¹ This assumption still underlies the most recent and in many ways the best study of the kind, Baxter, T.M.S., The Cratylus. Plato's Critique of Naming (Leiden 1992)Google Scholar, where the etymological section is interpreted as a sustained satire on attitudes, throughout the entire Greek cultural tradition down to Plato's day, to the relation between reality and language. Probably Baxter's most important forerunner in this tradition of interpretation is Goldschmidt, V., Essai sur le “Cratyle” (Paris 1940)Google Scholar, which detects in the etymological section an ‘encyclopédie’ of flux-based theories in cosmology, theology and ethics.

² Grote, G., Plato and the Other Companions of Sokrates (1st ed., London 1865; 3rd ed., 1875) 2, ch. 29Google Scholar. Grote's admirable arguments deserve to be read by everyone interested in the interpretation of this dialogue. Barney, Rachel, ‘Socrates agonistes: the case of the Cratylus etymologies’, OSAP 16 (1998, forthcoming)Google Scholar also regards the etymologies as exegetically serious; her conclusions, if very different from mine, are complementary rather than antithetical. The serious philosophical content of certain etymologies is recognised by Gaiser, K., Name und Sache in Platons Kratylos (Heidelberg 1974)Google Scholar; Montrasio, F., ‘Le etimologie del nome di Apollo nel “Cratilo”’, Rivista di storia della filosofia 43 (1988) 227–59Google Scholar; and Wohlfahrt, P., ‘L'etimologia del nome Hades nel “Cratilo”’. Contributo allo studio della religione in Platone’, Rivista di storia della filosofia 45 (1990) 5–35.Google Scholar

³ De int. 4.17a 1-2, where ὅργανον is a clear reference to Crat. 388b-c.

⁴ Translating the reading of the new OCT, edited by Nicoll, W.S.M. and Duke, E.A., in Platonis Opera vol. 1, ed. (Oxford 1995)Google Scholar, ϕαίνονται γὰρ ἕμοιγε αύτοὶ οὕτω διανοηθῆναι (αύτοί W: καὶ αύτοί BTQ: καὶ αύτῷ Heindorf, Burnet).

⁵ Baxter (n.1) ch. 5 is a very useful guide to this background.

⁶ Tim. 90c; see further, Sedley, D.N., ‘“Becoming like god” in the Timaeus and Aristotle’, in Calvo, T. and Brisson, L. (eds.), Interpreting the Timaeus-Critias (Sankt Augustin 1997) 327–39Google Scholar.

⁷ See n.17 below.

⁸ Rep. 369b-c.

⁹ The etymology proposed for himeros at Crat. 420a is partly different (although the ensuing etymology of erōs at 420a-b is very close to the Phaedrus analysis of himeros). There is no reason why consistency between the two dialogues should be demanded on this point, but in any case it is possible for two or more etymologies of the same word to be exegetically correct at the same time: see further p. 148 and n.32 below.

¹⁰ Thus Hackforth, R., Plato's Phaedrus (Cambridge 1952) 59Google Scholar. I resist any suggestion that Socrates, departing for once from his project of rhetorical persuasiveness, is here for reasons of his own resorting to an intentionally ridiculous and therefore unpersuasive device. Cf. especially Rowe, C.J., Plato's Phaedrus (Warminster 1986) 170–2Google Scholar, where the second group of etymologies is ‘deliberately fanciful’ (170) and ‘a kind of reductio ad absurdum’ of the Cratylus thesis that names are a guide to the truth (172).

