¹ Classical Quarterly, N.S. III = XLVII (1953), pp. 79–95 (referred to hereafter simply as ‘Owen’).

² Owen, p. 89.

³ Cornford, F. M., Plato's Cosmology, p. 98Google Scholar, n. 4: ‘“The non-existent” means (as in ordinary speech) the absolutely non-existent, of which, as the Sophist shows, nothing whatever can be truly asserted.’

⁴ To one of these I have already had occasion to refer in A.J.P., LXXV (1954), pp. 129–30. Thus far I have seen comments upon Owen's article by Profs.Skemp, J. B. (Plato's Statesman [1952], pp. 237–9)Google Scholar, Field, G. C. (who very generously sent me the text of his unpublished communication summarised in Proc. of the Classical Association, LI [1954], p. 52)Google Scholar, and Vlastos, Gregory (Philosophical Review, LXIII [1954], p. 334, n. 20, and p. 335, n. 29)Google Scholar; but in none of these is there any reference to Owen's, use of Timaeus 38B2–3.Google Scholar

⁵ Rhein. Mus., L (1895), p. 429, n. 2, reprinted in his Platonische Aufsätze (1912), p. 268, n. 1.

⁶ Cf. Fraccaroli, G., Platone: Il Sofista e l'Uomo Politico (1911), pp. 94–5.Google Scholar Among other interpretations which ‘reconcile’ the passage of the Timaeus with the doctrine of the Sophist, cf. especially Susemihl, F., Die Genetische Entwicklung der plat. Philosophie, II, 2 (1860), p. 376Google Scholar; Halévy, E., La Théorie Platonicienne des Sciences (1896), pp. 243–5Google Scholar and 268–70; Hartmann, N., Platos Logik des Seins (1909), p. 134Google Scholar, n. 1; Natorp, P., Platos Ideenlehre (1921), p. 364.Google Scholar

⁷ It is not mentioned in his edition of the Sophist (1897) where the relative chronology of that dialogue is discussed (pp. 37–41), and nothing is said of it in his later translations of the Timaeus (either in the Introduction [p. 20], where the Timaeus is declared to be earlier than the Sophist, or in the note on 38B [n. 73 on p. 161]) and the Sophist (p. 13), where that dialogue is dated c. 364 B.C. In the Introduction to his translation of the Parmenides (p. 13) Apelt states that both Timaeus 38B and the doctrine of τὸ μὴ ὄν in the Sophist are equally results of the same Platonic error, the conception of the copula as ‘Daseinsausdruck’; but it is not suggested that one of the two must be later than the other.

⁸ Sophist 260D1–3, cf. 240D6–241B3.

⁹ Earlier in the dialogue this tenet was ascribed to Protagoras in the defence that Socrates is made to pronounce for him (Theaetetus, 167A7–8: )

¹⁰ Sophist 263B and 263D.

¹¹ Theaetetus 185C–E; cf. Peipers, D., Die Erkenntnistheorie Platos (1874), pp. 535 f.Google Scholar

¹² Owen, pp. 82, 87, 94. Even among scholars who hold the ‘Orthodox opinion’ that the Timaeus is later than the Sophist which is closely preceded in order of composition by the Theaetetus and the Parmenides there has been lively debate as to whether the Theaetetus is earlier or later than the Parmenides: cf. e.g. Diés, A., Parménide, pp. XII–XIIIGoogle Scholar and Théétète, pp. 120–3; Stefanini, L., Platone, I, pp. LXXVIII–LXXXIGoogle Scholar and II, p. 133, n. 1; SirRoss, David, Plato's Theory of Ideas, pp. 6–9.Google Scholar

¹³ Cf. Parmenides 143B3–7 and Sophist 255E4–6.

¹⁴ Parmenides 163C2–164B4 (cf. 163C6–D1 and 164B1–2).

¹⁵ Parmenides 162A6–B4. The question of the text in 162A8-B2 does not affect the crucial point made here, i.e. (scil. μετήχον) (162B1–2). It may be said, however, that despite Dies in his edition and Calogero, (Studi sull' Eleatismo, p. 244, n. 1)Google Scholar the text which Burnet adopted from Shorey, (A.J.P., XII [1891], pp. 349–53)Google Scholar is surely right, for in A8 is guaranteed both by in B4 and by in A5–6 and this in turn requires in B1–2 (cf. Shorey's, reference to Anal. Prior. 51B36 ff.Google Scholar, op. cit., p. 353).

¹⁶ Cratylus 385B.

¹⁷ When later Cratylus in turn denies the possibility of false statement on the ground that whatever one says one says τὸ ὄν (429D), Socrates ironically remarks that this doctrine is too subtle for him and then elicits from Cratylus the grudging admission (430D–431C) that there can be false assertion consisting in the ascription to things of and their combinations, λόγοι, that are dissimilar or inappropriate to them (n.b. 430D5–7, 431B2–C1 and the similarity between the latter passage and Sophist 263D1–4).

