For all his help during the last four years, thank you very much to my supervisor, Chris Megone.

For various reasons, but very much in each case, thanks also to Claire Barraclough, Mike Beaney, Piers Benn, Céline Bergès, Christian Bergès, Lesley Brown, Anne-Marie Chaput, Rob Davies, John Divers, Fred and Marianne Gobé, Noreen Humble, Jennifer Jackson, Christine Jacobsen, Chris Janaway, Matthew Kieran, Jonathan Knowles, Warren Malone, Gideon Marcus, Cath Morgan, Seiriol Morgan, Mark Nelson, Jim Parry, Rebecca Roache, Bernhard Weiss, David Wiggins, Bill Wringe and Charlotte Wringe.

In the Crito, Socrates' refusal to escape from his death sentence is based on a certain account of what justice requires, distinct from the view that he owes unconditional obedience to the laws. This position is defended by Plato in later dialogues. The resulting account has famously been supposed to be one of the main targets of Nietzsche's critique of morality. I argue that on a proper understanding of the account of justice proposed in the Crito, and of Nietzsche's critique, it is not the case that Nietzsche rejects Socrates' position. In Part One I offer a novel interpretation of the Crito, and show how the assertions made in that dialogue are defended later by Plato. Socrates, I argue, refuses to escape because by doing so he would injure his fellow citizens, and thereby risk developing character traits which would eventually threaten to ruin his psychic health, but also, he would have to give up philosophy which is essential for producing and maintaining psychic health. In Part Two, I show that Nietzsche's critique is not targetted at Plato's account of justice thus understood, but that the positions of the two philosophers regarding the just life are in fact very close. In particular, I argue that the apparent resemblance between Nietzsche and Callicles vanishes when we look at their very different conceptions of power and the human good. I argue also that, contrary to traditional views of Nietzsche, he is not an anti-rationalist, and thus does not oppose Plato's view that reason plays an important role in the good life. Lastly, I compare Nietzsche' s concept of self-overcoming and Plato's psychic harmony, and show that the two resemble each other closely. I conclude that there is no substance to the claim that Nietzsche rejects Plato's account of justice, when this account is understood as an elaboration of the view put forward in the Crito.

My thesis is as follows. In the Crito, Socrates's refusal to escape from his death sentence is based on a certain account of what justice requires, distinct from the view that he owes unconditional obedience to the laws. This position is defended by Plato in later dialogues. The resulting account of justice has famously been supposed to be one of the main targets of Nietzsche's critique of morality. I argue that on a proper understanding of the account of justice proposed in the Crito and of Nietzsche's critique, it is not the case that Nietzsche rejects Socrates' position. In order to argue this I propose in the first part of my dissertation a novel interpretation of the assertions in the Crito which I suggest are developed in later dialogues. In a second part I show that Nietzsche's critique is not targetted at Plato's account of justice thus understood, but that the positions of the two philosophers regarding the just life are in fact very close.

Socrates' reasons for refusing to leave Athens and escape from his death sentence have been much debated, and the consensus seems to be that Socrates did not escape because he held the authoritarian belief that he owed unconditional obedience to the laws. In Chapter One I propose a different interpretation of the Crito on which my understanding of Plato's account of justice is based. Socrates, I argue, refuses to escape, not because he believes he owes unconditional obedience to the laws, but for the following, more complicated reason. Socrates believes that by escaping he would not only injure his fellow citizens, and thereby risk developing character traits which would eventually threaten to ruin his psychic health, but he would also be forced to give up the practice of philosophy, which he believes is necessary for producing and maintaining psychic health. The loss of psychic health, he claims, would render his life both unjust and not worth living.

In Chapters Two and Three I show how Socrates' position, which is merely asserted in the Crito, is defended in later dialogues. In Chapter Two I show how Plato develops and defends the views that (i) there is a certain healthy state of the soul which, like bodily health, can be explained in terms of functional analysis; and (ii) that the elenchos, central to the practice of philosophy according to Plato, is necessary and largely sufficient for achieving and maintaining psychic health. In Chapter Three I show how Plato defends the view that justice is psychic health in the Republic, and that the elenchos can produce psychic health understood as consisting of knowledge of the forms.

Nietzsche is often portrayed as rejecting everything Plato' Socrates stands for, and especially his belief that the good life is the just life. I explain briefly why I believe that this interpretation of Nietzsche's general attitude to Socrates is mistaken in the first part of my conclusion. In the last three chapters of this thesis, I show that it is not the case that Nietzsche rejects Socrates' beliefs about the just life as they are presented in the Crito, and as discussed in the first part of this thesis. In Chapter Four I investigate the claim that Nietzsche is restating Callicles's arguments in the Gorgias, to the effect that justice is a virtue suited for the weak, and that the virtue needed for the good life is in fact power. I argue that there is only a superficial similarity between Callicles' and Nietzsche's arguments, as Nietzsche's conception of power and the human good differ fundamentally from Callicles', and resemble more closely Plato's conception of self-mastery.

Chapters Five and Six investigate two other aspects of Nietzsche's supposed rejection of Plato's account of justice: Nietzsche's supposed anti-rationalism and rejection of asceticism. I argue that once again, in each case, the alleged critique is based on a fundamental misunderstanding of Nietzsche's views, in that he rejects neither the claim that reason has an important role to play in the good life, nor that some kind of control need to be exercised over one's appetites.

At the end of Part Two I will have shown that there is no substance to the claim that Nietzsche rejects Plato's account of justice, an account which is the elaboration of Socrates' position in the Crito. On the contrary, it will emerge that Nietzsche's and Plato's positions are surprisingly close.

In the passage which probably earns him the qualification of being 'free speaking', Callicles tells Socrates what he thinks of philosophy, and how he thinks Socrates will end up if he does not give up the practice for something more profitable.

Callicles's words summarise, albeit before its actual occurrence, Socrates' predicament as it is related in the Apology, the Crito and the Phaedo. Socrates is arrested because he is a philosopher. He is unjustly condemned to die, and he is unable to defend himself against his accusers.

Does Socrates then recall Callicles's advice? Does he realise that Callicles was right when he said that philosophy was really 'a craft which takes a man of good nature and makes him worse' by rendering him incapable of defending himself against injustice, and of preserving his own life? Does Socrates still believe that philosophy is a 'wise' craft, useful in the pursuit of justice and happiness?

Socrates no doubt recognises his situation as that once prophesised by the young orator. He knows he had been condemned unjustly, and he knows philosophy is the cause of his predicament. But recognising that Callicles was right would mean that he was ready to prevent the situation from continuing if it were in his power to do so. It is puzzling then that Socrates did not make a greater effort to convince the jury to let him go, did not offer to pay a suitable fine, and last, refused to escape from prison when his friends had arranged for him to do so.

By rejecting these alternatives to some extent, Socrates chose his predicament. He chose it believing that suffering an injustice was far less harmful to him than committing an injustice himself, or than giving up the practice of philosophy even though that practice was responsible for putting him in prison in the first place. Socrates, then, persisted in his disagreement with Callicles to the very end by placing justice and the practice of philosophy above any of the considerations Callicles took to determine the good life, including life itself.

The aim of the first part of my thesis is to build up an account of Plato's defence of justice as psychic health which explains Socrates' choice in the Crito, to clarify that concept of psychic health, and the role played by reason, and in particular philosophy in bringing about both justice and psychic health. In order to do this I place an unusual emphasis on the importance of the Crito in the development of Plato's moral and political philosophy. I claim that this dialogue offers an embryonic sketch of Plato's defence of justice as psychic health, even though it lacks the clarifications and arguments needed to make sense of this defence. I believe that this approach not only helps build a stronger case for Plato's arguments about justice, but also serves to compare these arguments to Nietzschean views - the object of the second part of my thesis - more fruitfully, by emphasising the importance for Plato of psychic health as opposed to adherence to any particular moral system. The interpretation I draw from the study of the Crito and which is backed up by a reading of the early and middle dialogues, highlights the central role of health and flourishing of the soul in Plato's moral philosophy. This emphasis on moral psychology makes Plato a virtue ethicist. My interpretation of Nietzsche's moral philosophy also leads to the conclusion that he is (some kind of) a virtue ethicist. This is not the most obvious way of understanding Plato if one does not read the dialogues as being part of the same project. Hence by choosing to begin my thesis with a study of the Crito, I effectively determine what kind of interpretation of Plato's defence I will put forward, and what I will have to say with respect to Nietzschean attacks on Plato's defence.