¹¹ For acceptance of the Cratylus etymologies as at least exegetically serious, see e.g. Plut. De Iside et Osiride 375C-D; Dionysius Hal. De comp. verborum 62.18-63.3 Usener-Radermacher; Proclus Schol. In Crat. passim; cf. Alcinous, Did. 159.44-160.30. Perhaps the most revealing acceptance of etymology is that of Sextus Empiricus, Adversus grammaticos (= Adversus mathematicos 1) 241-7. Sextus sets out to doubt everything he can about the grammarians’ ‘art’, including their use of etymology to establish that a word is authentically Greek. But the soundness of etymological analysis as such goes unquestioned. Indeed, to Sextus’ ear there is no difference between what we would consider sound etymologies and those we would think wildly fanciful: thus he accepts without evident discrimination both that προσκεϕάλαιον (‘pillow’) = something put close to (πρός) the head (κεϕαλή), and that λύχνος (‘lamp’) = something which dissolves (λύει) night (νύχος). The most severe ancient critic of etymology is Galen (PHP 2.2), who refers to a (lost) work of his, On Correctness of Names (1.104.20-1 De Lacy). However, even his objection is to the supposition that truth can be found through etymology. He never denies, but on the contrary seems to presuppose (1.116.25-31), that etymology offers access to the beliefs of the name-maker, provided that it is properly done (cf. 1.104.22).

¹² Hermogenes was an intimate enough member of the Socratic circle to be present at Socrates’ death, Phd. 59b. Note too that he is familiar with the theory of Forms (Crat. 389-90).

¹³ Respectively Phys. 197b 29-30, EN 1103a 17-18, 1132a 30-2, 1152b 7, DA 429a 2-4.

¹⁴ GA 736a 18-21; DC 1.9, 279a 18-b 1. In the former he follows the (fairly obvious) derivation from άϕρός at Crat. 406c, but differs from the explanation offered there that the name reflects the goddess's birth from the sea.

¹⁵ Metaphysics 12.1074a 38-b 14. Plato himself has a similar cataclysm theory in his late dialogues—at Timaeus 22b-23c, Critias 109d-110c, and Laws 677a-679e—but the latter two passages explicitly deny that any knowledge is preserved from one civilisation to the next beyond a bare record of some names of great dynasts.

¹⁶ Against the temptation to deny that Plato's Socrates has reverence for anybody's authority, but only for the unmediated truth itself, it is worth citing passages like Apology 29b and Crito 47a-48a, where, while denouncing the ignorant views of the many, he advocates paying the utmost respect to anyone, human or divine, who in terms of wisdom or expertise is one's own superior. This is not, of course, to deny that ultimately Socrates must make his own independent check on the information received, at least from any human authority, cf. Phdr. 21 4c 1-4, where Socrates clearly says that the ancients themselves know the truth, but that it is better for us, if we can, to find it out for ourselves and thereafter to remove our attention from these human authorities. (Thus Brisson, L., Platon, Phèdre (Paris 1989)Google Scholar, contrary to the more favoured but linguistically strained interpretation that the ancients alone know whether it is true or not; at Tim. 40d-e, in a case where the ancients’ word cannot be checked, Timaeus insists that it must simply be accepted as true.) In fact Socrates even set out to check the word of a divine authority, the Delphic oracle, despite an extreme reluctance to believe that it could be wrong (Apol. 21b-c).

¹⁷ Cf. also Laws 714a, 957c, where the association of νόμος with νοῦ διανομή, the ‘dispensation of (or ‘by’) intelligence’, is meant to invoke law's mythical origin in the age of Cronos as a divine benefaction to mankind, and therefore implicitly the original meaning of the word. Similarly, at Laws 654a the suggested derivation of the word χορός from χαρά is attributed to the gods who first instituted dance.

¹⁸ Charm. 173e-174a, Ion 531b, Lack 196d, 199a, cf. Tht. 179a; Rep. 389d, Laws 828b.

¹⁹ Pl. Crito 44a-b. Many scholars, following the lead of Lambinus, have thought that Socrates’ interpretation of the dream was itself based, if not on an etymology, at any rate on a linguistic decoding—that of ϕθίνειν from Φθία.

²⁰ Cf. Apol. 22b-c, Ion 534c-d, Meno 99c.

²¹ For a full list of occurrences, see Riginos, A.S., Platonica (Leiden 1976) 35-8Google Scholar.

²² Notopoulos, J., ‘The name of Plato’, CP 34 (1939) 135–45Google Scholar.

²³ Fraser, P.M. and Matthews, E., A Lexicon of Greek Personal Names, 2, ‘Attica’, ed. Osborne, M.J. and Byrne, S.G. (Oxford 1994)Google Scholar.

