Show Summary Details

p. 454.locked

Plato's Socrateslocked

  • C. C. W. Taylor


‘Plato's Socrates’ shows that Socrates is predominantly characterized, not as a teacher, but as an enquirer. He disclaims wisdom, and seeks, normally in vain, elucidation of problematic questions from his interlocutors, by the method of elenchus — that is, by critically examining their beliefs. All the dialogues are concerned with ethics in the broad sense of how one should live. Many focus on the attempt to define a virtue or other ethically significant concept. The discussion invariably ends in apparent failure. Others take the form of a confrontation between Socrates and various sophists and/or their pupils and associates over the aims of education and the proper qualifications of the educator.

As indicated at the end of the last chapter, we shall be considering the portrayal of Socrates' doctrines and methods of argument in twelve dialogues plus Apology. The following features are common to all or most of these dialogues.


Characterization of Socrates. Socrates is predominantly characterized, not as a teacher, but as an enquirer. He disclaims wisdom, and seeks, normally in vain, elucidation of problematic questions from his interlocutors, by the method of elenchus, that is, by critically examining their beliefs. In some dialogues, notably Protagoras and Gorgias, the questioning stance gives way to a more authoritative tone.


Definition. Many of the dialogues are concerned with the attempt to define a virtue or other ethically significant concept. Euthyphro asks ‘What is holiness or piety?, Charmides ‘What is temperance?’, Laches ‘What is courage?’, Hippias Major ‘What is fineness or beauty?’ Both Meno, explicitly, and Protagoras, implicitly, consider the general question ‘What is virtue or excellence?’ In all these dialogues the discussion ends in ostensible failure, with Socrates and his interlocutor(s) acknowledging that they have failed to find the answer to the central question; in some cases there are textual indications of what the correct answer is.


Ethics. All these dialogues are concerned with ethics in the broad p. 46sense of how one should live. Besides those dialogues which seek definitions, Crito deals with a practical ethical problem: should Socrates try to escape from prison after his sentence; and both Gorgias and Euthydemus examine what the aims of life should be. The only ostensible exception is Ion, which is an examination of the claim of a professional reciter of poetry to possess wisdom. But even that ties in closely with the general ethical interest of these dialogues, since the debunking of Ion's claims to wisdom has the implication that both poets and their interpreters are directed not by wisdom, but by non‐rational inspiration, and hence that poetry has no claim to the central educational role which Greek tradition ascribed to it. This little dialogue should be seen as an early essay on the topic which preoccupies much of Plato's writing, namely, the aims of education and the proper qualifications of the educator.


Sophists. In several of these dialogues, namely, the Hippias dialogues, Protagoras, Gorgias, Euthydemus, and Meno, that topic is pursued via the portrayal of a confrontation between Socrates on the one hand and various sophists and/or their pupils and associates on the other. These dialogues thereby develop the apologetic project enunciated in the Apology.

These topics will now be considered in more detail.

Socrates' Disavowal of Wisdom

That Socrates denied having any knowledge, except the knowledge that he had no knowledge, became a catchword in antiquity. But that paradoxical formulation is a clear misreading of Plato. Though Socrates frequently says that he does not know the answer to the particular question under discussion, he never says that he knows nothing whatever, and indeed he makes some emphatic claims to knowledge, most notably in the Apology, where he twice claims to know that abandoning his divine mission would be bad and disgraceful (29b, p. 4737b). What he does disavow is having any wisdom (Apol. 21b), and consequently he denies that he educates people, clearly understanding education as handing on a body of wisdom or learning (19d–20c). Given his assertions in the Apology that only god is truly wise and human wisdom is nothing in comparison to that true wisdom (23a–b), the denial of wisdom might be understood as simply the acceptance of human limitations. To possess wisdom would be to have the complete and totally perspicuous understanding of everything which is the prerogative of god. Neither Socrates nor anyone else can hope to aspire to that, and in denying that he has it Socrates is simply setting his face against a human arrogance which is none the less blasphemous for being virtually universal.

But while the devaluation of human wisdom as such is indeed a strain in the Apology, in denying that he possesses wisdom and, consequently, that he teaches people, Socrates is contrasting his own condition, not with the divine wisdom, but with a human paradigm of wisdom. This paradigm is realized by craftsmen such as builders and shoemakers who, he acknowledges (22d–e) do possess wisdom in the sense that they are masters of their craft, though they go wrong in thinking that their special expertise extends to matters outside the scope of the craft. This expertise is a structured body of knowledge which is systematically acquired and communicated to others, by possession of which the expert is able reliably to solve the practical problems posed by the craft and to explain the grounds of their solution. The sophists claimed to possess, and to teach to others, such an expertise applying to overall success in social and personal life, the ‘political craft’ (politikē technē) (Prot. 319a, Apol. 19d–20c). Though Socrates rejects these claims, it is not on the ground that such expertise is not available to human beings, but on the ground that the sophists' activity fails to meet the ordinary criteria for being a genuine expertise, for example, that of being systematically learned and taught (Prot. 319d–320b, Meno 89c–94e). He denies that he possesses this expertise himself (Apol. 20c), but p. 48does not say that it is impossible that he, or anyone, could possess it.

There is, then, no ground to assume that Socrates' disavowal of knowledge is an instance of what has become known as ‘Socratic irony’, that is, pretended ignorance for dialectical purposes. Socrates does indeed frequently pose as admiring the supposedly superior knowledge of the person he is talking to (e.g. Euthyph. 5a–b, where he says that he ought to take instruction from Euthyphro on how to defend himself against Meletus' accusation), but the reader, at any rate, is clearly not supposed to be taken in; on the contrary, these avowals serve to point up the particularly controversial character of what the interlocutor has said, or the dubiousness of his claim to authority. The context of the Apology, however, rules out any such dialectical function for the disavowal of knowledge. Socrates is not there posing as deferring to a supposed, but actually bogus, epistemic authority; he is with perfect sincerity matching his own epistemic state against an appropriate paradigm, and finding it wanting.

If the disavowal of knowledge is in fact the disavowal of wisdom or expertise, we can see how that disavowal is compatible with the particular claims to knowledge which Socrates makes. The non‐expert can know some particular things, but not in the way that the expert knows them; specifically those particular items of knowledge do not fit into a comprehensive web of knowledge which allows the expert to provide explanations of their truth by relating them to other items and/or to the structure as a whole. But how does the non‐expert know those things? Usually, by having been told, directly or indirectly, by an expert. Socrates does not, however, recognize any experts, at least human experts, in matters of morality. So how does he know, for example, that he must not abandon his mission to philosophize, whatever the cost? A possible answer is that he has been told this by god, who is an expert in morality. But, leaving aside questions (suggested by Euthyphro) of how he knows that god is an expert in p. 49morality, that is not in fact an answer which is given or even suggested in Apology or elsewhere.