¹⁸ Of Owen's statements concerning the relative chronology of the Cratylus (pp. 80, n. 3 [sub fin.]; 82, n. 3; 84, n. 3 [sub fin.]; 85, n. 6) only the last is of any importance; and that, as I have shown elsewhere (A.J.P., loc. cit., note 4 supra), depends upon a misinterpretation of Timaeus 49C–50B.

¹⁹ Euthydemus 284C7–8.

²⁰ As was recognised by Routh, M. J., Platonis Euthydemus et Gorgias (1784), p. 326.Google Scholar

²¹ Euthydemus 284B1–2.

²² When Dionysiodorus takes up the argument of Euthydemus again (Euthydemus 286A ff.), Socrates, ascribing it to the circle of Protagoras and still more ancient persons (cf. Theaetetus 167A7–8, Cratylus 429D2–3), says that he has always wondered at this self-refuting doctrine (286C, cf. 287E–288A).

²³ Cf. Friedländer, P., Platon, II, p. 188.Google Scholar Without citing the passages of the Euthydemus and the Cratylus, Bury, R. G. in 1895 (Journal of Philology, XXIII, pp. 196–7)Google Scholar had maintained that the doctrine of the Sophist concerning as is ‘fundamental for Platonism from first to last’.

²⁴ This chronology too has been proposed, of course. So, for example, Pfleiderer, E. (Sokrates und Plato [1896], pp. 318–20, 330, 333, 342)Google Scholar argued for the order Theaetetus, Cratylus, Sophist, Euthydemus and Natorp, P. (Platos Ideenlehre [1921], pp. 119 and 122–3)Google Scholar contended that both the Euthydemus and the Cratylus were written as appendices to the Theaetetus, though both believedthe Timaeus to be a much later composition than any of these.

²⁵ The same is true of Republic 478B6–C1. The argument there is so similar to that of Theaetetus 189A–B that according to Owen's method of interpretation the Theaetetus ought to be ranked with the Republic and both made to antedate the Euthydemus. Yet, since Republic 477A3–4 and 478D7 show that the of 478B6–C1 is it is possible to interpret this passage asasserting only what is asserted in Sophist 237E and 238C8–10 and as being perfectly compatible therefore with the later logical analysis of false statement in that dialogue; and that Theaetetus 188D–189B is to be explained by a similar interpretation rather than by the relative chronology of the dialogue is all the more probable because of the preceding passage, 185C–E (see note 11 supra), and the following one, 189C1–4 (cf. David Peipers, op. cit. [note 11 supra], p. 76).

²⁶ The Euthydemus, Cratylus, and Theaetetus belong to the large group of writings in which Plato paid no particular attention to the occurrence of hiatus, while the Timaeus belongs to the smaller group, consisting of the Laws, Philebus, Timaeus, Critias, Sophist, and Politicus, in which its occurrence is consistently avoided; and this is the best ‘objective’ evidence that all of the writings in this latter group are later than any of those in the former. I must reserve for another place discussion of Owen's attempt to circumvent this evidence as well as of the merits and shortcomings of the statistics of vocabulary, which he rejects, and of the statistics of prose-rhythm that he adopts.

²⁷ For this use of ἀκριβές see especially Republic 340D5–341A2, 341B5–6, 341C4–5, 342B6–7, 346B3 and Proclus, , In Timaeum 249D = III, p. 35, 24–5Google Scholar (Diehl):

²⁸ Not even for Aristotle is a statement illegitimate because it is not ἀκριβής (Rhetoric 1369B1–2; Eth. Nic. 1094B11–14 and 1104A1–10); cf. also Theophrastus, , Hist. Plant. I, iii, 5.Google Scholar

²⁹ Theaetetus 184C1–5 (cf. Metaphysics a, 995A10–12) and Politicus 261E; cf. Theaetetus 199A4–9, Laws 644A, and Euthydemus 277E–278C, this last an example of Plato's attitude towards the so-called ἀκριβολογία of Prodicus (for which cf. Radermacher, L., Artium Scriptores, pp. 67–8Google Scholar, nos. 6–10).

³⁰ Cornford, (Plato's Cosmology, p. 98, n. 4)Google Scholar took it to be a reference back to the Sophist. Teichmüller (Literarische Fehden, II, p. 360) insisted that it promises a later discussion, which in fact occurs in Parmenides 151E–137B, from which it follows that the Timaeus antedates the Parmenides. Pfleiderer, (Sokrates und Plato, p. 648)Google Scholar maintained that on the contrary it is a backward reference to the Parmenides (cf. Susemihl, , Genetische Entwicklung der plat. Philosophie, II, 2, p. 376).Google Scholar The ‘reference’ appears to have been a matter of debate among the ancient commentators also (cf. Proclus, , In Timaeum 253E–F [III, p. 48, 33 ff.Google Scholar, Diehl]).