²⁴ E.g. no one ascribes Aristotle's name—Άριστοτέλης = ‘best goal’?—to his pioneering work in teleology. Nor, when (exceptionally) Pythagoras’ name was etymologized (Diog. Laert. 8.21), did any name-change story result.

²⁵ Fr. 5A FHS&G.

²⁶ At Crat. 397a-b, the mechanical naming of a child after an ancestor is judged liable to be incorrect. This may have been one ground for rejecting ‘Aristocles’. Another may have been that any name with the ‘-cles’ termination, signifying a kind of ‘fame’, must fail to convey the subject's essence. Contrast the ‘crat-’ element in two names of which Cratylus does approve, Socrates and Cratylus, ‘power’ being an intrinsic property (I owe this last point to C.D.C. Reeve's draft introduction to his forthcoming translation of the Cratylus).

²⁷ Even if one shares the scepticism some have expressed about the accuracy of this report as concerns Plato's philosophical development, and sees it as an inference based largely on the Cratylus itself, there is no way that Aristotle could have inferred Plato's early association with Cratylus from the dialogue, and the overwhelmingly most probable explanation is that it is a simple biographical fact which he learnt directly from Plato.

²⁸ The only ancient suggestion as to who instigated the change is the one at Diog. Laert. 3.4, derived from Alexander Polyhistor (first century BC), that it was due to his gymnastics trainer Ariston of Argos; but this suggestion is an inference from the already dubious conjecture that the name reflected Plato's breadth of body.

²⁹ See esp. R. Barney (n.2) for a convincing interpretation of Socrates’ etymological performance as an ‘agonistic display’. Cf. also Dalimier, C., Platon, Cratyle (Paris 1998) 16–17.Google Scholar

³⁰ The Derveni commentator, like Euthyphro, combines being a mantis with being a purveyor of etymologies. Cf. Kahn, C.H., ‘Was Euthyphro the author of the Derveni Papyrus?’, in Laks, A. and Most, G.W. (eds.), Studies on the Derveni Papyrus (Oxford 1997) 55–63Google Scholar.

³¹ The ‘name-maker’ (όνοματουργός) is quickly identified with a ‘rule-maker’ (νομοθέτης) at 388e-389a. I suspect that this latter designation picks up the idea found at Hipp. De arte 2 that names are mere νομοθετήματα, while εϊδεα are βλαστήματα ϕύσιος, a contrast close to the views of Socrates’ current interlocutor Hermogenes. At all events, the casual and unexplained reference to the original name-maker as ό νομοθέτης at Charm. 175b 4 suggests that the designation was readily understood.

³² On the acceptability of multiple etymologies, cf. Dalimier (n. 29) 43-4. See 406b 5-6 for the principle applied to Artemis, and the multiple etymology of Apollo at 404e-406a. The same principle will explain why Cronos is allowed two etymologies (396b, 402b). Sometimes one decoding may commend itself as superior to another (399d-400b, 404b), but it is only when the two decodings of a word contradict each other that one must necessarily be rejected, as in the case of epistēmē, p. 151 below). Myles Burnyeat (in conversation) has illuminatingly compared the double meanings of many modern acronyms (cf. p. 142 above on ‘Basic’).

³³ 409d-410a.

³⁴ In the table which follows, square brackets indicate subordinate or digressive etymologies, round brackets indicate words which are listed but not etymologised (at least here). I have found no comparably full analysis in the modern literature, although Brumbaugh, R., ‘Plato's Cratylus: the order of the etymologies’, Review of Metaphysics 11 (1957-1958) 502–10Google Scholar does discern some natural philosophical sequences amounting to a ‘double dialectical progression from complex to simple, from thing to thought to name’.

³⁵ M 7.16.

³⁶ Xenocrates frr. 264-6 Isnardi Parente = fr. 53 Heinze: a verbatim quotation from Xenocrates’ Life of Plato preserved by Simpl. In Ar. Phys. 1165.33 ff., In Ar. De caelo 12.22 ff., 87.23 ff. For a survey of modern dismissals, see Isnardi Parente (Senocrate-Ermodoro, frammenti, Naples 1981) 433-5.

