^{page 65 note 1} Owen, G. E. L., ‘The Place of the Timaeus in Plato's Dialogues’, C.Q. n.s. III (1953)Google Scholar, repr. in R. E. Allen, Studies in Plato's Metaphysics.

^{page 65 note 2} On the subject of perception the most effective presentation known to me is provided by Bernard Williams's unpublished lectures on the Theaetetus; I owe much both to these and to private discussion with Professor Williams. Mention should also be made of Crombie's, I. M. important chapter on Aisthesis in An Examination of Plato's Doctrines, II, 1–32 Google Scholar. It is accounts such as these which make it necessary to go beyond H. Cherniss's monumental reply to Owen, , ‘The Relation of the Timaeus to Plato's Later Dialogues’, A.J.P. LXXVIII (1957)Google Scholar, repr. in Allen, op. cit.

^{page 65 note 3} ‘This argument [sc. Tht. 182 c] defeats the lame plea of the Tim. (49 de) that even if we cannot say what any mere γιγνόμɛνΞν is we can describe it as τὸ τΞιΞῦτΞν (cf. Tht. I52d6)' (Owen in Allen, p. 323 n. 3).

^{page 65 note 4} Note that the conditions are stringent – ὄντως, b 3, πιστῷ καὶ βɛβαίῳ… λόγῳ, b 5. See below §11.

^{page 65 note 5} Cf. n. 3.

^{page 65 note 6} This interpretation seems first to have been challenged in English by Cherniss, in A.J.P. LXXV (1954)Google Scholar, ‘A Much Misread Passage of the Timaeus (Timaeus 49C7-50 b5)’, who later pointed out the consequent weakness in Owen's attack (Cherniss in Allen, pp. 357 ff.). But cf. also p. 67 n. 2 below.

^{page 66 note 1} Cf. Kühner-Gerth, Ausführliche Grammatik d. Gr. Sprache, 3. Aufl., 11. i. 461 Anm. 3b, 4. I owe this argument to Cherniss's original paper, A.J.P. LXXV (1954), 116 Google Scholar.

^{page 66 note 2} Cf. Lee's criticism of Gulley's view as self-refuting ( Lee, E. N., ‘On Plato's Timaeus 49 d4-e7’, A.J.P. LXXXVIII (1967), 17)Google Scholar.

^{page 66 note 3} This argument derives from Lee's unpublished thesis, ‘The Concept of the “Image” in Plato's Metaphysics', Princeton, 1964.

^{page 66 note 4} De Caelo 637. 3-5 Heiberg: ɛἰπών δὴ πɛρί τῆς τῶν στΞιχɛἱων ɛἱς ἄλληλα μɛταβΞῆς καὶ ὄτι τὸ άɛὶ γινόμɛνΞν, ΞἴΞν τόδɛ τὸ πῦρ καὶ τόδɛ τὸ ὔδωρ, Ξὐκ ἄξιΞν ὤρισμένῳ ὄνόματι καλɛῖν ὡς ἀɛὶ μɛταπῖπτΞν, ἀλλὰ τὸ ἀɛὶ τΞιΞῦτΞν.

^{page 66 note 5} Not against those, like Owen, who claim precisely to find absurdity here.

^{page 66 note 6} It is noticeable that adherents of the traditional view (Jowett, Bury, Taylor et al.) tend to adopt Cherniss's construction at e 6 ff.

^{page 67 note 1} On the singular-plural shifts in these lines, to which Lee attaches great importance, it might be argued that, even on Lee's interpretation, there is a shift to be explained in e 2–4, and this follows more easily if the way has been prepared by a shift in the previous sentence. However, this may be thought special pleading, and it may well be felt that the ὄσα… ὑπΞμένΞν… ὡς ὄντα αὐτά shift is altogether too harsh ( Mills, K. W., ‘Some Aspects of Plato's Theory of Forms: Timaeus 49 c if.’, Phronesis XIII (1968)Google Scholar, seems to think that he has explained this, p. 15 5 n. 16, but in fact appears only to have drawn attention to it). In that case the problem might be solved by a simple emendation, reading ὑπΞμένΞντα and making the three formulations consistently plural. The ratio corruptelae would be the ambiguity of φɛύγɛι and the fact that it is immediately preceded by τι; also, perhaps, a kind of haplography, since ὑπΞμένΞν<τα> is followed by the odd collection of words τὴν τΞῦ τόδɛ.