The Crito is an important dialogue, and especially relevant to my project, in that it introduces the two central tenets of Plato's defence, i.e. that justice is psychic health, and that philosophical examination is essential for the latter. The dialogue offers no clarification of these claims - it will be our task to seek such clarifications in the arguments of the later dialogues Gorgias and Republic in the next two chapters. Instead the Crito shows that Socrates (and probably Plato) did believe such principles to be true, and moreover was prepared at least to attempt to apply them to a practical situation. I will argue that Socrates based his decision to obey the law and to drink the hemlock on his beliefs about justice.

If my interpretation of the argument of the dialogue is correct, then it must exclude another line of interpretation which has been adopted by several commentators, amongst them Bostock, to some extent Vlastos, and in a modified version, Kraut. This reading holds that Socrates decides to obey the laws and refuses to escape because he believes it is just to obey the laws no matter what. This means that Socrates' argument is in fact identical to that presented by the personified Laws of Athens at the end of the dialogue, i.e. that Socrates agrees with all their premises and has nothing to add to their argument.

Part of my task in defending my interpretation of the dialogue and setting up Plato's defence of justice as psychic health will be to argue that the interpretation outlined above is mistaken. I will show that it is by recalling Socrates' attitude to the opinions of 'the many' - voiced in the Crito and elsewhere. Since Socrates believes 'the many' are not trustworthy guides when it comes to matters of justice, the laws of Athens (democratic, hence the laws of 'the many') cannot be infallibly just. It follows that, Socrates cannot believe that it is always just to obey the laws in question.

I will need to address the following challenge to my argument. The opinions of 'the many' play an important role in the elenchos. This is exemplified in several dialogues in which Socrates appeals to 'the many' as a whole, or to individuals whom he considers to be part of 'the many'. Hence Socrates cannot be dismissing altogether the opinions of 'the many', nor can he be dismissing their laws which are the expression of those opinions.

My answer to this challenge will focus on the actual use which Socrates puts the opinions of the many to in his elenctic examinations. I will show that these appeals are never decisive, and that they certainly do not entail that the many are suitable guides in matters or justice, this because they do not submit their beliefs to the elenchos on a regular basis. Thus their beliefs remain at best true beliefs and cannot be knowledge.

I will also need to address, briefly, the problem of expertise. Socrates claims that the judgments of the many are not reliable guides in moral matters, and that we must turn to experts instead. But no experts appear in the Crito, and there seems to be grounds for saying that Plato did not believe there could be any moral experts. I will point to a solution to these problems.

Having refuted this conflicting interpretation of the argument of the Crito, I will show, through a study of the argument, that my proposed interpretation is correct. I will argue that the following claims play a major, if not always obvious role in Socrates' decision to obey the laws. Socrates believes both that the health of his soul depends on his not committing injustice, and that his life will not be worth living if he stops practicing philosophy. These claims, together with some (but not all, and only those which Socrates can independently verify) of the Laws' claims, lead Socrates to the conclusion that he must not escape. These claims need clarifying as well as defending, in particular it is a mystery how Socrates knows (or even how he has true beliefs) that injuring others and breaking agreements will harm his soul, or that practicing philosophy is beneficial. I conclude by proposing a programme which will achieve this clarification for the next two chapters.

That Socrates in the Crito acts from a belief that he owes obedience to the laws no matter what conflicts with my thesis that Plato believes justice is some kind of psychic health promoted by the philosophical life. My thesis entails that the moral character of actions is dependent on their consequences for the agent's soul. Just actions are those actions which promote psychic health (48b). As he believes that life with an unhealthy soul is not worth living (48b), Socrates is unwilling to act in a manner that will harm his soul: 'Then in no circumstances one must do wrong' (49b). But because Socrates knows that the laws will sometimes require one to act unjustly, as they did in the case of Leon and the Thirty (Apology, 30d), he cannot believe that he should obey them no matter what.

Thus I interpret Plato's defence of justice as agent-based or at least agent-prior, which entails that actions are just, not because they accord with rules, but because they contribute to psychic health. Hence for Plato, adherence to rules - and in this case laws - and justice are not necessarily co-extensive. He would not believe that his being just depended every time on his obeying the laws of Athens; as indeed it did not in the case of Leon and the Thirty, and would not in the hypothetical case of the jury ordering him to give up philosophy. This potential gap between what is just and what the laws order means that on each occasion where Socrates doubted whether obeying the law was the just thing to do, he would have to ascertain that it was through independent means. (We yet have to establish what these means are). He cannot rely on the assumption that if an action is in accordance with the laws, then it is just.

I will show that Socrates in the Crito does not in fact agree that he owes the laws of Athens unconditional obedience, because these laws are democratic, i.e. the laws of the many. Unconditional obedience would entail obeying even when it required him to act unjustly, and would conflict with my thesis. It would also in some way violate the principle stated at 48a that we should not 'consider popular opinion in questions of what is right, honourable and good, or the opposite'. Showing that this authoritarian interpretation of the Crito fails through a misunderstanding of Plato's attitude to the beliefs of the many, will be my first step towards a vindication of my thesis.

The principle which I believe is significant when trying to understand Socrates' attitude to the laws of Athens is stated at the very beginning of his argument. Socrates explains to Crito that where justice is concerned,

This could amount to no more than a rejection of ad populum inferences. Whether or not what they believe is right, the fact that 'most people' believe a proposition to be true is not sufficient reason to infer that this proposition is true. So the arguments put forward by Crito for Socrates' escape which appeal to what 'most people' will think (44c) are bad arguments.

This interpretation takes us no further than the recognition that Crito, if he is to convince Socrates, has to appeal to valid arguments. But it is tempting to read a stronger statement in Socrates' claim that we should not 'consider popular opinion in questions of what is right, honourable and good, or the opposite' - a statement that is not merely a rejection of ad populum argument forms, but of the content of Crito's arguments.

Whereas some forms of argument ad populum are acceptable, those that are not (most of them) are not because we do not believe that 'most people' are in a position to form correct judgments in the field which determines the content of the argument. For instance 'everyone says the summer of 1976 was very hot' is sufficient to infer that the summer of 1976 was hot. This is because no expertise or special shrewdness is required to determine that a summer was particularly hot. On the other hand, Socrates says that we should turn to experts in right and wrong, rather than popular opinion (48a). Clearly this implies he does not believe that 'the many' are experts in right and wrong, or that just anyone is in a position to form correct moral judgments. (This presupposes that Socrates believes there are experts in right and wrong, and it means that we must explain why no experts appear in the Crito. I will address these issues briefly in §2.4.)

This interpretation, of course, needs supporting and justifying, and I will devote a part of this section to explaining why Socrates believes the many lack moral expertise (in the sense that they cannot reliably tell right from wrong), as well as providing more textual evidence that he holds such a belief. I will finish by answering an objection in §2.3. to the effect that Socrates does respect the opinion of the many as he uses it in the elenchos, and in §2.4., by addressing some of the issues relating to the problem of expertise, and in particular, who the expert appealed to by Socrates at 48a may be.

The passage says that the many are not in a position to form correct judgments about right and wrong. It follows that as Athens is a democracy, and its laws are the laws of the many, Socrates has reason not to accept the laws' advice blindly in deciding what it is just for him to do. Hence it follows from that passage that Socrates does not believe it would be just for him to obey the laws no matter what. He would have to ascertain that what they did order him to do was just. As they are the laws of the many, he cannot be certain that they are just without pondering the question himself or turning to an expert. This is why I take this passage to refute the challenge presented by Bostock and Vlastos to my thesis. I will show that Socrates says plenty elsewhere which agrees with the passage at 48a, and confirms the thought that it is because they are not trained dialectically that the many are not reliable.

The following passage from the Apology expresses Socrates' mistrust of the many's fitness to form correct judgments about justice:

The content of this passage is clearly that Socrates finds it impossible to do politics with the Athenians. They will commit injustices, and they will not let any one correct them or tell them otherwise. The Athenians are not fit to govern themselves in a just manner, and if they have any authority, they will prevent any one from imposing justice in their city. This implies further that, unless it goes hand in hand with training in dialectic, there can be no such thing as a just democracy: the untrained many, when they are in charge, will not listen to justice.

Vlastos, in "The Historical Socrates and Athenian Democracy" disagrees with this interpretation of the passage:

Vlastos' argument may be right as far as the distinction between the laws and their application by the citizens are concerned. Even then, one might say that good laws make good citizens - and Socrates does say, in the Euthydemus that the point of politics is to make citizens wiser - in which case it would be unfair to blame the people of Athens entirely for their political life.

