¹ The view here criticised has been extracted, without I hope injustice, from two articles which are vital to the study of this disputed passage: Santas, G., Philosophical Review lxxv (1966) 3–33Google Scholar, and Vlastos, G., Phoenix xxiii (1967) 71–88Google Scholar. A similar version (in the main) is given by Gulley, N., The Philosophy of Socrates, (Macmillan 1968) 101 ffGoogle Scholar.

² 365a-b.

³ The gerundives in 356 are admitted on all hands to be (at least) ambiguous and supporting evidence for psychological hedonism is sought elsewhere. Possibly relevant considerations are listed by Sullivan, J. P., Phronesis vi (1961) 19–20Google Scholar. Both Santas (op. cit. note 12 and p. 10) and Vlastos (op. cit. p. 85) agree that the presentation is hardly conclusive. The dilemma is well represented by Gulley (op. cit. p. 107), who observes that ‘ought’ is the more natural meaning here, but thinks psychological necessity is intended because Socrates refers to human nature, 358d, and because the present argument requires more than ‘ought’. The correct view, I believe, is taken by Gallop, D., Phronesis ix (1964) 117–29Google Scholar, who insists that the only available meaning is ‘ought’.

⁴ Santas' account (op. cit. p. 14–18) is, in outline: ‘D2. Sometimes a man does something which contains good and bad where the bad outweighs the good, knowing that this is so, when he can avoid doing it’ because ‘E3. the man takes (chooses, prefers, decides to take) the (known) greater harm (evil) contained in what he does in return for securing (as the price of) the (known) lesser good contained in what he does’. Then, keeping λαμβάνειν in the sense of ‘chooses’, the thesis is similarly restated with the substitution of pain and pleasure. Next Socrates ‘proceeds to elaborate a principle which is implied by the hedonism of the hoi polloi (which we may remember is a premise of the whole argument) and which contradicts the explanation E3’, i.e. the principle that people always seek to maximise pleasure. The conclusion is that an action containing more pain than pleasure must be avoided, which contradicts the popular explanation with the second substitution, ‘and a similar principle, obtained by substituting “good” for “pleasant” and “bad” for “painful” in the above principle, contradicts directly the explanation of the hoi polloi that is obtained by the first substitution (that is, E3). This indeed is the absurdity that Socrates is talking’.

⁵ Vlastos argues that the popular thesis is absurd because no-one could knowingly choose the smaller of two goods offered him, a proposition which follows from two others: (S1) If one knows that X is better than Y, one will want X more than Y, and (S2) If one wants X more than Y, one will choose X rather than Y. Vlastos says that (S2) has been virtually taken for granted: Socrates ‘repeatedly speaks of “wanting” (ἐθέλειν) a given option to express the very notion of choosing it’, and (S1) follows from fundamental Socratic tenets, i.e. that all men desire welfare and desire anything else only as a means thereto. And since the better action is the one which secures the greater aggregate good to the agent, it is impossible to choose the lesser in preference to the greater. Likewise with the second substitution: taking the smaller pleasure-package would be preferring the lesser good, which is a patent impossibility. But the next comment seems to me, in the context, to give the game away: ‘If his adversaries had not seen the impossibility of that consequence, Socrates would stand ready to derive it from the principle of psychological hedonism to which they had agreed at an earlier stage of the debate.’ (op. cit. p. 83–5). This is tantamount to an admission that Socrates has not shown the impossibility.

⁶ This account is based on D. Gallop, op. cit. Arguments against Gallop are advanced by Santas (op. cit. p. 12, n. 14) and Vlastos (op. cit. p. 83, n. 38). The former says that Gallop's version does not hold without psychological hedonism, but that if this is added to the premises the proof is already complete by 356c. Still, I believe it would give the required explicitness, and that would be a very good reason for going beyond 356c. Vlastos says that the people have already abandoned their thesis before they are brought to accept Socrates' interpretation of being overcome by pleasure. With this I agree, except about the point at which and the reason why it is abandoned or, rather, reinterpreted so as to be defensible. However, unless the absurdity of whatever sort is clearly stated earlier, as I do not think it is in Vlastos' account, Socrates would be justified in showing that his doctrine of ignorance is implied in the popular hedonism, and in using it to make explicit the contradiction in the popular thesis.

⁷ The varìation is pointed out by Vlastos, op. cit. p. 81–2. The popular thesis as criticised always has the plural elsewhere, since Socrates is concerned with relative quantities, 353a, 355a etc. But the popular description ‘being overcome by pleasure’ may have either singular, 352d, 353a, 357e, or plural, 353c, 354e. They are regarded as equivalent for the purpose in hand; at 358c ‘giving in to oneself’, ῆττω αὑτοῦ, is admitted for the same idea. Cf. expressions used in Charmides without any point being made of their differences: ἐπιστήμη ἐπιστήμης and ἐπιστήμη ἐπιστημῶν, 170a6, c6.

⁸ Cf. Republic iii 403e—it is γελοῑον that a guardian should need a guardian (of forbidding guardians to get drunk); the aviary in Theaetetus 199d: it is πολλὴ ἀλογία to suppose that ignorance can be caused by the presence of knowledge.

⁹ If 355d–e and 356a–c do not help to show the absurdity, but follow it, why are they separated from each other by the intrusion of the absurdity with the second substitution? The reason is that Socrates analyses only the latter part of the popular thesis, the expression ‘Overcome by pleasure’. Thus he first substitutes ‘Overcome by the goods’, gaining a fresher impact by freeing the expression from the confusing associations in the original expression. Here we have ‘does bad because overcome by the goods’ and naturally this leads to an analysis of wrong action, taking the lesser good. But in the second substitution he wants to analyse the positive side, right action, and now talks only of pleasure and pains, freeing the expression ‘Overcome by …’ from confusing associations in ‘Overcome by the goods’; right action is to take the greater pleasure. The two passages, though of different length, are virtually balanced point by point (1. substitution, 2. goods/ pleasures not worthy to overcome, 3. ‘Obviously since’ … (the reason is only implied in the second substitution), 4. unworthiness is relative quantity, 5. wrong/right action). The exposition is affected by dialogue form, giving a sufficient coverage with its own orderliness, different though this is from that appropriate to other modes of exposition.

¹⁰ My views of the origin of the alternative is in substantial agreement with that of Gulley, N., Phoenix xxv (1971) 118–21CrossRefGoogle Scholar, (on p. 120). The distinction between knowledge and belief at 358b-c is the same as the distinction between knowledge and ignorance in the preceding argument, with ‘ignorance’ now defined as having a false belief that something is right when in fact it is not. Likewise I agree with his interpretation of the function of ‘belief’ in the proof that courage is knowledge. I think, however, that is is crucial to stress, as Gulley does not, that Socrates has no justification in the preceding argument for the way he defines ‘ignorance’ at 358b-c, so that the transition is illogical, whatever way one looks at it. My contention is that the illogicality is deliberate Platonic manipulation.

¹¹ Developed by D. Gallop, op. cit.