^{page 67 note 2} This point of the parallel argument in this passage seems to have been previously brought out only in a brief note by J. Moreau (n. 145, p. 1472 of the Pléiade complete translation, ed. L. Robin, Paris, 1950). The point rests, of course, on a general interpretation of the passage of the kind given above, in which Moreau therefore anticipated Cherniss.

^{page 67 note 3} This (49e 5) and parallel formulations in the passage (d 5-6, d6-7, e6-7) have been the subject of much dispute. For the sake of brevity here I simply assume these to be parallel, and that the point of them is given by the contrast in d5 with τΞῦτΞ (=ὄ καθΞρῶμɛν ἄλλΞτɛ ἄλλῃ γιγνόμɛνΞν, that is, the changing appearances of the physical world). These two points taken together should suffice to exclude any interpretation of τό διά παντὸς τΞιΞῦτΞν either in spatial terms (this would obscure the fundamental point of consistency through time which is central to the passage and revealed by the contrast in d 5), or in terms of aggregates of phenomena (this would scarcely do justice to the contrast with the inevitable condition of phenomena). The point is, as Lee insists, essentially an analytical one, with perhaps no clear metaphysical entity in view, but simply with the object of distinguishing the elements of qualitative recurrence in our sensations. Nevertheless, it is worth pointing out that Lee's insistence here on two kinds of λόγΞς has much in common with the argument for Forms at 51 e. The refusal of Cherniss and Lee to allow that τὸ τΞιΞῦτΞν ἄɛί may refer to Forms has been criticized by Mills, op. cit.-, but such criticism need not entail the general thesis which Mills is concerned to defend.

^{page 68 note 1} Owen in Allen, p. 323.

^{page 68 note 2} It is important to note how these twin problems over particulars dominate this whole area, not only in 49 de, but more summarily in 50 cd, and again less explicitly in 50d–51a (the lack of quality of the Receptacle) and 51 b if. (the forms as paradigms).

^{page 68 note 3} As by Williams in the lectures referred to above (p. 65 n. 2); also Cooper, J. M., ‘Plato on Sense Perception and Knowledge: Theaetetus 184-186’, Phronesis xv (1970), 125 n. 4Google Scholar.

^{page 68 note 4} E. N. Lee, ‘On the “Gold-Example” in Plato's Timaeus (50A5-B5)’ in Essays in Ancient Greek Philosophy, ed. J. P. Anton with G. L. Kustas.

^{page 69 note 1} I am much indebted to Lee's work for the account given in this paragraph; thorough discussion of points of detail, such as the interpretation of ἀγαπᾶν, will be found there.

^{page 69 note 2} Cf. Gulley, N., Plato's Theory of Knowledge, p. 80 Google Scholar on the exclusion of ɛῑναι in the Theaetetus. Gulley attempts in the following pages to set up a contrast between this Ξὐσία/γένɛσις disjunction and that which forms a regular part of accounts of the theory of Forms.

^{page 69 note 3} Cf. also Mills's extreme position (op. cit. pp. 156 fr.): ‘… we are driven to conclude about the gold that there only seem to be shapes in the gold, and that really they are not there at all’ (p. 158).

^{page 69 note 4} Cf. Tim. 49e 3-4.

^{page 69 note 5} Owen's case as published seems to depend heavily on πανταχόθɛν, though he would no doubt defend it strongly on the basis of the internal logic of the theory. Nevertheless, it does seem important to consider the reasons for the exclusion of ɛῖναι, how – if at all – this relates to its different uses, and what exactly the function is of the qualification at b1-3.

^{page 69 note 6} See below, § III of this paper.

^{page 70 note 1} Logically there is a third possibility (c): we really are stupid to lapse into this kind of talk. This interpretation is clearly unacceptable – the irony has been frequently noted from Campbell (ad loc.) onwards. The problem lies in determining the force and direction of the irony.

^{page 70 note 2} This interpretation makes no claims on the contentious issue of whether, how far and where Plato clarifies the distinctions in the use of ɛῖναι. The interpretation depends upon our making them, and attributing to Plato awareness that there are problems here. How far one may go beyond that is immaterial to the present investigation; but the veiled terms in which his qualifications are made in these two passages might suggest that clarification has still some way to go.