Moreover, Vlastos' argument cannot be applied to the passage from the Apology quoted above. It cannot be applied for the simple reason that if Socrates blames the Athenians, he does not stop at that. His words are 'you or any other multitude'. The blame is spread a lot farther, and we have to ask whether it can really refer to the bad character of particular populations, and their disrespect for the law. If we separate people's ability to make good laws from their ability to respect them, and attribute the latter to bad character, then the claim that any people is incapable of respecting laws amounts to 'misodemia', the dislike of the many. Whereas such a dislike could possibly be taken to a particular population, say that of Athens, it seems far more unlikely that it could be applied in general, to any multitude.

This unlikelihood means that we may have to look differently at Socrates' statement. Could he not mean, rather, that he believes that the many, whichever city they are from, are not capable of governing themselves, in the sense that if left to themselves, they will fail to see what is just and what is not? This would imply, not that they are fundamentally incapable of being good citizens, but that they lack the necessary expertise for founding and maintaining a just constitution. Also it might suggest that what the many lack is a method for sorting out their true from their false beliefs about justice, and, as we saw earlier, this could be remedied by their being taught to inquire elenctically.

In support of this conclusion I would like to point out a passage in the Gorgias in which Socrates is pointing to the evils of rhetoric by drawing an analogy between cookery and medicine on the one hand, and rhetoric and politics on the other:

Children are not fit to decide which foods are best for them, and without proper supervision, they would fall prey to the flatteries of cooks telling them that the cakes and sweets they sell are really the best foods for their bodies. Between the cook's cakes and the doctor's medicines, they would choose what pleased their appetites most: the cakes. If the comparison holds, and Socrates seems to believe it does, grown citizens are no more capable when it comes to choosing between the true politician who prescribes justice, and the rhetorician, who prescribes whatever is in his interest, by flattering the citizens' not so noble ambitions.

Later in the same dialogue, Socrates returns to this analogy to answer Callicles' accusation that he would not be able to defend himself in court, if he came to be wrongly accused: "For I will be judged as a doctor might be judged by a jury of children with a cook as prosecutor". This reiterates the idea at Apology 31d-e that it is not possible to engage in real political debate with 'the many'. It also adds to it the thought that this is due to the fact that the Athenian jury is not qualified to form correct judgments about justice, just as children are not qualified to form correct judgments about medicine. These texts support the conclusion of this sub-section: Socrates does not believe that moral decisions should be influenced by popular opinion. This is because many lack the training necessary for reliability in making correct moral judgements.

To understand the role of popular opinion in the Socratic elenchos, one has first to understand the elenchos. This is by no means an easy task and I can only pretend to touch on it in a later chapter. For the present we will assume that the elenchos is either one of two things. It is (a) either a method for refuting all claims to knowledge, or (b) a method for finding out the truth. Both interpretations find support in the texts.

The mechanisms of the elenchos can be summarised thus. Socrates commonly starts off by asking a definition (p) of his interlocutor. The interlocutor is then led on to make other related assertions (q&r). Socrates examines these and finds that the latter assertions contradict the original proposal. If (a) above is the right description of the Socratic method then Socrates has simply shown that his interlocutors held a set of inconsistent beliefs: if q&r are true, then p must be false, and vice-versa. The interlocutor turns out to be ignorant when he thought he had some knowledge. In this case, although the elenchos cannot work without appealing to the interlocutor's opinion, in no way does it compel us, or Socrates, to suppose that these opinions are trustworthy. Not only has nothing been said about their truth or falsity, but it has been found that they were held inconsistently.

If we accept (b) as the correct description of the elenchos, then Socrates is able to show that a given interlocutor holds some false beliefs and some which are true. I will not discuss here the assumptions which must be necessary if Socrates is to be able to tell which beliefs are true and which are not. Suffice it to say that there must be such assumptions. The elenchos, in this case, can only work if the interlocutor has some true beliefs with which he is able to refute his false beliefs.

If we accept that this is how the elenchos works, then we accept that it cannot function unless Socrates is assuming that popular opinion is partly constituted by true beliefs. The question then is why one should shun popular opinion if it is the case that it must sometimes be true. Is it because it is not always true? This surely must be rejected: we do not seek advice only from those persons we know are infallible sources of knowledge - that cannot be expected from anyone. Whereas on many points 'the many' may be entirely right, what matters here is that they do not know, and have no method for finding out which of their beliefs are true and which are not. Until, that is, they are subjected to the elenchos, the method for sorting out true from false beliefs. As the interview of the slave boy in the Meno is meant to suggest, there is no reason in principle why the elenctic method of inquiry shouldn't be accessible to the many.

I have tried to show that the importance placed on the opinions of the many in the practice of the elenchos in no way entailed that these opinions were reliable guides to just actions. The many cannot discriminate between their true and false beliefs, not unless, that is, they are subjected to the elenchos on a regular basis.

It seems right to infer, then, that if Socrates' reluctance to be guided solely by 'the many' holds one way, when they tell him he should escape, a priori it must also hold the other way, when they tell him, via their laws, that he should die. Whenever considering what the just thing to do is, one should not let oneself be guided solely by the advice of 'the many'. This is what Socrates implies at 48a.

This conclusion needs qualifying in two ways. First, it is not the case that the laws of the many are necessarily unjust, but that they are not reliably just. 'The many' would not know if their laws happened to be unjust, as they would not know if their beliefs about justice were false. In order to be credited with counter-factual reliability, according to Socrates, 'the many' must practice the elenchos. This is confirmed by his rejection of Crito's appeal to 'the many' in favour of an elenctic examination of his situation at 48c-d.

Secondly, it is not qua democratic laws that the laws are not reliably just; my conclusion takes it for granted that they are the laws of people who do not submit themselves to the elenchos. If every one in a democracy practiced elenctic examinations, then presumably they would be more reliable as far as justice is concerned.

The argument that the citizens of Athens cannot make just laws because they lack political or moral expertise, if we are to attribute it to Socrates, requires an explanation of its premise; i.e. what exactly is moral expertise, and does Socrates believe that there are moral experts, or at least that there could be? It is not the purpose of this thesis to explore the problem of moral experts in any detail. However, for the sake of clarification, I propose to try and answer at least the question who the expert appealed to at 48a is supposed to be.

My answer, that Socrates himself, as an elenctic teacher, is the expert he appeals to in the Crito, will detach the idea of a moral expert from its obvious interpretation as someone who simply knows what is just and what is not, and is able to teach it, while on the other hand it will link it to the practice of the elenchos. I will try to show that what Socrates means by a moral expert is someone who is able to inquire methodically into what is just and what is not, and who can help others engage in the same inquiry. In these terms, I will argue, Socrates himself is a moral expert.

Socrates appeals to experts in a way which seems to show that he believes they exist in at least two dialogues: the Crito, 47c-48a, and the Laches, 184e-185e. But in neither of these passages is the expert called by name, nor seen to arrive on the scene. Bostock suggests that in the Crito, the personified Laws of Athens are the expert: they arrive after Socrates has called an expert, and when they leave, Socrates claims that he has heard a message from the gods (54d). Allen on the other hand, believes that the expert appealed to is none other than the 'truth itself'.

Both suggestions have in common that they make out the expert to be an abstract entity, rather than an actual person. In the first case, it is Socrates himself who gives birth to these entities, the Laws, through his elenctic reasoning. The Laws' speech is a product of his imagination, and insofar as they express the truth, they do so only as the result of Socrates' thought processes. 'Truth itself' on the other hand exists whether Socrates looks for it or not, and the expert would be more likely to stand for a metaphor of truth itself if it was not for Socrates' claim that the expert represents the truth: clearly the expert is a mediator and not truth itself.

Might such a mediator be an elenctic teacher, someone who shows others how to find out what is true and what is not in moral matters? For this is exactly what Socrates is doing every time he engages in a dialogue with someone, when he is, according to his own description "practising philosophy and exhorting you and elucidating the truth for everyone that I meet"(Apology 29d) Is Socrates the expert he invokes in the Crito? This again will not be easily conceded as the expert in question is spoken of quite clearly as someone other than Socrates or Crito.

In the Apology, Socrates, calling himself "the true champion of justice" (32a) defines his god-assigned role amongst the Athenians as follows: to make them understand that it is wrong that they should "give no attention or thought to truth, understanding, and the perfection of your soul" (29e), and "urging you to set your thoughts on goodness" (31b). One way to interpret these passages would be to note their focus on 'thinking' about virtue and say that the role of the teacher is not so much to teach the citizens how to become virtuous, i.e. how to get them to find out what is just and what is not, but to get them thinking about virtue.

