¹ Havelock, E. A. (Preface to Plato, Oxford 1963Google Scholar, especially in chapters 2, 3 and 9), describes the deep-seated causes for psychosomatic participation of classical Greek audiences in poetic performances or recitals. But he accepts (p. 22) the conventional interpretation that Socrates is referring to the imitation of ‘the growls and squeaks of animals’ in the passages under discussion here. See further in note 13 below.

² An example of this not cited by editors is in the fragment of Diphilos cited by Athenaios, , Deipn. 14Google Scholar, 657e where pipe-players are said to cackle like geese (χηνίζϵιν).

³ Such ex machina solutions are not required in view of Aeschylus' reference to the thundering sound of drums in a fragment of his Edonoi: Cf. LSJ at μυκάομαι (last entry).

⁴ E.g. Laws 665a where Plato refers to a τάξις φωνῆς which is called ἁρμονία and consists of acute and grave tones mixed together. Even μέλος is used in this sense by Dionysios of Halicarnassos (De comp. 11), there also with reference to the Greek pitch-accent: cf. Aristoxenos, Harm. 1.18.

It is the chronic ambiguity of words like μουσική, ἁρμονία, φωνή and μέλος that makes discussion of literary and musical mimesis so complicated. (On φωνή see Leroy, M. in REG lxxx [1967] 234—7.CrossRefGoogle Scholar) The only safe way of referring to them is to use the Greek terms, since most translations are question-begging. There is also the uncertainty caused by the fact that in antiquity non-lyrical poetry was delivered in a manner intermediate between singing and normal speaking (Aristoxenos, , Harmonics 1.3Google Scholar, Aristides Quintilianus, , De mus. 1.4Google Scholar, Cicero, , De oratore 17Google Scholar, schol. on Dionysios Thrax 744.32B and 746.1).

⁵ The evidence is summarised by Havelock (see note 1) pp. 57–60.

⁶ This is the only recorded instance of χρϵμϵτίζω in Greek poetry before Plato. (Hesiod uses the shorter form, χρέμισαν, of horses in Shield 348.) Herodotos uses it in what looks like an oracular phrase (3, 86 and 87). Aristophanes has the noun χρϵμϵτισμός in a dithyrambic passage (Knights 553).

⁷ The favourable view of Homer's use of onomatopoeia is supported by Demetrios, (De eloc. 94Google Scholar, cf. 72), the Plutarchan, Life of Homer 16Google Scholar, and the Homeric scholiasts and Eustathios (see J. Barr, Index zu den Ilias-Scholien and the index to the Leipzig edition of Eustathios at ὀνοματοποιἱα, ἔμφασις and μίμησις and cognates). Philodemos (Sudhaus 1.33.14 and 2.257–8: cf. 1.150.15) cites the opposite view that such sound-effects (ἦχοι) distract the attention of hearers from the contents of literature and from the truth (n. 10 below).

⁸ The confusion of Solon with Plato here may result from the fact that both of them had a distaste for the ‘lies’ of drama, if the story of Solon's rebuke to Thespis (Plutarch, , Solon 29Google Scholar) is true.

⁹ In 423b an ὄνομα is defined as a μίμημα φωνῇ which implies an important distinction between a word when it has been accepted as a current term and a word when it is being ‘invented’ by a ‘word-maker’, the first being an ἔργον a product, the second an ἐνέργϵια a process, which is often a mimesis. Cf. Vicaire, Paul, Platon, critique littéraire (Paris, 1960) 221Google Scholar. In Cratylos 423c Socrates says that both those who mimic other noises and also performers of musical mimeseis are not ‘word-makers’ in the same sense as those who coin mimetic (i.e. onomatopoeic) words. He goes on to make a subtle analysis (426c–27d) of some of the phonetic elements in speech which underlie onomatopoeic effects. (On early views, going back to Demokritos, about the development of music from imitation of bird-song see Webster, T. B. L., CQ xxxiii [1939] 168Google Scholar.)

¹⁰ This prejudice against onomatopoeia survives into modern times, as e.g., when a writer in CQ xxxvi (1942), p. 39, remarked, ‘The attempt to be ono-matopoetic seems just a shade below the dignity of great composition’.

¹¹ The phrase τοῦ τοιούτου ῥήτορος in 396e implies that Socrates is also thinking of ‘the good man’ as a public speaker in contrast with ‘the inferior man’ who will use vocal mimesis freely in his speeches.

¹² This seems to be the general sense of Aristotle's statement in Poetics 1455a 21–30 that in poetic composition one must συναπϵργάζϵσθαι both in diction and in gestures. The passage is much disputed, but the verb could mean ‘working out with’ in the sense both of using appropriate tones of voice in composing the words and of using appropriate gestures when describing actions and emotions. Commentators have noted that Aristophanes may be burlesquing the second process in Acharnians 410–14 and Thesmophoriazousai 148–67.

¹³ For a recent discussion of the evidence see Knox, B. M. W. in GRBS ix (1968) 421–35Google Scholar. If the prevalence of reading literature aloud, even when reading alone for oneself, is accepted, it removes one of the main difficulties in Havelock's very proper insistence on the oral nature of Greek poetry down to the fourth century B.C. (see n. 1 above), namely, why does Plato apply the term mimesis to the learner or reader of poetry as well as to the poet and the performer (pp. 24 and 37 ff.)? His solution is that in fact the learning and personal enjoyment of poetry remained an oral process, by recital and performance, down to Plato's time. But he admits that Plato sometimes uses language that implies taking up a manuscript and reading it for oneself (Repub. 6064e, Apology 22b), and his attempts to rebut the strong case made by Turner, E. G. (Athenian books in the fifth and fourth centuries B.C., London 1951Google Scholar) for some degree of private reading in that period is hardly successful. Yet, whatever the prevalence of literacy may have been at that time, Havelock's main contention, that poetry continued to be an oral and mimetic experience into the fourth century, remains valid because even the private reading of poetry was normally aloud (not silent as Havelock seems to assume) until long after Plato's time.

¹⁴ I am grateful to Professors J. V. Luce and D. E. W. Wormell for helpful comments on this article, and to my wife for proof-reading.