1.

The Socrates of the Platonic dialogues is represented as doing many things: he takes a walk (Phaedr. 227a ff.), goes to see a festival (Rep. 327a), defends himself in court (Ap.), and wakes up from a deep sleep (Crit. 43b). Of the many things that Socrates is represented as doing, most involve discourse: he fabricates etymologies (Crat.), constructs similes (e.g., Rep. 506e-517c), makes jokes (e.g., Phaed. 117b), and tells stories of the afterlife (e.g., Gorg. 523a-524a and Rep. 614b-621b). But far and away the larger part of what Socrates does in speech is ask questions about virtue; in this regard, the elenchus is indisputably the “most striking aspect” of Socrates’ behaviour as depicted in Plato’s dialogues. ^{Footnote 1}

Given the amount of time Socrates spends on elenchus, he must deem it to be a very important activity. So the question arises as to what he hopes to achieve by it. There are in the scholarship two basic answers to this question. According to the first, Socrates’ method is constructive—it aims to establish definite ethical or philosophical propositions; according to the second, Socrates’ elenchus serves mainly to refute an interlocutor’s claim to knowledge of the topic under discussion. ^{Footnote 2} In this paper, I want to sketch out an alternative interpretation of Socrates’ method. On the view I shall propose, Socrates’ formal goal in elenchus is neither to establish a positive doctrine nor to refute his interlocutor, but to know his interlocutor’s knowledge of virtue. The Socratic method can then be described as a kind of ‘epistemic exegesis,’ which aims to draw out and give discursive shape to knowledge of virtue that it assumes that the interlocutor already possesses.

The results of this essay are limited in scope: I show the application of the proposed model to only one elenchus, viz. that of Laches in the eponymous Platonic dialogue. ^{Footnote 3} Although this is admittedly a narrow conclusion, it is not, therefore, unimportant. For, if it can be established that Socrates practices the method of epistemic exegesis in only one elenchus, then it will have been shown that he practices a method that is markedly different from what has been ascribed to him in the scholarship. Thus, even if the proposed exegetical model cannot be generalised to other interlocutors and dialogues, it will represent a substantial contribution to the project of understanding Socrates’ methods, and how these relate to one another.

2.

I shall begin by sketching out a simple conversational model that I use as a device for understanding Socrates’ elenchus. In this model, two participants, A and B, are engaged in a ‘talk exchange’ about a given topic, X, and A’s primary goal in the conversation is to gain knowledge of what B thinks or believes about X. ^{Footnote 4} In this respect, A’s attention is not focused directly on the topic, but on B’s thoughts about it. The goal of the talk exchange is from A’s point of view to draw out what B thinks about X.

Given that A’s goal in the talk exchange is determining what B thinks about the topic, it is reasonable for A to ask B what she thinks, that is, question her about X. A’s approach might involve questions of various kinds and different levels of generality. For example, if X refers to the causes of the first Persian invasion of Greece in 492 BC, A might ask open questions such as, ‘What were the primary causes of the Persian invasion?’ Or he might use his own knowledge of the topic to formulate closed questions such as ‘Was the Persian invasion a response to the Ionian revolt?’ It is important to notice that, although neither of these questions refer directly to B’s beliefs, A can determine what B thinks about X by asking her questions about X; he need not formulate the questions as questions about her beliefs. ^{Footnote 5}

I now want to augment this model by adding an epistemic dimension. I shall do this by requiring that A assume that B has knowledge of the topic of the conversation. ^{Footnote 6} Given this stipulation, A’s primary goal of determining what B thinks about X should be understood as an attempt to gain knowledge of X by drawing out, or ‘leading’ out (exēgeisthai), what B knows. For this reason, I shall refer to the pattern of inquiry exhibited in the augmented conversational model as ‘epistemic exegesis.’ Epistemic exegesis has as its formal goal the discursive articulation of some answerer’s knowledge of a topic.

Although the term ‘epistemic exegesis’ is a neologism, the basic pattern of conversation is quite familiar. Consider one illustration. Suppose that A is a science journalist, B a physicist, and the topic for conversation the nature of the Higgs particle. Given this interpretation of the variables, A’s interview of B can be understood as an attempt by A to gain knowledge of the Higgs boson by figuring out what B knows about it. And there is nothing strange about this. Insofar as the interviewee is an expert on the topic, the journalist can know the nature of the Higgs particle by determining what the physicist thinks about it. So the journalist is able to acquire knowledge of the world by means of epistemic exegesis.

The idea of a talk exchange in which one participant ascribes knowledge to another is not uncommon, being broadly characteristic of any conversation between a layperson and an acknowledged expert. However, I wish to distinguish epistemic exegesis from this familiar type of conversation by requiring that the questioner’s attribution of knowledge of X to the interlocutor be methodological in character, by which I mean not be open to question in the course of the conversation. So, whereas the science journalist would normally be permitted to withdraw the attribution of expertise to the physicist on grounds of her having made a number of glaring errors in conversation, the interpreter is strictly prohibited from making this move in epistemic exegesis. A may not consider any evidence against the assumption that B has knowledge of X, no matter what the quality of B’s answers to his questions.

