¹ West, M. L., ‘The rise of the Greek epic’, JHS cviii (1988) 151–72.CrossRefGoogle Scholar

² My arguments against an Aeolic epic, of a purely linguistic nature, can be found in Ἐπιστιμονικὴ Ἐπετηρὶς τῆς Φιλοσοφικῆς Σχολῆς τοῦ Ἀριστοτελείου Πανεπιστημίου Θεσσαλονίκης xiv (1975) 133–47. Cf. also Gary Miller, D., Homer and the Ionian epic tradition (Innsbruck 1982)Google Scholar; and Horrocks, G., Minos xx–xxii (1987) 269–94Google Scholar for a measured and skilful treatment of the subject.

³ I note that Wathelet, P., AC I (1981) 833 n.Google Scholar 65 states that only by ‘la rencontre de données de divers ordres’ will the history of epic be clarified. I concur, but would insist that all the data be examined separately and be securely based before one proceeds to global conclusions. I do not feel that this precondition has been met in the matter in question.

⁴ One reason why Aeolisms are easily assumed is that Ionic has developed and diverged from earlier forms of Greek more than most dialects, and hence Aeolisms, if assumed, are very likely to correspond at least metrically with earlier, non-Ionic,stages of the language. Aeolic is more conservative of older forms, and thus earlier stages of epic are inevitably going to appear more Aeolic.

⁵ Such a descriptive grammar is as yet lacking, as noted by Forssman, B. in Latacz, J. (ed.), Zweihundert Jahre Homer-Forschung (= Colloquium Rauricum II, Stuttgart and Leipzig 1991) 287 n. 104.Google Scholar

⁶ Chantraine, P., Grammaire homérique I, (³Paris 1959).Google Scholar

⁷ Cf. my Metrical lengthening in Homer (Incunabula Graeca 35, Rome 1969).

⁸ Cf. Chantraine (n. 6) 75–83.

⁹ As we shall see below, however, history is perhaps less important in this regard than nature. Classical scholars are inevitably inclined to seek the explanation of events and states in history, whereas a more nearly synchronie explanation may in many cases suffice.

¹⁰ This view, not really to be seriously maintained any more in this form, constitutes a kind of maximal hypothesis according to which all features in the Homeric poems can be mapped onto a Mycenaean grid. And by this theory therefore there can have been a straight development from Mycenaean Greek to Ionic. The major supporter of this theory is Georgiev, V., Mycenaean studies 125–39 (Bennett, E. L. Jr,ed., Madison 1964)Google Scholar, Minos xii (1971) 361–79, par. 371. Strunk, K., Die soganannten Aeolismen der homerischen Sprache (Köln 1957)Google Scholar approaches this view, as do all those who deny an Aeolic phase and assume a Mycenaean, in that he denies that there are any specifically Aeolic forms in the Homeric poems. By both these views the Ionic epic can be the direct lineal descendant of a Mycenaean (Achaean) epic.

¹¹ Parry, M., ‘The epic technique of oral verse-making II’, HSCP xliii (1932) 1–47Google Scholar = The making of Homeric verse (Oxford 1971) 325–361; Whitman, C., Homer and the heroic tradition (Cambridge 1958) esp. 60CrossRefGoogle Scholar; Page, D.L., History and the Homeric Iliad (Berkeley and Los Angeles 1959) 218–96, esp. 219–221Google Scholar; Kirk, G.S., CAH 11.2 (Cambridge 1975) 828Google Scholar; Hammond, N.G.L., A history of Greece to 322 BC (Oxford 1967) 57.Google Scholar

¹² The list of scholars subscribing to this view is enormous, and includes M. L. West. I cite only the most recently authoritative book-length treatments: Wathelet, P., Les traits éoliens dans la langue de l'épopée grecque (Incunabula Graeca 37, Rome 1970)Google Scholar; and Durante, M., Sulle preistoria della tradizione poetica grece I (Incunabula Graeca 50, Rome 1971).Google Scholar For schematic summaries of the views of various linguists on the whole question of Mycenaean and Aeolic elements in Homer, cf. Studia Mycenaea (ed. Bartoněk, A.) Brno 1968, 198–201.Google Scholar The view in its origins goes back well into the nineteenth century, as West observes, and found its most outspoken and extravagant exponent in A. Fick who attempted to recover the original Aeolic text in which the poems were first composed. For a history of scholarship on this matter cf. Wathelet 50–6.