My immediate concern at this point, in order to show that we can take the appeal to experts at 48a seriously, is to determine whether elenctic teachers can or could be what Socrates calls experts in right and wrong. If we find that they could, then we must depart from the traditional image of the Socratic experts, the one that we (and no doubt Socrates' interlocutors also) picture when Socrates appeals to them, i.e. someone who will decide one way or the other on a moral question. Instead, our expert will be no more than someone who facilitates the search for truth by teaching a method and helping in the practice of it.

Irwin in Plato's Moral Theory argues that elenctic teaching does not constitute moral expertise:

This assessment obviously depends on Irwin's understanding of the terms 'expert procedure'. Irwin's account of Socrates' definition of experts is as follows:

"The expert in a particular craft offers authoritative guidance, supported by a rational account".

Irwin gives no textual evidence for this definition. We can however recover the definition by putting together passages from several dialogues. That an expert offers authoritative guidance is evident from the Crito (47c-48a) and the Laches (184e-185e) (Irwin quotes these passages with reference to moral expertise). That an expert must offer a rational account can be seen from Socrates' claim in the Gorgias (464e-465a) that rhetoric is not a craft because it can give no rational account of itself; and obviously, if one is not even a craftsman, one cannot be an expert.

Whereas there is no question that Socrates believes the elenchos can give a rational account of itself (it is the elenchos which he contrasts to rhetoric in the Gorgias), we might question whether the elenchos is in any way authoritative. If authoritative means to have infallible knowledge, then Socrates is not authoritative. Even if we discount his claims that he knows nothing, we are left with the fact that he believes he must constantly examine himself as well as others (28e). Someone whose beliefs are always up for revision cannot be said to have infallible knowledge. But authoritative need not mean anything more than being dependable. Socrates' authority as a teacher does not lie in his ability to transmit knowledge about virtue, but to provide a reliable method for sorting out true from false beliefs about virtue, and hence, for becoming a virtuous person.

I suggest that the expert in right and wrong appealed to in the Crito is no more nor less than someone who teaches using the elenchos, i.e. Socrates himself. I find that this suggestion - it can be no more than a suggestion - is the more sensible one, as it is the only one which makes sense of Socrates' several appeals to experts who seem never to appear. It also confirms the idea discussed in §2.2. that what the citizens of Athens lack is 'elenchein', and that this is why their laws are unlikely to be just laws.

In this section I have argued that if Socrates decides to obey the laws' order, it is in spite of his belief that the Laws of Athens are not to be relied on where justice is concerned. He will obey a law only if it does not require him to commit an injustice. This belief, I argued, stems from Socrates' conviction that we should not take the advice of the many blindly - when these are not trained elenctically.

In §2 I ruled out the interpretation of Socrates' behaviour in the Crito as based on a principle of unconditional obedience to the laws. I will now defend an alternative interpretation by arguing that Socrates' decision to obey the laws and drink the hemlock is grounded in his acceptance of two principles introduced in that dialogue. The principles are (i) that unjust behaviour harms the soul and just behaviour benefits it, and (ii) that the practice of philosophy is a necessary element of the good life. I will show that these two principles occupy a salient place in Socrates' argument and thereby justify my decision to focus on them in Chapters II and III.

In order to make clear the role played by the two principles in question, I will analyse the dialogue and show that it contains not one but two arguments put forward by Socrates for the conclusion that on this occasion, he should not disobey the laws. Both arguments are distinct from the argument of the Personified Laws of Athens.

The first argument runs from 48a to 51e, and comprises part of the Laws' speech. The Laws, however, contribute only two of the ten premises, and the main part of the argument is contained in Socrates' dialogue with Crito. This argument depends on an understanding of justice in terms of psychic health - the first principle. The relation between unjust actions and harm to the soul and just actions and benefit to the soul, however, is assumed rather than argued or explained. Thus the argument remains largely mysterious.

The second argument is offered in the last part of the Laws' speech at 53c-54c, and must be read together with part of the first argument (48b-49a). It is independent from the Laws' main argument, and is attributed by Socrates to the Laws as an afterthought, as it were. This argument depends on the premise that if he must give up philosophy, Socrates' life will not be worth living. This, in turn, must be understood in relation to a tacit premise (which is made explicit in the Apology) that the non-philosophical life is not worth living. This second principle, like the linking of justice and psychic health, is neither explained nor defended. So again, we don't seem to be offered much of a reason why we should accept Socrates' argument.

It will have become apparent, in the course of this section that the principles which constitute the major premises of Socrates' two arguments for the conclusion that he must do as the Laws order and drink the hemlock, are not in any way defended further in the Crito. Accordingly, I will conclude this Chapter by outlining a plan for Chapters II and III, the aim of which will be to explain the relations between psychic health, philosophy and justice, and thereby dispel these mysteries.

The first and most important step of Socrates' first argument is expressed in the following passages at 47e-49a:

To say that life is not worth living with an unhealthy soul (48a) is to say that one cannot live well in those circumstances. As it is of first importance that one should live well, then it is equally important that one should have a healthy soul. In order to have a healthy soul, one must avoid wrong actions and perform right ones (47e). This entails that in order to live well one must live 'rightly or honourably' (48b), and that one must avoid doing wrong (49b).

Although this part of the argument is set out fairly clearly, it is probably the one which provokes the most questions. What is it for the soul to be healthy? Why is psychic health a desirable state, so desirable that one may not want to live without it? How can a state of the soul be linked to moral behaviour? There is no attempt on Socrates' part to answer any of these questions in this dialogue. In fact, the rest of the argument raises more questions of a related nature, and it will become clearer as we progress in the analysis of the argument, that these questions must be answered if we are to make sense of Plato's defence of justice.

The claim that one must never do an injury to someone else, is in itself problematic. What can Socrates mean by 'injury'? Surely he does not mean 'to inflict something unpleasant on someone else'. Socrates has been to war, and there he considered it his duty to kill as many enemies as he could. Following Socrates' belief that the only real injury one can sustain is injury to one's soul, it would seem that to injure others is to harm their souls. But is this what Socrates means? If so, it would follow that to steal, kill, etc. does not by itself constitute an injury. As long as one can perform those acts without harming anybody's soul, then one is not acting wrongly. We can rule out this interpretation of 'injury', both from the point of view of common sense, and because the rest of the text does not warrant it. Socrates believes that to undermine the laws' authority would constitute an injury to the state (50ab). In saying so he does not mention harm to any souls. However, the following weaker claim is plausible, i.e. that Socrates understands by injury much the same thing as every one else (physical harm, loss of property, harm to one's relatives and friends, etc.), but believes that the worst kind of injury is harm to one's soul. This follows from his claims at 48a that the soul is much more precious than the body, and that life is not worth living if the soul is ruined.

In an attempt to answer the first of these questions Kraut points out that there are two ways in which an agreement may be unjust. An agreement may be unjust at the outset, i.e. someone may be coerced into an agreement, or badly enlightened as to its implications. Or an agreement may be unjust in that it requires one to do something wrong. This last category is somehow dubious: although in some cases it is clear that an agreement demands that a party does something unjust (e.g. a contract killer's agreement to kill somebody in exchange for money), in many cases the consequences of an agreement are unknown at the time the agreement is made. How, in this case, is one supposed to know whether an agreement is just before entering into it? If one cannot, then it must be the case that agreements can become unjust after they have been entered into, i.e. even if they were just at the outset.

The following example from Book I of the Republic illustrates the problem just outlined. If, Socrates claims, one has borrowed weapons from a friend who has subsequently become mad, then it would not be just to return these weapons, even if one had promised to do so (331c). But how could one know that the friend from whom one had borrowed weapons would subsequently become mad?

Given an identical agreement, it will sometimes be right to abide by it, as it turns out to be in the Crito, and sometimes wrong, as it was in the case of Socrates' involvement with Leon and the Thirty. Socrates' agreement with the Laws is the same agreement on each occasion, i.e. that he will obey them, when the Laws ask him to co-operate against Leon, in the hypothetical situation when they ask to give up philosophy, and when they ask him to die. But only in one case does Socrates deem it just to abide by his agreement. What makes the difference is an evaluation of the justice or injustice inherent in the situation. The implications of the agreement turn out to be unjust in two of the situations refered to above, and it is Socrates' perception of the injustice inherent in them which dictates to him his behaviour. However, Socrates, as in the case of his injunction that we should not inflict injury, elaborates neither on the reasons for that 'rule', nor on its implications. It is by reading between the lines that we are able to propose the following: an agreement is a just agreement only if it does not require one to commit an injustice. In the case of Leon and the Thirty, Socrates' agreement with the City of Athens clearly did require him to commit an injustice. Hence it ceased to be, on that occasion, a just agreement.

