^{page 194 note 1} Cornford, , Plato's Cosmology, p. 190.Google Scholar

^{page 195 note 1} Mugler, Charles, Platon et la recherche mathématique de son époque (Strasbourg: Heitz, 1948).Google Scholar It is in substantiating a general hypothesis (that the development of Plato's theory of knowledge is to be interpreted in the light of a change in his conception of geometrical method) that M. advances his empiricalinterpretation of the Meno doctrine. It is impossible to discuss here the amazing elaboration of his attempt to show that the Meno doctrine reflects the ‘old synthetic method’ of geometry. Ch. vi of his book is devoted exclusively to an examination of the dialogue with the slave.

^{page 196 note 1} Plato's Theory of Ideas (Oxford, 1951), p. 35.Google Scholar

^{page 196 note 2} Ibid., p. 25.

^{page 196 note 3} Ibid., p. 18. Note that Ross, in stating that there is ‘no reference, explicit or implicit, to the Ideas in the passage dealing with anamnesis’, has only the passage 81 a 5–86 b 5 in mind (p. 18, n. 3), and does not refer at all to the later introduction of anamnesis in connexion with the distinction between knowledge and true opinion.

^{page 197 note 1} An. Pr. 67 ^a 21–30, An. Post. 71 ^a 27 ff., cf. 99 ^b 26–34.

^{page 197 note 2} It seems to be used in the latter sense in Crat. 436 d, and Aristotle so uses it (e.g. An. Pr. 41 ^b 14).

^{page 197 note 3} (73 b 2) must refer to the fact that ‘if questions are put well, people of themselves answer correctly about everything’.

^{page 197 note 4} Note on Phaedo 73 a 7, agreeing with Bury, R. G. (C.R. xx. 13).Google Scholar Bury's argument was based on the use of (73 a 10).

^{page 198 note 1} 76 b. The reference is presumably to the Forms.

^{page 198 note 2} The argument does not concern itself with the question of the extent of the world of Forms. The form of the argument almost implies that there is no limit to it.

^{page 198 note 3} 75 b–e, 76 c–77 a, 92 d–e. The ‘primary postulate’ is the existence of Forms, and this is the of 92 d. The pre-existence of the rational soul is deduced from this postulate, and from the postulate of the theory of anamnesis introduced in 73 c ft (76 d–e).

^{page 199 note 1} Cornford, , Plato's Theory of Knowledge, p. 6 (commenting on Phaedo 78 b–79 e).Google Scholar

^{page 199 note 2} Ibid., p. 108.

^{page 199 note 3} Cf. now his Principium Sapientiae (Cambridge, 1952), p. 51,Google Scholar where a sharp contrast is made between the doctrine of the Meno, with its emphasis on dialectic as the stimulus to recollection, and that of the Phaedo, where ‘Plato now offers another proof, which dispenses with the intervention of a questioner, recollection being occasioned by acts of perception’. The earlier ‘proof’ is the one afforded by the dialogue with the slave in the Meno.

^{page 200 note 1} The Theaetetus similarly appears to assume, in its description of ‘Socratic’ midwifery, a theory of anamnesis. Here again it is not the ‘suggestiveness’ of sense-experience which is in point but the efficacy of systematic questioning and discussion as an aid to the recognition of truth. See Cornford, , P.T.K., pp. 27–28,Google Scholar for the relations between anamnesis and ‘midwifery’.

^{page 200 note 2} Stenzel, (Plato's Method of Dialectic, trans. Allan, D. J., 1940, p. 154)Google Scholar rightly stresses this.

^{page 200 note 3} Plato's Phaedrus (Cambridge, 1952), p. 85.Google Scholar

^{page 200 note 4} Cf. Grote's remarks on Tht. 184–7 (Plato, ed. 1865, vol. ii, p. 374,Google Scholar n. i). His quotation from Hamilton remains very apt. ‘Sense he [M. Royer Collard] so limits that, if rigorously carried out, no sensible perception, as no consciousness, could be brought to bear.’ This, Grote adds, ‘is exactly true about Plato's doctrine narrowing ’. Bertrand Russell remarks (History of Western Philosophy, p. 175) on the argument of the Theaetetus: ‘The proposition “Knowledge is perception” must be interpreted as meaning “Knowledge is judgements of perception”. It is only in this form that it is grammatically capable of being correct.’ To do this would, of course, merely forestall Plato's intention to show that the former proposition is a paradox.

^{page 201 note 1} See Stenzel, , op. cit., pp. 108 ff.,Google Scholar for references and discussion.

^{page 201 note 2} While realizing that in the process of ‘collection’ Plato never has in mind a transition from ‘sensations’ to conceptual thinking, it is important not to resort to the opposite extreme of denying that the procedure is ever concerned with ‘individuals’, and ‘deals with Forms only’ (Cornford, , P.T.K., p. 267,Google Scholar n. 2, on Soph. 253 d 6). This view led C. to translate Phdr. 265 d as ‘taking a synoptic survey of widely scattered Forms (species) and bringing them into a single (generic) Form’. C. was misled by Stenzel (op. cit., particularly p. 156), who saw clearly enough (pp. 150–6) the difficulties in the interpretation of Phdr. 249 b, but was wrong, or at least misleading, in stating (p. 156) that ‘the only “individual” (Individuum) recognised by him (Plato) is the “atomic Form”, the object of true opinion’. Plato could still regard the ‘many’ objects of sense-perception as ‘individuals’, while stressing that any apprehension of what they are is conceptual.