The methodological attribution of knowledge to B controls the exegesis by imposing constraints on how A interprets B’s answers to questions about X. The main point of significance is that A cannot, given the methodological attribution of knowledge to B, infer that B’s answers are in error. Insofar as B makes claims about X that seem to A to be in error, A must find a way of avoiding the conclusion that B is really in error. For example, suppose that B answers the question ‘what is the transference?’ by saying that the transference is an oriental rug. A cannot accept this answer at face value; it seems to involve a fundamental misunderstanding of the question and of the whole conversational context. But since A cannot on account of the attribution of knowledge judge that B fails to know the answer to the question, he must infer that B has not for some reason expressed her knowledge of the transference.

When confronted by an answer that appears to be false or nonsensical, A should attempt to diagnose why B appears to have spoken incorrectly. On account of the assumption of knowledge, it is not open to him to judge that B does not have knowledge of X. Two basic possibilities present themselves here. First, A could judge that B is deliberately flouting a maxim of cooperation; she is, for example, being ironical and deliberately obscuring her true meaning. Secondly, A could judge that B is incompetent as a speaker. Although B’s overt claim misrepresented her knowledge of X, this misrepresentation was not intentional, but due to some error in the performance of speech. This is a permissible explanation of B’s error since the attribution of knowledge of X to B does not imply knowledge of how to effectively communicate this knowledge.

Whichever of the above explanations of apparent error he might adopt, it is clear that A should infer in response to an answer that seems false or nonsensical that B has not expressed her knowledge of X. A must either divorce B’s true meaning from the meaning of her utterance, or find an interpretation of that utterance on which it is not untrue. And, in both cases, A looks below the surface of the overt claim in order to discover what B really thinks that the transference is (or what an oriental rug is), given that her overt meaning cannot be her true meaning. ^{Footnote 7} The basic form of inference is as follows: you said that p, but since you are a knower, and p is false or confused, your real meaning is other than the overt meaning of p.

Given that A’s goal in conversation is to determine what B knows about X, the investigation should now move to a deeper level. Since B’s overt meaning fails to express her knowledge, A must find some way of getting at her real meaning. The precise strategy adopted will depend on why A thinks that B did not express her knowledge of the topic. If B is assumed to be flouting conversational maxims in order to communicate irony, then A could try to understand what the problematic overt content is intended to convey. Is B making a joke? Is she attempting to get the interpreter to understand something about the transference by making the absurd claim about rugs? On the other hand, if the judgement is that B is incompetent as a speaker, a different set of maxims will have to be adopted, depending on the nature of the incompetence concerned. Is B prone to slips of the tongue? Is the topic one that it is very difficult to explain such that the speaker’s powers are not up to the task? No further discussion of these matters is necessary in the present context. The basic point is that the pattern of A’s attempt to elicit B’s knowledge—moving beyond the overt meaning to the true meaning—will be shaped by the explanation he chooses for why she does not say what she knows. From the perspective of A, the goal of the conversation is to get B to say what she means in order to know what she knows. The process of exegesis will come to an end when A believes that he knows B’s knowledge.

Insofar as A attempts to gain knowledge of X by finding out what B knows, A presupposes not only that B has knowledge of X, but also that B is capable of communicating this knowledge by answering his questions. B’s capacity to communicate her knowledge is then dependent both on her competence as a speaker and the nature of her knowledge of the subject matter. With regard to the second component, I have thus far left the topic of conversation unrestricted; I now want to make some remarks about how material differences in what it is that B is assumed to know will affect the exegesis.

In epistemic exegesis, A attempts to draw out and give discursive shape to B’s knowledge of X. The question then arises as to whether B’s ‘pre-discursive’ knowledge of the topic is in principle communicable by speech. ^{Footnote 8} For many interpretations of X, such as the atomic number of gold or the causes of the Persian invasion, it will be reasonable to think that B qua knower will be capable of communicating her knowledge to A. In these instances B’s pre-discursive knowledge of X is simply propositional knowledge of the topic, which is either present to B’s mind or capable of being easily brought to mind. On the other hand, if X refers to a topic the knowledge of which consists in art or skill, such as kithary or carpentry, then A will not be able to know B’s knowledge of X by questioning. ^{Footnote 9} One cannot learn how to do something merely by having a conversation about it. ^{Footnote 10} And something similar is true of qualities for which knowledge requires experiential acquaintance, such as, perhaps, goodness or love. Bringing these points together, it is clear that epistemic exegesis can succeed only if what B is assumed to know pre-discursively is in fact capable of being communicated to A by speech. It is only when this presupposition holds that A will be in a position to know B’s knowledge by interrogating her. ^{Footnote 11}

In summary, then, epistemic exegesis is an attempt by an inquirer to draw out by questioning an assumed knower’s knowledge of some topic. Give that the inquirer’s attribution of knowledge is not open to question in the conversation, the interrogation is governed by specific rules for the interpretation of apparent error. The exegesis can succeed only if the answerer is competent in the performance of speech, willing to speak her mind, and the subject matter such that the answerer’s assumed pre-discursive knowledge of the topic of a sort capable of being articulated in speech.