In the third step of the argument, Socrates attempts to deduce a conclusion about his particular situation from the moral guidelines he has established. This is where the Personified Laws of Athens come in. They argue that by escaping, Socrates will be both injuring them (50a-b) and breaking a just agreement he entered with them (51d-e). Together with the conclusions of the first and the second step of the argument, this means that Socrates must not escape.

The Laws' speech, then, plays the following role: if confirms that escaping would constitute an injustice on Socrates' part according to the principles established earlier in the dialogue, i.e. that it is unjust to injure people and to break just agreements. So what the Laws' speech is supposed to establish is that by escaping, Socrates would either injure them, or break a just agreement, or both. The Laws do in fact establish this, as I will show. But they claim to establish more. According to them, they are offering Socrates three self-standing reasons why he should not escape : "because he disobeys us and we gave him birth, because he disobeys us and we nurtured him; because he agreed to obey us and neither obeys nor persuades us that we are doing something incorrect" (51e-52a). The Laws' claims re the wrongness of disobeying parents and nurturers plays no role in Socrates' argument. The claim that Socrates neither persuades nor obeys the Laws goes some way towards explaining why he would be breaking a just agreement by escaping, but is not complete. Socrates believes that it is only on condition that the Laws do not bid him to commit an injustice that he is bound to persuade or obey them. In what follows I show how on the one hand some of the Laws' speech completes Socrates' argument for the conclusion that he should not escape, and how on the other the Laws make claims which do not and could not feature in Socrates' argument.

At the very beginning of their speech, the Laws claim that Socrates, by escaping will invalidate the law which states that judgments, once rendered, must be authoritative, and that by doing this, he will attempt 'as far as he has the power' to destroy the Laws (50a-b). Socrates' escape would constitute an attempt to destroy the Laws, presumably because it would undermine their authority in the eyes of the Athenians, and open the door to further acts of disobedience. The Laws have grounds for worry, because if an influential citizen - such as Socrates - were to defy the system, it may well be perceived by others as an example to follow. And by taking part in the destruction of the Laws, Socrates would be injuring his fellow citizens, who are undoubtedly better off with a less than perfect legal system than in a state of chaos such as would be brought about by the disintegration of that system. After all, Socrates' escape would engender destruction, not reform.

Although Crito does not recognise it, it is clear that to attempt to destroy the Laws because they have attempted to destroy Socrates would be to render injury for injury. According to Socrates' principles, this alone would be grounds for not acting, for remaining in jail, and the Laws need go no further in their argument. We must presume that if they do, it is at least partly because Crito is not yet ready to accept Socrates' argument; he needs further persuasion.

At 50c the notion of an agreement between Socrates and the Laws is introduced. This, however, is not developed in the following part of the speech. Instead the Laws refer to Socrates' relationship to them, comparing it with that of a child to its parents. It is this part of their speech, the Laws' second claim, which may most rightly be called 'offensively authoritarian'. The Laws declare that obedience and respect is due to them as from a child to its parents, or worse, as from a slave to his master. From the point of view of Socrates' principles it would appear that this argument, appealing to Socrates' special relationship to his country, is somewhat redundant: if it is wrong to return injury for injury, it is no more wrong to do so when one has a particular relationship to the person one intends to injure. Or at least, even if there is a difference in degree between injuring one's father and injuring one's cousin, or a stranger, it nonetheless remains that injuring a stranger is an injustice, i.e. something that Socrates would commit on no account.

It is also in this passage that the Laws introduce the 'persuade or obey' principle: "you must do whatever your city and your country command, or else persuade them in accordance with universal justice" (51c). It is slightly odd that this point should be made here, amongst the arguments for comparing the citizen/city relationship to the child/parent one. A child on a properly authoritarian regime, that is in most traditional families, is certainly not given the alternative to persuade her parents. 'Obey' is the only option, on the grounds, not only that the parents know better, but also that it is this kind of authority which is required to sustain the authoritarian family unit. Of course, this objection does not apply, and the analogy does hold, in the case of an older child, say twenty, who is expected both to respect parents' decisions, and engage in rational dialogue with them when he or she disagrees. That this particular kind of child parent relationship is what we should be concerned with here is confirmed by what follows.

The function of the 'persuade or obey' principle becomes clear when we put it together with another of the Laws' statements, their third claim, at 51d-e. The Laws point out to Socrates that he was given the choice, when he became an adult and could observe for himself the legal system of the city, to remain and obey the laws as a citizen, or to emigrate. Together with the 'persuade' clause, this amounts to saying that the agreement between Socrates and the Laws was a fair one.

The agreement, they claim, was just on two accounts. First, because Socrates was not coerced into it, he was given the chance to leave Athens; secondly because the persuade clause meant that there was always an alternative to obeying, if for some reason he did not want to obey. But even though this limits the possibility of someone being bound by the agreement to commit an injustice, it in no ways guarantees that this will never happen: a citizen may fail to persuade the Laws that what they order is unjust. Nonetheless, according to Socrates' argument, the agreement is just enough to be binding on this particular occasion if it does not require Socrates to commit an injustice. As it only requires him to die, and not to harm his soul in any way, (which means he injures himself less than if he were to live but harm his soul by committing an injustice), nor injure anyone else, then the agreement must be binding.

Socrates, the Laws go on to argue, did enter into such an agreement with them by not leaving Athens when he had the chance. If he were to leave now, he would be breaking this agreement. This then, is the second way in which Socrates, from his own point of view, would be committing an injustice if he escaped: by breaking a just agreement. The first way, we saw, would be by attempting to destroy the Laws, an act which would amount to rendering injury for injury.

Let us recapitulate briefly the three steps of Socrates' argument for the conclusion that on this occasion he must obey the laws, and the questions which have emerged from the discussion. This is the argument:

P1: The most important thing is to live well 48b

P2: One cannot live well with an unhealthy soul 48a

C1/P3: The most important thing is to have a healthy soul (P1,P2)

P4: Wrong actions harm the soul 48a

C2/P5: One must avoid doing wrong in any circumstances 49b (P3, P4)

C3/P6: One must not inflict injury 49c (P5)

C4/P7: One must not break agreements when they are just 49e (P5)

(both would constitute an injustice)

C5/P8: If escaping means inflicting injury or breaking a just agreement, then Socrates must not escape 50a (P6,P7)

P9: Escaping injures the state 50a-b

P10: Escaping breaks just agreement with the state 51d-e

Socrates must not escape (P8,P9,P10).

It seems clear from this summary that when Socrates agrees with the Laws' conclusion that he should not escape, it is for different reasons from these the Laws themselves rely on, which are furnished by his moral perspective on action, not by the Laws' claims to absolute authority. He does not therefore, agree with the Laws' stronger implicit conclusion (as identified by Bostock) that it would always be wrong for any citizen to disobey any law. He agrees merely with their weaker conclusion that it would be wrong for him on this occasion to disobey the law that once judgments are rendered they are authoritative.

As to the moral perspective which dictates Socrates' behaviour, we are left very much in the dark. We know that the most important criterion in deciding on a course of action is whether it is right or wrong, and that this is the case because being right or wrong is linked very closely with the effect actions have on the soul. But we do not know what the desirable state of psychic health consists in, nor how it is linked to moral behaviour. We cannot understand Socrates' decision to drink the hemlock until we can answer these questions.

The second argument offered by Socrates in the Crito is important for the role played in it by Socrates' second principle, i.e. that the practice of philosophy is a necessary element of the good life. This argument is interpreted by Colson in "Crito 54a-c: to what does Socrates owe obedience?". Colson notices that in their speech, the Laws do not exclusively present Socrates with arguments on the nature of his duty to the Laws. In fact, he says referring to the last part of their speech which deals with what Socrates' situation would be if he chose to disobey, "I would even go so far as to say that the most important reasons have been saved for last". The passage Colson refers to, 53a-54c, starts with the following claim: "We invite you to consider what good you will do to yourself or your friends if you commit this breach of faith and stain your conscience".

The Laws then go on to enunciate what will become of Socrates if he escapes from his sentence. First, if he goes to a well-governed state, he will not be well-received, since he will have the reputation of a law-breaker. More importantly, with such a reputation, his discussions about justice will be ill-received. Secondly, if Socrates escapes to a not so well-governed city, then in an atmosphere of 'indiscipline and laxity' there will be no occasion for him to discuss philosophy, and no one to talk with. Thirdly and lastly, his friends and children will suffer if he escapes, since they will either have to answer to the law for aiding him to do so, or else they will have to go into exile with him. From these considerations, the Laws go on to say that

thereby showing that they believe Socrates will be the first to suffer from his decision. It is worth noting here that it does not follow from this that it would always be wrong for anyone to disobey: the conclusion is contingent on this particular situation. Had Socrates been condemned in a Barbarian country, and had he been welcome as an exile in Athens, it might have been right then for him to escape.

