As might be expected, Socrates refuses to accept Critias’ claim that temperance is knowing oneself for the reason that he believes himself to be ignorant about the topic (165b5–c1), and expresses the wish to consider that definition further in order to decide whether or not it seems acceptable (165c1–2).Footnote ¹ At the very outset, then, Socrates distances himself from the notion of self-knowledge that Critias has in mind, whatever the latter may be. And he begins the investigation in typical fashion, by asking his interlocutor to clarify exactly what knowing oneself amounts to. To contribute to this task, Socrates will introduce one of the most familiar features of Plato’s so-called Socratic dialogues, namely a set of analogies between virtue and technê or the technai, art or the arts. In the debate that will follow, he will use ‘technê’ and ‘epistêmê’ interchangeablyFootnote ² to refer to all sorts of first-order branches of expertise, including medicine, architecture, mathematics, geometry, and weaving.Footnote ³ And assuming, as he often does in other dialogues, that these forms of expertise are relevantly analogous to virtue (presently, sôphrosynê), he will attempt to draw certain implications concerning a particular aspect of sôphrosynê, namely what it is of and what it may be good for. And he will suggest that these implications also concern the people who possess sôphrosynê and are temperate. Nonetheless, for the first and only time in Plato’s Socratic dialogues, Socrates’ reliance on the technê analogy will be seriously challenged. Critias will deny that temperance is art-like in respect of its object and function, and eventually will appear to prevail: Socrates will back down and will assist Critias to fully articulate what he takes to be unique about temperance and submit it to dialectical scrutiny. At present, our aim will be to lay out in detail this methodological debate between the two interlocutors, highlight what is involved in their respective stances, and indicate what is at stake. To begin, it seems apposite to say a few things about the nature of the analogy and its philosophical importance.

1

While the Socratic dialogues are interspersed with analogies between the virtues and the technai, Socrates rarely suggests an argument in defence of the contention that virtue is a sort of technê.Footnote ⁴ Rather, he usually appears to assume that the former is identical to or closely resembles the latter. And he often compares the possessor of virtue or of some particular virtue with an expert in the arts and sciences: a doctor, mathematician, geometer, architect, grammarian, and musical performer, but also a cobbler, weaver, carpenter, or some other lowly artisan. Typically, Socrates draws inferences about the virtuous person on the basis of features exhibited by first-order experts. For instance, he argues against Meletus that, as the horse-trainer is the only person who benefits horses while the many harm them, so he himself may be the only person who benefits the Athenians while the many corrupt them (Ap. 24e1–25c4). Or he counters Thrasymachus’ claim that the true ruler unfailingly operates in his own interest by arguing that, in fact, the true ruler acts like the experts in first-order arts: as doctors, sea-captains, horse-breeders, etc. direct their activities to the good of others, so the ruler, insofar as he is a ruler, seeks what is advantageous to his subjects rather than what is advantageous to himself (Rep. 340c–342e). There is vast disagreement about the nature and scope of such arguments and, generally, about what the craft model amounts to and what purpose it serves.Footnote ⁵ But almost everyone agrees on this point: regardless of whether or to what extent the character Socrates is committed to the craft model in the so-called early dialogues of Plato, there comes a time when he subjects it to scrutiny, demonstrates its weaknesses, and abandons it altogether.Footnote ⁶

Traditional readings frequently suggest the following outline. The early Socrates uses the technê model to elucidate the intellectualist thesis that virtue is a kind of knowledge and to lend support to a series of paradoxes related to that thesis: virtue or knowledge is sufficient for happiness, all error is due to ignorance, all desire is for the good, and weakness of the will is impossible. To achieve his goal, Socrates focuses on the intellectual and cognitive elements of technêFootnote ⁷ to match his rationalistic conception of virtue. Namely, he suggests that, like every genuine expertise, virtue should be supposed to consist in the expert mastery of a body of knowledge that is governed by rules, uses a particular set of methods and tools, has a distinctive function or work (ergon), and pursues in a systematic manner its own proprietary goal.Footnote ⁸ Importantly, like every other technê, virtue is just the sort of knowledge susceptible to giving a logos – a rational explanation of its own practices. And because of the latter feature, one might expect that virtue, like every other technê, is transmissible from one person to another and can be taught.

