¹ Portrait of Socrates, being the Apology, Crito, and Phaedo of Plato in an English translation (that of Benjamin Jowett, 3rd edn, Oxford 1892) with introd. and notes (Oxford 1938) 26.

² Plato's Apology of Socrates: A Literary and Philosophical Study with a Running Commentary, ed. and completed from the papers of the late De Strycker, E., SJ by Slings, S.R.Google Scholar (Mnemosyne Suppl. 137, Leiden, New York and Cologne 1994) 140Google Scholar.

³ Burnyeat, M.F., ‘Virtues in action’, in Vlastos, Gregory (ed.), The Philosophy of Socrates: A Collection of Critical Essays (New York 1971) 210Google Scholar.

⁴ Cf. the contrast between mortal gold, which the Guards of the ideal city are not allowed to possess, and the divine gold they have in their souls from the gods (Rep. 3.416e-417a), a contrast echoed later as their being not poor (save financially) but by nature rich (8.547b).

⁵ Vlastos, Gregory, Socrates: Ironist and Moral Philosopher (Cambridge 1991) 219CrossRefGoogle Scholar with n.73.

⁶ Burnet, John, Plato's Euthyphro, Apology of Socrates, and Crito, ed. with notes (Oxford 1924) 124Google Scholar.

⁷ Oeuvres de Platon 1 (Paris 1956)Google Scholar.

⁸ Platon: Apologie de Socrate, Criton, Traductions inédites, introductions et notes par Luc Brisson (Paris 1997) n.173, referring to Vlastos (n.5) 303-8. A similar account of the traditional translation in Brickhouse, Thomas C. and Smith, Nicholas D., Plato's Socrates (New York and Oxford 1994) 20Google Scholar with n.33.

⁹ Leipzig 1936. The ominous title heralds Hildebrandt's long introduction, where he enlists both Socrates and Plato for the Fascist cause.

¹⁰ Protreptik und Paränese bei Platon: Untersuchungen zur Form des platonischen Dialogs (Stuttgart 1959) 109Google Scholar with n.113.

¹¹ Cornford, F.M., Before and After Socrates (Cambridge 1932) 36CrossRefGoogle Scholar; he does not cite Burnet, because he is writing for a non-scholarly audience.

¹² Taylor, John Hammond, SJ, ‘Virtue and wealth according to Socrates (Apol. 30b)’, Classical Bulletin 49 (1973) 49–52Google Scholar.

¹³ Indianapolis 1989. See pp. 124-5 with n.21.

¹⁴ Indianapolis 1997. This will be the standard complete works in English translation for a good while to come.

¹⁵ The Trial and Death of Socrates (Indianapolis 1980)Google Scholar.

¹⁶ Plato: Apology, with introd., tr. and comm. (Warminster 1997), note ad loc. See also Stokes's review of De Strycker, and Slings, , Archiv für Geschichte der Philosophie 78 (1996) 192–8Google Scholar.

¹⁷ De Strycker and Slings (n.2) 334.

¹⁸ Several of these texts are cited by Vlastos (n.5) 214-32 to argue that in Socrates' own view wealth is a ‘non-moral good’ whose value, however, is minuscule compared to the good of virtue. His argument, which has been influential (see nn.8 above and 25 below), ignores the dramatic contexts within which wealth is called good.

¹⁹ De Strycker and Slings (n.2) 234-5; I extend their quotation by one further sentence.

²⁰ On the nuances of the combination ἢ καί (which De Strycker and Slings render ‘or else’), see Denniston, J.D., The Greek Particles (2nd edn, Oxford 1954) 306Google Scholar: ‘Sometimes καί means “also”, or marks a climax, “even”.’

²¹ De Strycker and Slings (n.2) 139.

²² De Strycker and Slings (n.2) 235. For the qualification ‘supposed’, cf. also 3.406c 7: τῶν πλουσίων τε καὶ εὑδαιμόνων δοκούντων εἶναι.

²³ Elsewhere, Plato locates the origin of war in the desire for money and possessions: Rep. 2.373d-374a, Phd. 66cd.

²⁴ P. 138 n.39.

²⁵ De Strycker and Slings' position on what they call ‘minor goods’ is close to that of Vlastos (n.5) ch.8 on ‘mini-goods’; my objection is indicated at n.18 above.