¹³ This view is not new, and has been upheld by Monro, D. B., A grammar of the Homeric dialect (²Oxford 1891) 386–96)Google Scholar; Pisani, V., Enciclopedia classica Sec. II Vol. v Pt. I (1960) 25–47Google Scholar, esp. 42–43; Szemerényi, O. J. L., JHS lxxix (1959) 193Google Scholar; SMEA i (1966) 31–5.

¹⁴ He here cites as authorities Chadwick, J., G&R iii (1956) 47Google Scholar; Ruijgh, C. J., Lingua xviii (1967) 96 f.Google Scholar, Wathelet 104, 180 f., 290; Durante 28, 34, 54; Janko, R., Homer, Hesiod and the Hymns (Cambridge 1982) 89 f.Google Scholar

¹⁵ In fact I believe that the conditions for the passage of αο to -ηο in Ionic were not met in Homeric epic. The change is one of contraction, not metathesis, and so -αο remained because contraction could not take place. Cf. Metrical lengthening in Homer (n. 6) 100.

¹⁶ Blümel, W., Die aiolischen Dialekte (Ergänzungsheft zur Zeitschrift für vergleichende Sprachforschung xxx, Göttingen 1982) 30.Google Scholar

¹⁷ The table is derived in the main from matter presented by Chantraine(n. 6) 496–513. It is divided into six columns, the afirst two presenting the linguisic evidence of the Homeric text itself. In column one I have listed the words and forms which appear both in the Homeric poems and in the Ionic dialect. In the second column appear forms which are semantically identical with forms of the first column, but which are—in most cases—metrically different. These forms appear both in the Homeric poems and also in one or more of the Aeolic dialects, aa fact which is noted in column three: T stands for Thessalian, B for Boeotian, L for Lesbian. Many of the forms of column two are characteristic alsoof other dialects, and I have listed in column four some of the dialects which share these features. Column five lists, wherever possible, the earliest reconstructed form of the Ionic dialect, a reconstruction based not only on Attic and Ionic, but on those dialects plus Homer. At the risk of begging the question I wish to argue that the earliest form of East Greek reconstructible leads to forms which are both Homeric and are also close to the metrical shape ofthe earliest Greek reconstructible. In column six I give the presumed earliest form within the Greek language as a whole. A fuller treatment of these matters would entail much discussion, e.g., of my assumed early πότι vs. Mycenaean po-si (cf. SMEA xix [1978] 117–231], the relation of παρ and παρά, and the source of the -μ- in forms such as άγέμεν. (Cf. TAPA ci[1970] 557–632, ‘The prehistory of the Greek dialects’, for an overall view of the history of Greek dialectal history). The table is schematic and suggestive, and is designed to show that an Aeolic phase of epic composition is not necessary: it does not prove the point.

¹⁸ The essence of this argument is to be found already presented by Szemerényi, O.J.L., JHS lxxix (1959) 193Google Scholar: ‘wherever the normal Ionic development would have led to forms non-existent in Ionic, the speech-form of the neighbouring Aeolic was adopted. This is to assume a certain amount of influence from Aeolic, but not a full-scale Aeolic stage in the development of epic poetry.’