¹² Illusory pain-pleasure values are compared with the deceptive effects of distance on visual objects at Philebus 41e–42c. Here Plato seems to waver between saying that when someone feels pain and pleasure simultaneously, the magnitude of the one may make the other appear less than it is (i.e. than it would be if experienced in a context not dominated by its opposite?), and e.g. that an immediate pain may make the pleasure of anticipation less than the pleasure is when realised (in which case there are two separate pleasure feelings). Only the latter suits the analogy of visual distance at all closely, but only the former suits the context of the argument, so that Plato perhaps ought to say that the effect of a strong immediate pain or pleasure on its simultaneously felt opposite is deceptive in much the same way as an object which dominates one's vision may make an object half the size but further away appear less than it is and as it would appear if not seen at such a disadvantage. The immediate pain is exaggerated correspondingly by its association with a pleasure of anticipation which is less because the anticipated pleasure is distant in time. Thus there are two separate effects, which Plato presents as one: (a) mutual distorting effects of different degrees of pain and pleasure experienced simultaneously; (b) the importance of proximity of a pleasurable or painful experience in producing the particular combination felt. The degree of falsity in a pleasure is the extent to which it is exaggerated as in (a), to which a contributory causal factor is proximity as in (b). Although Plato suggests that theoretically such an analysis reveals measurable quantities of true and false pleasure or pain, he does not imply that there is a technique available for making the measurement, or recommend such a technique to the hedonist. But this is what (ironically) he does propose in the Protagoras, and further, not only in connection with estimating an immediate experience, but a whole undertaking, project, outcome and all. Measurement remained for Plato the basis of accurate knowledge (as far as it is possible) of the sensible world, cf. Republic x 602d ff., and R. G. Bury, Appendix E to his edition of the Philebus (CUP 1897). Pleasure is measurable in a special sense at Republic ix 582 ff., by means of ‘experience, practical wisdom and reason’, ἐμπειρία, φρόνησις, λόγος; but there the question is about the type of pleasure to be found in each of the main ways of life, i.e. an assessment of available data, not, as the hedonist's position in the Protagoras requires, a way of calculating present facts and future eventualities.

¹³ It is not sufficient to refer Socrates' phrase ὑπò ἡδονῆς ἡττώμενος to a familiar Greek compendious expression meaning ‘Overcome by desire for pleasure’, because of the complete agreement at all stages of the entire argument that we are concerned not with strength of desire, but with evaluation of actions.

¹⁴ 354c–d.

¹⁵ The detail proves the important point that Plato is not concerned at all with strength of desire: the objector at 356a would in this case be suggesting that an immediate pleasure creates a stronger desire than a delayed one does, which would make Socrates' reply nonsensical. This reply, i.e. that an immediate pleasure differs from a delayed one only in being more or less pleasant, shows that the objection is to be referred to the evaluation of an action, not to its psychological motivation. So also with 355e: surely what one means by ‘being overcome’, ἡττᾶσθαι, in these contexts is not opting for an alternative which is in fact worse, as Socrates interprets it. This is part of what one means, of course, but the other part is a question of strengths of desire and willpower. Socrates' interpretation only makes sense if this latter aspect is omitted; if, that is, the issue is only how to evaluate a stock situation in hedonistic terms. This being so, any attempt to make the Socratic paradox watertight by introducing a premise about the relationship of strength of desire to estimates of the amount of good contained in a projected action, is bound to be talking about something else than the argument in the Protagoras.

¹⁶ Some of the difficulty in finding a precise interpretation is due to the tendency of the Greek language to drop the definite article in a predicate, even when, as here, the article is crucial to the sense. It is curious that Plato is not always on the alert for this even in passages where the presence or absence of the article is the central point, e.g. Hipp. Maj. 293e, cf. 294e; and compare Phil. 11b, with 60a; cf. also Euthyphro 8d–e. But there is no real possibility of misunderstanding in any of these passages, since the contexts make them clear. What makes one suspect that Plato is providing a loop-hole for Socrates here is that the context of the ensuing argument establishes acceptance of the hedonistic identification only for the sophists and not for Socrates.

¹⁷ Socrates could hold, as he does, that pleasure is good, and that good is pleasant (in the sense that the best life produces the most pleasure in the long run) and therefore that, in the long run, good is materially equivalent to pleasant, while at the same time denying that pleasure and the good are identical (he might assert p ≡ g and deny p = g without contradiction). Plato could express this distinction by use of the terms essence, οὐσία, and accident, πάθος, as he does in connection with the proposed definition of piety as what is loved by the gods, Euthyphro 10e–1 1a. It is hard to see why Plato should hold back from committing Socrates to the outright hedonistic identification unless he means to suggest some significance in the distinction between Socrates' view and that ascribed to the people and accepted by the sophists. If Plato is going on to criticise hedonism obliquely, then the distinction is very much to the point.