²⁶ See Mansfeld, Jaap, ‘Notes on the Didaskalicus’, with a mass of references to the relevant texts, in Joyal, Mark (ed.), Studies in Plato and the Platonic Tradition: Essays Presented to John Whittaker (Aldershot 1997) at 248–54Google Scholar.

²⁷ Here I expand a tiny bit, guided by Cooper, John, ‘Aristotle on the goods of fortune’, Philosophical Review 94 (1985) 173–96CrossRefGoogle Scholar, repr. in his Reason and Emotion: Essays on Ancient Moral Psychology and Ethical Theory (Princeton 1999) ch.13Google Scholar.

²⁸ They do cite the passage (p. 140 n.44), but as a parallel for their understanding of the Apology. It was Newman, W.L., The Politics of Aristotle 3 (Oxford 1902)Google Scholarad 1323a 36 and 40, who suggested that Aristotle is actually drawing on Apol. 30b.

²⁹ Quite possibly the Protrepticus: Ingemar Düring, Aristotle's Protrepticus: An Attempt at Reconstruction (Göteborg 1961) 254–6Google Scholar.

³⁰ Here I am indebted to correspondence with John Cooper.

³¹ LSJ s.v. III 6.

³² Taylor (n. 12) 51; Stokes (n. 16) 150.

³³ From a voluminous literature, I pick out for their excellence two writers in particular. First, Kahn, Charles H., ‘Why existence does not emerge as a distinct concept in Greek philosophy’, Archiv für Geschichte der Philosophie 58 (1976) 323–34Google Scholar, and ‘Some philosophical uses of “to be” in Plato’, Phronesis 26 (1981) 105–34CrossRefGoogle Scholar, which distils much of his previous work on the subject (listed at his n.45) going back to his massive study The Verb ‘Be’ in Ancient Greek (Foundations of Language, Supplementary Series 16 (1973), repr. Minneapolis 2003). And more recently, Brown, Lesley, ‘The verb “to be” in Greek philosophy: some remarks’, in Everson, Stephen (ed.), Companions to Ancient Thought 3: Language (Cambridge 1994) 212–36Google Scholar, which generalizes the lessons of her pioneering ‘Being in the Sophist: a syntactical enquiry’, Oxford Studies in Ancient Philosophy 4 (1986) 49–70Google Scholar. The idea of extending their approach from εἶναι to γίγνεσθαι is my own initiative.

³⁴ For critical remarks about standard editorial practice, see Barrett, W.S., Euripides: Hippolytus, ed. with introd. and comm. (Oxford 1964) 424–7Google Scholar, and Kahn (n.33: 1973) Appendix A, 420-5.

³⁵ The example is gratefully borrowed from Brown's first article (n.33: 1986), but my use of it is more limited than hers. Her thesis that ‘to teach’ is ‘a verb of variable polyadicity’, in that it can be added to indefinitely (‘I am teaching French to small children’, ‘with enthusiasm’, etc., etc.), implies commitments in semantic theory which I do not wish to incur, let alone extend to εἶναι and γίγνεσθαι in ancient Greek.

³⁶ For obvious reasons, Plato's language is the main focus of this study. Kahn (n.33: 1973) gives a broader treatment of εἶναι in Homer and authors of the Classical period, which establishes beyond doubt that Plato's use of the verb is typical, however novel the philosophical theory he builds on it. We shall see that the same is true of Aristotle and the other philosophical authors considered below.

³⁷ In Russian (the language for which this essay on cross-cultural translation was originally written), as in some other Indo-European languages, the present indicative of the verb ‘to be’ is ‘unmarked’. Between two nouns it may be indicated by a dash; in conversation, one simply goes straight from subject to predicate without a word between, just like Plato's Greek in the passage quoted. Note that omission is not restricted to (and so is no criterion for) non-existential uses of εἶναι: Kahn (n.33: 1973) 264 n.32, where an example like Horn. Od. 13.1023 could go over into Russian word-for-word.

³⁸ I have discussed the philosophical significance of this argument in ‘Plato on the grammar of perceiving’, CQ n.s. 26 (1976) 29-51, and its place in an overall interpretation of the dialogue in The Theaetetus of Plato (Indianapolis 1990) 52–65Google Scholar.

³⁹ For scepticism about the standard dating of this tract to around 400 BC, see my entry ‘Dissoi Logoi’ in Craig, Edward (ed.), The Routledge Encyclopedia of Philosophy (London 1998) 106–7Google Scholar.

