2. Socrates and the unity of virtue

At a crucial stage in the Protagoras (329cd), Socrates asks his interlocutor whether (1) ‘wisdom, ’ ‘temperance, ’ ‘courage, ’ ‘justice, ’ and ‘piety’ are all names for one and the same thing, or (2) whether they are parts of a single whole. When Protagoras opts for the latter, he is confronted with the question of whether (a) their relation resembles the parts of a face, each possessing a distinct capacity, or whether (b) they resemble the parts of a piece of gold, ‘which do not differ from one another, or from the whole, except in size.’ At this point, Protagoras argues that the different virtues constitute a whole just as the parts of a face constitute a whole. It is a matter of controversy which position Socrates endorses. On the biconditional interpretation Socrates endorses (2); on the identity interpretation, Socrates endorses (1).

Vlastos asserts that the second claim, that the virtues are distinct parts of a single whole, is ‘standard Socratic doctrine.’ He cites passages from the Meno and Laches in support of his assertion.Footnote ¹ According to Vlastos, Socrates himself would have opted for (2) if he were confronted with the same question – the two alternatives provided by Socrates in response to Protagoras’ answer, (a) and (b), seem to indicate that (1) cannot be the unity relation that Socrates has in mind. However, it is common in the early dialogues for Socrates to proceed with his questioning using only those premises that are granted by his interlocutor. If this is the case in the Protagoras, the alternatives that Socrates presents to his interlocutor may not reflect his own view at all. It is certainly true that some passages in the Meno and Laches present problems for the idea that Socrates endorses (1). But leaving other dialogues aside, a careful reading of the Protagoras will suggest that this is precisely the view that Socrates defends in our dialogue. In fact, Socrates states his own view at 331b: ‘On my own behalf, I myself would assert that justice is pious and that piety is just… that justice is the same thing as piety or as similar as possible …’Footnote ² Protagoras understands Socrates to be defending the position that the virtues are all ‘one and the same thing.’ At 350d4, he recognizes that Socrates is trying to draw this conclusion, as Socrates recapitulates the two positions that are in play. Socrates asks Protagoras whether the virtues stand for a single thing, or whether each virtue has a particular thing underlying it, ‘a thing with its own separate capacity’ (349b). The context plainly indicates that they are resuming the previous discussion with the same two positions, and that Protagoras will continue to support (2). If (1) is still the active alternative to (2), then it would seem that Socrates is the one maintaining it. The conclusion of the dialogue appears to confirm this hypothesis, since Socrates reaches the conclusion that all the virtues are ‘one and the same’ knowledge (361a).Footnote ³

If Socrates does defend the identity view in the Protagoras, then it is necessary to explore the deeper question of how this identity relation should be understood. An intriguing suggestion is made by Penner. His explanation begins with the primary question of the early dialogues. He explains that the ‘what is x?’ question (e.g. ‘what is courage?’ or ‘what is temperance?’ or ‘what is piety?’) is generally understood to be a request for either the meaning of the word ‘x, ’ or for the essence of x. The biconditional interpretation will typically take the primary question to be a request for the essence of x. It is a request for a conceptual analysis, which requires a definition that picks out all and only instances of ‘x, ’ so that the underlying essence (οὐσία) can be grasped.Footnote ⁴ It is in this way that the thesis of the biconditional interpretation (the thesis that the virtues are distinct parts of a whole) becomes agreeable, seeing that the virtue-names correspond to distinct essences.

This, however, is not how Penner understands the primary question. According to Penner, the ‘what is x?’ question is a request for the psychological state, or the capacity (δύναμις), responsible for making a virtuous person virtuous. When we encounter the question ‘what is temperance?’ it is a request for an account of the psychological state of the temperate person, the very psychological state which makes the temperate person temperate (Reference Penner1973, 80). For this reason, Penner refers to such questions as ‘the general’s question’ – a question designed for the purpose of discovering that quality, the imparting of which will make a person virtuous. Once Penner makes this clarification, it is much easier for him to defend the identity interpretation. It amounts to the claim that the very same psychological state (δύναμις) lies behind each of the individual virtues. This underlying psychological state is a kind of knowledge, more precisely, knowledge concerning good and bad. Therefore, strictly speaking, the virtues are one and the same thing – knowledge of good and bad.