¹⁸ That the Protagoras is in part an ironical attack on the hedonism implicit in the teaching of the sophists is, of course, one of the main contenders for the title of the standard view of the dialogue, cf P. Shorey, The Unity of Plato's Thought, ch. 1 (reprinted in Plato ed. Vlastos, G., [Macmillan 1972] ii 26Google Scholar), and in more detail, Grube, G. M. A., CQ xxvii (1933) 203 ffGoogle Scholar. My interpretation adds what I believe is a vital reinforcement of this position, namely that the dialogue offers an oblique criticism of the hedonism attributed to the sophists. Without this we have to rely on other dialogues to know that Socrates is in fact criticising, not just explaining, the sophistic ethics.

¹⁹ The difference between the overt run of the argument and the substantial point of it may be put more formally. The argument, on the surface, goes:

1. Pleasure is identical to good.

2. Right action is action which contains most pleasure.

3. Knowledge determines which actions contain most pleasure.

Therefore 4. Knowledge determines right action. From which it follows, if we ignore the logical gap obscured by the ambiguous word ‘determines’ in 4, that wrong action is due to ignorance. But Socrates leaves us, though not without clues, to detect the falsity of 3 and to reconstruct the argument:

1. Knowledge determines right action.

2. Pleasure is identical to good.

3. Right action is action which contains most pleasure.

Therefore 4. Knowledge determines which actions contain most pleasure.

But 5. Knowledge does not determine which actions contain most pleasure.

Therefore, either 6. Knowledge does not determine right action.

Or 7. Pleasure is not identical to good.

Now, since Socrates has enthusiastically committed himself to 1, and has carefully avoided endorsing 2, the sting of the criticism is turned against 2. Socrates has not proved 1, but Plato uses his conviction of the truth of 1 to draw out the falsity of 2.

²⁰ For the criticism of the sophists, see csp. Shorey and Grube quoted in note 18 above. For the view that Plato is serious about hedonism, see Hackforth, R., CQ xxii (1928) 39Google Scholar ff., Crombie, I. M., An Examination of Plato's Doctrines (Routledge and Kegan Paul 1962) i 240Google Scholar ff., Vlastos, G., intro, to Protagoras (Bobbs-Merrill 1956) p. xl ff.Google Scholar, esp. n. 50. Vlastos' main argument for the view that Socrates sponsors the identification of pleasure and good is that it would be misleading to base the important Socratic thesis that wrongdoing is ignorance upon a falsehood, unless sufficient signs that this is being done are given, which is not the case. There is some force in this, though it is not clear how far Plato felt himself entitled to go within the limits of Socratic elenchus. For instance, the argument from opposites to prove the identity of temperance and wisdom, 322a ff., is accepted by Protagoras and allowed to stand, though it is hard to think that Plato felt it to be much more than suggestive. Likewise at Euthydemus 279e ff., Socrates uses ambiguities in the word εὐτυχία to encourage young Kleinias to study philosophy; the ambiguity is never acknowledged, although the reader is prepared to be on the look out by explicit comments in the preceding discussion, 277e–278b. Similarly, though more subtly, the ironical description of the sophists and their entourage 315a ff., the treatment of Simonides’ poem 340a ff., the attempted proof that courage is wisdom via the importance of knowledge in confidence (about which Socrates can hardly be entirely sincere, cf. Laches 193c) 349e–350c, and the airing of the notion of illicit conversion 350C–351b, all put the reader on his guard. Admittedly there is no explicit declaration that Socrates has reservations about hedonism, yet it is surely odd that Socrates does not declare for it either, but for something similar but different, while hedonism is several times expressly said to be implicit in popular values. Assuming Plato to be in control, this too would be misleading unless it is a sign that Socrates is not committed to hedonism and is not going to commit himself to it.