One might attempt to dissolve the problem by suggesting that Socrates does not intend to claim knowledge of these things, but merely to express his beliefs. But Plato makes him say that he knows them, so why should we suppose that Plato does not represent him as meaning what he says? As we have seen, Socrates does indeed recognize an ideal epistemic paradigm which he fails to satisfy, yet he claims knowledge in particular cases. The suggestion being considered amounts to this, that satisfaction of the paradigm is to be equated with knowledge, while the epistemically less satisfactory state which Socrates is in is to be relegated to that of belief. But the distinction between paradigm‐satisfying and epistemically inferior states can be maintained without denying the latter the title of knowledge, by using the distinction between the expert's integrated knowledge and the non‐expert's fragmentary knowledge. (We might, if we choose, talk of the former as knowledge ‘strictly speaking’ and the latter as knowledge ‘for ordinary purposes’ or ‘in a loose and popular sense’. Plato does not in fact use such locutions, but the essential distinction is unaltered.) We are, then, still left with the question how Socrates, an avowed non‐expert in matters of morality, knows the particular moral truths which he claims to know.

The straightforward, though perhaps disappointing, answer is that Socrates does not say how he knows those truths. Consideration of his argumentative practice may give us some clues. Often his arguments seem intended to do no more than reveal that his interlocutor has inconsistent beliefs about some matter on which he purports to have knowledge, and thereby to undermine that claim to knowledge, as Socrates describes himself in the Apology as doing. But at least sometimes he clearly thinks that, provided his interlocutor maintains nothing but what he sincerely believes, the critical examination of those beliefs will reveal, not merely p. 50inconsistency among them, but the falsehood of some belief. A particularly clear case is the claim of Polus and Callicles in Gorgias that it is better to do wrong than to suffer it. Socrates claims (479e) that the critical arguments by which he has led Polus to accept the contrary thesis that it is worse to do wrong than to suffer it have proved that the latter is true, and asserts even more emphatically at the end of the argument with Callicles (508e–509a) that that conclusion has been ‘tied down with arguments of iron and adamant’ (i.e. of irresistible force). Yet this very strong claim is conjoined with a disavowal of knowledge: ‘My position is always the same, that I do not know how these things are, but no one I have ever met, as in the present case, has been able to deny them without making himself ridiculous.’

Here we have a contrast between expert knowledge, which Socrates disavows, and a favourable epistemic position produced by repeated application of the elenchus. There are some propositions which repeated experiment shows no one to be capable of denying without self‐contradiction. Commitment to these is always in principle provisional, since there is always the theoretical possibility that someone might come up with a new argument which might allow escape even from the ‘arguments of iron and adamant’, as Socrates acknowledges (509a2–4). But realistically, Socrates clearly believes, the arguments rely on principles which are so firmly entrenched that there is no practical possibility of anyone's denying them. Might the truths which Socrates knows non‐expertly be truths which he has thus established via the elenchus? While that is an attractive suggestion, we have to acknowledge that it has no clear textual confirmation. In Crito (49a) the fundamental proposition that one must never act unjustly is said to be one which Socrates and Crito have often agreed on, and that agreement is to bind them in considering the propriety of Socrates' attempting to escape from prison. The implication is, surely, that the agreement was based on reasons which are still in force; otherwise why should Socrates and Crito not change their minds? But there is nothing p. 51to suggest that those reasons took the form of elenchus of Socrates' and Crito's beliefs.

Our conclusion has to be that, though Socrates treats elenchus of the interlocutor's belief as sometimes revealing truth, and though the achieving of truth by that means provides a possible model for non‐expert knowledge, we are not justified in attributing to Socrates the claim that all non‐expert moral knowledge is in fact achieved via that method. He gives some indication that he knows some moral truths on the strength of having a good argument for them, but he gives no general account of the conditions for non‐expert moral knowledge.

Gorgias is the dialogue which provides the clearest cases in which the elenchus is seen as leading to the discovery of truth, and it is probably not coincidental that in the same dialogue we find Socrates abandoning his stance as a non‐expert questioner and claiming expertise. One of the themes of the dialogue is the role of rhetoric in education, that is, in promoting the good life. Socrates sets up a taxonomy of genuine crafts concerned respectively with the good of the soul and that of the body, and of counterfeits corresponding to each (463a–465a). The generic name for the craft concerned with the good of the soul is politikē, the art of life, subdivided into legislation, which promotes the good of the soul (as gymnastics promotes the good of the body), and justice, which preserves it (as medicine preserves the good of the body). Rhetoric is the bogus counterpart of politikē, since the aim of the orator is not to promote people's good, but to pander to their wishes by enabling them to get what they want through the power of persuasion. It thus promotes, not the genuinely good life, but a spurious appearance of it, as cosmetics is the skill, not of making people actually healthy, but of making them look healthy (465c). Politikē is thus a genuine expertise, and in striking contrast to his stance in the Apology we find Socrates not merely claiming that he practises it, but that no one else does (521d), since he alone cares for the good of his fellow‐citizens.

p. 52This conception of Socrates as the only genuine practitioner of politikē recurs in an image at the conclusion of Meno (99e–100a), where Socrates sums up the conclusion of the argument that goodness cannot be taught, but is acquired by a divine gift without intelligence ‘unless there were one of the politikoi who was capable of making someone else politikos' (i.e. unless there were someone who could pass his expertise in the art of life on to another, as conventional politicians have shown themselves incapable of doing). He goes on to say that such a man would be like Homer's description of Tiresias in the underworld (in the Odyssey): ‘He alone of those in Hades is alive, and the rest flit about like shadows.’ This reference to Odysseus' visit to the underworld in Odyssey 11 picks up the description of Socrates' meeting with the sophists in Protagoras, where Socrates refers to the sophists by quoting the words of Odysseus (315b–c), thereby casting himself as a living man and the sophists as shadows (i.e. ghosts). He is then the real expert in the art of life ‘the real thing with respect to goodness, compared with shadows' (Meno 100a), who has (in Meno and Protagoras) a positive conception of the nature of goodness and (in Meno) a new method of transmitting that conception to others. This is the method of recollection, in which knowledge which the soul has possessed from all eternity but forgotten in the process of reincarnation is revived via the process of critical examination.

The development of this more authoritative figure of Socrates is a feature of dialogues which we identified as transitional between the earlier ‘Socratic’ dialogues and the dialogues of Plato's middle period. It is a particular instance of the gradual metamorphosis of the figure of Socrates into the representative of Plato which we noted earlier.