3.

I shall now argue that the exegetical model of conversation gives an accurate picture of Socrates’ elenchus of Laches in the eponymous Platonic dialogue. I shall develop the argument by showing that the model both explains the textual data and explains it better than competing accounts. If my reasoning is correct, then Socrates’ elenchus of Laches should be understood as a kind of epistemic exegesis. ^{Footnote 12}

If Socrates practices a method of epistemic exegesis, then he must attribute knowledge of the topic of conversation to the interlocutor. Although the assumption of knowledge need not be indicated, but might become manifest in the way in which the conversation is conducted, it is clearly indicated in the present instance. For Socrates begins the elenchus by asking Laches whether they should say that they know what virtue is (190c4; Φαμὲν ἄρα, ὦ Λάχης, εἰδέναι αὐτὸ ὅτι ἔστιν;). ^{Footnote 13} When Laches answers with an emphatic affirmative (190c5; Φαμὲν μέντοι), the expectation that the interlocutor will be assumed to be a knower is confirmed. ^{Footnote 14}

If Socrates practices a method of epistemic exegesis, then he will attempt to ‘draw out’ what his interlocutor is assumed to know about the topic by means of questioning; this ‘drawing out’ of knowledge is what the exegesis consists in. Now, as explained in §2, an attempt to draw out an answerer’s assumed knowledge of a topic by means of questions can succeed only if this knowledge is communicable by speech. The exegetical model, therefore, offers an explanation for why Socrates checks to see whether Laches will agree that those who know what virtue is are able to state what they know (190c6; Οὐκοῦν ὅ γε ἴσμεν, κἂν εἴποιμεν δήπου τί ἐστιν;). It is noteworthy and surprising that Laches, a man of action with a marked suspicion of talk (cf. 188d f.), strongly affirms the possibility of ‘saying’ his knowledge of virtue (190c7; Πῶς γὰρ οὔ;).

Socrates’ and Laches’ self-attribution of knowledge at the beginning of the search is in conflict with the constructive interpretation of elenchus as first-order inquiry, that is, inquiry into the nature of virtue per se, rather than inquiry into what somebody knows about it. ^{Footnote 15} In contrast with inquiry as exegesis, which aims to articulate knowledge already possessed, one cannot conduct a first-order inquiry into what virtue is if one assumes that one already knows its nature (cf. Men. 80d-e). On the other hand, Laches’ attribution of discursive knowledge of courage to himself is compatible with an account of elenchus as refutation. ^{Footnote 16} On this view, Socrates’ preliminary questions would allow Laches to saddle himself with a claim to knowledge, ^{Footnote 17} which subsequent argumentation would aim to overturn. What needs to be considered in the ensuing discussion is whether Socrates does attempt to overturn the claim to knowledge, or whether, as the exegetical model would predict, he treats it as indefeasible during the course of the conversation.

After narrowing the discussion from virtue to courage, ^{Footnote 18} Socrates moves away from the deliberative perspective marked by the first person plural, and concentrates on Laches qua knower of courage, urging him “to state … what courage is” (190e1). Given that Laches is assumed to be a knower, who is capable of articulating his knowledge, the ‘what is x?’ question must apparently be understood at the level of Laches’ cognitions—Socrates is asking Laches qua assumed expert to say what he knows courage to be. ^{Footnote 19} The discourse is not argumentative, as the standard interpretations would have it, ^{Footnote 20} but exegetical, since the goal of questioning is to know Laches’ knowledge. In fact, Socrates is nearly explicit on this point: he says that he wants to learn from Laches (βουλόμενος σου πυθέσθαι; 191d1) “what constitutes courage” in various domains (191c8-d2; cf. 191e9). ^{Footnote 21}

On the exegetical model, Socrates is not permitted to attribute error to his interlocutor in the course of the elenchus. He must on account of the methodological attribution of knowledge (190c4-5) interpret error as merely apparent, indicative not of any epistemic deficiency, but of the interlocutor’s unwillingness or inability to express his knowledge. This prediction is amply confirmed by the text. Laches’ first answer to Socrates is that “anyone who is willing to remain in rank and defend against the enemy, and does not run away” is courageous (190e5-7; my trans.). This answer is obviously unsatisfactory as a definition, giving neither necessary nor sufficient conditions for courage (191a8-c5). Thus, if Socrates were aiming to test or refute Laches’ claim to knowledge, as the standard interpretations maintain, ^{Footnote 22} then the elenchus could now end. But Socrates does not say that Laches’ definition has come up short. Instead, he says that, while Laches has “spoken well,” he has not answered as desired, because the question has been poorly formulated (190e7-9; 191a1-c6; cf. 191c7-d1). In other words, Socrates does not challenge Laches’ credentials as a knower—just as the exegetical model would predict. ^{Footnote 23} Laches has failed to express his knowledge because of Socrates’ poor exegesis.