Colson links the Laws' comment at 53c "And if you do this [go to places where he cannot talk about virtue], will your life be worth living?" to Socrates' claim in the Apology at 38a that "the unexamined life is not worth living", and in the Crito at 48b that "it is not living, but living well which we ought to consider most important". Thus, the Laws are telling Socrates, that according to his own principles, he will be better off dead than in exile, even if he has been unjustly condemned - a claim that Socrates is too wise not to agree with.

Using the evidence put forward by Colson, it is possible to reconstruct Socrates' argument in the following way. The first premise (given by the Laws) states that if Socrates were to escape, he would have, for contingent reasons, to give up philosophy (53c). The second premise states that a life which is not at least partly devoted to philosophy is not worth living (53c, Apology 38a). This leads us to the first conclusion that by escaping Socrates will injure himself, i.e. render his life not worth living (54c). This conclusion, taken together with the further premise that it is most important to live well (48b), takes us to the conclusion that Socrates must not escape.

This second argument is strikingly different from the first one. It takes little heed of the effect of Socrates' behaviour on others, and does not mention the rightness or wrongness of his actions. In fact, it relies on a different principle from those invoked in the first argument, i.e. the principle that philosophy is necessary for the good life. Unfortunately, as was the case with the premises of the first argument, no reason is offered why we should believe that philosophy is as important as Socrates claims.

In the first part of this chapter, we saw that the reason why the many weren't experts in justice was because they did not practice philosophy, which would imply that there is a link between justice, the good life, and philosophy, just as there is a link between justice, the good life, and psychic health. This is all we have to go on, so far, to understand Socrates ' second argument. In later chapters, if we are to make sense of Socrates' claims in the Crito, we will need to investigate these links.

I argued that Socrates' beliefs about justice certainly did not include the authoritarian belief that he owes obedience to the laws no matter what, even when they condemn him to die unjustly. Therefore I had to look elsewhere for an answer, and as I did, two ideas transpired from a study of the arguments of the Crito. The first was that judgements about moral right and wrong must focus on what goes on in the agent's own soul. As just actions benefit the soul and unjust ones harm it, to know what is just amounts to knowing what makes our soul healthy.

The second idea to arise from the study of the arguments is related to the importance placed by Socrates on the practice of philosophy in relation to justice. We saw that it was at least partly because he would have to give up philosophy that he decided it would be unjust for him to escape. And it was because they had received no dialectic training that the many were deemed unfit to make reliably correct moral judgments.

We are now in a position to hazard a formulation of the reason behind Socrates' refusal to escape. If he escapes, he will be doing some things which are harmful to his soul i.e. unjust actions. In particular he will be committing two injustices, injuring the City of Athens by attempting to destroy its laws, and breaking a just agreement with the City - on Socrates' own definition of a just agreement, i.e. one which requires him to commit no injustice. He will also be giving up philosophy which would also render his life not worth living. Hence he must not escape. In order to make more sense of this, we must answer the following questions:

I will address questions 1 and 2 in the next chapter. Question 3, which requires a more involved and complex answer, will make the subject of my third chapter. When we are able to give an answer to these three questions, we will be in a position to say why Socrates decided to drink the hemlock, and we will have a more complete picture of Plato' s defence of justice to weigh against the critiques presented in Part Two.

I will argue the following:

(i) that there is a certain healthy state of the soul which is desirable, and which, like bodily health, can be explained in terms of a functional analysis;

(ii) that the elenchos, central to the practice of philosophy according to Plato, is necessary and largely sufficient for achieving and maintaining psychic health. (This argument is incomplete as I make a further point in Chapter Three).

In order to argue (i), I will turn to Plato's analysis of psychic health as harmony in a soul which is ruled by reason in the Republic. I will show that this analysis depends on a functional argument, i.e. the claim that a soul is healthy if and only if it fulfils its natural function.

I will address two objections to Plato's account. The first is that natural kinds, such as the human soul, cannot be said to have functions, as these belong specifically to designed objects. I will refute this objection by examining more closely what it is for something to have a function. Secondly, it can be objected that the concept of health serves only to evaluate bodies, and hence that it is misapplied by Plato. I will argue that in fact, we cannot evaluate bodies if we don't have a general concept of human flourishing, psychic health being a necessary part of this concept.

(ii) relates to the role played by the practice of philosophy and why it is beneficial to the extent that Plato had rather die than give it up. I will argue that the practice of philosophy, in particular of the elenchos, is held by Plato to contribute to psychic health. First, by examining the soul it ensures that it is and remains psychically healthy. As in the Republic Plato says that music, and gymnastics are also needed to create a harmonious soul, we can conclude that elenchos is largely, but not completely sufficient for psychic health. This raises the following worry: if life is only worth living when one's soul is harmonious, then it seems that not only should one prefer to die than give up philosophy, but one should also choose death rather than give up the practice of music and gymnastics. One way to resolve this is to suggest that the roles played by music and gymnastics in bringing about harmony, i.e. the training of the thumos, can be played by other practices, so that if one was for instance tone deaf one could take up some other artistic activity such as painting. Secondly, I will argue that the second way in which the elenchos contributes to psychic health is as follows: failure to submit to the elenchos leads to internal conflict and an unhealthy soul. This entails that the elenchos is necessary for psychic health.

Again there will be an objection to address. An initial objection questions the hypothesis that the elenchos examines souls, since the elenchos examines primarily propositions. The answer to this objection lies in discerning what kind of propositions the elenchos tends to concentrate on. Because it focuses on sincerely held beliefs, it is right to say, I will argue, that the elenchos in some ways examines souls.

Secondly, my defence will highlight the following problem to be taken up in Chapter Three. If, as it seems, the elenchos can only produce consistency, not truth, then there will be problems for the overall defence of justice, as psychic health will no longer be co-extensive with virtue. This worry arises because it looks as though the elenchos prevents internal conflict by bringing about consistency. I will only comment on this problem briefly here, and will address it fully in Chapter III.

By the end of this chapter, we will be in a position to understand why Socrates in the Crito will not act in a way which would harm his soul, and why this means he will not give up the practice of philosophy. We will not, however, be able to say why such behaviour as Socrates wants to avoid would in fact be unjust. This is because we'll have only elucidated part of Plato's defence. Plato says that because unjust actions harm the soul, it is in one's interest to avoid committing injustice. So far we have only focussed on the claim that it is in one's interest to avoid harming one's soul, and that this entails we ought to practice philosophy.

The full account, which we have not yet investigated, supposes that one cannot be psychically healthy and unjust. This supposition is highly contentious, however, since most of us firmly believe that some people are evil or immoral who are, in some important sense, perfectly sane (and one would be inclined to say this amounts to being psychically healthy), people to whom a psychiatrist would not hesitate to give a clean bill of health.

In Chapter Three, I will tackle this objection and answer the third of the questions I raised at the end of Chapter One by showing why, according to Plato, it is not possible to be psychically healthy, practice philosophy, and at the same time be unjust.

The concept of psychic health which we encountered at Crito 48a is recurrent throughout the early and middle dialogues (see Protagoras 313a-c, Gorgias 464a, 480a-b, Republic 444c-e), where it serves to build an account of virtue. In the Republic, for example, Plato writes that justice and injustice

In fact, not until the Republic does Plato elaborate at all on what psychic health actually consists in. In Book IV Plato tells us that for someone to be psychically healthy, he must have

This harmony is attained when each part fulfils its proper role, which means that the rational part of the soul, aided by the thumos, must rule over the entire soul 'being wise and exercising forethought in behalf' of it. (441d-e).

That this is what psychic health consists in is deduced by Plato from a consideration that health in general depends on the establishment in something of 'natural relations' amongst its parts. Just as a body becomes diseased if its parts behave 'contrary to nature', the soul will not be healthy if the rational part does not rule over the appetites and the thumos. Because of this we should understand the analysis of the soul as tripartite in Book IV as an attempt to describe the mechanics of psychic health in the same way that a treatise on anatomy can serve to explain what it is to have a healthy body. Only by understanding how the parts relate to each other can we hope to understand how the whole functions.

The constraint that a healthy soul must be ruled by its rational part in no way entails, as I will argue in Chapter Six, that reason should repress the appetites or the emotions. The necessary condition for psychic harmony is that these are 'neither starved nor indulged to repletion' (R. 572a). This means that the three parts of the soul must be organised in harmonious rule, and that the health of each part is for it to fulfil its own function.