⁴⁰ As a case where translation in terms of existence is entirely appropriate, Brown (n.33: 1986) 63-4 aptly cites the Sophist's review of theories about what there is, beginning at 242c.

⁴¹ This is but a brief summary of the points relevant to my discussion. For a fuller treatment of the way τὸ ὄν in (i) unpacks into the explicitly predicative εἶναι of (ii), see Kahn (n.33: 1981) 112-14.

⁴² Just this absurdity is found in the very first English translation of the Republic (Spens 1763: ‘between existence and non-existence’), and occasionally in its modern successors (e.g. Lee 1955). Some English translators prefer ‘between unreality and perfect reality’, vel sim., because degrees of reality make better sense than degrees of existence (Cornford 1941, Sterling and Scott 1985, Halliwell 1993, Waterfield 1993). Those who offer ‘between nonbeing and pure being’, vel sim. (Jowett 1875, Bloom 1968, Grube 1974), do so in a context where ‘being’ need not be understood as ‘existence’, because they have used the indeterminate ‘what is’ and ‘what is not’ since 476e 10. But the two best translations to date (Shorey 1930-35, Lindsay 1935) stick close to the Greek: ‘between that which is not and that which purely is’, vel sim. (likewise Reeve's 1992 revision of Grube). This does justice to the fact that που at 479d 4 picks up the earlier locative designation of τὸ ὅν as the domain or province (the ἐφ' ὦι an) of knowledge, and of τὸ ὅν as the domain or province of ignorance (477a 9-10).

⁴³ This choice example (later echoed in both content and syntax by Epictetus, Diss. I 12.1) arrived in a letter from Lesley Brown. Earlier in the same discussion, the Athenian undertook to defend the thesis ὡς θεοί τε εἰσὶν καὶ ἀγαθοί, δίκην τιμῶντες διαφερόντως ἀνθρώπων (887b 7-8).

⁴⁴ Such exceptions are well discussed by Brown (n.33: 1994) 233-6. The first systematic challenge to the general rule came from the Stoics, who distinguish the class of beings (ὄντα), provocatively restricted to bodies, from the most general class of'somethings’ (τινά), which additionally includes certain incorporeal items like void, place, time and λεκτά (‘sayables’). In effect, the Stoics allow ‘x ἐστι F’ to range more widely than ‘x ἐστι’, blocking the inference from the first to the second. For a valuable discussion of this doctrine and its anti-Platonic import, see Jacques Brunschwig, ‘The Stoic theory of the supreme genus and Platonic ontology’, in his Papers in Hellenistic Philosophy (Cambridge 1994) 92–157CrossRefGoogle Scholar. The inference was defended (question-beggingly) by later Aristotelians (Alex. Aphr. In Top. 301.19) as well as Platonists (Plot. VI 1 [42] 25, 9-10).

⁴⁵ Wiedemann, Hermann, Aristoteles: Peri Hermeneias, tr. and comm. (Berlin 1994) 178–87Google Scholar, provides an exhaustive account of the debate from antiquity into modern times.

⁴⁶ So Whitaker, C.W.A., Aristotle's De Interpretatione: Contradiction and Dialectic (Oxford 1996) 55–9Google Scholar (cf. 30-2), arguing against Ackrill, J.L., Aristotle's Categories and De Interpretatione, tr. with notes and comm. (Oxford 1963)Google Scholarad loc.

⁴⁷ For reason to doubt that the differences between predicative and identity statements are due to different meanings of ‘is’ or ἐστι, see Mates, Benson, ‘Identity and predication in Plato’, Phronesis 24 (1979) 216–20CrossRefGoogle Scholar.

⁴⁸ Distinguished from it by ἤ at line 3, alongside it because ἔνιαι μέν contrasts with ἔνιαι δέ at line 6, where quantity and other dependent categories come in.

⁴⁹ De Dis. 3, col. 10, 35 Diels, De Pietate col.22, 628 Obbink, both about the existence of gods, ὕπαρξις = property comes earlier.

⁵⁰ A different use of the phrase occurs in Galen's near-contemporary Alexander of Aphrodisias, In Top. 52.25-53.10, where ἁπλῆ ὕπαρξις is belonging to something simpliciter as opposed to belonging as its genus; it contrasts with the different ways of being or belonging to a subject (τρόποι ὑπάρξεως) determined, not by the theory of categories, but by the Topics doctrine of predicates. In general, ἁπλοῦς, and ἁπλοῶς are devices for setting aside whatever qualifications are relevant in a given context.