Perhaps this explanation sheds light on the Protagoras, where Socrates defends the view that the virtues are ‘one and the same thing.’ Unfortunately, it appears to conflict with Socrates’ position concerning the virtues in other dialogues. As Vlastos points out, parts of the Meno and Laches seem to indicate that Socrates endorses the notion that the virtues are parts of a single whole – (2). And Penner’s explanation doesn’t seem to account for the fact that it is often a definition that Socrates is after. In such cases, it would seem that the essence (οὐσία) is being pursued, not the psychological state. Of course, given the complexity of the early dialogues, it is certainly possible that both features are present. In other words, it is possible that the primary question serves to introduce two distinct searches. This is precisely what I intend to uphold. I will suggest, in keeping with Penner, that there are passages in the early dialogues where the psychological state (δύναμις) of the virtuous person is the object of inquiry. (The Protagoras would seem to be an example of this). It is with respect to these inquiries that Socrates endorses the identity view (1), so that the identity interpretation is correct to this extent.Footnote ⁵ In other places, however, Socrates and his interlocutor(s) are engaged in a search for the essence (οὐσία) of a particular virtue. Such inquiries aim to establish a definition capable of picking out all and only instances of the virtue in question.Footnote ⁶ It is with respect to their definitions, as well as their essences, that the virtues are considered distinct parts of a single whole (2), and the biconditional interpretation is correct to this extent. Previous commentators have attempted to reconcile Socrates’ claims about virtue by invoking similar distinctions. Brickhouse and Smith, for example, have suggested that (1) holds for the extension of the individual virtue-names, while (2) holds for their intensions. But such commentators have failed to identify the two distinct levels of inquiry that would give rise to such a distinction in the text.Footnote ⁷ The presence of two distinct levels of inquiry will explain the complexity of the unity thesis, and allow readers to understand why Socrates is making claims that appear to conflict. In what follows, I intend to bring these two levels of inquiry to the surface. I will call the search for an underlying psychological state the psychological level of inquiry, or simply the ‘psychological search.’ Socrates endorses the identity view here, at the psychological level. Meanwhile, I will call the search for definition the conceptual level of inquiry, or simply the ‘conceptual search.’ Socrates endorses the biconditional view here, at the conceptual level. I will demonstrate that the psychological search can be outlined through various passages in the Protagoras and Laches (Sections 3–5), that the conceptual search can be outlined through various passages in the Euthyphro and Meno (Sections 6–7), and that the Meno finally verifies that these two searches are in fact distinct (Section 8).

3. Protagoras and the psychological search

The clearest formulation of Socrates’ concern in the Protagoras occurs at 349b:

The first disjunct represents Socrates’ position in the Protagoras. But for the moment, let us focus our attention upon the second. By asking whether the virtue-names have (a) a distinct essence (οὐσία) as well as (b) a distinct thing (πρᾶγμα) underlying them with its own particular capacity (δύναμις), Socrates is making a distinction between two potentially separate questions. Throughout the Protagoras, the discussion focuses on (b), the particular thing (πρᾶγμα) underlying virtue, or the particular capacity (δύναμις). The essence (οὐσία) of virtue is never pursued in the Protagoras. Of course, the passage could be interpreted in such a way that οὐσία and πρᾶγμα are merely synonymous, or somewhat redundant, so that the occurrence of οὐσία adds nothing significant to the content of the question. This is how commentators have typically interpreted (P).Footnote ⁸ However, the passage might be interpreted quite differently. It might be read as suggesting that the essences (οὐσίαι) of the virtues are in fact distinct, and that the only remaining question is whether the underlying πράγματα (and corresponding δυνάμεις) are distinct as well.Footnote ⁹ I think this alternate interpretation is correct. After all, (P) is presented as a recapitulation of an earlier passage (329c-330c) in which Protagoras had assured Socrates that each of the individual virtues are ‘distinct from the others’ (331a5). Remarkably, Socrates responded with a follow-up question: ‘Does each [virtue] also have its own particular capacity (δύναμις)?’Footnote ¹⁰ In this way, it is made clear that, even after the virtues are deemed distinct somehow, another question remains – whether the virtues also refer to distinct psychological states or capacities. We will return to this suggestion shortly. For now, let us note that the Protagoras examines the πρᾶγμα and δύναμις, with hardly a mention of essence (οὐσία).Footnote ¹¹ Socrates and Protagoras are engaged in the psychological search. They are seeking the thing (πρᾶγμα) in the world to which the virtue-names refer. This involves either a single capacity (δύναμις) or multiple capacities. The question of how virtue is imparted (παραγίγνεσθαι) will depend entirely upon the nature of these capacities. In the Protagoras, Socrates equates virtue with a single capacity, the knowledge of good and bad, which is later described as the science of measurement. By doing so, he apparently abandons (2), the notion that the virtues are distinct parts of a single whole, the central notion of the biconditional interpretation.