³¹ Cf. λέγομεν in 37E5 (which governs τὰ τοιάδε in 38A8) and 38B3; cf. Proclus, , In Timaeum 253D (III, p. 47, 28 ff. [Diehl])Google Scholar:

³² Owen, p. 89 and p. 89, n. 6. For Cornford's interpretation see note 3 supra.

³³ Boeckh's addition of and in 258C1–2 is highly improbable, and Cornford was right in rejecting it (Plato's Cosmology, p. 292, n. 2); but perhaps instead of construing as Cornford does one should take καὶ before in C1 as introducing a new clause depending upon

³⁴ It is improbable that this is meant to refer to Plato's Sophist, as most modern commentators suppose it is. In the Sophist (258C3) is expressly said to be whereas the τινες referred to here, as Pseudo-Alexander, points out (Metaph., p. 473, 17–19)Google Scholar, do not according to Aristotle say that Asclepius, (Metaph., p. 385 30–31)Google Scholar takes the reference to be to ‘the sophists’, was used by Gorgias, ([Aristotle], De M.X.G. 979A26Google Scholar; cf. Empiricus, Sextus, Adv. Math. VII, 67)Google Scholar; and Aristotle may have in mind such arguments as his, in fact the same kind of expressions as those to which Plato refers in Timaeus 38B.

³⁵ Or ‘that which is not’ or even ‘the false’ (cf. Aristotle, , Metaphysics 1017A31–32Google Scholar, 1026A35, 1051A34–B2).

³⁶ Cf. Calogero, G., Studi sull' Eleatismo, p. 174Google Scholar but also the note of Lovedayand Foster in the Oxford Translation, Vol. VI ad loc.

³⁷ Metaph., p. 385, 31–4: Hayduck saw that the last apodosis requires the excision of the second [τὸ], but the alternative is an argument only if the former [τὸ] also is excised: ‘if it exists, it is obviously absurd to say that it is non-existent’. For μὴ ὄν used predicatively in this sense, cf. Plutarch, , Adv. Coloten 1115E (p. 190, 28[Pohlenz]).Google Scholar

³⁸ Cf. Porphyry apud Simplicium, , Phys., p. 135, 1–2Google Scholar:

³⁹ De M.X.G. 979A25–28 (on which see Calogero, op. cit. [note 36 supra], p. 161). Cf. Sextus, , Adv. Math. VII, 67Google Scholar (p. 204, 10–11 [Bekker]): (scil. ) and Aristotle, , Rhetoric 1402A4–5: Google Scholar

⁴⁰ These periphrastic forms of the perfect and present tenses are common; cf. e.g. for Philebus 33C6, Symposium 204D2 and for Politicus 301D8, Laws 800C5–6. The periphrasis with the future participle is less common, but it occurs along with the periphrastic present and aorist in Laws 888E4–6 (cf. England, E. B., The Laws of Plato, II, p. 452Google Scholarad loc. and Apelt, 's translation, Platons Gesetze, p. 402).Google ScholarCf. also Plutarch, , De Communibus Notitiis 1082D (p. 114, 16–17 [Pohlenz])Google Scholar where and

⁴¹ In Parmenides 141D–E Plato avoids the former ambiguity by distinguishing between and and the latter by coining which he distinguishes from cf. Proclus, , In Parmenidem, col. 1237, 23–41Google Scholar (Cousin²) and Meillet, A., Rev. de Philologie, XLVIII (1924), pp. 44–9.Google Scholar

⁴² So Proclus, , In Timaeum, III, p. 48, 8–10 and 23–25Google Scholar; cf. Empiricus, Sextus, Pyrrh. Hyp. III, 142Google Scholar and Parmenides, Frag. B8, 20

⁴³ Cf. Empiricus, Sextus, Adv. Math. X, 208Google Scholar Antiphanes, Frag. 122, 6 (Kock, , Comic. Att. Frag., II, p. 59Google ScholarPlato, , Parmenides 152C6–D2Google Scholar; Proclus, , In Timaeum, I, p. 239, 29 and p. 243, 19–20.Google Scholar

⁴⁴ Cf. Aristotle, , Physics 235B28Google ScholarProclus, , In Timaeum, I, p. 290, 25–26 Google Scholar

⁴⁵ Cf. Antiphanes, Frag. 122, 8–9 Aristotle, , De Generatione 336A 22–23Google Scholar Physics 263B26–28; Simplicius, , Phys., p. 1297, 15–19.Google Scholar