^{page 70 note 3} Op. cit., Anton and Kustas, p. 225; it might nevertheless be argued that Plato would bar statements of existence regarding what was radically unstable.

^{page 70 note 4} Lee's account in terms of ostensive definition depends partly upon a more general account of this whole area of the Timaeus which is defended in his thesis (p. 66 n. 3 above); cf. in particular 48 b 5-7. I am grateful also to Gregory Vlastos for pointing me in a similar direction.

^{page 70 note 5} I suspect that this interpretation parts company with Lee in this area; cf. his account (Anton and Kustas, p. 229) of the dialectically less able pointing at some fire and asking ‘what is that?’, where he gives no reason why the answer ‘it is fire’, in the minimal sense where ἐστί functions purely as copula, should be seen as improper rather than, as suggested here, simply dangerous.

^{page 71 note 1} Gulley (op. cit. p. 86) mentions a resolution along these lines with some approval, although he goes on to reject it, implicitly, in favour of simply assuming an unconscious inconsistency on Plato's part here. However, his rejection of the present solution is weakened by the implausibility of his discussion of the earlier ɛῖναι passage (cf. p. 69 n. 2 above).

^{page 71 note 2} Cf. Crombie ii. 15 on 185 ab; Cooper, op. cit.

^{page 71 note 3} Cooper, op. cit. p. 140 and n. 22 (on 186a 10).

^{page 71 note 4} The charge of inconsistency might be met differently, by allowing the complete elimination of ɛῖναι from objective accounts of the physical processes which generate sensation, while retaining it for judgements of sense experience; cf. Crombie's distinction (ii. 12) between ‘normal’ and ‘rampant’ Heracliteanism (a similar view seems to be held by Nakhnikian, G., ‘Plato's Theory of Sensation’, Review of Metaphysics ix (1955), 134 Google Scholar). This interpretation is susceptible to two attacks: (1) it does scant justice to the point of the perception theory in the Theaetetus, which is to rule out perceptual disagreements and corrigibility, cf. 154a, 1596-160a; (2) it is certainly inconsistent with the Timaeus passage, where perceptual disagreements are precisely what is in question. Crombie discusses these objections (ii. 31-2), the first to some effect; but on the second, despite the more realistic view of the Timaeus which he later expresses (ii. 39), his defence is nullified by misinterpretation of Tim. 49-50.

^{page 71 note 5} Crombie develops a position which to some extent points to a resolution of this dichotomy: while arguing finally against the ‘traditional’ interpretation (‘false for some things’), his own conclusion in terms of ‘something is not in flux’ is formally closely akin to that which he rejects. I have suggested (p. 71 n. 4 above) that the distinction of ‘normal’ from ‘rampant’ Heracliteanism will not hold, and it is on this basis that Crombie argues that the theory as originally stated did not exclude stable properties. Nevertheless, I have found Crombie's distinction, and his concentration on the status of properties in this refutation, most illuminating, and the account which follows owes much to his discussion.

^{page 72 note 1} This seems to make much the best sense of 182d 1-3. Cf. also Nakhnikian's πΞιότης₁ and πΞιότης₂ (op. cit. p. 143).

^{page 72 note 2} In other words the whiteness in flux is either whiteness-P or whiteness-S; and for the purposes of the present argument it does not matter which. In view of the restriction of whiteness to whiteness-P in the earlier passages (and analogously sight to sight-P) and the fact that whiteness and sight are treated closely together here (d2~3, d8 ff.) it might seem that whiteness-P is the subject here; on the other hand C9-11 in particular suggests that the difficulty concerns the grip of language on the world as it presents itself to us (that is, whiteness-S). Moreover, notions of ἀλλΞίωσις and so forth find no ready application to whiteness-P.

^{page 72 note 3} Cf. DeLacy's, P. important article ‘ὐ μᾶλλΞν and the antecedents of ancient scepticism’, Phronesis in (1958)Google Scholar.

