1 Page references given here to this article will be to its appearance in Gregory Vlastos (ed.), Plato II (London: Mamillan, 1971).

2 See ‘Plato's Modern Friends and Enemies’, Philosophy 37(1962) 97–113. This and ‘Plato's Political Analogies’ are also printed in Plato, Popper and Politics, edited by Renford Bambrough (Cambridge: W. Heffer & Sons, 1967). See also Moral Scepticism and Moral Knowledge (London and Henley: Routledge and Kegan Paul, 1969).

3 Bambrough continues to disagree with Plato over the ‘question whether moral knowledge is a specialized technē like medicine or mathematics’ (Moral Scepticism and Moral Knowledge, op. cit. 48).

4 E.g. John Gould, The Development of Plato's Ethics (Cambridge University Press, 1955), Chapter II, and R. C. Cross and A. D. Woozley, Plato's Republic, A Philosophical Commentary (London: Macmillan, 1964), 12–16.

5 ‘Plato's Political Analogies’, op. cit. For an examination of the importance of the analogy for Plato without emphasis on his philosophic errors, see Rosamond Kent Sprague, Plato's Philosopher-King (Columbia, South Carolina; University of South Carolina Press, 1976). On the connection between epistēmē and technē I am greatly indebted to J. C. B. Gosling, Plato (London: Routledge and Kegan Paul, 1973), esp. Chapter IV.

6 Op. cit. Introduction.

7 The Concept of Mind (London: Hutchinson, 1949), Chapter II.

8 Op. cit. xiii.

9 Ibid. 32.

10 Ibid. Chapter II.

11 The argument Gould cites from the Greater Hippias has the same point as Republic 333e-334c, and can be treated as I am about to treat that argument. On the argument Gould cites from the Charmides, as well as on Republic 332c-333e, I would refer the reader to the interpretation of Sprague, op. cit., esp. pp. 62–63.

12 Julia Annas gives an account of technē similar to that being argued here (An Introduction to Plato's Republic (Oxford: Clarendon Press, 1981), 24–26) but she sees the point of Republic 333e-334c as the reverse of that suggested here. Annas, p. 28, traces Polemarchus’ embarrassment to his assuming the denial of assumption (I) above, and finding that his notion of justice does not include assumption (2). But this is surely the wrong way round; otherwise Polemarchus would not havebeen made to agree to (I) at the outset of the argument.

13 Gorgias 465a translated (with the reintroduction of some Greek words) by Walter Hamilton (Harmondsworth, Middlesex: Penguin Books, 1960), 46.

14 ‘Plato's Political Analogies’, op. cit. 202. Aristotle, it should be observed, did not consistently classify technē as a two-way power. Nicomachean Ethics VI.4 puts technē in thegenus hexis, which in the Metaphysics was said to be a one-way power.

15 On the ‘craft’ analogy in Plato's ethical theory see Terence Irwin, Plato's Moral Theory, The Early and Middle Dialogues (Oxford: Clarendon Press, 1977), Chapter HI, Sections 9–11, 17. Irwin (p. 133) takes the disappearance in the middle dialogues of the analogy betwee technē and virtue as evidence of Plato breaking with his Socratic inheritance. That the analogy continues to be used in connection with epistēmē is to my mind evidence of continuity as the centre of Plato's interests shifts.

16 ‘Plato's Political Analogies’, op. cit. 202.

17 Annas, op. cit. 337.

18 Annas, op. cit. 230. It is significant that Socrates speaks here of the carpenter looking to the eidos of what he is making and of what ‘we should properly call auto ho estin kerkis’ (Cratylus 389b).

19 ‘Plato's Political Analogies’, op. cit. 201.

20 I have used the topic of Socrates’ conversation with Pistias (as reported by Xenophon) to provide a concrete illustration with a period flavour. In what follows, however, I am forced to take some historical liberties with the illustration in order to make my point clearer. I am aware that most hoplites did not (and could not afford to) change the style of their weapons and armour to suit the tactics to be used on a given campaign. The general had to take the armour his soldiers already possessed as a material constraint comparable to those material constraints imposed on the armour maker by the metallurgy of his day. The liberties I am taking, however, are thoroughly in keeping with the spirit of Plato's philosophy; for Plato tended to ignore material constraints. Presumably he would argue that one must first of all become clear about the proper objective of one's activity,for it is this which determines how new developments in materials and design are to be employed.

21 Although Gorgias 464e-465a stigmatizes the cook's activities as amounting to no more than an empetria, there is no reason why the activity of preparing food cannot be subordinated to the Good and principles of procedure discovered which would raise cookery to the level ofa technē. Statesman 288e-289a includes the provision offood among the technai. The Phaedrus seems to mark a change in Plato's attitude towards another of the empetria of the Gorgias, rhetoric, so that Socrates may be referring to Gorgias’ technē without inverted commas at Philebus 58c.

23 ‘Plato's Political Analogies’, op. cit. 197.

24 Printed in The Listener 104 (6 November-i 1 December 1980).

25 ‘Plato's Political Analogies’, op. cit. 196.

26 In the analogy of 488a-389a, of which Bambrough makes so much, Plato does not speak of a navigator’, a ‘kubernētēs’, but a ‘nauklēros, which strictly means ‘ship-owner’ as Lee observes (The Republic, 2nd edn, translated by Desmond Lee (Harmondsworth, Middlesex: Penguin Books, 1974), 282, n. 1).

27 ‘Plato's Political Analogies’, op. cit. 195.

28 The ambiguity of ‘why something should be the case’ (i.e. between ‘why it is better that way’ and ‘why it happens that way’) mirrors an ambiguity in a wide range of Greek idioms which Plato uses, making it hard for him to observe this distinction carefully and exposing him to the criticisms of Aristotle, who pushed hard on the inability of the Forms to provide explanation in the second sense (Metaphysics M5). Nevertheless Aristotle confirms the point that in the philosophic usage of Plato's circle, epistēmē is not ‘knowledge’ without qualification but ‘knowledge of reasons-why (aitiai)’; see Metaphysics Ai, Physics Ai, or Posterior Analytics A2.

28 ‘Plato's Political Analogies’, op. cit. 202.

29 Ibid.

30 Epistēmē not only requires a grasp of reasons-why, but is communicable to other people. This is why it is axiomatic that a person who has epistēmē can teach. See Meno87c, 89c, and Metaphysics AI, 98Ib8—Io.

31 There is now, however, a great deal on this subject which is both sound and stimulating, in Moral Scepticism and Moral Knowledge,op. cit., esp. Chapters 7-9.

32 ‘Plato's Political Analogies’, op. cit. 199. In ‘Plato's Modern Friends and Enemies’, op. cit., Plato is said to share with his opponents ‘the false assumption that all true knowledge is of one kind, or of a small number of specifiable kinds’ (p. 109) and to differ from them in inferring that as there is moral knowledge, it must be of this (or one ofthese) kind(s).

33 ‘Plato's Political Analogies’, op. cit. 198.

34 In other words the Republic contains resources to meetthe threat of a vicious regress in the subordination of technēto technē, which Irwin (op. cit. 77) notes.

35 ‘Plato's Political Analogies’, op. cit. 198.I see no reason not to take Socrates’ disavowals of this claim in Book VI, e.g. 505a, and 506c, at face value. Indeed, the following more recent statement by Bambrough is more cautious: ‘Plato believed that there was such a thing as moral knowledge, and he also believed that it was accessible to his philosopher-kings…’ (Moral Scepticism and Moral Knowledge, op. cit. 48).

36 Bambrough now regards the claim that ‘Moral disagreement is more widespread, more radical and more persistent than disagreement about matters of fact’ as ‘almost certainly untrue’ and ‘quite certainly irrelevant’ (Moral Scepticism and Moral Knowledge, op. cit. 48).

37 I would not, in other words, endorse the claim that ‘wheneverone man denies what another man asserts at least one of them will be found on full enquiry to have committed himself to a position that he cannot reasonably combine with his initial assertion’ (Moral Scepticism and Moral Knowledge, op. cit. 88).

38 ‘Plato's Political Analogies’, op. cit. 201.

39 I am grateful to the Editor of Philosophy and to four colleagues, John Cottingham, Rosemary Harriott, Geoffrey Harrison and Professor G. H. R. Parkinson, whose criticisms of earlier versions of this paperhelped me to remove obscurities in some places and to tread more cautiouslyin other places.