A soul which is dominated by either the appetites or the thumos is then, according to Plato, diseased. In Books VIII and IX Plato explains the different ways in which a soul may be diseased. For example, the type Plato refers to as tyrannical, which applies to those who are ruled by their appetites, will inevitably 'always be needy and suffer from unfulfilled desire', 'be full of terrors and alarms', and be 'maddened by [...]desires and passions' (578a).

The question which concerns us here is whether Plato is in fact entitled to refer to psychic harmony as health. It is important that he should be able to identify some state of the soul as psychic health as it follows from that that this state is desirable (We can take it for granted here that health is always desirable: in Book II (357b-c) it is said to belong to the second class of goods, the 'fairest', which consists of things which are good both in themselves and for their consequences). As Plato wants to argue that psychic health is the same as virtue and that virtue is desirable, it follows that the identification of psychic health is a crucial step in his defence. In the next sub-section I will examine Plato's arguments for the identification of psychic harmony and health of the soul.

Such a claim is open to the following objection, that it is wrong to attribute a function to the human soul, as only artefacts can be said to have functions. If the objection stands, then Plato will not have grounds for arguing that psychic harmony such as he describes in Book IV is health of the soul. And if he cannot argue this, then it will be more difficult for him to sustain the claim that psychic harmony is desirable.

In what follows I will show that there is indeed a difficulty inherent to functional analyses regarding the assignment of functions to non-artificial kinds, such as bodies and souls, and I will argue that we can disperse that difficulty.

The functional argument is traditionally attributed to Aristotle. Although there are resources in Plato for a view of this sort, especially in his discussion of virtue as psychic health, there is no explicit argument such as can be found in Aristotle.

At the end of Republic I, Plato states that objects, ranging from horses, pruning knife, bodily parts, bodies, and souls all have functions (erga),and he defines these as whatever these things can perform better than something which was not designed for that purpose. Hence the function of a pruning knife is to trim vine branches. It does this better than an object which was not 'fashioned for that purpose'.

This part of the argument is unclear. What if a particular pen knife of mine was better for trimming vine branches than any pruning knife, however well made? Would that mean that the function of the pruning knife was not to trim vine branches? No, but we might be tempted to argue that pruning knives were not well designed - that some knife (like my pen knife) could be designed which would fulfil the required function better than pruning knives. So if pruning knives did not fulfil their function as well as they might, they would be abandoned and replaced. That is because they are supposed to perform their function better than anything else which is not designed for that purpose.

We expect an object such as a pruning knife to fulfil its function better than anything else which does not share that function. This is because pruning knives are designed to do just that - and a failure to do it better than something else represents a failure in the design. But what of functional objects which are not designed? How can we determine their function? And if we can't, can we say that they do have a function? Thus we observe that there will be problems to deal with in the attribution of a function to a natural kind, such as a human soul.

Everything which has a function, Plato goes on to say, must also have a corresponding virtue or excellence (arete), which enables it to fulfil its function. Thus, degrees in good performance can be measured according to the presence of the required virtue or excellence in the object observed. If the function of a knife is to cut, then one of its virtues will be sharpness. If the knife is not sharp enough, then it lacks the necessary virtue to perform its function. In other words, the good is determined by the thing's function.

The function of the soul, according to Plato, is to live well - to achieve and maintain a state of harmony which is necessary for the good life, a 'beautiful order' amongst the parts of the soul, whereby each part does its own work (fulfil its own function) and does not exceed its role. Thus, Plato argues, we have to be rational if we are to be happy or live well: just as a pruning knife must be sharp in order to cut well, a human being must be just in order to be happy.

The objection which concerns us here, namely that human life cannot have a function may strike one as obvious: an argument which applies to pruning knives may not easily be made to work for human life and happiness. This implies no derogatory view of pruning knives, no comment on the grandness and elusiveness of human life - for various reasons, human beings, or their lives, cannot be said to have functions. Human beings are natural kinds, and, unlike artificial kinds, they have not been created for a purpose, they are not the product of intentional design. But 'function' is a term which belongs with the idea of design and purpose.

In general it seems that something only has a function if it has been given one. A pruning knife, or a carpenter have functions which have been attributed to them. No such thing is true, on the other hand, of a human being. Hence, when Aristotle protests that if a carpenter has a function, then there is no reason why a man should not have one, he is mistaken. Artificial kinds, not natural ones have functions.

If this were the answer, then one might wonder at some of Plato's examples of things which have functions. Plato says an eye has a function, i.e. to see. An eye is not an artificial kind. It was not designed in the sense that a knife was. Hence, according to the above argument, it should not have a function. He also mentions that a horse has a function, although he does not specify what that function is.

Despite strong intuitions that only designed objects may have a function, it is a fact that we are ready to impose a functional analysis on bodily parts. We believe that a heart, for example, has a function, because we believe it must behave a certain way in order to be a successful member of its kind: it must pump blood through the body. We believe that there is such a function because we know how it can be realised. In fact, we know several ways in which it can be realised. A pace-maker can fulfil this function, as can a cardiomyoplasty operation, the linking of a back muscle to the heart and to a battery lodged in the stomach. It is because we can identify the function of the heart, by studying its contribution to the function of the whole organism, that we can decide when it is failing to perform, and that we can figure out how to redress this failure.

For each bodily organ we can identify what it is for it to be successful or not, and we can try to find ways in which the organ can be replaced if it fails. This means that we can identify a function for these organs - a function which determines success and failure, and a function which can be realised by substituting something else for a failing organ. In this case the objection that only artificial kinds have functions fails. There is no reason why a natural kind, including a human being as a whole, should not have a function.

Our imaginary objector might reply that the reason a heart has a function is that it is a part of a natural kind, and that it can be replaced by an artificial kind, e.g. a pace maker, without the entire body becoming an artificial kind. This is a tenuous objection as the fact that a heart can be replaced by a pace maker does not entail that it is not a natural kind itself.

Moreover, it makes no sense to say that parts of something have a function if the whole doesn't. In the case of a heart, we can only decide whether it is functioning well by referring to its contribution to the functioning of the body as a whole. We need to know whether it is better for the body to be lifeless, or for it to be alive and mobile for a period of, say, seventy years. Only then will we know what a good heart is supposed to be like. I will discuss this problem further in the next sub-section.

The fact that we are so ready to attribute functions to bodily organs means that we must revise our original assumption that bodies and souls cannot have functions because they are natural kinds, and that only artificial kinds, which are designed, can have functions. We suspected that it was wrong to attribute functions to human beings, be it their bodies or their souls. We then discovered that the source of that suspicion was the belief that only artificial kinds, because they are designed, can have functions. Now this belief has been shown to be false, we have no further reason to hold on to our suspicion, i.e. we can welcome the hypothesis (crucial for our thesis) that bodies and souls may have functions.

In order to dispel this prejudice entirely, we need to ask what it is for something to have a function. Having a function does not depend on having been designed, but on there obtaining two conditions:

1. That we can identify a set of lawlike principles which dictate its behaviour.

2. That we can identify a good member of the kind.

That the first condition must obtain is obvious as there would be no point in assigning a function to something which behaved erratically. A heart can be said to have a function because its behaviour is related in a lawlike manner to that of the other elements of the body. A carbuncle, on the other hand, does not have lawlike relations to the rest of the body (or only to a very small part of it) - if it does have a function at all, it is distinct from, and not constitutive of the function of the body as a whole.

The first condition, although it is necessary for anything to have a function is not sufficient. If all we knew about the heart was how it related to the rest of the body, we would not know enough to assign a function to it. We would not know, given the principles ruling its behaviour, whether it is supposed to sustain life in the body or not, and for how long, and under what conditions. We would not be able to distinguish between a healthy heart and a failing heart. This is why in assigning a function, we must decide what a healthy, or successful member of the kind is supposed to be like. We need to be able to say that a good heart will sustain life in the body for such or such a period of time.

It is clear that in the case of the body, at least, both conditions obtain. We are able to identify lawlike principles to describe the body's behaviour, and we have a fair idea of what a healthy body is supposed to be like. There should thus be no prejudice in favour of artificial kinds. And if it is part of Plato's analysis that he can make sense of the concept of a good soul, as well as identify the laws upon which it behaves, then a soul also may have a function. In that case the onus is on those who believe that natural kinds cannot have functions to show that it does not make sense to evaluate human lives.

There are thus no a priori reasons why a functional analysis of the the human soul, one which appeals to the concept of psychic health, cannot be supported. If then there is an objection to the postulation of a healthy state for the soul, it is not supported by a critique of the functional analysis of the soul. In the next sub-section, I will consider and reject a second objection against the application of the concept of health to the soul.