^{page 73 note 1} Recently revived by Luce, J. V., ‘The Date of the Cratylus ’, A.J.P. LXXXV (1964)Google Scholar, cf. also his ‘The Theory of Ideas in the Cratylus ’, Phronesis x (1965)Google Scholar; for earlier adherents of the same view cf. Goldschmidt, V., Essai sur le Cratyle, pp. 33 ff., 69 ff., 175 ff.Google Scholar; Ross, W. D., ‘The Date of Plato's Cratylus ’, Rev. Int. Phil., XXXII (1955), fasc. 2Google Scholar. Luce points out the marked development between the Cratylus and Theaetetus in the refutation of subjectivism, but cannot deny their affinity; and many of his arguments on the tentative presentation of the ‘Theory of Ideas’ in the Cratylus may well be regarded as two-edged, in an attempt to prove that the dialogue does not belong to the critical group. Much of Ross's argument is similarly self-defeating. Neither comes seriously to grips with Weerts's account of the Flusslehre as an essentially late and critical concern of Plato, (Philologus, Suppl.bd. XXIII, I, 1931)Google Scholar. Against an early dating cf. also Kirk, G. S., ‘The Problem of Cratylus’, A.J.P. LXXII (1951)Google Scholar, Allan, D. J., ‘The Problem of Cratylus’, A.J.P. LXXV (1954)Google Scholar.

^{page 73 note 2} Cf. preceding note on affinity between the refutations of subjectivism in the Cratylus and Theaetetus; also the metaphor of letters and syllables (Crat. 421 ff.), on which cf. Crombie II, 376 and n. 1, Runciman, W. G., Plato's Later Epistemology, p. 129 Google Scholar.

^{page 73 note 3} In Allen p. 323 n. 3. It is interesting to note that, given the true argument of Tim. 49-50 (§ I above), Owen would be obliged to extend the same eager hospitality to the Timaeus.

^{page 74 note 1} Nor is this a passing, half-glimpsed insight: it is reiterated in this passage of the Cratylus at 440 a 1-3 (and alluded to at 439 e 1-5), and might be implicit in the dual role of language – διδάσκɛιν and δισκρίνɛιν – at 388 b (cf. Lorenz, K. and Mittelstrass, J., ‘On Rational Philosophy of Language: The Programme in Plato's Cratylus Reconsidered’, Mind n.s. LXXVI (1967)CrossRefGoogle Scholar).

^{page 74 note 2} Calvert, B. (‘Forms and Flux in Plato's Cratylus’, Phronesis xv (1970)Google Scholar) misinterprets seriously here, taking τΞιΞῦτΞν (d9) as correlative with ἔκɛίνΞ, and πρῶτΞν…ἔπɛιτα temporally; cave also his reading of μηκέτι Ξὔτως ἔχɛιν. He does nevertheless pinpoint some real problems in e 1-4.

^{page 74 note 3} Cf. Luce, , Phronesis x (1965), 24 Google Scholar: this passage ‘develops the implications of referring to any object as existing with a character of its own’, and later ‘αὐτΞ τὸΞ X is the natural and necessary correlate of γνῶσις’. This seems to confine the later stages of the argument too narrowly. Cf. also Gulley, op. cit. p. 72.

^{page 74 note 4} Rejected on insufficient grounds by Luce, ibid. p. 33. If it is retained there is dispute whether it is to be taken assertively (cf. Gulley, ibid., Jowett, 1970 paperback ed., p. 193) or as a continuation of the hypothetical (cf. Calvert, op. cit. p. 35, Kirk, op. cit. p. 232).

^{page 74 note 5} Cf. Calvert's distinction between what is contingently and what is necessarily unknowable (ibid. p. 40).

^{page 75 note 1} Cf. Gulle's attack (op. cit. pp. 74-5) on the Robinson-Owen position.

^{page 75 note 2} This is close to (i) above. It differs by playing down the assertive element and seeing the argument essentially as a dilemma, as Weerts and Kirk maintain.

^{page 75 note 3} The former seems to be the case with an account which I have found otherwise very helpful, that of Yolton, J. W., ‘The Ontological Status of Sense-Data in Plato's Theory of Perception’, Review of Metaphysics III (1949)Google Scholar. Nakhnikian offers an extended defence of this position, but his arguments depend crucially upon a unitarian view of Plato's work which would not carry universal conviction.