⁵¹ Examples: Arist. Meteor. 2.8.365b 24, Part. an. 4.10.688a 21; Sext. Emp. Pyr. 2.5, Math. 8.305, 9.182; Alex. Aphr. In APr. 275.21; Plot. 14 [46] 3, 28-9, II 1 [40] 2, 27-8.

⁵² For the good reason that ‘being there already’ is what the verb expresses on its second extant occurrence (Pind. Pyth. 4.205) and frequently thereafter. In a valuable (and humorous) article, ‘The origin of ὑπάρχω and ὕπαρξις as philosophical terms’, in F. Romano and D.P. Taormina (eds), Hyparxis e Hypostasis nel Neoplatonismo (Atti del I Colloquio Internationale del Centro di Ricerca sul Neoplatonismo, Florence 1994) 1-23, John Glucker classifies the various uses of ὑπάρχειν in both philosophical and non-philosophical authors of the fifth and fourth centuries BC, and shows how they all relate in one way or another to the idea of being there already. He speaks more readily than I would of different ‘senses’ of the verb, overestimates the extent to which its range narrows later under Stoic influence, and wrongly assigns the first existential use of the noun ὕπαρξις to Philo Judaeus. But these minor disagreements still leave me in debt to his helpful contribution.

⁵³ Opif. 170. Cf. Opif. 172; Spec. 1.41, 2.225; Aet. 53 and 70.

⁵⁴ Praem. 40, reading ὀνόματι with Colson and all MSS except A, against Cohn, who prints A's ὄνομα.

⁵⁵ Post. 168.

⁵⁶ Examples: Plut. De E 387c; Apollonius Dyscolus, Conj. 216.11-16 (where ὕπαρξις; contrasts with ἀναίρεσις, something's not being the case); Sext. Emp. Pyr. 2.5, Math. 8.304. So too in Galen himself ὕπαρξις (without ἁπλῆ) sometimes represents the use of ὑπάρχειν to signify the obtaining of a whole state of affairs: Inst. Log. 3, p.7.13; p.8.8-9; 4, p.9.21; 5, p. 12.17. In the Stoic definition of a true proposition under attack at Sext. Emp. Math. 8.85-6 (φασὶ γὰρ ἀληθὲς μὲν εἶναι ἀξίωμα ὂ ὑπάρχει τε καὶ ἀντικεῖταί τινι) ὑπάρχειν cannot mean ‘to exist’ because there are false propositions as well as true. The truth of a Stoic ἀξίωμα (a non-linguistic item expressible by a sentence) is something's being the case, a whole state of affairs obtaining; Gal. Inst. Log. 15, p.35.12 and 17, or Sext. Emp. Pyr. 1.14 may serve to illustrate ὑπάρχειν used in accordance with this definition. There is even a word ὑπακτικός, glossed by Ammonius, In Int. 27.12-13 as ‘expressive of one thing's belonging or not belonging to another’, which corresponds (for subject-predicate propositions) to the modern logician's ‘having truth-value’. It is then no surprise to find that in modal logic ὑπάρχουσα πρότασις is an ‘assertoric’ premise of the form ‘x belongs to y’ (with ὑπαΡχόντως, not in LSJ, the associated adverb), as opposed to an ‘apodeictic’ one of the form ‘x necessarily belongs to y’: e.g. Alex. Aphr. In APr. 124.21-8.

⁵⁷ A very clear example is Alex. Aphr. In Top. 375.1617, 23-4.

⁵⁸ Porph. ap. Simpl. In Cat. 34.21-3; Ammon. In Cat. 20.26-21.1.

⁵⁹ Alex. Aphr. In Metaph. 399.14-16: ἡ γάρ ἑκάστου οἰκεία ὕπαρξις οὐσία ἐκείνου.

⁶⁰ In Cat. 67.27-36, where it is handy to have a term which is neutral between the various Aristotelian categories in a way that ὄντα cannot officially be.

⁶¹ On the importance of distinguishing the ‘input question’ (what meaning a word brings to its sentential context) from the ‘output question’ (what meaning it has in that context), see David Wiggins, ‘On sentence-sense, word-sense and difference of word-sense. Towards a philosophical theory of dictionaries’, in Steinberg, Danny D. and Jakobovits, Leon A. (eds), Semantics: An Interdisciplinary Reader in Philosophy, Linguistics and Psychology (Cambridge 1971) 14–34Google Scholar.