After extending the application of ἀνδρεία well beyond the domain of war (191d3-e2), Socrates asks Laches to “try again to say what courage is” (191e10). But since his interlocutor does not yet understand, Socrates gives assistance by offering a model definition of swiftness (192a9-b3) before reformulating his question:

Laches answers that courage seems to him to be “a sort of endurance of the soul” (192b9; Δοκεῖ τοίνυν μοι καρτερία τις εἶναι τῆς ψυχῆς). ^{Footnote 24} Although this answer is a reasonable one, it is both too broad and too narrow, since some forms of endurance are not courageous and not all forms of courage require endurance. ^{Footnote 25} But again Socrates does not hint that Laches’ answer is in any way deficient. Instead, he says:

Socrates’ behaviour here is precisely to be expected if he were engaged in epistemic exegesis—he interprets the unsatisfactory answer in terms of a distinction between what his interlocutor has said and what he really thinks. Laches has not, Socrates suggests, adequately expressed his thoughts on the topic; for he does not truly think that all forms of endurance are courageous. ^{Footnote 26} By contrast, if Socrates were a first-order inquirer searching for a definition of courage, his response to Laches’ answer would make little sense; for why should he be so concerned with Laches’ thoughts on the matter, rather than on the matter at hand? An account of the elenchus as a first-order search for a definition of courage does not explain why the inquiry is centred on what Laches thinks about the subject, rather than the subject matter itself. ^{Footnote 27} But this centring of the inquiry on the thoughts of the interlocutor would be expected if Socrates were practicing epistemic exegesis. Inquiry as exegesis is in fact distinguished from first-order inquiry on exactly this point: it assumes that a respondent has knowledge of a topic, and then focusses on what she knows about the topic, rather than the topic itself.

On the model of epistemic exegesis, insofar as Laches’ knowledge of courage remains hidden, Socrates must attempt to excavate it. This is what happens. After Laches admits that he does indeed consider courage to be one of the most beautiful things (192c7), Socrates introduces two hypotheses: courage is prudent endurance, and courage is foolish endurance. He conjectures that Laches does not think that courage is endurance with foolishness (192d7-8; ἡ μετ’ ἀφροσύνης) but that it is, by his account (κατὰ τὸν σὸν λόγον), prudent endurance (192d10-11; φρόνιμος καρτερία). “Apparently,” says Laches (192d12). This is the first attempted reconstruction of Laches’ knowledge of courage.

In the next stretch of the dialogue, Socrates interrogates the nature of the prudent endurance that, according to his reconstruction of Laches’ meaning, defines courage (192e1-2). Would Laches call a man courageous who endures in spending his money prudently “because he knows that by spending he will gain more?” (192e2-4). Laches swears an oath: “by Zeus, not I” (192e5). This pattern is repeated with respect to a doctor who refuses his patient something to drink or eat. Socrates then switches from absolute to comparative judgements of relative courage. Who would Laches say (φαίης) is more courageous—the man who endures in war on account of knowing that he is in the winning position, or the man in the losing position who is willing to endure? The latter, says Laches. Would Laches agree that his endurance is more foolish? He does. This model is repeated thrice more: cavalry, slinging and shooting, and well-diving. In each instance, Laches says that the man who endures without knowledge is more courageous than the man who endures with prudence, even though his endurance is more foolish (193a-c). ^{Footnote 28}

Socrates has offered inconsistent reconstructions of—what is assumed to be—Laches’ knowledge of courage. Initially Laches had accepted the suggestion that courage is prudent endurance. Now he has been reduced to saying that courage is foolish endurance (193d6-7). According to the exegetical theory, Socrates will interpret this inconsistency as evidence for thinking that Laches has not managed to state his knowledge of courage. This is exactly what occurs:

Socrates takes apparent inconsistency to imply that Laches has failed to say what courage is, rather than his failing to know what courage is. ^{Footnote 29} Thus he invites Laches to conclude that he has not articulated his knowledge of courage in speech. While this behaviour is predicted by the exegetical model, it is quite unexpected given the standard interpretations of the elenchus. If Socrates’ goal were to show the interlocutor up as inconsistent, ^{Footnote 30} he should not interpret inconsistency as merely apparent and as indicating a failure in the articulation of speech.

After Laches and Socrates have been found not to have spoken beautifully (193d1-10), Socrates remarks that they are not “tuned in the Dorian mode” (193d11; Allen trans.; cf. 188d f.). Someone would likely say, he suggests, that he and Laches have a share of courage in deed; but, if this person overheard their discussion, he would not say that they had a share of courage in speech (193e2-4). And, since it is not beautiful to be in this state (193e6-7), Socrates asks whether Laches is willing to be persuaded by their earlier logos, the one that commanded them to endure (193e8). He says:

Socrates’ remarks echo and modify Laches’ first definition of courage (190e5-7; cf. Ap. 28d-e). The need for willingness has been retained, but remaining in rank has become an exhortation to stay on in the search (ἐπὶ τῇ ζητήσει ἐπιμείνωμέν). It seems that the passive state of remaining in rank to defend against the enemy (ἀμύνεσθαι τοὺς πολεμίους) must be reconceived as an active pursuit (194a4). ^{Footnote 31} There is in addition a reference to a higher goal, not that of the polis, but of avoiding being shamed by courage herself, who might laugh at them if they were to give up, exhibiting a lack of courage in the search for courage. ^{Footnote 32}

Laches is willing to persevere (194a6), motivated by a love of victory and sense of anger at his inability to speak his mind (194b1). Socrates re-asserts the active nature of the courage that is required by introducing a metaphor of pursuit: “the good huntsman must follow the hounds and not give up the chase” (194b5-6; Lamb trans.). After Laches agrees to this, Socrates asks whether he would be willing to have them call Nicias to the hunt—in case he might be more resourceful (194b8-9). Laches is willing and Socrates asks Nicias to help his friends who are “storm-tossed by the argument [and in aporia]” (194c2-3). Socrates thinks that his work with Laches is done.

Most commentators suppose that Socrates intends by his elenchus to disabuse his interlocutor of the claim to knowledge. ^{Footnote 33} If this were the case, then Socrates’ action of drawing the conversation to a close (194c2-3) would be practically irrational, since Laches has not inferred that he does not know what courage is, but that his knowledge of courage has not been adequately conveyed by his proposed definitions. This point is expressed by Laches himself in vivid terms. He says:

As is clearly evident from this passage, Laches still thinks he knows what courage is (194b1-2), but is unable to “gather it together in logos and say what it is” (194b3-4; my trans.; cf. Theaet. 148d7). And Socrates does nothing to contradict this interpretation, even though he could easily do so by referring back to the proposition that was agreed upon at the beginning of the discussion, viz. that one who knows will be able to state what he knows (190c6). Hence Socrates does not here intend to disabuse his interlocutor of the claim to know what courage is.

Does the exegetical model fare any better in explaining Socrates’ behaviour? It might initially seem that it does not, since Socrates’ act of bringing the conversation to a close is not conducive to the formal goal of knowing his interlocutor’s knowledge of courage (194a6-b4). It appears that Socrates qua exegete should either continue the search for Laches’ knowledge or conclude that Laches does not know what courage is.

I suggest that the central point in need of explanation is why Socrates upholds the attribution of knowledge while bringing the conversation to a close. Since he voluntarily ends the discussion with Laches, and forms an alliance with him in the subsequent dialogue with Nicias (e.g., 197e6-8.), his ascription of knowledge to Laches appears to be methodological in character, upheld independently of whether Laches really does have knowledge of courage. ^{Footnote 35} Thus it seems that Socrates’ practice of his method is ironical: while his formal goal is to know his interlocutor’s knowledge, his material goal is to get Laches to infer that he has not been able to say what he knows. Socrates pretends that Laches is a knower in order to get him to interpret his dialectical failures as indicative of an inability to speak his mind. ^{Footnote 36}

After Laches expresses frustration at his inability to pin down what he takes himself, and continues to take himself, to know, Socrates encourages him to endure in the search for knowledge of courage (194a1-a5). Hence Socrates links Laches’ experience of frustration at being unable to say what he knows with the requirement to endure in the pursuit of knowledge of courage. But what difference would it make if Socrates were to abandon the attribution of knowledge and guide Laches to the conclusion that he did not know what courage was? Could he not also then continue the hunt for knowledge of courage? The answer to these questions is to be found in a distinction between first- and second-order inquiry, that is, first-order inquiry into a topic and second-order exegesis of one’s knowledge of a topic. Inasmuch as Socrates wants Laches to search for knowledge of courage from the perspective of one who already possesses it, the inquiry Socrates is recommending must take place at the second-order level. Laches should continue the attempt to attain discursive knowledge of what he takes himself to know pre-discursively. In short, the kind of inquiry that Socrates is recommending is epistemic self-exegesis.

4.

The model of elenchus as epistemic exegesis fits the pattern of conversation exhibited in Socrates’ interrogation of Laches up until the point at which he voluntarily brings it to an end. The fact that Socrates maintains the attribution of knowledge even as he breaks off the conversation suggests that, while his formal goal is to know his interlocutor’s knowledge of courage, his material goal is different. I have suggested that Socrates wants Laches to experience himself as unable to state his knowledge in order to get him to endure in epistemic self-exegesis. Socrates wants Laches to apply the exegete’s formal goal to himself; he should continue the search for discursive knowledge of what he takes himself to know pre-discursively.