Two further features of the technê analogy are significant. First, in the same way that every first-order technê is set over a distinct domain and governs whatever falls within it, virtue too must be set over a distinct if greatly extended sphere and must be taken to govern everything belonging to that sphere.Footnote ⁹ In other words, just as medicine is of health, arithmetic of number, divination of foretelling the future, and carpentry of producing wooden artefacts, and these objects or subject-matters determine the function and goal of the corresponding forms of expertise and experts, so also virtue must be of something (tinos), i.e. it must govern a domain that determines the ergon of virtue and, consequently, of the virtuous agent. Second, in tandem with fifth-century attitudes towards the first-order arts and their products, Plato’s Socrates underscores the beneficial character of the technai and the difference that they make to the preservation and quality of human life.Footnote ¹⁰ And he repeatedly suggests that, likewise but much more so, insofar as virtue is an expertise directed towards a certain goal, its possession and achievement must be of the greatest benefit to us. Thus Plato’s Socrates develops, many believe, a conception of virtue as the crowning achievement of human rationality and the essential component of the good life: virtue as an expertise whose distinct domain is the realm of value, whose function or work consists in the fulfilment of our peculiarly human capacities, and whose goal is nothing less than happiness for both the individual and society. Indeed, it seems plausible to infer that Plato’s Socrates uses the technê analogy to pursue and reframe an aspiration initially expressed by the great sophists of the Classical era.Footnote ¹¹ That is, he argues, in different dialogues and from different perspectives, that the pursuit of virtue coincides with the exercise of the true political art and that those devoting their lives to it are the only people fit to rule the state.

To complete this fairly traditional scenario of the trajectory of the technê model in Plato, we must turn to the dialogues that developmentalists and unitarians treat, for different reasons, as transitional. These dialogues are taken to show Socrates challenging central features of the technê analogy and drastically undermining the idea that virtue is relevantly art-like. The main reason given is that Plato gradually comes to acknowledge that human beings are motivated by non-rational as well as rational forces and that, therefore, virtue cannot be merely a matter of knowledge but also depends on natural inclination, training, and habit. Thus, it has been claimed that the arguments for and against the teachability of virtue in the Protagoras and the Meno, and the refutation of the hypothesis of a kingly art in the Euthydemus, mark Plato’s decisive turn away from the Socratic craft analogy and towards his own substantive ethics.Footnote ¹² The same conclusion has been drawn, with even greater confidence,Footnote ¹³ with regard to the central argument of the Charmides. The reason lies in the passage that we shall now discuss: a rare instance of explicit and sustained criticism against a particular aspect of the analogy between the virtue of sôphrosynê and various branches of technical expertise. As mentioned, the almost unanimous consensus is that the criticisms exercised by Critias against Socrates’ use of the technê model are successfulFootnote ¹⁴ and reveal Plato’s readiness to shake off the spell of Socrates and open his own wings.Footnote ¹⁵

While we must bear in mind that hypothesis, we shan’t be able to fully assess its merits until we reach the end of the search. In this chapter, we shall take the first step towards that goal. We shall closely follow the moves that the two interlocutors make in debating the technê analogy and determine their respective dialectical positions at the end of this exchange.Footnote ¹⁶ But, first, let me briefly explain why I believe that it is important to set the record straight regarding the stance of the Charmides vis-à-vis the analogy between virtue and the arts. Historically, this question bears on one’s overall interpretation of the Charmides and its position in the Platonic corpus. If the Charmides is a transitional dialogue, it should be classified as such for the right reasons and not for the wrong ones. Exegetically, the widespread assumption that, in the passage to be discussed, Socrates abandons the technê analogy is largely responsible for the tendency of many scholars to interpret the elenchus occupying roughly the second half of the dialogue, and notably the Argument from Relatives, in a subversive manner: while the ostensible point of the latter is that there probably cannot be a science that, unlike all other sciences, is orientated solely towards itself, the examples that Socrates brings up in order to defend that contention do in fact undermine it. As I have indicated,Footnote ¹⁷ I reject that reading because it misconstrues Socrates’ own position, violates the principle of charity, and disregards the say-what-you-believe rule of the Socratic method. Therefore, in my view, there is strong motivation for us to reconsider whether it is true that the Charmides does imply or strongly suggest the rejection of the technê model of virtue.