⁶² From a paragraph which is itself a vivid illustration of the impossibility of distinguishing an existential as opposed to copulative meaning for either verb.

⁶³ Here again I follow Brown (n.33: 1994). Compare Kahn (n.33: 1976) 333: ‘Thus for Aristotle, as for Plato, existence is always εἶναί τι, being something or other, being something definite. There is no concept of existence as such, for subjects of an indeterminate nature … Platonic Greek for “X exists” is “X is something”, εἶναί τι.’ For examples of this use of εἶναί τι see Phd. 74a 912, 102b 1, Rep. 9.583c 5, 584d 3, Tht. 157a 3 and 5, Soph. 246e 5, 247a 9, Tim. 51b 7-8, Phlb. 37a 2-9. The idiom is less frequent in Aristotle, because of his technical contrast between εἶναί τι and εἶναί ἁπλῶς, but examples abound in his discussion of place at Phys.5.1-5 (210a12, etc.).

⁶⁴ Fr. 27 Wehrli = Schol. In Ar. APr. I cod. 1917 in margine p. 146a 24-7 Brandis.

⁶⁵ Cf. APr. 1.1.24b 16-18; 3.25b 22.

⁶⁶ Wiedemann's lengthy review of rival interpretations of the passage (n.45, 327-38) nowhere pauses to defend the assumption (written into his translation) that Aristotle switches from ἐστι as ‘Existenzprädikat’ to ἐστι as ‘Kopula’; nor does he record anyone else doing so. About γίγνεται both he and Ackrill (n.46) remain silent.

⁶⁷ This last is the use that Aristotle invokes for a list of the three types of non-substantial change at Cael. 1.7.274b 15-16 (εἰ αδύνατον γενέσθαι λευκὸν ἢ πηχυαῖον ἢ ἐν Αἰγύπτωι) so please do not think that γίγνεσθαι ἐν Λυκείωι would have to mean ‘be born in the Lyceum’. At Hdt. 5.33 Megabates set sail and in due course ἐγένετο ἐν Χίωι. Against the idea that ‘to be born’ is the root meaning of γίγνεσθαι, see Kahn (n.33: 1973) 384-5.

⁶⁸ At 317a 33 Aristotle uses κυρίως as a synonym for ἁπλῶς; at 317a 17 he speaks of ἡ ἁπλῆ καὶ τελεία γένεσις. Briefer treatments of the contrast can be found at Phys. 5.1.225a 12-17, Metaph. 8.1.1042a 32-b 8.

⁶⁹ Here I am indebted to Sarah Broadie.

⁷⁰ Aristotle's De Generatione et Corruptione, tr. with notes (Clarendon Aristotle Series, Oxford 1982) ix-xv, cf. p.83Google Scholar.

⁷¹ Gottlob Frege, Die Grundlagen der Arithmetik (1884) §53; my example ‘Tame tigers exist’ comes from an often cited debate between Kneale, W. and Moore, G.E. on the question ‘Is existence a predicate?’ (Proceedings of the Aristotelian Society, Supplementary vol. 15 (1936) 154-88CrossRefGoogle Scholar; Moore, G.E., Philosophical Papers (London and New York 1959) 115–26Google Scholar).

⁷² The Works of Aristotle Translated into English, 8: Metaphysica (2nd edn, Oxford 1928)Google Scholar.

⁷³ Much less happy is H. Tredennick's Loeb translation (London and Cambridge, MA 1933): ‘Everything which is generated is generated by something and from something and becomes something.’ The best German translation, that of Frede, Michael and Patzig, Günther, Aristoteles ‘Metaphysik Z’: Text, Übersetzung und Kommentar (2 vols, Munich 1988)Google Scholar, has ‘Aber alles, was entsteht, entsteht unter Einwirkung von etwas, wird aus etwas und wird zu etwas.’ They too find it necessary to change the verb. The most widely cited French translation is Tricot, J., Aristote: La Métaphysique, new tr. and notes (new edn, Paris 1953)Google Scholar, where we read ‘Tout ce qui devient, devient, par quelque chose et à partir de quelque chose, quelque chose.’ This mirrors the Greek nicely, thanks to the fact that ‘devenir’ is rarely existential in meaning (here I am indebted to advice from Francis Wolff). A similar effect can be achieved in English by changing the word-order: ‘Everything that comes to be comes to be something, from something, by the agency of something’, which, like the French, is most naturally heard as confined to non-substantial change. The drawback is that this English, like that French, fails to capture the whole of what Aristotle intends.

⁷⁴ And continued to do so into late antiquity, beyond the authors cited n.51 above: witness Hesychius s.v. σῶς: ὁ ὁλόκληρος, καὶ τέλειος. ὁ σωζόμενος, καὶ σῶς ὑπάρχων (example owed to Michael Frede).

⁷⁵ Examples: Lucr. 3.97; Cic. De Orat. 2.54.217.

⁷⁶ As at Math. 10.4 he substitutes predicative ὑπάρχειν for the εἶναι he used in virtually the same sentence at Pyr. 3.124. For this and other examples of the interchangeability of the two verbs in Sextus, see Janáček, Karel, Prolegomena to Sextus Empiricus (Acta Universitatis Palackianae Olomucensis, 1948) 42–4Google Scholar.

⁷⁷ Is Galen updating Aristotle's statement (APr. 1.36.48b 2-4) that ὑπάΡχειν is said in as many ways as εἶναι?

⁷⁸ Cf. the translations ‘real purpose’ in Bailey, Cyril, Epicurus: The Extant Remains (Oxford 1926)Google Scholar, ‘il fine realmente dato’ in Arrighetti, Graziano, Epicuro, Opere: Introduzione, traduzione e note (Turin 1967)Google Scholar.

⁷⁹ Admittedly, Plotinus is attacking the Stoics here, not expounding his own philosophy, and I know no other place in Plotinus where ύπόστασις admits of degrees. But compare, in the same treatise, μᾶλλον ὄν (26, 8), μᾶλλον οὐσίας (27, 37), μᾶλλον εἴη (28, 16), μᾶλλον οὐκ ὄν (29, 24). My concern is with language, not philosophy, and one case is enough to establish the grammatical possibility of grading ὐπόστασις in the same way as οὐσία.

⁸⁰ Cf. n.42 above.

⁸¹ Stob. Ecl. 1.106.18-23W = SVF 2.509 = LS 51B (4); Plut. Comm. not. 1081f = SVF 2.518 = LS 51C (5).

⁸² For a judicious treatment of the texts and the debate they have prompted, see Schofield, Malcolm, ‘The retrenchable present’, in Barnes, Jonathan and Mignucci, Mario (eds), Matter and Metaphysics (Proceedings of the Fourth Symposium Hellenisticum, Naples 1988) 329–74Google Scholar.

⁸³ The basic study by H. Dörrie, ‘Ύπόστασις: Wort-und Bedeutungsgeschichte’, Gött. Nachr., Philol.-hist. Kl. 1955, 35-92, is not superseded by the essays in Romano and Taormina (n.52), which mostly focus on what the words are applied to rather than on the prior issue of the meaning in virtue of which they can be so applied; in Fregean terms, on reference instead of sense. As a result, they deliver much arcane metaphysico-theological doctrine, but (apart from Glucker) scant linguistic analysis; doctrinal differences are one thing, semantic differences another. More promising is the approach taken by Damascius, De Principiis 2.74.23-77.24 Westerink (= Ruelle Vol. 1, ch.62): he compares εἶναι with six other Greek verbs for ‘to be’ (ὑφεστάναι ὑπάρχειν, τελέθειν, πέλειν, σώιζεσθαι, τυγχάνειν), arguing that they differ in meaning both from εἶναι and from each other even though they share uses, including an existential use, in common. Better still would be treatment by the methods of transformational grammar, which replace (and thereby illuminate) ontology in Kahn (n.33: 1973) and epistemology in Lyons, John, Structural Semantics: An Analysis of Part of the Vocabulary of Plato (Oxford 1963)Google Scholar.

⁸⁴ OED (2nd edn, Oxford 1989)Google Scholar s.v.

⁸⁵ Paris 1948, 15. I thank Michael Screech for checking French lexicographical resources to confirm that Gilson's claim is substantially correct.

⁸⁶ The Theory of Forms itself was first introduced earlier in Book 5 at 475e-476d. But that passage has none of the subsequent emphasis on the changeability of sensible things.

⁸⁷ I say ‘deliberately’ because the Book 5 argument was designed to soothe the lovers of sights and sounds and persuade them that they lack knowledge, without blatantly telling them that, from an epistemological point of view, they are sick (476d 8-e 2). To this end, Socrates kept his hand close to his chest, not revealing until later the full import of the various admissions he secured from his interlocutors, who refuse to accept the existence of Forms. For a pioneering account of what the argument with the lovers of sights and sounds does and does not presuppose, see Gosling, J.C., ‘Δόξα and δύναμις in Plato's Republic’, Phronesis 13 (1968) 119–30CrossRefGoogle Scholar.

⁸⁸ 6.508d 7, 7.521d 4, e 3, 525b 5, 526e 7, 527b 5-6, 534a 3.

⁸⁹ For more on this subject, from a different angle, see Frede, Michael, ‘Being and becoming in Plato’, Oxford Studies in Ancient Philosophy, Supplementary vol. 1988, 37–52Google Scholar.

⁹⁰ There are parallels at Parm. 153c 7-e 3. My construal of the Sophist passage follows Cornford, F.M., Plato's Theory of Knowledge (London 1935) 225Google Scholar n.2 and 226.

⁹¹ Compare the varying translations and notes in, e.g., Léon Robin's Budé edition (Paris 1947); de Vries, G.J., A Commentary on the Phaedrus of Plato (Amsterdam 1969) 122-3Google Scholar; Rowe, C.J., Plato: Phaedrus with tr. and comm. (Warminster 1986)Google Scholar.

⁹² The context makes clear that the principle we are discussing is the ultimate first principle of all movement.

⁹³ See Burnet's apparatus, or the fuller one in Robin, which quotes several proposed emendations. Cicero translates, ‘nec enim esset id principium, quod gigneretur aliunde’. Note how he puts the consequent of the conditional first, because the passive ‘gigneretur’ cannot take a complement to match Plato's ἀρχὴ γίγνοιτο.

⁹⁴ There is a section on hyperbaton in the ‘Digest of Platonic idioms’ affixed to the Rev. James Riddell's edition of the Apology (Oxford 1877; repr. separatim Amsterdam 1967), and many examples from Plato in J.D. Denniston, Greek Prose Style (Oxford 1952) ch.3, ‘The order of words’; a recent, more theoretical treatment, again with numerous examples from Plato, is A.M. Devine and L.D. Stephens, Discontinuous Syntax: Hyperbaton in Greek (New York and Oxford 2000). From antiquity we have a superb account of the rhetorical effectiveness of hyperbaton in Longinus, On the Sublime 22. The technical term ὑπερβατόν occurs already in Plato, Prot. 343e 3, in a context which assumes that readers need no elaborate explanation of what it is.

⁹⁵ On the Apology as both defence and counter-accusation, see my ‘The impiety of Socrates’, Ancient Philosophy 17 (1997) 1–12CrossRefGoogle Scholar, repr. in Brickhouse, Thomas C. and Smith, Nicholas D. (eds), The Trial and Execution of Socrates: Sources and Controversies (New York and Oxford 2002) 133–45Google Scholar.

⁹⁶ So Taylor (n. 12) 51.

⁹⁷ Adam, James, The Republic of Plato, ed. with critical notes, comm. and appendices (Cambridge 1902)Google Scholarad loc.

⁹⁸ As already intimated (n.37 above), this essay was originally designed for a Russian audience more familiar with the standard rendering from the translation by Michail Solov'ev than with Burnet's commentary or recent scholarly literature on the philosophically all-important Greek verb ‘to be’. I owe a very great debt to my translator, Irina Levinskaya, for making me explain each unfamiliar point as clearly as possible, to produce an argument that would be satisfactory to us both in either language. (The Russian version, attentive to their translations of all the crucial texts, is due to appear in 2003 in a collection of articles put out by the Philosophy Faculty of St Petersburg University.) At a later stage I benefited from discussion and correspondence with J.N. Adams, Susanne Bobzien, Luc Brisson, Lesley Brown, David Charles, John Cooper, Bruce Fraser, Jaako Hintikka, Edward Hussey, Charles Kahn, Calvin Normore, Dominic Scott, Lucas Siorvanes, Michael Stokes and William Taschek, and from M.L. West's lectures on Greek accentuation; from the opportunity to present the sections on Aristotle at the Fifteenth Symposium Aristotelicum (devoted to Gen. corr. 1) in Deurne, Holland in 1999; from a wide-ranging discussion the same year in the Classics Department at Toronto, followed up by a vigorous letter from Brad Inwood; and from a friendly exchange of views with S.R. Slings.