The nature of the discursive knowledge that Laches does not have, takes himself not to have, and must search for, seems to involve having an articulate account, that is, a definition, of courage. ^{Footnote 37} But what is the nature of the pre-discursive knowledge that Socrates attributes to Laches and that Laches attributes to himself at the end of the discussion (194a6-b4)? Does he really have pre-discursive knowledge of courage? Or is Socrates’ attribution of knowledge to Laches a matter of pure pretence?

It is clear that Laches does not have pre-discursive propositional knowledge of courage, of the sort, perhaps, that Nicias has (194d1-2), and as might be expressed in a definition. This is evident from the flow of the conversation, as well as Laches’ admission that he is unacquainted with philosophical discourse (188d f.). However, this is not the only kind of knowledge that might justify or explain why Socrates upholds the attribution of knowledge at the end of the elenchus. For all of his dialectical errors, Laches does have a share of courage in action (191a and 193e; cf. 195c); he is capable of enduring in battle, of remaining in rank, and warding off the enemy. This capacity to respond appropriately in conventional patterns of hoplite warfare can be described in epistemic terms: Laches knows how to do what courage requires of him in certain circumstances.

Laches’ knowledge of courage is not restricted to forms of action. He is also, as would seem to be required for courage in action, able to identify courage in hoplite warfare and other types of fighting (181b1-4; 190e4-6), whether these identifications refer to his own actions, or those of others. And he displays in the dialogue with Socrates significant powers of conceptualisation, proving capable of distinguishing courage from forms of technical accomplishment that imitate it (cf. 183c-184a) and generalising from courage in hoplite warfare to other domains in which courage may be displayed (191d-e). Moreover, Laches is able to locate courage in a nexus of relevant ‘concepts,’ rightly associating courage with beauty or nobility (192c5-6), linking it to endurance (192b9-c1), and contrasting it with foolishness (192d1-6). If he did not have dialectical abilities of this sort, he might well answer Socrates’ questions by saying that courage is equality distinguished by sympathy, and which stands in opposition to pride. ^{Footnote 38}

In response to the question of whether Laches really does have pre-discursive knowledge of courage, there are then reasons for answering in the affirmative. This pre-discursive, non-propositional knowledge of courage is presumably what Laches is referring to when he says that he still thinks he “has the concept of what courage is” (194b1-2; Allen trans.; νοεῖν μὲν γὰρ ἔμοιγε δοκῶ περὶ ἀνδρείας ὅτι ἔστιν) even though he is unable to “say what it is” (194b3-4). ^{Footnote 39} And it follows that Socrates’ attribution of knowledge of courage to Laches is not pure pretence.

Although there are grounds for attributing knowledge of courage to Laches, the knowledge that he is assumed to have is incomplete or limited in several respects. Inasmuch as his attempted definitions express his ‘concept’ of courage, it appears that he is for a general rather insensitive to the possibilities for courage in flight or pursuit, that is, those instances in which courage requires neither remaining in rank nor enduring in battle formation (cf. 182a-b). ^{Footnote 40} It is in this respect ironical and amusing that Laches is both effusive in his praise of Socrates’ courage in the flight from Delium (181b) and sufficiently neglectful of its possibility so as to fail to accommodate it in his proposed definitions. ^{Footnote 41}

Laches’ failure to be properly mindful of the ways that courage can be expressed in retreat is also expressed in his martial deeds. For, in the Symposium, Alcibiades reports that during the Athenian flight from Delium, Laches proved to be much less composed (ἔμφρων) than Socrates when called upon to abandon the battle line (Symp. 221b1). ^{Footnote 42} Hence it can be said that, whereas Laches knows how to be courageous when in battle array, he does not know how to be courageous when the formations have been broken up. His conception of courage is not well suited to his station as a general, that is, someone who must decide when to hold ground, when to pursue, and when to take flight.

Laches exhibits similar weaknesses in another sort of deed, viz. the activity of philosophical dialogue. For, while he is bold in answering the ‘what is x?’ question, supposing it not to be difficult to say that courage is holding one’s ground (190e4), he is initially flummoxed by Socrates’ expansion of the modes of action and contexts in which courage can be displayed (191d-e), thus failing to hold his ground in defending his account of courage as holding one’s ground. Apparently he has the courage to stand by his convictions but not the courage to abandon them. ^{Footnote 43} And his courage in pursuit is similarly limited, since, after venturing that courage is endurance, he becomes tangled up in logoi, exemplifying a diminished mindfulness and even anger that mirrors his lack of composure in the flight from Delium (192d7-19310; cf. 194d-196b).

It is important to notice that Laches’ deficiencies in speech and action do not exactly reveal a ‘gap’ in his conception of courage; for, to take the most conspicuous example, Laches is able to recognise Socrates’ courage in retreat (181b), despite omitting this possibility from his definitions, and failing to embody it in his actions. Thus, while knowing pre-discursively that courage can be expressed in retreat and pursuit, he for some reason does not say what he knows. His failure to articulate his own recognition of the possibilities for courage in retreat implies that his understanding of courage is disjointed. And this is presumably one reason that he cannot properly express it in discourse or action.