^{page 202 note 1} Plato's Statesman (1952), p. 76.Google Scholar The ‘earlier doctrine’ is presumably that of the Phaedo.

^{page 202 note 2} Ibid., p. 76. On 265 b he says (p. 137, n. 1): ‘The translation is meant to suggest that here external features, obvious to the eye, give a strong lead to the dialectical philosopher in discovering the fundamentum divisionis required.’

^{page 202 note 3} Cornford, , Plato's Cosmology, p. 144.Google Scholar

^{page 203 note 1} Cf. Phdr. 248 b: the incarnate soul feeds on opinion the contrast being with the of 247 d.

^{page 203 note 2} ed. Phaedrus (1868), p. 56.Google Scholar

^{page 204 note 1} Cornford, (Plato's Cosmology, p. 41)Google Scholar rightly takes Phdr. 250 d to mean that ‘there are no sensible images’ of these Forms.

^{page 204 note 2} Professor G. C. Field, who kindly read the typescript of this article and made many helpful comments, considers this identification to be too specific. He writes: ‘We come to recognize an act as just partly by hearing how people talk and partly by imagining in ourselves how they are feeling, and so on. I do not think it is simply a question of words as images, though that may come in, but rather of words as revealing a state of mind which we understand by imagining our own feelings, etc. It is a very confused and indistinct process.’

^{page 205 note 1} He does not, of course, take into account here that class of which describe This distinction was made only later.

^{page 205 note 2} See my references in C.Q., N.s., vol. ii, 1958, pp. 78–80.Google Scholar

^{page 206 note 1} And thus apparently answering the question of Hipp. Maj. 298 b–d whether the perception of is the kind of perception given through sight or hearing, or quite different from it

^{page 207 note 1} It is unwise, of course, to rely too much on the implications of the method of the earliest dialogues in assessing Plato's view of the ‘image’. In particular we should beware of basing the analysis of the structure of those dialogues on metaphysical distinctions drawn from Plato's later thought. This was attempted by Goldschmidt, V. (Les Dialogues de Platon, Paris, 1947, pp. 1–90),Google Scholar who interpreted Plato's appeal there to cases as an appeal to ‘images’, and saw in the use of cases for suggesting or correcting hypotheses an application of the formula of Phdr. 265 d (p. 40). There is certainly a wide use of cases in the earliest dialogues (see Robinson, Richard, Plato's Earlier Dialectic, 2nd ed., Oxford, 1953,Google Scholar ch. iv, for a full discussion of this), but here the search for a definition does not appear to have been regarded as of metaphysical significance, nor the appeal to cases as a strictly empirical appeal. It was only with the recognition of an antithesis between empirical and a priori knowledge that Plato came to see in an appeal to cases an appeal to sensible images.

^{page 207 note 2} And even here the passage (523–4) noted by Robinson should be excepted. The argument there is that apparent contradictions in sense-experience act as a stimulus to the mind to seek a solution in a realm of abstract thought where no contradiction is found; it is used to emphasize the special value of arithmetic as one of the studies introductory to Dialectic. But I see no reason to suppose that Plato credits sense–experience with this useful purpose merely as a device in teaching.

^{page 208 note 1} In my remarks about this type of image I am indebted to Robinson's valuable dis cussion (op. cit., pp. 209 ff.) which gives full references.

^{page 208 note 2} Robinson gives Phaedo 87 b, Rep. 375 d, 487 e, 509 a, 517 b, 538c, Pol. 297 e as references for Plato's use of for his own figures (op. cit., p. 221).

^{page 208 note 3} He argues that the inconsistency is between Plato's practice and principles with regard to images in general, but he fails to distinguish properly the different kinds of image recognized by Plato.

^{page 209 note 1} Robin, L., Platon, pp. 89–90.Google Scholar This book (see esp. pp. 88–99; see also Robin's remarks in R.E.G. xxxii. 458–60)Google Scholar and Stenzel's Studien zur Entwicklung der platonischen Dia–lektik (Eng. trans, by Allan, D. J., 1940; see esp. pp. 55–72, 135–56,Google Scholar with Introd. xxxvi– xli, of this edition) have been widely influential in gaining acceptance for the view that (and the doctrine of ‘communion of kinds’ associated with it) replaces anamnesis. Both Robin and Stenzel make the same general assumptions about the significance of later developments in Plato's theory of knowledge; their argument against the view that Plato retains anamnesis is the only argument worth serious consideration, and it is exclusively this argument, and the assumptions on which it is based, which I shall criticize.

^{page 213 note 1} Cf. Festugière, , Contemplation el vie contemplative selon Platon (ed. 2, Paris, 1950), p. 189, n. 2.Google Scholar