¹⁹ Shipp, G.P., Studies in the language of Homer (Cambridge 1972) 4CrossRefGoogle Scholar also assumes Dorisms. West's main example is the one I cite, and his others are equally compelling, but one cannot instance the forms in dialect texts. His (1) and (2) involve pronouns of the second and third person singular, and such are very rare on inscriptions (Buck, C.D., Greek Dialects [Chicago 1955] 97 n. 1).Google Scholar His (4) πρότι can, he holds, be very old: I tend to think of it as being purely epic. His (5) ὲσσεῖται (Chaintraine [n. 5] 290–91) is too irregular to sustain much weight. 1–3, though, are definitely Doric, and the only question is when they entered the tradition. I cannot point to inscriptional examples from Asia Minor, save in the case of άμός, which occurs in Rhodes (Buck Greek dialects § 105, 106.17). Homer certainly acquired this form at least in Asia Minor.

²⁰ Many scholars feel that this word merely represents a later spelling of ἄμμος; Floyd, E.G., Glotta xlvii (1970) 119Google Scholar with refs. Wackernagel, J., Sprachliche Untersuchungen zu Homer (Göttingen 1916) 210–12Google Scholar, determined the motive for the appearance of άμός but assumes that ἄμμος (not *ήμός) was replaced by the Doric form through the influence of Attic.

²¹ Morpurgo Davies, A. (Linear B: a 1984 survey, ed. Morpurgo Davies, A. and Duhoux, Y., Eds. [Louvain-le-neuve 1985] 86Google Scholar interestingly suggests that epic, among other forms of literature, avoided ‘some types of linguistic variation and in particular types tied to geographical distinctions.’

²² On the developments involving labiovelars, cf. my article in GRBS xvi (1975) 251–62.

²³ This ending has much been discussed: Cf. Schwyzer, E., Griechische Grammatik i (Munich 1939) 564Google Scholar, and more recently C. J. Ruijgh (n. 9) 14–17, and my article, to appear in Essays in linguistics offered to Oswald Szemerényi onthe occasion of his seventy-fifth birthday, ed. Broyanyi, B. and Lipp, R. (Amsterdam 1992).Google Scholar

²⁴ It might be well here to point out that in fact -εσσι can have been generated by epic poets purely out of epic resources without recourse to the Aeolic explanation. They possessed the form ἔπεσσι beside their colloquial ἔπεσι and ποσί. Analysis of the longer form can easily—by means which cannot and need not be specified—have yielded the ending –εσσι, and that, considered to be added to the stem of the noun, can have led directly to πόδεσσι. To be sure, the occurrence of –εσσι in neighbouring Lesbos will have strengthened the poets' conviction of the antiquity of the form.

²⁵ This is his only real argument, since the other cases are too weak to support any weight. They are instances such as ὲνάτη ἄνοιτο ξενίη in which the digamma once following the nasal disappears, but without the compensatory lengthening of the previous syllable seen elsewhere in epic: this characteris tic is West Ionic and of course Attic. (Cf. Wathelet [n. 12] 154–57, AC 1 (1981) 819–33, Chantraine [n. 6] 161–63). Attic correption, ordinarily avoided in epic, is allowed in (e.g.) προσηύδα (which replaces Aeolic *ποταύδα). But it seems clear—at least to me—that correption before this word indicates rather that an earlier form was (Ionic) *ποσηύδα: cf. SMEA xix (1978) 89–123. His concern that West Ionic has –ττ– and not –σσ– and –ρρ– for Ionic –ρσ–, can be dismissed easily: even Attic authors avoided these clusters in their formal writings because they were too parochial. Even if—as I believe they were not—epic poets had been Euboean, they still would have avoided these sounds.

²⁶ Cf. Buck (n. 17) 63, Bechtel, F., Die griechischen Dialekte iii (Berlin 1924) 87–89.Google Scholar Bechtel holds that -k- develops regularly in the indefinite relative between two /ο/'s. And it may well be that this is where the development began. See now Chadwick, J., JHS ex (1990) 174–77.CrossRefGoogle Scholar