Socrates’ practice of exegesis works on the supposition that Laches knows what courage is. I have suggested that, inasmuch as Laches really does have knowledge of courage, or a degree of it, this is not propositional knowledge, but more akin to forms of know-how manifested in deed and in powers of conceptual collection and division (cf. Phaedr. 266b ff.). On the other hand, Laches’ ‘conception’ of courage appears to be malformed. Hence it can be said that, while Socrates’ attribution of knowledge to Laches is not pure pretence, it does contain an element of pretence, since it involves attributing to him complete knowledge of courage that he does not really possess.

5.

Why does Socrates want Laches to endure in epistemic self-exegesis? One possible answer is that he wants him to express his pre-discursive understanding of courage in a definition; this seems to be, after all, the formal goal of the exegesis. But this cannot be correct, since, as discussed in §4, Laches does not have a pre-discursive propositional account of courage in place that need only be verbalised; and, moreover, his non-propositional knowledge of courage is incomplete or disjointed in several respects. Hence the activity of self-exegesis cannot be understood as articulating a knowledge that is already fully formed and only in need of correct enunciation. It must be understood as having a formative component. Moreover, after Laches fails to say what he knows, Socrates refers back to the ideal of the truly musical man who has “tuned himself in the finest harmony, not on a lyre but in his life” (188d f.), ^{Footnote 44} recommending, on this basis, that they endure and remain in the search (194a2-5) so that “courage herself won’t ridicule [them] for not searching for her courageously” (194a1-a5; Sprague trans. with revisions). And this would suggest, strikingly, that Socrates wants Laches to endure in self-exegesis because he would by this process facilitate a Dorian harmony of courage in speech and deed.

Although Laches himself introduces the ideal of a Dorian harmony of speech and deed, his position would be more accurately described as one that prioritises deed over speech. For, without ascribing this priority to him, one would be unable to understand why he would be aggrieved at hearing fine speeches about virtue from those who do not manifest virtue in deed, but not be aggrieved at hearing no beautiful speeches about virtue from those who do manifest virtue in deed (188c3-188e5). ^{Footnote 45} Laches’ speech in praise of a Dorian harmony of speech and deed does not itself harmonise with his deeds.

In his response to Laches’ aporia, Socrates emphasises the self-reflexivity of philosophical inquiry into courage (194a1-a5), and, in so doing, questions Laches’ prioritisation of deeds over speeches in at least two ways. First, he points to a context in which virtue in speech and deed cannot be separated; for, in the activity of self-exegesis, beautiful speech about courage is the beautiful deed. Secondly, and this is the point I want to focus on, he implies that perseverance in exegetical discussion is relevant to the formation of a courage that is genuine and complete. How could this be?

Laches’ emphasis on taking a stand and endurance in his attempt to articulate his knowledge of courage manifests genuine insight into courage of a sort that is embodied in his actions. For inasmuch as courage is oftentimes endurance (194a4-a5) it is plain that Laches does have a share of courage: he is capable of enduring in the fight against the enemy (193e2-4). In this regard, Laches exhibits a harmony of speech and action, for both his words and martial deeds faithfully express his pre-discursive understanding of courage as endurance (cf. 183c-184b). In fact, it seems that they are in harmony with one another because they both manifest his pre-discursive understanding of courage.

Within this context, it is evident that Socrates’ exhortation of Laches to endure in philosophical inquiry refigures Laches’ conception of courage by shifting his focus from hoplite warfare to philosophical dialogue (194b5-6). The disposition to endure by holding one’s ground in battle has been expanded so as to include the endurance required for the active pursuit of knowledge. This non-propositional enlargement of Laches’ understanding of the role of endurance in courage might contribute to his forming a Dorian harmony of speech and deed. For, if Laches were to endure in philosophical dialogue, the quality with which he identifies courage would come be reflected in his discursive actions (194a1-a5). He would no longer, absurdly, fail to exhibit endurance in the search for courage (194a1-a5).

In the previous section, I identified certain weaknesses in Laches’ pre-discursive knowledge of courage. I suggested that Laches thinks of courage in terms of the hoplite’s ability to hold his ground, while at the same time recognising, as would befit a general, that courage can also be manifested in pursuit and retreat (cf. 191a-c). His failure to express discursively his own recognition of the possibilities for courage in pursuit and retreat implies that his pre-discursive understanding of courage is fragmented. And this is one reason that he cannot properly express it in discourse. Moreover, as mentioned in the previous section, the disunity inherent in Laches’ understanding of courage is also expressed in his actions. He knows that pursuit and retreat are required in certain circumstances, and he knows that courage can be manifested in these forms of action, but he does not exhibit the mindfulness needed for courage when structure has been lost, whether this be in war or in philosophical dialogue.