^{page 75 note 4} This might well seem to create a problem over what there is in the world, and possibly a conflict with the Timaeus. However, there can strictly be no motion unless something moves, and it seems that this point is accepted by the ‘completion’ of the theory in the following paragraph (156c7), where the account is framed entirely in terms of things moving. The earlier remark, then, must be interpreted, not as the absurd exclusion of there being things that move, but as excluding the possibility that something does not move. All that we really know of these objects is that they are loci of δυνάμɛις to produce certain effects, a view which would happily accommodate the theory of the Timaeus in terms of space, structures and motion.

^{page 76 note 1} As Crombie rightly sees (ii. 21 ff.) sense can be made of this only by supplementing it from the Timaeus account.

^{page 76 note 2} It might be argued that this and other passages present physical objects as having sensible qualities, at least at the moment of sensation, a view inconsistent with the causal account given here and in the Timaeus; one may compare 15964-5: the conjunction of two motions makes one, not perception but percipient, and the other, not a quality but qualified. It is difficult to see precisely what point Plato has in mind here, but note (a) he is not banishing every use of general terms for qualities since here, as in 156e, these are used in an account of the physical processes involved; (b) whatever point he is making is one which applies equally to the organ as to the object of sensation. Now it seems that the problem arises at the point at which Plato wishes to effect a transition from the mechanics to the content of a sense act: having used the general terms of quality and sensation to refer to the physical processes involved, in order to avoid blurring the vital distinction between the physical and the empirical he has to move to the particular terms to describe the transient content of the sensation. In other words, the object which is spoken of as qualified is not the physical but the sensed object; and this is confirmed by the passage at 159 d 3-5. Here it is important to note that the force of ɛῖναι is doubly qualified, first by the dative of the sentient organ, secondly by being paired with φαῖνɛσθαι (in the context of a theory which holds that what appears to a man is for that man). In other words, the force of ɛῖναι is confined within the sensation, so that the quality is attributed only to the sense object. One may perhaps compare Tim. 62 ab. (These arguments should dispose of Yolton's difficulty, op. cit. pp. 28 ff., over the senses other than vision, and his consequent objective interpretation of sense qualities.)

^{page 76 note 3} Cf. 58ac, 56d ff.; also Miller, H. W., ‘The Flux of the Body in Plato's Timaeus ’, T.A.P.A. LXXXVIII (1957).Google Scholar

^{page 76 note 4} It should be noted that sensation is in general confined to the most mobile of the particles in an animate structure, since these alone readily admit the stimulation necessary for the external motions to be transmitted to the brain; cf. 64be, 74e ff.

^{page 76 note 5} Cf. 65 a ff., 66 dff., 67b ff.

^{page 77 note 1} It may be useful to deal summarily here with a doubt which may be felt on the subject of motion. This is central to the Theaetetus, where it serves to guarantee the uniqueness and incorrigibility of sense experiences, first by generating interactions and thus ensuring the relativity of these events, secondly by ruling out any continuity even within the experience of a ‘single’ subject (cf. 159e7 ff.). For the Timaeus the point of discontinuity has already been adequately dealt with; but it might be thought, on the first point, that this account tends to generate sensation from only one motion, that of the object, except in the case of sight. However, as has been argued above, the motion of the sentient subject is sufficiently established for the Timaeus by the repeated assertions of the constant motion in which the physical body is maintained ; moreover, Plato here shows explicit awareness of at least part of the epistemological significance of the role played by a changing subject in sensation when he draws attention to differing sensations generated by similar atomic influxes in different organs (67 de). There is one qualification to this which should be mentioned, namely the tendency to speak of the subject's mobility, rather than its motion, and its capacity for being affected (cf. 64a ff.); but here again the Theaetetus has similar formulations (156a5-7; and cf. 157a 3-7 with Tim. 57e 3-6), and in both dialogues the one really clear exception is the case of sight (Tht. 156 de, Tim. 68a, 45 b ff.). On this question of passivity cf. also Nakhnikian, op. cit. p. 141.

^{page 77 note 2} Cf. 61 de, 64d ff., 65c ff., 67de.

^{page 77 note 3} Of course this point could be quite independent – and is so in the Timaeus account – of any theory concerning the nature and processes of sensation. One of the crucial themes in discussion of the Theatetus concerns the nature and propriety of this relationship, cf. in particular Crombie's discussion and nn. 4, 5 on p. 71 above.