Socrates' interest in definitions arises from his quest for expertise. The expert knows about his or her subject, and according to Socrates the primary knowledge concerning any subject is precisely knowledge of p. 53what that subject is. The connection with expertise is made explicit in Hippias Major (286c–d), where Socrates tells Hippias how, when he was praising some things as fine and condemning others as disgraceful, he was rudely challenged by someone who said, ‘How do you know what sorts of thing are fine and what sorts disgraceful? Tell me, could you say what fineness is?’ Being unable to meet this challenge he consults Hippias, whose universal expertise includes, as ‘a small and unimportant part’, knowledge of what fineness is; if Hippias were unable to answer that question his activity would be ‘worthless and inexpert’ (286e).

The primacy of the ‘What is such‐and‐such?’ question is emphasized in a number of dialogues. The general pattern of argument is that some specific question concerning a subject, which is the actual starting‐point of discussion, for example, how one is to acquire goodness, is problematic in the absence of an agreed conception of what that subject, in this case goodness, is. Hence, though the specific question is psychologically prior, in that that is where one actually begins the enquiry, the ‘What is X?’ question is epistemologically prior, in the sense that it is impossible to answer the former without having answered the latter but not vice versa. The problematic question may be of various kinds. In Laches (189d–190d) it is how a particular virtue, courage, is to be inculcated, while in Meno (70a–71b) and Protagoras (329a–d, 360e–361a) it is the generalization of that question to goodness as such. In Republic 1 (354b–c) it is whether justice is advantageous to its possessor. In Euthyphro (4b–5d) it is whether a particular disputed case, Euthyphro's prosecution of his father for homicide, is or is not an instance of piety or holiness. Similarly, at Charm. 158c–59a the question of whether Charmides has self‐control is treated as problematic, and therefore as requiring prior consideration of what self‐control is.

The pattern exhibited by the last two examples, in which the question ‘Is this an instance of property F?’ is said to be unsettleable without a p. 54prior answer to the question ‘What is E?’, has given rise to the accusation that Socrates is guilty of what has been dubbed ‘the Socratic fallacy’, namely, maintaining that it is impossible to tell whether anything is an instance of any property unless one is in possession of a definition of that property. That thesis would be disastrous for Socrates to maintain, not merely because it is open to countless counter‐examples (e.g. we can all tell that a five‐pound note is an instance of money even if we are unable to give a definition of money), but because Socrates' approved strategy for reaching a definition is to consider what instances of the kind or property in question have in common (e.g. Meno 72a–c). Obviously, if we cannot tell which are the instances of the kind or property in question in advance of giving the definition that procedure is futile, as is the procedure of rejecting a definition by producing a counter‐example, for if you cannot tell whether any instance is an instance without a definition, equally you cannot tell whether any instance is not an instance without a definition. But since the production of counter‐examples is one of the standard procedures of Socratic elenchus, the fallacy would be wholly destructive of Socrates' argumentative method.

In fact, Socrates is not committed to that methodologically self‐destructive position. The most that the examples in Euthyphro and Charmides commit him to is that there are some, disputed, instances, where the question ‘Is this an instance of E?’ cannot be settled without answering the prior question ‘What is E?’. That claim does not commit him to maintaining that there are no undisputed cases, and so leaves it open to him to look for a property present in all the undisputed cases of F and absent from all the undisputed cases of non‐F, and then to settle the disputed cases by determining whether that property applies to them. (In fact that procedure is bound to leave the dispute unsettled, because the original dispute is now transformed into a dispute over the propriety of widening the extension of the property from the undisputed to the disputed cases. That, however, is another question.)

p. 55Socrates' rude challenger in Hippias Major does, however, appear to go so far as to claim that it is impossible to tell whether any particular thing is fine before one has given a definition of fineness. When all Socrates' and Hippias' attempts at defining fineness have failed, Socrates imagines himself being confronted again by the challenger and asked, ‘How will you know whether any speech has been finely put together, or any action whatever finely done, if you are ignorant of fineness? And if you are in that state, do you think you are better off alive than dead?’ (304d–e). We cannot avoid the difficulty by saying simply that this is someone else's view, not Socrates’, since Socrates makes it clear that the rude challenger is an alter ego; ‘he happens to be very nearly related to me and lives in the same house’ (304d). Yet the rude challenger's view is not one which Socrates simply endorses, for he concludes (304e) by saying that he thinks he knows that the proverb ‘Fine things are difficult’ is true; but on the challenger's account he could not be in a position to know even that. The challenger's view, then, is not after all Socrates' own; it is very closely related to it, indeed (and thereby likely to be confused with it), and constitutes a challenge in that, if accepted, it would overthrow Socrates' entire argumentative methodology. Hence the challenge is to distinguish that view from Socrates' actual, more modest view that there are some difficult cases which cannot be settled without the ability to give a definition. To be an expert in an area is to be able to tell reliably, for disputed and undisputed cases alike, whether any case is an instance of the property or kind in question, and for that, according to Socrates, it is necessary, as well as sufficient, to be able to say what the property or kind is.

The examples from Laches, Meno, Protagoras, and Republic 1 exhibit another pattern; here the question which gives rise to the quest for the definition of a property is not whether a given, disputed, instance falls under it, but whether that property itself has some further property, specifically whether justice is beneficial to its possessor, and whether courage and overall goodness (i.e. the possession of all the virtues, p. 56courage, self‐control, justice, wisdom, etc.) can be taught. At Meno 71b Socrates gives an analogy for this pattern of the priority of definition which suggests that it is the most basic platitude. If I don't know at all who Meno is, I can't know whether he has any property, for example, whether he is rich or handsome. Similarly, if I don't know at all what goodness is, there is no possibility of my knowing anything about it, including how it is to be acquired.

Understood in a particular way, this is indeed a platitude. If I have never heard of Meno, the appropriate reply to the question ‘Is Meno handsome?’ is ‘Sorry, I don't know whom you mean.’ Similarly, if I have no idea what goodness is, the appropriate reply to ‘Can goodness be taught?’ is ‘Sorry, I don't know what you're talking about.’ Here we have cases where a prerequisite of intelligible speech about a subject, that one should be able to identify the subject, is not fulfilled. Clearly, that prerequisite of intelligible speech does not require the ability to give a definition of the subject. In the case of an individual subject such as Meno one does not have to be in possession of any specification of Meno which uniquely specifies him independently of context; one might, for instance, be able to identify him only ostensively as ‘That man over there’, or indefinitely as ‘Someone I met in a pub last year’. The analogue in the case of a universal such as goodness is no more than the minimal requirement to know what we are talking about when we use the word; but that again does not presuppose the ability to give a verbal specification (i.e. a definition) of the universal. To return to our earlier example, I can know what I am talking about when I use the word ‘money’, even if I am unable to give a definition of money; it is clearly enough that I can, for instance, recognize standard instances. Now, in that sense it is clear that Meno knows what he is talking about from the very start; otherwise he could not even raise his initial question ‘Can goodness be taught?’ So the platitude that intelligible speech about any subject requires the ability to identify that subject does not point towards the priority of definition. Why, then, does Socrates insist on p. 57that priority even though the condition which the platitude specifies is satisfied?