³⁷ E.g. 402b, 404d.

³⁸ There is an obvious parallel at Tht. 152e-153d, where the universality and creativeness of flux are argued with frequent appeal to early philosophers and, especially, poets. There is no doubt much irony in the suggestion that these people must have hit on the truth, but I see no reason to think Plato is being ironic in interpreting them as holding such a view. Thus in establishing his predecessors’ emphasis on flux Plato finds an encouraging harmony between the findings of etymology and those of textual exegesis and allegoresis.

³⁹ The new list, at 437a-c, is: έπιστήμη, βέβαιον, ίστορία, πιστόν, μνήμη, άμαρτία, συμϕορά, άμαθία, άκολασία.

⁴⁰ Despite the apparent lack of parallels, έμβάλλειν here gives the impression of meaning, not simply ‘insert' (as elsewhere), but ‘insert an aspiration’. Thus 412a 3-4, διδ δὴ έμβάλλοντας δεῖ τὸ εί έπιστήμην αύτὴν καλεῖν, should mean ‘So one should aspirate the e and call it “hepisteme”.’ And 437a 5-8, καὶ όρθότερόν έστιν ὥσπερ νῦν αύτοῦ τὴν άρχὴν λέγειν μᾶλλον ή έμβάλλοντας τὸ εί έπιστήμην, άλλὰ τὴν έμβολὴν ποιήσασθαι, άντὶ τῆς έν τῶι εί, ὲν τῶι ίῶτα, should mean ‘And it is more correct to say the beginning of it as we now do, rather than to aspirate the e and say ‘hepisteme’, and to make the aspiration in the i instead of the e [i.e. to say ‘ep-histeme’].’ If however this is thought unacceptable, see the revised OCT (n.4) for H. Schmidt's simple emendation (represented there by an ad hoc typographical device), which yields much the same sense. (I am grateful to Malcolm Schofield, Reviel Netz and David Robinson for help in understanding these two passages.)

⁴¹ That the earlier impression of consistency, now being overturned, was itself one of philosophical consistency is confirmed by 418e-419a.

⁴² This is well argued by David Robinson, ′Kρόνος, Kορόνους and Kρουνός in Plato's Cratylus’, in Ayres, L. (ed.), The Passionate Intellect (New Brunswick and London 1995) 57–66Google Scholar. For the association of Cronos with nous, see also n.17 above on Laws 714a, 957c.

⁴³ I take it that, as at 404c-d, the alpha prefix is meant to be the one which signifies ‘(all) together’, rather than the privative.

⁴⁴ 403e-404a. At 403e 4-5, Hades is ‘a τέλεος σοϕιστής, and a great benefactor of those who are with him’. The context makes it clear that σοϕιστής, here is used in a primarily positive sense, ‘sage’, as at Rep. 10.596d 1, Smp. 203d 8, 208c 1; this is well noted by Dalimier (n.29) 234, and fully argued by Wohlfahrt (n.2).

⁴⁵ The etymology is implicit in Philolaus A27, where ϕλέγμα likewise is explicitly etymologised as also having something to do with temperature.

⁴⁶ See e.g. Montrasio (n.2) for a thoroughly Platonising reading of the multiple Apollo etymologies (404d-406a).

⁴⁷ I owe this last suggestion to Myles Burnyeat, whose ideas were a constant support and inspiration to me in developing the present interpretation at a Cambridge seminar on the Cratylus held in 1994-5. I thank all participants in that unusually embattled seminar—I should mention in particular Malcolm Schofield, Robert Wardy, Geoffrey Lloyd and Reviel Netz—and also audiences at the Center for Hellenic Studies (Washington) in March 1996, at Princeton University in April 1996, and at UC Berkeley and at Pomona College in April 1997, for further discussion of the issues. Thanks for helpful comments on earlier drafts are owed to M.M. McCabe, Myles Burnyeat, David Reeve, Voula Tsouna, Francesco Ademollo, Barbara Anceschi, Gabor Betegh, Fernanda Decleva Caizzi, Malcolm Schofield, Catherine Dalimier, and the Editor and anonymous referee of this journal.