⁴⁶ Cf Melissus, Frag. B2 ( Anaxagoras appears to have said of god or νοῦς (Frag. A48 = Philodemus, De Pietate, c.4 a [cf. Reinhardt, K., Parmenides, p. 176, n. 2])Google Scholar; and Heraclitus had used the formula of the cosmos (Frag. B30). Plato makes his Parmenides conclude (Parmenides 155D3–4) on the assumption that, if exists, it must partake of time (155C8, cf. 151E7–152A3) and on the other hand that, if it does not partake of time, and so (141E3–10). Parmenides himself had written (Frag.B 8, 5) J and it has often been said that Plato's criticism of the formula in Timaeus 37E5–38A2 is an echo of this line. It may have been in Plato's mind; but, if so, he probably took the as an indication of Parmenides' failure to grasp clearly the notion of atemporal eternity (cf. [Parmenides 152B3–4] and 141E2–3), as in fact it seems to be (cf. also Frag. B8, 26–28 and Albertelli, P., Gli Eleati, pp. 143–4Google Scholar [note 11“]).

⁴⁷ It is difficult to refrain from taking this as a direct reference to the arguments whereby Parmenides is made to conclude that (Parmenides 155C4–7; cf. 141C8–D5, 152E2–3). Parmenides is made to assert that whatever exists is temporal (141E7–10, 151E7–152A3) and is located somewhere (145E1, 151A4–5); both assumptions are denied in the Timaeus, the former here in 37E–38A, the latter in 52B–C.

⁴⁸ Cf.38A1–2: Here, Owen says in another context (p. 86), ‘it is allowed only to say what a γιγνόμενον was and will be’. What the sentence infact says is that it is proper to say of γένεσις that it was and will be, whereas of eternal being it is proper to say only that it is; it is not even forbidden here to say of γένεσις also that it is.

⁴⁹ In interpretations such as Cornford's (see note 3 supra), where τὸ μὴ ὄν is taken to mean precisely ‘the absolute non-existent’, it is forgotten that the statement about is criticised not as being untrue or meaning less but as being imprecise. The specific interpretations of the passage that I have seen (see note 6 supra) tend to fall into this error or into one or both of two others: the expression in question is treated as if it were instead of (so even Proclus, , In Timaeum, III, p. 48, 12)Google Scholar or the relation of this expression to the other three is neglected. This last is true even of P. Shorey who takes the expression to refer specifically to the tenet of the Sophist (What Plato Said, p. 300).

⁵⁰ 28B7, 36E5–6, 92C8–9

⁵¹ 31B3; for γεγονός cf. 29A5, 37C7.

⁵² 28A3–4 (cf. 27D6f., 28C1, 52A6). Cf. Diogenes Laertius, III, 64 Proclus, , In Parmenidem, col. 999, 27–29Google ScholarIn Timaeum, I, p. 263, 25

⁵³ 38C2–3. For the subject of the clause and its construction, concerning which Cornford is right against Fraccaroli and Taylor, cf. besides Proclus, , In Timaeum, III, p. 50, 29–31Google Scholar and p. 51, 7–8 Simplicius, , Phys., p. 1155Google Scholar, 13–14 and De Caelo, p. 105, 25.

⁵⁴ Cf. Proclus, , In Timaeum, I, p. 282, 1–9Google Scholar; I, pp. 290, 24–291, 12; III, pp. 50, 31–51, 12.

⁵⁵ See note 43 supra.

⁵⁶ Cf. Timaeus 52C2–5. So in a sense γένεσις as well as real being and space exists (52D3): it is not non-existent even though it is not Cf. Proclus, , In Timaeum, I, p. 277Google Scholar, 29–30

⁵⁷ Sophist 258C3, cf. 258B1–2. A. L. Peck has contended that or and and are not meant to be taken seriously as ideas, in fact that the Sophist is meant to prove that they are not genuine ideas (Class. Quarterly, N.S. II = XLVI [1952], pp. 32–56 [cf. pp. 52–53] and N.S. III = XLVII [1953], pp. 146–8). His argument, which seems to me to be entirely mis taken, cannot be examined here. Since, however, he takes the Timaeus to be a later work than the Sophist and an exposition of Plato's genuine doctrine, it is enough to point out that the ideas of and appear in Timaeus 35A and 37A–B (cf. Cherniss, , Aristotle's Criticism of Plato and the Academy, I, pp. 409–11).Google Scholar

⁵⁸ As Owen appears to think (p. 85, n. 1). He does not say when ‘Plato came to see’ that the adoption of the supposed reform ‘is ruled out by logical absurdities’; but the text would require us to conclude that it was in the interval between finishing Timaeus 38A8–B5 and writing 38C2–3.

⁵⁹ Timaeus 29C5–6.

⁶⁰ Timaeus 34C2–4.