To answer that question we need to observe that in Laches, Meno, and Protagoras the search for the definition of particular virtues and inclusive goodness is prompted by the practical question of how those qualities are to be acquired. What kind of definition of those qualities is demanded by the practical question? Clearly, something more than the bare ability to know what one is talking about is demanded, because, as we have seen, that ability is presupposed by the asking of the practical question itself. It is tempting to suggest that what more is required is just the ability to elucidate the dictionary meaning of the term designating the quality under discussion. In the case of the Greek term which I have rendered distributively as ‘virtue’ and collectively as ‘goodness’ (aretē), a reasonably accurate specification of its meaning would be:


An attribute of an agent, one of a set of attributes severally necessary and jointly sufficient for the attainment of overall success in life.


The set of attributes specified under 1.

How is the ability to give that elucidation demanded by the practical question? It does indeed advance the enquiry to the extent of making it clear that the search is for properties which promote success in life, but it gives no indication what properties those are, nor, crucially, how those properties are to be acquired. People could agree on that definition but disagree radically in their answers to the practical question, if, for instance, some thought that the properties which bring success in life are all gifts of nature such as intelligence and noble lineage, while others thought that they could all be acquired through practice like practical abilities. The practical question thus appears to demand a different kind of definition from the elucidation of the meaning of the term which designates the property; it demands a p. 58substantive specification of what that property is. A substantive specification will include both the decomposition of a complex of properties into the components of that complex (e.g. goodness consists of justice, self‐control, etc.) and explanatory accounts of those properties (e.g. self‐control consists in the control of the bodily appetites by reason). That is to say, it provides a theory of goodness, which explains it by identifying its constituents and causes, and thereby indicates appropriate methods of acquiring it.

That the definitions sought are of this substantive kind chimes in well with the demand that the giving of definitions is what characterizes the expert. The expert on goodness should be able to explain what goodness is with a view to providing reliable guidance on how to acquire and maintain goodness, just as the expert on health should be able to explain what health is with a view to providing reliable guidance on how to become and stay healthy. The texts of the dialogues mentioned above provide some confirmation that the definitions sought are of this kind, though it would be an oversimplification to pretend that they are distinguished with total clarity from elucidations of the meanings of the terms designating the properties in question.

That Socrates' search is for substantive rather than purely conceptual or ‘analytic’ definitions is indicated by those dialogues which either explicitly identify or suggest the identification of goodness with knowledge or some other cognitive state. The most detailed discussion occurs in Meno (suggested above to be transitional between ‘Socratic’ and ‘Platonic’). At 75–6 Socrates attempts to explain to Meno that he is looking not for lists of specific virtues such as courage and self‐control but for a specification of what those virtues have in common, and illustrates this by giving two model specifications, first of shape and then of colour. Of these, the former is a conceptual elucidation, namely, that shape is the limit of a solid, and the second a ‘scientific’ account of colour (based on the theory of the fifth‐century philosopher p. 59Empedocles) as a stream of particles flowing out from the perceived object, of appropriate size and shape to pass through channels in the eye to the internal perceptive organ. Socrates gives no clear indication that he regards these specifications as of different kinds; he says that he prefers the former, but does not indicate why, except that he describes the latter as ‘high‐flown’, perhaps indicating that it is inferior because it is couched in over‐elaborate technical terminology. Despite this expressed preference for what is in fact a conceptual elucidation over a substantive definition, Socrates then goes on to propose an account of goodness of the latter kind, namely, that goodness is knowledge. This is not itself an elucidation of the concept of goodness, as specified above, though it does depend on a conceptual thesis, that goodness is advantageous to its possessor (in Greek, that aretē is ōphelimon, 87e). Rather, it is the identification of knowledge as that state which is in fact necessary and/or sufficient for success in life, and it is arrived at not purely by considering the meanings of words but by the adducing of a highly general thesis about how success is achieved. The thesis is that since every other desirable property, such as strength or boldness, can lead to disaster, the only unconditionally good thing is that which provides the proper direction of those qualities, namely, intelligence, which is equated with knowledge (87d–89c). Again, Socrates is led ostensibly to abandon that account in favour of the revised suggestion that goodness is not knowledge but true belief (89c–97c) by consideration of the alleged empirical fact that there are no experts in goodness, as there would have to be if goodness were some kind of knowledge (another conceptual thesis). In Socrates' arguments conceptual theses and general empirical claims about human nature mesh to provide the best available theory of what goodness really is, that is, of what property best fits the specification set out in the elucidation of the concept given above.

In Meno, then, the practical question of how goodness is acquired leads to a substantive account of goodness as a cognitive state. It is no coincidence that the two other dialogues which begin from that p. 60question, either about goodness in general (Protagoras) or about a particular virtue (courage in Laches), exhibit a similar pattern of development. In Protagoras Socrates' young friend Hippocrates begins by assuming that the way to acquire goodness is to be taught it by Protagoras, but the sophist's conception of goodness as a cluster of only contingently connected attributes is rejected in favour of what is in effect a version of the theory proposed in Meno, that goodness is knowledge. In Laches the question of how courage is to be acquired leads, after the rejection of various alternative suggestions, to a specific version of the theory that goodness is knowledge, namely, that courage is knowledge of what is and what is not to be feared (194e–195a). This is eventually rejected on the grounds that, since what is and what is not to be feared is identical with what is and what is not bad, courage will then just be the knowledge of what is and what is not bad. But since, on this cognitive account, that is precisely what goodness as a whole is, courage will be identical with goodness as a whole, instead of a part of it, as was the original hypothesis (198a–199e). Hence the dialogue ends with the admission that the participants have failed in their search for what courage is. Commentators disagree on whether this inconclusive outcome is to be taken at face value, and, if not, which of the assumptions which lead to it should be abandoned. The significant point to observe is that here again the practical question leads not merely to a substantive account of the property in question but towards the same account as is canvassed in Meno and Protagoras.

I do not wish to suggest that at the time of writing these dialogues Plato had a clear grasp of the distinction between purely conceptual definitions and the substantive type of account exemplified by the cognitive theory. The fact that even in the dialogue which discusses definition in greatest detail, Meno, which I assume to have been one of the latest of the dialogues I discuss, he gives as model definitions an example of either kind without any explicit differentiation suggests that he had not arrived at any theoretical discrimination between the two. My suggestion is rather that his practice shows him favouring a p. 61kind of definition which we can characterize as substantive rather than conceptual, and that the practical orientation of the discussions leading to those definitions provides an explanation of that fact.