Laches is capable of two important insights into the nature of courage: it can be manifested by remaining in rank and by breaking it, by enduring in the fight and giving it up. But he does not bring these insights together in dialectical definition or martial or philosophical action. Hence it can be said that Laches neither says what he knows, nor does what he knows; and this is why he fails to manifest courage in word and deed. In order to attain a Dorian harmony, Laches would need to draw together his disparate insights into the nature of courage and express this unified understanding in speech and action. Effecting this unification would apparently require bringing together opposing aspects, viz. enduring in the fight and giving up the fight, into a single conception. Presumably this is why Socrates introduces the concept of phronesis in the course of his dialectical questioning (192d10-11). The power that would unify Laches’ pre-discursive understanding of courage is to be identified with the power that is missing from both his definitions of courage and his martial and philosophical actions, viz. the wisdom of intelligent adaptability that would facilitate appropriate responses to changing circumstances. ^{Footnote 46}

In the previous section, I argued that Socrates’ goal in the dialectic is to get Laches to endure in self-exegesis. In the present section, I proposed that his reason for recommending this course of action is that self-exegesis would contribute towards Laches’ development of a Dorian harmony of speech and deed. In order to develop this harmony, Laches would have to integrate his pre-discursive insights into courage and express this unified understanding in definitions and actions. If I am right to suggest that Socrates thinks that this unification requires the power of wisdom (phronesis), then it follows that he believes that Laches would develop wisdom—the wisdom involved in courage—by enduring in self-exegesis.

There is a potential obstacle to this type of account. To see this, notice that it makes sense that Socrates should appeal to Laches’ endurance when encouraging him to continue the effort to articulate his knowledge of courage (194a1-a5); for this is a power that he already has. But how could he give an account of his knowledge of wise endurance when he does not have wisdom? ^{Footnote 47} More generally, if courage is a power of the soul, as Socrates implies (192b6; cf. 192b1, 194c3), then one would already have to be courageous in order to know one’s courage. It seems impossible to know by self-exegesis a power that one does not already have. Consider, for example, someone who knows how to play the kithara. It is obvious that she would be able to describe the power of playing the kithara only if she had this power. Without it, no second-order inquiry into the nature of her musical power can be conducted.

While the dependence of second-order knowledge of a given power on the first-order possession of this power seems to me uncontroversial, matters are considerably more difficult in the case of courage than they are in the case of music. For, while the attempt to describe and articulate a musical power requires that the person be in possession of that musical power, the very attempt to give a discursive account of a musical power does not itself require a musical power. By contrast, not only is second-order inquiry into the nature of the power of courage impossible without courage, the attempt to search courageously for an account of one’s knowledge of courage is itself impossible without courage. ^{Footnote 48} Thus the power of courage is presupposed by the very attempt to articulate one’s knowledge of courage in a way that musical powers are not presupposed by the attempt to articulate the nature of one’s musical powers. ^{Footnote 49}

The way to bypass this difficulty is to say that the power of courage and the power of articulating one’s knowledge of courage are jointly realised in the activity of philosophical self-exegesis. This claim fits well with the results of earlier discussion in Laches (192c-193d), where Socrates compares the relationship between virtue and the soul to the relationship between sight and the eye (189e-190a). For as sight is inherent in a properly functioning eye, wisdom is inherent in a properly functioning soul (cf. Aristotle, E.N., 1144a26-33). This claim is also in keeping with the Platonic or Socratic view that “the power to learn is present in everyone’s soul,” being natural to a human being and only needing activation and development; “education isn’t … putting knowledge into souls that lack it, like putting sight into blind eyes,” but turning the eye of the soul so that it can see (Rep. 518b-c; Grube trans.).

I have suggested that Socrates’ exhortation of Laches to remain in the search for discursive knowledge of his own knowledge of courage expresses the view that continued self-exegesis would actualise Laches’ dialectical power as imperfectly embodied in his pre-discursive understanding, where the full actualisation of this power would be identified with wisdom (phronesis). It is then important to notice that Laches’ dialectical power does improve on account of the activity of self-exegesis, since he, the man unaccustomed to Socratic discourse (194a), follows Socrates’ lead in using exegesis to mount an impressive, albeit unsuccessful, attack on Nicias (194d-196b). In this dialectical movement, he displays, or begins to display, the courage involved in pursuit. ^{Footnote 50}

6.

My goal in the paper has been to offer an original account of Socrates’ method in his conversation with Laches. The proposed account of the elenchus is intended to be illuminating on its own terms, and a critique of the standard interpretations of the elenchus of Laches. It is important to notice, however, that I have not attempted to generalise the proposed model to other Socratic conversations, not even the conversation with Nicias in the same dialogue. Socrates’ conversations with sophistic interlocutors, or those influenced by sophists, always involve a fundamental deviation from the pattern of epistemic exegesis. And, since Nicias is in the relevant sense a sophistic interlocutor, on account of his association with Damon and Prodicus (197d3), a different model is required to explain the second part of the dialogue. This model is, as I attempt to show in other writings, parasitic on the exegetical account inasmuch as it involves a projectable departure from it. ^{Footnote 51}