Philosophically, the ethical intellectualism inherent in that model has, in fact, its own attractions. Not only does it highlight certain structural features of virtue as a sort of knowledge or understanding, but it can also have important implications for politics and society. However, the power and appeal of the technê model cannot be properly measured if we do not explore further the issue of what virtue as a technê might be really like. In the Charmides, Plato’s Socrates makes some moves in that direction. He intimates that the technê of virtue, if one may call it that, is not merely a matter of cognitive mastery, but also entails that the virtuous people will dedicate themselves to the single-minded, disinterested, and life-long pursuit of their goal in much the same way as that in which the best experts endeavour to fulfil their respective tasks.Footnote ¹⁸ If we consider virtue in such terms, we need to think about rationality and cognition, functions and norms, systematicity and method, success and failure, and the price-tags attached to each of them. Let us keep these reflections alive as we walk from this point onwards, together with Socrates and Critias, from one passage to another and from one argument to the next until we reach the end of the search.

2

Even though Socrates calls temperance an epistêmê,Footnote ¹⁹ the terms in which he argues belong unmistakably to the technê analogy. If temperance is knowing something, then it is an epistêmê. And if it is an epistêmê, like every other epistêmê or technê (kai tôn allôn technôn: 165d6), it will have a work or function (ergon) and a domainFootnote ²⁰ distinct and different from itself.Footnote ²¹ To put it a different way, assuming that temperance is a science and also that it is a relative, it follows that, like all the other sciences and arts, temperance too will be aliorelative: it will be of something – i.e., as Socrates seems to think, a correlative other than itself.Footnote ²² Socrates takes it that what a science is of constitutes the object or subject-matterFootnote ²³ of that science and determines its own domain and the benefit that it brings.Footnote ²⁴ As medicine is of health and its beneficial function consists in making us healthy, so temperance must be of something other than itself and its beneficial work must be determined by reference to that something. On these grounds, Socrates now asks Critias what the proprietary object of temperance might be. The question is far more difficult than it might initially appear. For, according to Critias’ definition, temperance is knowing oneself, and the relation between the epistêmê responsible for the knowing and the thing known is tricky. Grammatically and syntactically, the epistêmê under consideration appears to be aliorelative: the phrase ‘epistêmê heautou’ appears to imply that this epistêmê is of something distinct from itself, namely oneself. Philosophically, however, it would seem that the science doing the knowing and the object of knowledge occur in one and the same person: the temperate knower both is the subject of knowing himself/herself and constitutes the object of that capacity, or, on an alternative reading, it both engages in the activity of knowing himself/herself and constitutes the object of that activity.

Even so, Socrates contends that a distinction needs to be drawn between the science equivalent to temperance and what that science is of. An obvious move would be to claim that temperance is knowledge of the self. For although the self is a notoriously elusive item, it is arguably distinct from the epistêmê that knows it. Thus, in the terms of the technê analogy, as health is the correlative of medicine and buildings the correlative of the building art, so the self could be posited as the correlative of temperance.Footnote ²⁵ It is important to note that Socrates has high expectations regarding the work of temperance and the benefits that it yields. Whatever its ergon is, he says that it must be ‘worthy of the name’ (165e2) – an expectation that Critias probably shares. At present, however, neither interlocutor gives us information about the sorts of benefits he may have in mind.

Critias responds to Socrates as follows:

Dialectically, Critias’ move is squarely within the rules. For, as Aristotle would remark (Top VIII 157b34–6),Footnote ²⁶ in his role as questioner Socrates has every right to demand that Critias should accept the inductive generalisation that every art and science has a distinct object, unless Critias can bring some counterexample. And Critias does bring quite effective counterexamples: calculation and geometry are arts which do not have distinct products in the sense in which housebuilding and weaving do. Again, Critias is portrayed as having consummate dialectical skills that find no close parallel in any other character of Plato’s Socratic dialogues other than Socrates himself.Footnote ²⁷ Also, from an intuitive point of view, Critias’ criticism seems eminently plausible.Footnote ²⁸ Not all arts and sciences seem to be alike in respect of their function and outcome. The arts of building and weaving aim at the production of ontologically self-standing products, whereas the arts of mathematics and geometry do not have such an aim. Nonetheless, the latter do not for that reason qualify any less as arts or sciences. Critias suggests that temperance is more similar to mathematics and geometry than it is to building and weaving: like the former pair, but unlike the latter, it is not of something distinct and separate from itself, such as a cloak or a house. All the same, it qualifies as an expertise at least as much as the so-called productive arts do.

Up to this point, Critias’ criticism of the technê analogy has a narrow focus. On the one hand, he takes Socrates’ use of the technê analogy to suggest that temperance must produce some ontologically independent thing and retorts that not all arts and sciences aim at such things. On the other, he has given no indication as yet that he objects to what we may call the requirement of aliorelativity: the idea lying at the core of the technê analogy, that every art or science is related to a proprietary object distinct from the art or science itself. Nor does Critias yet challenge an assumption that Socrates builds into the technê analogy, namely that whatever benefit derives from the exercise of an art or science has to do with the aliorelative object of this latter. Because medicine has health as its own object, it is beneficial in just that regard. And because the art of weaving aims to produce cloth, it is profitable in exactly that respect. In sum, Critias’ initial moves do not affect the art model at its core. Consider Socrates’ answer:

Socrates is not concerned with the distinction that Critias introduced between productive and non-productive arts, for it is irrelevant to his point. Clearly, he intends the contention that every form of expertise has an object or governs a domain distinct from itself to be applicable to arts or sciences as different as medicine, housebuilding, weaving, arithmetic, measurement, dancing, boxing, and lyre-playing.Footnote ²⁹ And his commitment to the technê model of virtue inclines him to infer that temperance too, if it is a kind of knowledge, must have, likewise, an object other than itself and extend over the domain determined by that object. Having thus clarified his meaning, Socrates puts his question to Critias again in clearer terms: assuming, on the grounds of the analogy with the first-order arts, that temperance is of something other than itself, what is this? It does not make any difference whether the latter is an ontologically independent product or a distinct subject-matter. What Socrates is asking Critias to do is identify what temperance is a science of in terms that do not comprise reference to that science itself. The following reply by Critias provides the main textual basis for the virtually unanimous agreement that, here, Plato parts company with Socrates by denouncing the flaws of the technê model and by suggesting that it should be completely abandoned.

At first glance, Critias appears to be on the right track when he contends that Socrates’ methodological procedure is at odds with the purpose of the search. His argument is this: since they want to examine what temperance is, they should focus on what is distinctive about it, not on what it shares in common with other sciences; but the technê model relies on commonalities, not differences between virtue and the first-order arts; hence, it impedes rather than advances the investigation underway. Worse, Critias appears to think that Socrates’ misguided application of the technê model is deliberate on his part (166c3–4) and accuses him of intellectual dishonesty: even though Socrates knows that the art model is misleading in this context, he is using it merely in order to win the debate (166c4–6). Critias therefore steers the argument onto what he believes to be the right track, by specifying what he takes to be distinct about temperance in relation to all the other sciences or arts: ‘while all the others are sciences of something other than themselves and not of themselves, this one alone is the science both of all the other sciences and of itself (epistêmê autê heautês)’ (166c1–3).

In the important passage quoted immediately below, Socrates rejects summarily the accusation of contentiousness and explains why he wishes to pursue the investigation. He throws new light on the nature of the elenchus and on his conception of his own self-knowledge.

The first thing to note is Socrates’ implicit admission that he sometimes thinks he knows something. But he also realises that he has to check again, and he wishes to continue the enquiry precisely because he fears that he might suppose that he knows (eidenai: 166d2) something that he doesn’t know. Furthermore, he makes the significant remark that he wishes to conduct an investigation primarily for his own sake, but also ‘perhaps’ (isôs: 166d4) for the sake of his friends. This confirms that, although Socratic cross-examinations are typically ad hominem and proceed on premises conceded by the interlocutor, they somehow benefit the questioner as well as the answerer: Socrates engages in them primarily in order to gain self-knowledge and only secondarily in order to help his interlocutor scrutinise his own beliefs. Presumably, the converse holds true of the interlocutor, if the latter is a right-thinking person. In this sense, the Socratic elenchus is genuinely a joint search: each of the two participants has something important to gain, even though, formally speaking, the contention under scrutiny as well as the premises of the argument belong to the answerer and not to Socrates himself. Yet another comment by Socrates is revelatory about his own view of the goal of the elenchus, and may tell us something about his relation to Critias as well: he asks Critias, rhetorically, whether he doesn’t share with him the conviction that ‘the discovery of the truth about everything there is is a common good for almost all men’ (166d4–6). His tone seems to me to suggest that he is merely reminding Critias of something that has been commonplace in their conversations, much as in the Crito he reminds Crito of their ‘serious discussions’ regarding the principle of justice and the rejection of retaliation (Crit. 49a–b). Socrates appears confident that Critias will readily acknowledge the value of discovering the truth about each of the things there areFootnote ³⁰ and, consequently, will withdraw his accusations and agree to continue the conversation. This is exactly what happens (166d7).

On the other hand, Critias appears to have gained the upper hand in the debate. For, as it seems at present, he has successfully met the challenge issued by Socrates that, if temperance is an epistêmê, then, like every other science or art, it must be of something other than itself. Critias contends that, on the contrary, temperance differs from the other sciences in just this respect: while these latter are of something other than themselves and not of themselves, temperance alone is of itself and the other sciences and not of any other object. As we shall see later, this amounts to the claim that temperance alone is a science of science (and of its privation) and of nothing else, with the consequence that temperance alone governs the other sciences, whereas each of them governs only its own specific domain. In sum, Critias’ position implies that the technê model for virtue is misleading: there is really no similarity between temperance and the first-order sciences or arts in respect of their corresponding correlative objects. For his own part, Socrates chooses to put an end to that dispute by conceding his interlocutor’s claim and inviting him to attend to its investigation (166e1–2). At this early point, we simply do not have enough information to decide whether Socrates makes this move in earnest, tacitly acknowledging that the technê model is flawed and that Critias is right to insist that temperance alone, unlike every other science or art, is directed towards science itself and no other object distinct from itself. I urge that we keep an open mind about this matter and monitor it in the chapters that follow.

Nonetheless, it is worth mentioning at the outset that the Republic arguably offers grounds for being cautious about accepting too readily the traditional view according to which, in the Charmides, Socrates pushes the craft model aside and, in particular, abandons the assumption that virtue has an object or governs a domain distinct from itself. In order to undermine Thrasymachus’ claims that justice is the advantage of the stronger (Rep. 340c) and, moreover, that the stronger or the ruler, insofar as he is a ruler, does not err about his own advantage (340d–341a), Socrates argues that in fact every technê (342c4–6) or epistêmê (342c11) is orientated towards the good of the object that it governs and not towards its own good. The doctor’s art entails that he governs and seeks the good of his patients and not his own (341c), the art of seamanship entails that the captain rules over and looks after the good of his sailors and not his own (341c–d), and the same holds for all other arts and sciences. No art or science needs the help of another art in order to determine the good of its own proprietary object and no art or science is orientated towards itself (342a–b). Medicine does not judge its own interest but the interest of the human body (342c), the equestrian art does not seek its own interest but the interest of the horses (342c), and, generally, no art aims at its own interest (because it is complete and does not need to) but rather at the interest of what it is an art of (342c). And since each science (epistêmê: 342c11) rules over its own object and is stronger in that respect, it follows that the ruler, who possesses the art of ruling and is stronger, looks after the interest of the objects of his art, i.e. his subjects, who are weaker, and not after his own interest (342c).

Notwithstanding the considerable differences between the two dialogues, it is striking that in the Republic as well as in the Charmides Socrates argues for a very similar view: all the sciences or arts have as their object something distinct from themselves, not (or perhaps not mainly) themselves.Footnote ³¹ Besides, as in the former dialogue, so in the latter (Rep. 346a), Socrates argues that every art or science is individuated precisely by reference to its own distinct object. Every time we judge that an art is different from others, we make this judgement in virtue of the fact that it has a different power (dynamis). And this power is typically related to the distinct object, work, and benefit peculiar to the art in question and not to any other art (346a).Footnote ³² To summarise this point, I suggest that, in the first book of the Republic (as well as later in that work) Plato’s Socrates argues for a cluster of views about the technai, including the technê of ruling, which are closely similar to the position that he initially defends vis-à-vis Critias (Charm. 165c4–166b6) but eventually appears to give up. In both cases the central idea is that every technê,Footnote ³³ including temperance and ruling, is typically or exclusivelyFootnote ³⁴ related to and directed towards a proprietary object distinct from the art or science itself. This parallel should give us pause regarding the concession that Socrates makes to Critias about temperance, i.e. the concession that the latter is a science directed only towards science (i.e. towards itself and the other sciences) and no other object. We should not assume, without further examination, that the concession in question represents Socrates’ better judgement rather than a strategic move on the chessboard.

Moving on, I shall briefly address the vexed issue of the transition effected by Critias from the fairly innocuous claims that temperance is knowing oneself (gignôskein heauton: 165b4) and that it is knowledge or science of oneself (epistêmê heautou: 165e1) to the peculiar contention that temperance alone is a science of itself (epistêmê autê heautês: 166c3), as well as of the other sciences. There is extensive literature on this topic and I shall therefore restrict my comments to the origins of the notion of a ‘science of science’, the hypothesis that it may reflect a view held by the historical Critias, and some speculations as to how it may be related to Socrates’ conception of virtue.

T. G. Tuckey’s useful survey of nineteenth- and early twentieth-century scholarshipFootnote ³⁵ gives a sense of the range and quality of interpretations that have been on offer. For example, according to Grube, Critias posits that temperance is knowledge of the self and problematises the question of how the knowing subject can be the object of his/her own knowledge. Hence, the question of whether a knowledge of itself is possible amounts to the query whether self-knowledge is possible, and Critias’ inability to establish the former entails his inability to defend the latter. In a similar vein, von Arnim maintains that the transition under discussion proves that Critias is entirely confused. Worse, Bonitz believes that this transition marks the beginning of a long digression, since, in his view, the investigation concerning the ‘science of science’ has no relation to the main goal of the dialogue, which is to define sôphrosynê. On the other hand, Schirlitz argues that the ‘knowledge of oneself’ is closely connected with the ‘knowledge of knowledge’ so that the former notion entails the latter. Likewise, Susemihl suggests that the ‘knowledge of itself and everything else that is knowledge’ constitutes an explanation of what it is to know oneself. In his view, ‘knowledge of knowledge’ is equivalent to ‘knowledge of oneself’, because the true self just is knowledge. In sharp contrast, Pohlenz argues that the transition in question is a fallacy and Plato flags that fact. In reality, according to Pohlenz, there are no conceptual links between the ‘knowledge of oneself’ and the ‘knowledge or science of itself and the other sciences’. The latter is merely a theory that Plato wishes to shoot down, whereas the former conception remains intact. On an entirely different wavelength, Taylor maintains that Critias’ identification of self-knowledge with the ‘science of itself and the other sciences’ is an effort to turn moral psychology into epistemology.