Sometimes the course of the dialogue is even less clearly indicated. In Euthyphro the initial question is ‘What property is it in virtue of which things (especially kinds of actions) are holy?’ When Euthyphro suggests (6e–7a) that it is the property of being approved of by the gods (which is very close to an elucidation of the ordinary Greek conception of to hosion), Socrates elicits from him the assertion that the gods approve of holy things because they are holy (10d). This excludes the possibility that holiness should be that very property of being approved of by the gods, and points the rest of the discussion in the direction of a search for the kind of conduct which attracts the gods' approval. Here too we may say that Socrates is groping towards a substantive account of holiness, in that the answer would have to be given in terms of a theory of human nature and its relation to the divine, but the dialogue provides no more than hints as to the detailed form of such a theory. The situation in Charmides is even less clear‐cut, partly because the virtue under discussion, sōphrosunē (conventionally translated ‘self‐control’, but sometimes better rendered ‘soundness of mind’), is genuinely indeterminate between a style of behaviour and the mental and motivational state directing it. Hence the various suggestions that it is one kind or another of knowledge are less easy to classify as either conceptual elucidations or substantive accounts than the suggested definitions in Laches, Meno, and Protagoras.


The search for definitions, then, is the search for expertise, and the possessor of expertise possesses a theory of the subject‐matter of that expertise, a grasp of its nature which delivers answers to further questions, both theoretical and practical, about it. In the dialogues discussed in the previous section we see Socrates searching for such a p. 62theory applied to human goodness, in some cases a theory of one of the constituents of goodness, that is, an individual virtue (piety in Euthyphro, courage in Laches, and self‐control in Charmides), in others (Meno, Protagoras) a theory of goodness as a whole. In all of these the search is, at least ostensibly, unsuccessful, in that each dialogue ends with the acknowledgement by Socrates and his interlocutors that they have not arrived at the account of goodness or of its parts which they were seeking. But there are some discernible differences. In the three dialogues dealing with individual virtues the discussion is more tentative, Socrates is not readily identified with any positive position, and it is at least plausible to accept the final impasse at its face value. In Meno and Protagoras, on the other hand, Socrates argues firmly for the thesis that virtue is knowledge, and it is plausible to think that the ostensibly aporetic conclusions are to be interpreted as not detracting from his commitment to that thesis. In these dialogues, it seems to me, Plato depicts Socrates not indeed as possessing the fully developed theory of goodness which is his goal but at least as having a grasp of its general shape. There is, then, even within the dialogues of definition, a development in the portrayal of Socrates from that of purely critical searcher to the proponent of theory (though not expert in the fullest sense). It is an open question whether this development is one within Plato's perception of the historical Socrates, or the first stage of a development from that perception to a presentation containing more of his own views.

The basis of the theory is the combination of the conception of goodness as that property which guarantees overall success in life with the substantive thesis that what in fact guarantees that success is knowledge of what is best for the agent. This in turn rests on a single comprehensive theory of human motivation, namely, that the agent's conception of what is overall best for him‐ or herself (i.e. what best promotes eudaimonia, overall success in life) is sufficient to motivate action with a view to its own realization. This motivation involves desire as well as belief; Socrates maintains (Meno 77c, 78b) that p. 63everyone desires good things, which in context has to be interpreted as the strong thesis that the desire for good is a standing motive, which requires to be focused in one direction or another via a conception of the overall good. Given that focus, desire is locked onto the target which is picked out by the conception, without the possibility of interference by conflicting desires. Hence all that is required for correct conduct is the correct focus, which has to be a correct conception of the agent's overall good.

On this theory motivation is uniform, and uniformly self‐interested; every agent always aims at what he or she takes to be best for him‐ or herself, and failure to achieve that aim has to be explained by failure to grasp it properly, that is, by a cognitive defect, not by any defect of motivation. Socrates spells this out in Protagoras, on the assumption, which he attributes to people generally, that the agent's overall interest is to be defined in hedonistic terms, as the life which gives the best available balance of pleasure over distress. Given that assumption, it is nonsense to explain doing wrong by being overcome by pleasure or by any kind of desire; one must simply have made a mistake in one's estimation of what would bring the most pleasure. As Socrates says (358d), ‘It is not in human nature to be prepared to go for what you think to be bad in preference to what is good.’ There is considerable disagreement among commentators as to whether Socrates is represented as accepting the hedonistic assumption himself or merely as assuming it ad hominem to show that Protagoras has no view other than common opinion, but there is no doubt that, independently of that question, the view that the agent's conception of the good is the unique focus of motivation (maintained also in Meno) is Socrates' own. This account of goodness as knowledge thus issues directly in one of the claims for which Socrates was notorious in antiquity, the denial of the possibility of action against the agent's better judgement (akrasia); in Aristotle's words (Nicomachean Ethics 1145b26–7) Socrates used to maintain that ‘no one acts contrary to what is best in the belief that he is doing so, but through error’, a thesis expressed more concisely in p. 64the slogan ‘No one goes wrong intentionally’ (oudeis hekōn hamartanei (Prot. 345e)).

Thus far the theory identifies goodness with the property which guarantees overall success in life, and identifies that property, via the motivational theory just described, with knowledge of what is best for the agent. But that theory lacks moral content; nothing in it shows or even suggests that what is best for the agent is to live a morally good life, as defined by the practice of the traditional virtues, including justice, with its implications of regard for others, and self‐control, with its implications of the sacrifice of self‐gratification. But if anything is characteristic of Socrates it is his insistence on the pre‐eminence of morality. We saw that in the Apology he says that he knows that, come what may, he must not do wrong by disobeying the divine command to philosophize, and in Crito the fundamental thesis that one must never do wrong (or ‘commit injustice’ (adikein)) is the determining principle of his decision not to attempt escape from prison (49a–b). The link with the motivational thesis is established by the thesis that the best life for the agent is a life lived in accordance with the requirements of morality. Given that thesis, the slogan that no one goes wrong intentionally takes on the moral dimension that ‘no one willingly does wrong (or ‘acts unjustly’), but all who do wrong do so involuntarily’ (or ‘unintentionally’) (Gorg. 509e), the full moral version of what has become known as the ‘Socratic paradox’.

The thesis that the moral life is the best life for the agent thus has the central role of linking Socrates' intuitions of the pre‐eminence of morality with the theory of uniform self‐interested motivation which is the foundation of the identification of goodness with knowledge. It is the keystone of the entire arch. Given that centrality, it is surprising how little argumentative support it receives. At Crito 47e justice and injustice are described as respectively the health and sickness of the soul; hence, just as it is not worth living with a diseased and corrupted body, so it is not worth living with a diseased and corrupted soul. But p. 65that is not an argument. Even granted that health is an intrinsically desirable and disease an intrinsically undesirable state, the crucial claims that justice is the health of the soul, and injustice its disease, require defence, not mere assertion.