The next objection I wish to consider starts with one of the last points we made: the question whether we can evaluate human life. That we can is a necessary condition for assigning a function to the soul. Plato claims that souls can be healthy or unhealthy, just as the body can. But those who say that we cannot evaluate lives or souls will claim that the concept of health is a purely physical one, and that it can only serve to evaluate states of the body. I will argue, against this objection that we cannot make sense of bodily health if we don't also have a concept of the good life (or healthy soul).

Current debates on the philosophical problem of mental health question whether it is right to apply a medical analogy to the mental. 'Health' and its counterpart, 'illness' are perceived as medical terms, i.e. scientific terms, and it is inferred from this that they belong exclusively to the physical realm. The proposition that virtue is some form of psychic health, on the other hand, assumes that it can be enlightening to apply those terms to the psychological realm.

In what follows, I will review the argument against applying the concept of health to the mental, and show that it fails because of a misunderstanding of that concept. I will do so by showing that we cannot make sense of the concept of physical health unless we also have a concept of what it is for a human being to be healthy as a whole. For this purpose I will use an example of a medical model which relies on a functional analysis of health.

The central feature of the medical model I propose to focus on is the concept of homoeostasis. Claude Bernard, its first proponent (though it wasn't named until after his death by Walter Cannon in 1926) described it as follows:

Homoeostasis is the activity of an organism whereby it maintains a stable equilibrium in its internal environment. The mechanism of homeoestasis means that the organism is self-regulating through negative and positive feedback loops which inhibit or amplify the response of the organism to its external environment. For example, the cellular temperature of the organs has to be protected against changes in external temperature when, for instance, we have a hot bath. Homoeostasis at the cellular level helps our bodies fulfil this function.

Medical health seems to be on this interpretation a physical phenomenon related to the proper functioning of an organism. A healthy organism 'works' properly - it goes through the regular motions. If it's unhealthy, it stops working, and the malfunction is there for the physicians to examine.

The problem is, Bernard isn't particularly clear in the above quote about what the body's equilibrium in question is, how we can tell that the way it's working is the proper way for it to work. We can't, as we do in the case of machines, refer to a design which tells us that the human body is supposed to behave in certain ways until a certain age and then behave a different way, etc. How do we know that we're not 'supposed' to carry on growing all our lives, and maybe live forever? How do we know that reproduction for example isn't in fact a disease (as labour pains and over-population certainly seem to indicate it is)?

Bernard speaks of an equilibrium, and a 'social interdependence' of the organs, but this is not helpful here as the very same words would constitute an adequate description of an organism coping with disease. The heart will pump more blood in order to deal with clogged arteries (thus responding to their needs by modifying its behaviour) in order to maintain 'stability' thus realising the 'social interdependence' of the organs - but clogged arteries and high blood pressure do not make for a healthy body. Hence, stability or maintaining an 'equilibrium' cannot be the function of the healthy body, or at least, it needs to be qualified.

In order to understand why we can't get to the concept of physical health just by studying the behaviour of the organs, it will help to take a look at servo-mechanisms, i.e machines in which activity analogous to homoeostasis can be observed. A simple one would be the thermo-stat of a central heating system. The function of the thermo-stat is to control the activity of the boiler, and maintain the temperature as close as possible to the desired level of warmth. However, the control of the thermo-stat is limited in that it is not immediate, nor precise, and more importantly, won't regulate the 'desired level of warmth' according to seasons, which means that this level has to be imposed from outside.

The thermo-stat's function is determined not just by the existence of lawlike principles which dictate its behaviour but because we can identify a good state for the central heating system to be in, i.e. such that it keeps the room warm but not too warm. In the same way, unless we can say what a good state for a human being to be in would be, we will not be able to say whether a heart or a kidney is functioning properly. Criteria we will invoke to identify this state will usually comprise the ability to recover from illness, or to resist viruses, and germs, and more generally, the ability to perform the functions of a normal human life. These will include having a job, providing for oneself and one's family - probably raising a family - pursuing hobbies, physical, intellectual or artistic, etc. What this shows is that the function of health cannot be elucidated purely in physical terms, i.e. the optimum state that a healthy body must sustain has to be appraised in terms other than physical, in terms of the good life and of general human flourishing.

This explains why it is not possible to decide, purely on physiological grounds, what makes a particular state desirable. Imagine coming across an alien species which is totally different from any terrestrial species. How will you know whether the creature is in good health or not? You cannot know that without knowing how it is supposed to behave, how it prefers to behave. How do we know whether the state of its skin indicates normality or, say, a severe case of eczema (think ET)? We can't even decide by finding out whether it is in pain or not, as pain does not always indicate illness. Cancer is painless at early stages, and giving birth, which we do not consider a disease, is painful. Thus there is no purely physiological or anatomical basis for determining health and unhealth. And in order to determine whether a being is healthy or not we have to appeal to a conception of what the good life for it is.

In this section I have argued that the objection against using the concept of health in the mental domain was based on a misunderstanding of that concept. Although medical health is explained predominantly by using physical descriptions, it cannot be defined without invoking values, i.e. values which serve to decide what constitutes a desirable state for the body to be in. Medical health must, like psychic health, be defined in terms of the inner mechanisms which constitute the function of the parts, and of the function of the body as a whole, i.e. the end of human life. Hence, far from refuting the possibility that the concept of health can serve to evaluate souls, the analysis of the concept of medical health confirms that we need a concept of human flourishing, such as the one 'psychic health' partly constitutes.

We can therefore conclude this section by saying that there is such a thing as psychic health, that it is indeed determined by the function of the soul's parts in accordance with the function of the soul as a whole, and that, like bodily health, it is a desirable state to be in. We cannot conclude, however, that to be psychically healthy is the same as to be virtuous. In order to reach that conclusion, we need first to establish that the end of the soul as a whole is virtue, that its function is to be virtuous. Before we can do that, we cannot rule out certain alternatives, for example that the function of the soul is to be consistent, whether or not it is virtuous, and that one can be evil and psychically healthy at the same time provided one is consistent. The work needed to rule out these alternatives will be carried out in the next chapter.

A tempting answer to the above question is that philosophy is beneficial just for Socrates, because he is a philosopher by vocation. It would not matter for a musician if she had to give up philosophy, or indeed never take it up, but it would diminish the quality of her life (maybe even to the extent that it is not worth living any more) if she had to give up music. The assumption behind these intuitions is that any calling through which one flourishes becomes part of one's identity, and cannot be given up without influencing the quality of one's life.

Although the above position is defensible, and indeed, Socrates probably would have been reluctant to give up music on the jury's request had he been a musician, it does not apply here. For Socrates is not claiming that it is only he who must not give up philosophy, his claim is universal: "the unexamined life is not worth living' he says in the Apology, referring not just to himself, but to the whole of humankind.

We must thus explain why the practice of philosophy is bound to be beneficial for everyone, philosopher by profession, painter, or redundant factory worker. Following the other argument of the Crito, i.e. that life is not worth living if we are not psychically healthy, I wish to put forward this hypothesis: the practice of philosophy is beneficial, because it contributes to psychic health, and psychic health is beneficial. I will show that the claims put forward by Plato in the Gorgias, most of which are echoed in other dialogues, support this hypothesis.

The Gorgias, more openly that any other dialogue, focuses on the question which kind of life one ought to live (472c-d, 481c, 500c-d). It is also highly explicit in its answer. One should live a just life (527e), because only then can one live a happy life ("For I say that the fine and good man and woman is happy, and the unjust and base is wretched" 470e), and the best way to do this is to live the life of a philosopher. The choice presented by Socrates is between "doing what a real man does, speaking in the people's Assembly, practising rhetoric, conducting politics the way you [Callicles] conduct it now" or "the life spent in philosophy" (500c). And according to the final myth, it is the latter which enables one to be judged favourably in the after life (526c). The life spent philosophising, examining oneself and others (Apology 28e,38a) is a life lived in accordance with justice and virtue, and it is a happy life.

The defence of the philosophical way of life in the Gorgias either goes unnoticed - it is mentioned by neither Dodds nor Irwin- or when it is recognised, as in Monique Canto-Sperber's introduction to the dialogue, it is left unmotivated, and the link between the defence of philosophy and psychic health is not pointed out. Without claiming that this is the only or main theme of the dialogue, I wish to defend the view that the elenchos, as the main feature of the philosophical life, is held by Plato in the Gorgias to be a necessary condition for psychic health, and that as such it is beneficial and desirable.

My defence focuses on the analysis of two sets of of Platonic claims which can be summarized thus: