^{page 116 note 1} ‘Pleasure’ seems to be used, especially by the older utilitarians, as if it had a variety of meanings, while δον is more or less free from ambiguity; cp. Cicero, , de Finibus, II, 3Google Scholar. 8. In this paper ‘pleasure’ is being taken as a simple equivalent of the Greek term— ‘motus referiucundus quo sensus hilaretur’.

^{page 116 note 2} The Philebus treats this question much more carefully than the Republic, and avoids its naive realism; but in proportion to the care with which pleasure is distinguished from any beliefs or acts of imagination which may accompany or occasion it (cp. 37b and 40d), it becomes more difficult to draw any direct ethical conclusions from the investigation of argutheir truth. In fact Plato makes little attempt to use that section of the dialogue for his main theme, and when it comes to the selection of ingredients for the good life, no pleasure is excluded on the score of unreality. 61e professes to employ the principle but 62e abandons it. I am returning to this point later,

^{page 216 note 3} When Plato is attacked by modern critics for a lack of rigour in his ethics or for not having formed a clear conception of obligation as a source of imperatives free from any reference to self-interest, we should remember that more in most of his writings he thinks of himself as vindicating morality against various types of hedonism and egoism. To be told that he was not rigorous enough would no doubt have surprised him. But because of this feature in his work it is difficult to extract from it anything in the nature of a direct answer to those ‘Kantian’ critics against whom his arguments were not aimed. The Philebus perhaps comes closer than other dialogues to providing such material because in it he is dealing, among others, with those whose views were stricter than his own, particularly in the denial of value to pleasure. In the interests of ‘moderation’ he there undertakes a defence of pleasure, the worth of which he had previously either denied, when its source was vicious, or simply taken for granted (when virtuous), presumably on the ground that the opponents against whom he was arguing would not have denied it.

^{page 117 ntoe 1} Cp. 12C–13d and 60b. The good cannot be wisdom or pleasure, whereas Protarchus's position at the beginning of the dialogue seems to exemplify in a simple form what Professor G. E. Moore calls the ‘naturalistic fallacy’.

^{page 118 note 1} E.g. 520, 540. But these passages (on the ‘return to the Cave’) indicate that Plato had not fully thought out the relation between τ λογιστικν and its own impulses. In accepting it as his duty to postpone his speculations the philosopher is exercising self-control (τ κρεττω αὑτο), and therefore according to 430e and 431a τ λογιστικν, if both the control and the impulse come from it, requires a subdivision to which the Republic has not attended.

^{page 119 note 1} Rep. 441e.

^{page 119 note 2} Critique of Practical Reason, ch. II; Abbott's, Translation, p. 152Google Scholar.

^{page 119 note 3} Rep. 581b.

^{page 119 note 4} 443d and e … φλον γενμενον αντῷ κα σɛναρμσαντα τρα ντα … κα παντπασιν ἔνα γενμεννον κ πολλν

^{page 119 note 5} Contrast 61e with 62d.

^{page 120 note 1} See Joseph, H. W. B., Essays in Ancient and Modern Philosophy, VIGoogle Scholar, for a comparison of Aristotle's use of the notion of a ‘mean state’ with Plato's. I am assuming that Mr. Joseph is right as to the reason for the unsatisfactoriness of Aristotle's treatment.

^{page 120 note 2} Plato's ground for treating reluctance to accept pleasure by itself as evidence that we do not think it good can be stated simply by saying that there is nothing better than what is good. The comparative ‘better’ means ‘closer to good’ and it is the positive which is absolute. So if I think something good I also think it better than its alternatives (unless they are just equal) and there therefore seems no reason why I should prefer them to it.

^{page 120 note 3} Its use for the conscientiousness of behaviour would require special examination.

^{page 120 note 4} The ‘purities’ of pleasurable and of cognitive experience are not really parallel cases so far as their value is concerned. The former seems comparatively unimportant in itself and 63 recognizes that in effect. Not so with intellectual experience as treated either here (55–59) or in the Republic. The aim of the classification of the sciences here is to distinguish the pure experience of knowing (qualified by τ σαφς κα τκριβς) from other cognate states, and for that experience ‘pure’ objects are necessary. Plato, is going over again the ground traversed in Rep. 504–34Google Scholar. The highest experience (which the Republic calls νησις combines certainty with insight, and it is only of a selected class of objects or problems that that combination is possible. No one, therefore, who had not made that selection (i.e. who had not attended to the sort of problems about which vdrjais is possible) would have given his mind free enough play (cp. Rep. 508d) to have realized the nature of cognitive experience at its best. Cp. Phil. 50a–c. It is not in all sciences that deductive insight can be had. Such considerations are therefore quite relevant to assessing the differential value of particular intellectual pursuits as ingredients of an actual life. The ‘mixing’, however (61–3), is rather hurriedly done, and all kinds of cognition are in the end bundled into one class (62b), while the pleasures, paradoxically, seem to be allotted two classes (presumably), although relative purity seems in fact far less important in their case. Plato seems reluctant to admit the last point, although his argument requires it.

^{page 121 note 1} His remark at 20c (that perhaps the specification of pleasure will not after all be necessary for the purpose of awarding the second prize) may imply a recognition of that fact, although 23b 8 ((ἔστι δἔ ἴσως ἔνια κα ταὔτ) seems to imply that such specifications may be required even for the revised form of the enquiry. When he does begin the task of specification (31b) he seems to treat it as arising from the preceding discussion of ‘the four classes’. I confess I find the sequence of thought in that section of the dialogue extremely obscure.

^{page 122 note 1} See the valuable Appendix B to Bury's Edition.

^{page 122 note 2} Cp. 65a 2, οἶον ἔν (τ τρα).

^{page 122 note 3} The latter argument may be stated in this form:—

the rival of δον is νος κα φρνησις

the rival of δον is above it because of its closer affinity with the good

therefore whatever is put above δον on that ground ought to be νος κα φρνησις, and

therefore also the fourth class to which the argument from affinity is attached ought to consist in νοσ κα πρνησις, whereas it is in fact called ‘the sciences, true beliefs’, etc. (66b).

In other words the attachment to class IV of the affinity argument almost irresistibly suggests (in spite of the undistributed middle term above) that whatever comes above class IV is not the original rival of pleasure but belongs to the description of the good life and its qualities, kinship with which is to settle that rivalry. I, II, and III, therefore, should all be confined to the regulative principles, etc., and νος κα φρνησις have no place there. And that Plato did not intend to put them there is suggested by the words τ πρς τοῖς τρις τταρτα (66, b 5) which seem to allude to I, II, and III as τλλος, λθεια) already therefore classified (cp. 65a).

^{page 122 note 4} (ii) does not seem to have attracted equal attention, but 59d (if nothing else) seems to make the separatin of νος κα φρνησις from the πιςτμαι intolerable.

^{page 122 ntoe 5} Badham notes a ‘bye-purpose’ in the introduction of λθεια into the sentence which defines the third class, but understands it differently—‘the bye-purpose of showing that this νος owes its place to the Truth of which it is the realization’.

^{page 123 note 1} Two further comments on this passage may be added:—

(i) Plato does not expressly mention here the sixth class of ‘necessary pleasures’; but they were definitely admitted (as their name implies) at 62e, and the allusion to ‘Orpheus’, as Professor Taylor, points out (Plato, p. 433Google Scholar, note 2), suggests that they are to be understood as included here. ‘The theogonic poet quoted must have described his “sixth generation” of deities as well as the preceding five’.

(ii) In the upper classes Plato does not seem to care much about niceties in the form of statement. The first form of good discovered (64d) is μετριτη ς κα σνμμετρα (as in one breath); these notions then suggest κλλος as the second form, and λθεια is discovered independently as the third. In 65a the three forms are restated as κλλος, ςɛμμμετρα and λθεια: the variation in order seems to lend some slight support to the view that Plato is not much concerned with the relative ranking of these top classes. Later (66a and b) τ σμμετρον is separated from τ μτριον and joined with τ καλν, possibly in order to suggest symmetry as the connecting link between moderation and beauty.

^{page 124 note 1} It is not that ‘wisdom’, so far as it is a virtue, is a matter of degree; the word is of course being used ambiguously in such discussions: primarily it is the name of a psychological quality at a high level of excellence, and it is impossible to use it as the name of a virtue without artificiality; but if it is so used it can only refer to the conscientious employment of the faculty.