7 above); historical context, discussion of problems, bibliography: Beaton, R. and Ricks, D. (eds.), Digenes Acrites. New approaches to Byzantine heroic poetry (London 1993)Google Scholar.

²³ This metaphor was, of course, used by Ferdinand de Saussure in his original definition of the linguistic sign in Cours de linguistique générale (Paris 1916;Google Scholar many reprints).

²⁴ MHV 13-14.

²⁵ This is another way of referring to what Foley describes as ‘traditional referentiality’, see his Immanent art: From structure to meaning in traditional oral epic (Bloomington and Indianapolis 1991), chps. 1-2, esp. pp. 6-10, 38–60Google Scholar.

²⁶ Cf. Sifakis, , Greece & Rome xxxix (1992) 151Google Scholar.

²⁷ On formulas as idioms of poetic language, see Kiparsky's, P. very interesting article on ‘Oral poetry: Some linguistic and typological considerations’ in Stolz, B.A & Shannon, R.S. III (eds.), Oral literature and the formula (Ann Arbor 1976) 73–106Google Scholar.

²⁸ This unfortunate metaphor was used by van Gennep, A., La question d'Homère (Paris 1909) 52,Google Scholar quoted by Hainsworth, The flexibility of the Homeric formula 15, n. 1.

²⁹ As Hainsworth, says, the ‘ἀοιδοί were masters of a special form of language, not jugglers of formulas’, The Iliad: A commentary iii (Cambridge 1993) 16CrossRefGoogle Scholar.

³⁰ ST 5.

³¹ Lord did not use Saussure's terms, langue and parole (but Cf. ST 279, n. 7), and could not have used N. Chomsky's notions of competence, creativity and performance (which corresponds to Saussure's parole), see Aspects of the theory of syntax (Boston 1965) 6–8,Google ScholarLanguage and mind (New York² 1972),Google ScholarCf. Lyons, J., Semantics (Cambridge 1977) 77, 107, 265)Google Scholar. He shows no awareness of Hjelmslev, L.'s connotative semiotics (‘Pour une sémantique structurale’ in Essais linguistiques (Copenhagen 1959)),Google Scholar but was able to point out that ‘in studying the patterns and systems of oral narrative verse we are in reality observing the “grammar” of the poetry, a grammar superimposed, as it were, on the grammar of the language concerned’ (ST 36).

³² ‘A’ stands for the Academy of Athens standard anthology of folksongs: Έλληνικἀ δημοτικὰ τραγούδια Spyridakis, G.K. et al. (ed.) , i (Athens 1962)Google Scholar. References are to pages followed by the place of origin of the variant quoted (or, if not known, the name of one of the early folksong editors).

³³ Politis, N.G., Έκλογαὶ άπὸ τὰ τραγούδια τοῦ ἑλληνικοῦ λαοῦ (Athens 1914) 214Google Scholar.

³⁴ ‘Syntagmatic’ coincides here with ‘syntactic’, but the former is more general than the latter. Syntagmatic and paradigmatic relations in structural linguistics constitute the two axes of relationships of a verbal utterance. The former are relations in praesentia and hold together the terms of an utterance (a linguistic syntagm); the latter (actually called ‘rapports mémoriels’ or ‘associatifs’ by Saussure) are relations in absentia and refer to synonyms, homonyms, antonyms, etc., of each term in the syntagm, which may be recalled or connoted in the process of verbal communication.

³⁵ All translations quoted below are by Richmond Lattimore. I made no changes to them even when the same Homeric phrases, quoted next to each other, appear to differ in the translations.

³⁶ Or ‘A addressed him with qualification,’ see Edwards, M.W., ‘Homeric speech introductions’, HSCP lxxiv (1970) 7Google Scholar. Eight more categories of speech introductions are studied by Edwards (pp. 1-36).

³⁷ Some of Joseph A. Russo's ‘phrase-patterns’ would be useful here. However, in his well known paper on ‘The structural formula in Homeric verse’ (Yale Classical Studies xx (1966) 2191240),Google Scholar Russo is mainly interested in the localization of various types of grammatical/syntactical phrase-patterns to particular parts of the hexameter, and not in a common meaning underlying them. So his groups of examples do not constitute paradigmatic matrices in the sense indicated above.

³⁸ I suppose that this is a way of explaining the genesis of unique ‘formulaic’ expressions which Hainsworth rightly refuses to call formulas (Cf. The Iliad: A commentary iii 17).

³⁹ ‘He drew his knife from its sheath, threw it up into the air, but the knife [reaching its originally intended destination] pierced through his/her body.’

⁴⁰ The term ‘allomorph’ is borrowed from linguistics, where it actually means an alternative morpheme. But it is a convenient term to suggest an alternative morphe (form), also, and in this limited sense it is a much more specific term than Nagler's allomorph (Cf. pp. 138, 139 above). On the other hand, my allomorphs and their respective generative matrix constitute a much wider category than Hainsworth's examples of substitution (The Iliad: A commentary iii 15).

⁴¹ Cf. p. 146 above. As we shall see a little later (p. 149 and Table of Ship-setting-sail Theme), poetic synonyms are flexible and come in a variety of forms.

⁴² Hainsworth, The flexibility of the Homeric formula 82.

⁴³ Cf. Chantraine's, P. pioneering work on ‘Remarques sur l'emploi des formules dans le premier chant de l'Iliade’ (RÉG xlv (1932) 121–54),Google Scholar ‘an article on the “play of formulae”,’ which however ‘remains the only work of its kind, an excellent source for observing how formulae are used’, as Edwards, M.W. says (Oral Tradition i/ii (1986) 198)Google Scholar.

⁴⁴ Arend also discusses ship departures, op. cit. (n. 21) 81-86.

⁴⁵ Epic singers and oral tradition 78.

⁴⁶ Poet. 1451^a 28.

⁴⁷ Essays, ed. by Suchoff, B. (New York and London 1976) 346Google Scholar.