Plato supplies some arguments in Gorgias, but they are weak. Against Polus Socrates argues that successful tyrants, who, it is agreed, manifest the extremes of injustice, do not secure the best life for themselves, as Polus claims. On the contrary, they never get what they really want, because what they want is to do well for themselves, whereas their injustice is bad for them. The proof that it is bad for them (473e–475c) starts from Polus' admission that acting unjustly, while good (agathon) for the agent, is disgraceful (aischron). Socrates then secures agreement to the principle that whatever is disgraceful is so either because it is unpleasant, or because it is disadvantageous. Acting unjustly is clearly not unpleasant; hence by the above premisses it must be disadvantageous. Hence a life of injustice is bad for the agent. Of the many weaknesses of this argument the crucial one is its neglect of the relativity of the concepts. To be acceptable the first premiss must be read as ‘Whatever is disgraceful to anyone, is so either because it is unpleasant to someone or because it is disadvantageous to someone.’ Given that premiss, it obviously does not follow that, because injustice is not unpleasant to the unjust person it must be disadvantageous to that person; it could be disadvantageous to someone else, and its being so could be the ground of its being disgraceful to the unjust person. (Indeed, one of the main reasons why we think that injustice is disgraceful to the perpetrator is that it is typically harmful to someone else.) Later in the dialogue (503e–504d) Socrates argues against Callicles that, since the goodness of anything (e.g. a boat or a house) depends on the proper proportion and order of its components, the goodness of both body and soul must depend on the proper proportion and order of their components, respectively health for the body and justice and self‐control for the soul. The parallelism of bodily health and virtue, which was simply asserted in p. 66Crito, is here supported by the general principle that goodness depends on organization of components, but that principle is insufficient to establish the parallelism. For the proper organization of components is itself determined by the function of the kind of thing in question; it is by considering that the function of a boat is to convey its occupants safely and conveniently by water that we determine whether its parts are put together well or badly. So in order to know which arrangement of psychological components such as intellect and bodily desires is optimum we need first to know what our aims in life ought to be. One conception of those aims may indeed identify the optimum organization as that defined by the conventional virtues, but another, for example, that of Don Juan or Gauguin, may identify a quite different organization, such as one which affords the maximum play to certain kinds of self‐expression, as optimum.

The doctrine that virtue is knowledge is the key to understanding the so‐called thesis of the Unity of the Virtues, maintained by Socrates in Protagoras. In that dialogue Protagoras assumes a broadly traditional picture of the virtues as a set of attributes distinct from one another, as, for example, the different bodily senses are distinct. A properly functioning human being has to have them all in proper working order, but it is possible to have some while lacking others; most notably, it is possible to possess conspicuous courage while being grossly deficient in respect of the other virtues (329d–e). Socrates suggests that, on the contrary, the names of the individual virtues, courage, self‐control, etc., are all ‘names of one and the same thing’ (329c–d), and later in the dialogue makes it clear how that is to be understood by claiming (361b) that he has been ‘trying to show that all things, justice, self‐control, and courage, are knowledge’. The sense in which each of the virtues is knowledge is that, given the motivational theory sketched above, knowledge of what is best for the agent is necessary and sufficient to guarantee right conduct in whatever aspect of life that knowledge is applied to. We should not think of the individual virtues as different species of a generic knowledge; on that model piety is p. 67knowledge of religious matters and courage is knowledge to do with what is dangerous, and the two are as different as, for example, knowledge of arithmetic and knowledge of geometry, which are distinct species of mathematical knowledge, allowing the possibility that one might have one without the other. The Socratic picture is that there is a single integrated knowledge, knowledge of what is best for the agent, which is applied in various areas of life, and to which the different names are applied with reference to those different areas. Thus, courage is the virtue which reliably produces appropriate conduct in situations of danger, piety the virtue which reliably produces appropriate conduct in relation to the gods, etc., and the virtue in question is the same in every case, namely, the agent's grasp of his or her good.

It has been objected7 that this integrated picture is inconsistent with Socrates' acceptance in Laches and Meno that the individual virtues are parts of total virtue. In Laches, indeed, the proposed definition of courage as knowledge of what is fearful and not (194e–195a) is rejected on the ground that on that account courage would just be the knowledge of what is good and bad. But then courage would be identical with virtue as a whole, whereas ex hypothesi courage is not the whole, but a part of virtue (198a–199e). Given the aporetic nature of the dialogue, it is unclear whether at the time of writing Plato himself believed that the definition of courage was incompatible with the thesis that courage is a part of virtue, and, if so, whether he had a clear view on which should be abandoned. It is perfectly conceivable that he himself believed that they were not incompatible, and that the reader is being challenged to see that the rejection of the definition is not in fact required. What is clear is that the talk of parts of virtue can be given a straightforward interpretation which is compatible with the integrated picture. This is simply that total virtue extends over the whole of life, while ‘courage’, ‘piety’, etc. designate that virtue, not in respect of its total application, but in respect of its application to a restricted area. Similarly, coastal navigation and oceanic navigation are p. 68not two sciences, but a single science applied to different situations. Yet they can count as parts of navigation, in that competence in navigation requires mastery of both.8

The theory that virtue is knowledge is, as we have seen, flawed, in that one of its central propositions, that virtue is always in the agent's interest, is nowhere adequately supported in the Socratic dialogues. It also has a deeper flaw in that it is incoherent. The incoherence emerges when we ask ‘What is virtue knowledge of?’ The answer indicated by Meno and Protagoras is that virtue is knowledge of the agent's good, in that, given the standing motivation to achieve one's good, knowledge of what that good is will be necessary if one is to pursue it reliably, and sufficient to guarantee that the pursuit is successful. But that requires that the agent's good is something distinct from the knowledge which guarantees that one will achieve that good. ‘Virtue is knowledge of the agent's good’ is parallel to ‘Medicine is knowledge of health’. Given that parallel, the value of virtue, the knowledge which guarantees the achievement of the good, will be purely instrumental, as the value of medicine is, and derivative from the intrinsic value of what it guarantees, that is, success in life (eudaimonia). But Socrates, as we saw, regards virtue as intrinsically, not merely instrumentally, valuable, and explicitly treats it as parallel, not to medicine, but to health itself. Virtue is, then, not a means to some independently specifiable condition of life which we can identify as eudaimonia; rather, it is a constituent of it (indeed, one of the trickiest questions about Socratic ethics is whether Socrates recognizes any other constituents). So, far from its being the case that virtue is worth pursuing because it is a means to a fully worthwhile life (e.g. a life of happiness), the order of explanation is reversed, in that a life is a life worth living either solely or at least primarily in virtue of the fact that it is a life of virtue.

The incoherence of the theory thus consists in the fact that Socrates maintains both that virtue is knowledge of what the agent's good is and that it is that good itself, whereas those two theses are p. 69inconsistent with one another. It could, of course, be the case both that virtue is knowledge of what the agent's good is, and that the agent's good is knowledge, but in that case the knowledge which is the agent's good has to be a distinct item or body of knowledge from the knowledge of what the agent's good is. Otherwise we have the situation that the knowledge of what the agent's good is is the knowledge that the agent's good is the knowledge of what the agent's good is, and that that knowledge (i.e. the knowledge of what the agent's good is) is in turn the knowledge that the agent's good is the knowledge of what the agent's good is, and so on ad infinitum. So, if Socrates wishes to stick to the claim that virtue is knowledge he must either specify that knowledge as knowledge of something other than what the agent's good is, or he must give up the thesis that virtue is the agent's good.

Plato represents Socrates as grappling with this problem in Euthydemus. This dialogue presents a confrontation between two conceptions of philosophy, represented respectively by Socrates and by a pair of sophists, the brothers Euthydemus and Dionysodorus. The latter demonstrate their conception by putting on a dazzling display of the techniques of fallacious argument which enable them to ‘combat in argument and refute whatever anyone says, whether it is true or false’ (272a–b). For his part Socrates seeks to argue for the central role of wisdom in the achievement of eudaimonia. The first part of his argument (278e–281e) is in essence the same as that used in Meno 87d–89a to establish that virtue is knowledge; knowledge or wisdom (the terms are interchangeable) is the only unconditionally good thing, since all other goods, whether goods of fortune or desirable traits of character, are good for the agent only if they are properly used, and they are properly used only if they are directed by wisdom. Thus far Socrates reproduces the position of Meno, but in the second part of his argument (288d–292e) he goes beyond it. Here he points out that the previous argument has shown that the skill which secures the overall good of the agent is one which co‐ordinates the production and use of p. 70all subordinate goods, including the products of all other skills. It is thus a directive or governing skill, which is appropriately termed the political or kingly (basilikē) art. But now what is the goal of the kingly art? Not to provide goods such as wealth or freedom for people, for the previous argument has shown that those are good only on the condition that they are directed by wisdom. So the goal of the kingly art can only be to make people wise. But wise at what? Not wise (= skilled) at shoemaking or building, for the same reason, that those skills are good only if they are directed by the supreme skill. The goal of the kingly art can therefore be none other than to make people skilled in the kingly art itself. But, as Socrates admits (292d–e), that is completely uninformative, since we lack any conception of what the kingly art is.

Socrates leaves the puzzle unresolved, and it may well be that at that point Plato did not see his way out of the puzzle. What this dialogue does show is that Plato had become aware of the incoherence of the system of Socratic ethics whose two central tenets are that virtue is knowledge (sc. of human good) and that virtue is human good. If human good is to be identified with both knowledge and virtue, then that knowledge must have some object other than itself. Plato's eventual solution was to develop (in the Republic) a conception of human good as consisting in a state of the personality in which the non‐rational impulses are directed by the intellect informed by knowledge, not of human good, but of goodness itself, a universal principle of rationality. On this conception (i) human good is virtue, (ii) virtue is, not identical with, but directed by, knowledge, and (iii) the knowledge in question is knowledge of the universal good. It is highly plausible to see Euthydemus as indicating the transition from the Socratic position set out most explicitly in Meno to that developed Platonic position.

Protagoras may be seen as an exploration of another solution to this puzzle, since in that dialogue Socrates sets out an account of goodness p. 71whose central theses are: (i) virtue is knowledge of human good (as in Meno); (ii) human good is an overall pleasant life. The significance of this is independent of whether Socrates is represented as adopting that solution in his own person, or merely as proposing it as a theory which ordinary people and Protagoras ought to accept. Either way, it represents a way out of the impasse which blocks the original form of the Socratic theory, though not a way which Plato was himself to adopt. Having experimented with this theory, which retains the identity of virtue with knowledge while abandoning the identity of virtue with human good, he settled for the alternative just described, which maintains the latter identity while abandoning the former.

Socrates and the Sophists

The confrontation of Socrates with sophists is central to Plato's apologetic project. Socrates, as we have seen, had been tarred with the sophistic brush, and it was therefore central to the defence of his memory to show how wide the gap was between his activity and that of the sophists. Since Socrates represents in Plato's presentation the ideal philosopher, the confrontation can also be seen more abstractly, as a clash between genuine philosophy and its counterfeit.

Plato depicts Socrates in confrontation with sophists and their associates in the three longest and dramatically most complex dialogues of the group which we are considering: Gorgias, Protagoras, and Euthydemus. I shall consider those together with Republic 1, which may originally have been a separate dialogue; even if it was not, it certainly looks back to the aporetic and elenctic style of the earlier dialogues, while there are obvious similarities between the positions of Callicles in Gorgias and Thrasymachus in Republic 1. As well as these major dramatic dialogues, Socrates is presented in one‐to‐one discussion with a sophist in the two Hippias dialogues.

The Greek word sophistēs (formed from the adjective sophos ‘wise’ or p. 72‘learned’) originally meant ‘expert’ or ‘sage’; thus the famous Seven Sages were referred to as the ‘Seven Sophistai’. In the fifth century it came to be applied particularly to the new class of itinerant intellectuals, such as Protagoras and Hippias, whom we find depicted in the Socratic dialogues. We saw earlier that sophists were regarded in some quarters as dangerous subversives, overthrowing conventional religion and morality by a combination of naturalistic science and argumentative trickery. Plato presents a much more nuanced picture. There are indeed elements of subversion, in that both Callicles and Thrasymachus mount powerful attacks on conventional morality. As for argumentative trickery, Euthydemus and Dionysodorus are shameless in their deliberate bamboozling of opponents. But Plato is far from presenting sophists as a class as either moral subversives or argumentative charlatans, much less as both. In Protagoras the sophist represents his own teaching of the art of life not as critical of conventional social morality but as continuous with it, since he takes over where traditional education leaves off. He defends traditional morality, and in particular the central role which it assigns to the basic social virtues of justice and self‐control, by a story designed to show how it is a natural development, determined by the necessity of social co‐operation if humans are to survive in a hostile world. He argues sensibly and in some places effectively for his views. Interestingly, neither his claim to make the weaker argument the stronger nor his agnosticism on the existence and nature of the gods gets any mention in this portrayal. Prodicus, who also appears in Protagoras and is mentioned fairly often in other Platonic dialogues, is said to have given naturalistic accounts of the origin of religion and was accounted an atheist by some ancient writers, but this is nowhere mentioned by Plato, whose primary interest is in making fun of his penchant for nice verbal distinctions. Hippias is presented both in Protagoras and in the Hippias dialogues as a polymath, whose interests range from science and astronomy to history, literary criticism, and mnemonics. In Hippias Major he has little capacity for following an argument, and there is no suggestion in any of these dialogues of p. 73radical views on anything. Gorgias starts out by claiming that rhetoric, his field of expertise, is a value‐free discipline (455a), but is trapped by Socrates into acknowledging that a good orator must know what is just and unjust, and that if his pupils do not know this already they will learn it from him (460a). There is no indication of what his substantive views on justice and injustice may have been; specifically, there is no suggestion in the dialogue that Callicles has derived his immoralism from Gorgias. It would give a better fit with what is plainly meant to be Gorgias' real position if any influence that Gorgias may have had on Callicles were restricted to the rhetorical force which he manifests in such abundance in expressing his atrocious views. In Plato's eyes that influence was no less dangerous than positive indoctrination.

It is worth pointing out that Plato's presentation of the personalities of the sophists is as nuanced as his treatment of their doctrines. At least, they are not portrayed in a tone of uniform hostility. Thrasymachus, indeed, is a thoroughly nasty piece of work: arrogant, rude, and aggressive (he even tells Socrates to get his nurse to wipe his nose and stop his drivelling (343a)), and Hippias is a learned and conceited blockhead, but the others are treated more gently. The charlatanry of the brothers in Euthydemus is so transparent as to be almost endearing, while Prodicus is a figure of rather gentle fun. Protagoras, on the other hand, is a much more considerable figure; he is certainly pompous and complacent, and he does get ruffled when he loses the argument, but he quickly recovers his poise and concludes with a generous, if slightly patronizing, compliment to Socrates. More significantly, Plato presents him as someone to be taken seriously intellectually. The speech which sets out his defence of social morality and his role as an educator is a serious piece of work, and up to the concluding argument he is represented as holding his own in debate with Socrates. When we add to this the lengthy critique of his doctrines in Theaetetus (something which has no parallel in the case of any other sophist) it is clear that Plato took him very seriously indeed.

p. 74Plato's Socrates is not interested in the religious unorthodoxy of the sophists. (Later, in book 10 of the Laws, Plato argues strongly that atheism leads to immorality, and recommends institutional means of suppressing it – including the death penalty for those who persist in it – but that is a stance foreign to the Platonic Socrates.) He faces a serious challenge from one strand of sophistic moral thinking, represented by Thrasymachus, who is himself a sophist, and Callicles, who is an associate of Gorgias. The basis of those views, explicit in Callicles, implicit in Thrasymachus, is the dichotomy between what is natural and what is merely conventional. Both assume an egoistic view of human nature, maintaining that, in common with other animals, humans have a natural tendency to seek the maximum self‐gratification, from which they conclude that, for the individual, success in life (eudaimonia) consists in giving that tendency free play. Law and morality they see as conventional devices for restricting that natural tendency with a view to promoting the good of others; their effect is to force people to sacrifice their own eudaimonia in favour of that of others. But since everyone has more reason to favour their own eudaimonia over that of others, the rational course for everyone is to free themselves from the shackles of law and morality. (Callicles goes a step further in claiming that that is not merely rational but in reality right or just (phusei dikaion), since the individual who is strong enough to exploit others is thereby entitled to do so, and is wronged by laws or conventions which seek to prevent him.)

The moral theory sketched in the previous chapter provided a response to this challenge, though a weak one, since the crucial link between morality and the agent's good was not established. But in addition to this radical challenge to conventional morality, the sophistic tradition provided an argument in support of it, and thereby an answer to the challenge, in the form of the theory of the social origin of morality expounded by Protagoras in the dialogue (see above). This theory rejects the fundamental thesis of the radicals that nature and convention are opposed. On the contrary, convention, in the form of p. 75social morality, is itself a product of nature, since it naturally comes about when human beings are obliged to adapt (by forming communities) in order to survive. So far from its being the case that convention stultifies the development of human nature, it is only via convention that human nature is able to survive and flourish, in the sense of developing civilization.

To the extent that Protagoras upholds conventional morality, especially justice and self‐control, he is an ally against Callicles and Thrasymachus. For all that, Socrates finds his theory inadequate. He could have made the point, though he does not in fact, that Protagoras' account makes justice and self‐control only instrumentally instead of intrinsically desirable; their value lies in their necessity as prerequisites for the benefits of communal life, but what is necessary is that those virtues should be generally, rather than universally, cultivated. Hence someone who can get away with wrongdoing on a particular occasion without endangering the social fabric has no reason not to do so (the ‘free‐rider’ problem). That issue is addressed in book 2 of the Republic. In Protagoras Socrates' criticism is that, in assuming the separateness of the individual virtues (see above), Protagoras manifests an inadequate grasp of the nature of goodness. Hence his claim to expertise about goodness (in other words, to teach politikē technē (319a)) is fraudulent, and those, like Hippocrates, who flock to him in the expectation of acquiring goodness, are not merely wasting their time and money, but are risking the positive harm of acquiring a mistaken view of goodness and hence a mistaken conception of their proper goal in life (312b–314b).

Sophists, then, are dangerous, but not in the way that they are conceived in the popular caricature. They are a threat, not primarily because they peddle atheism or immorality (though some sophists did promote one or the other), but because they set themselves up as experts on the most important question, ‘How is one to live?’ without actually having the requisite knowledge. This is the recurrent theme of p. 76Socrates' confrontations with them. Protagoras claims to teach people how to acquire goodness, but proves to have no grasp of what it is. Euthydemus and Dionysodorus make precisely the same claim (275a), but all they actually have to teach is verbal trickery. (Protagoras is clearly represented as making his claim in good faith, but the same can hardly be said for the brothers. The point is immaterial; whether or not the sophist believes his claim, the important point is that it is unfounded.) Hippias claims universal expertise, including expertise on the nature of the fine or beautiful, an aspect of goodness, but his claim proves as hollow as those of the others. Socrates, by contrast, does not normally claim to have expertise. What he represents is the true conception of the task of philosophy, which is to search for genuine expertise in the art of life. What that expertise is is the possession of the true account of goodness, and hence the true account of our proper aim in life.

This conception of philosophy is emphasized in Gorgias via the contrast with rhetoric. The art of life (politikē) seeks the good, which requires knowledge of what the good is, whereas rhetoric aims merely at gratifying the desires of people who lack knowledge of whether the satisfaction of those desires is good or not. Hence the true expert in the art of life is the philosopher, represented by Socrates, who here, exceptionally, does claim expertise. If, instead of being guided by philosophy, people's lives are ruled by rhetoric, the result is the substitution of the pursuit of pleasure for that of the good, a situation which can lead to the moral chaos represented by Callicles, for whom the good is the indiscriminate pursuit of every pleasure. Gorgias, it seems, does not himself claim to teach goodness, unlike the sophists; the dialogue is then, unlike the others we have discussed, a critique not of an unfounded claim to expertise, but of the misguided practice (characteristic, in Plato's view, of Athenian democracy) of assigning to the technique of persuasion the role which properly belongs to philosophical enquiry, that of identifying fundamental values.