Such views have been developed by recent interpretations as well. These dwell also on the question of where the notion of ‘a science of itself and all the other sciences’ may derive from, and the answer differs according to each author’s negative or positive view about Plato’s portrayal of Critias. For instance, according to one approach, the ‘science of itself and the other sciences’ has a sophistic ringFootnote ³⁶ and corroborates the picture of Critias as a representative of the new learning taught by the sophists.Footnote ³⁷ According to another, the notion of a ‘science of itself and the other sciences’ derives from the intellectualism detected in the extant remains of Critias’ writings and alluded to in the Charmides as well.Footnote ³⁸ A group of interpretations propose that Critias’ final definition of temperance is a competent response to the ‘what is X?’ question, but also conveys Critias’ aristocratic values and attachments – especially, Critias’ deep-seated belief that temperance essentially consists in doing noble deeds including, first and foremost, deeds related to the successful governance of the city.Footnote ³⁹ Each of these positions has merits, but none of them is founded on firm evidence. Whether or not Plato borrowed the expression ‘science of science’ or ‘science of itself and the other sciences’ from somewhere or simply invented it, it is a successful choice. Dramatically, it corroborates Socrates’ insinuation that Critias is a sophos, wise man or sophist (161b8–c1). Philosophically, there can be no doubt that the elaboration of the notion of a ‘science of science’ is Plato’s own. As has been shown, however, there is no agreement about the philosophical content and implication of the latter.Footnote ⁴⁰ As a preamble to the argument that will soon follow, recall that, on one sort of view, Critias misunderstood Socrates’ contention that temperance must be an epistêmê of something, i.e. it must be a rule-governed form of expert knowledge, and instead took epistêmê to be equivalent to gignôskein: knowing in a non-expert manner or knowing how.Footnote ⁴¹ Hence, ‘the science of itself and the other sciences’ is intended to coincide with Socratic self-knowledge, and the elenchus targets either the ‘scientific’ body of knowledge identical to temperance,Footnote ⁴² or a ‘knowing how the mind knows itself’.Footnote ⁴³ On the other hand, many other readings argue, in vastly divergent ways, that the ‘science of itself and the other sciences’ is different from or even incompatible with Socratic self-knowledge.Footnote ⁴⁴ It should be clear by now that my own interpretation lies closer to this latter camp, since, first, I distinguish the Socratic conception of self-knowledge from the counterpart developed by Critias and, second, I take it that, from this point of the debate onwards, the sole direct target of the elenchus is Critias’ ‘science of itself and the other sciences’. If the elenchus will also raise problems for Socratic self-knowledge, it will do so in some oblique if significant way.

Ending this chapter, I wish to stress again that what prompted Critias to define sôphrosynê in the way he did was his rejection of an assumption central to the technê analogy,Footnote ⁴⁵ namely that every epistêmê or technê, including temperance, is relative to an object other than itself and governs a distinct domain determined by that object. Only because he refused to accept that temperance is thus analogous to the other sciences was he able to come up with the definition of the virtue as ‘the only epistêmê that is both of the other epistêmai and of itself’ (166c2–3). Socrates, on the other hand, initially appeared entirely committed to that assumption and defended it, even though he eventually backed down. We cannot know for certain whether he did this for the sake of the argument or because he was persuaded by his interlocutor’s criticisms. But we may suspect that he would not abandon his own position so easily. If his retreat is merely strategic, he can’t be expected to find congenial the idea of temperance as a unique sort of science solely directed towards science (i.e. itself or any science). Be this as it may, from the moment that he has conceded to Critias his point, he will act as a well-intended, constructive, and superbly skilful debater. At the outset, however, he will take the liberty of articulating Critias’ conception of temperance and of the temperate person in his own strikingly Socratic terms.