4. The Devereux discrepancy

Devereux acknowledges Socrates’ commitment to the identity view in the Protagoras; but he notices that, in the Laches, Socrates clearly states that the virtues are parts of a whole. This favors the biconditional interpretation over the identity interpretation, since the part-whole relationship is compatible with the former, not with the latter.Footnote ¹² It is made explicit at 190c9-11, where Socrates suggests that their task would be overwhelming if they began by considering the whole of virtue. They ought instead to consider a part of it first.

Socrates remains loyal to this view throughout the Laches. In fact, he turns down Nicias’ suggestion (194e) that courage is a kind of knowledge (knowledge of what is to be feared and dared), which turns out to be identical to the knowledge of good and bad. According to Devereux, if Socrates were still committed to the identity view, he would have availed himself of this definition.Footnote ¹³ But even if we employ the biconditional interpretation of the Protagoras, Devereux argues, Socrates’ position in the Laches would remain incompatible. In the Protagoras, each of the five virtues is given an equal status, Footnote ¹⁴ and they are all somehow associated with knowledge. In the Laches, however, Socrates treats four of the virtues (justice, courage, temperance, and piety) equally as parts of a whole, while wisdom is said to be the elite virtue that constitutes the whole itself. This is precisely what leads Devereux (Reference Devereux1992, 788) to conclude that, in the Laches, ‘Plato … articulates a more coherent doctrine … he drops the claim that the virtues are identical … and develops and refines the idea that the virtues are parts of a whole, and that wisdom is the key to their unity.’

5. Laches and the psychological search

If the Devereux discrepancy can be explained away, Devereux’s conclusion will not be required. My suggestion is that the Laches features a Socratic endorsement of both (1) and (2); Socrates is committed to both the identity view and the biconditional view. In the lengthy introduction of the Laches, Socrates joins a conversation about the correct way to train, teach, and raise young men. He then motivates the question of how virtue can be imparted (παραγίγνεσθαι). This is a fitting prelude to the general’s question. In fact, once the primary question (what is courage?) has been posed at 190d, Socrates narrows the investigation by providing an example. According to Socrates, quickness is ‘the capacity (δύναμις) to get a great deal done in a short amount of time.’ He then summons Laches to provide a similar answer for courage:

Socrates makes it perfectly clear that they are seeking a capacity (δύναμις) responsible for making the courageous person courageous. Moments later, this is described as a natural quality (πεφυκὸς) in the soul (192b), and also as the noble thing (πρᾶγμα) designated by the name ‘courage’ (192d). Socrates never once asks for the essence (οὐσία) of courage in the Laches. In this respect, the investigation resembles the Protagoras.Footnote ¹⁵

But in the Protagoras, Socrates argued that the virtues are one and the same thing. Thus, he appeared to abandon the notion that the virtues are distinct parts of a whole. In the Laches (L1), he endorses this notion. In fact, when Nicias proposes the identity relation (195c) he had once heard from Socrates (namely, that courage and wisdom are the same thing), this allegedly Socratic account of virtue is met with strong resistance from Laches, who says ‘surely wisdom is distinct (χωρίς) from courage.’ And curiously, Socrates appears sympathetic to Laches’ concern:

It appears that Socrates’ interlocutors must either abandon (a) the notion that courage and wisdom are distinct parts of virtue, or (b) the account of courage as ‘knowledge of what is to be feared and dared, ’ which is identical to the knowledge of good and bad (the whole of virtue). These two options create an apparent dilemma. The dilemma is compelling enough to generate the dialogue’s aporetic ending. But the dilemma is false.Footnote ¹⁶ The two options are in fact compatible, and the Socratic solution is to accept both (a) and (b).

To retrace this solution in the dialogues, we must return to (P), where it was originally implied that the virtue-names have distinct essences (οὐσίαι), but where Socrates was also defending the identity view, which says that the virtues refer to a single capacity (δύναμις). Now we find that the Laches also features a Socratic identity statement together with a Socratic endorsement of the notion that the virtues are distinct parts of a single whole. My suggestion is that Socrates retains the identity view at the psychological level, while holding that the virtues are definitionally distinct, each possessing a distinct essence (οὐσία).Footnote ¹⁷ In order to support this interpretation, it will be necessary to explore the notion of essence (οὐσία), and to clarify the conceptual search, which I take to be a search for definition.

6. Euthyphro and the conceptual search

It is not uncommon for Plato scholars to describe Socrates’ preoccupation with the primary question as a search for definition. Even Aristotle understood the question this way:

Socrates’ preoccupation with definition is most prominent in the so-called ‘dialogues of definition.’ As Charles Kahn has pointed out, there are three dialogues in particular, Laches-Euthyphro-Meno, which form ‘a unified exposition on the logic of definition.’Footnote ¹⁸ About this, there seems to be general agreement among scholars. Still, some have resisted the temptation to speak in terms of definition. Hugh Benson, for instance, is quite cautious:

I sympathize with Benson’s caution. But I also believe that Plato provides a unified exposition on the logic of definition, and that a careful examination will reveal those features which, for Plato, are necessary for an adequate definition.Footnote ¹⁹

The Euthyphro contains a search for definition. It exhibits an interlocutor who inquires with Socrates at the conceptual level, rather than the psychological level. The primary question (what is piety?) is introduced at 5d, arising in response to Euthyphro’s profession of knowledge. Euthyphro thinks he has expert knowledge of divine matters; and thus, he is unafraid that, in prosecuting his father for murder, he is doing wrong. But when Euthyphro provides his first answer that piety is ‘what [he is] doing now, ’ Socrates responds promptly, clarifying the object of inquiry:

Socrates is asking for the essential quality (εἶδος) common to all pious things. He is seeking the distinctive feature of both pious acts and pious persons (7a).Footnote ²⁰ Thus, the inquiry takes a different description than that of the Laches and Protagoras, where Socrates had sought the particular capacity (δύναμις) of the virtuous person, and where Socrates never once asked for the quality (εἶδος) unique to courageous things in general. Nor was Socrates seeking there (in the Laches - Protagoras) to establish a model (παραδείγματι) upon which to fix the eye, as he does here (E1). Conversely, here in the Euthyphro, the language of the psychological search is nowhere to be found: Socrates does not request a thing (πρᾶγμα) in the world, Footnote ²¹ nor does he request the underlying capacity (δύναμις) of the pious person, nor does he request a natural quality (πεφυκὸς) in the soul. Bearing these facts in mind, it appears that Socrates is requesting something very different in the Euthyphro. I contend that he is requesting a definition, and seeking to clarify the concept of piety.

Socrates usually rejects an answer to the primary question by showing that it is either too broad or too narrow. He wants to capture all and only instances of the virtue. This is true of the Euthyphro most of all. But why should we think that Socrates is requesting a definition? I suggest that we should, because Plato tells us to. Upon hearing Euthyphro’s second answer, that piety is ‘what is dear to the gods, ’ Socrates commends his interlocutor for his vast improvement.Footnote ²² He then explains that piety cannot be what is dear to the gods, since the gods are often in disagreement. And here, Plato further informs his reader about the object of inquiry.

The verb ὁρίζω (or ὁρίζεσθαι) serves as a convenient verb for definition – to mark out the boundaries of a word or concept. Also, when referring to these boundaries, Plato occasionally uses the term ὅρος.Footnote ²³ Aristotle would later coin his own technical term for definition, ὁρισμός, from these words, and he would do so in a way that links the notion of essence to that of definition. In Topics, Aristotle explains that ‘a definition (ὁρισμός) is an account (λόγος) that signifies an essence (οὐσία).’Footnote ²⁴ In (E2), Plato makes it clear that he too will sometimes use the term λόγος to denote the definiens. And after amending the definition so that piety is ‘what all the gods love, ’ Plato puts the verb of definition (ἠμῖν ὡρίσθαι) back into Socrates’ mouth: ‘Do you wish this now to be our definition (ἠμῖν ὡρίσθαι) of piety and impiety?’Footnote ²⁵

As Benson notes, appealing to definitions pushes the question back, so that it becomes necessary to determine what constitutes an adequate definition; but before we declare this a problem, let us decide whether a theory of definition emerges. We have witnessed that, for Socrates, an adequate definition will reveal the common quality (εἶδος) belonging to all and only instances of the virtue being defined. The definition must also serve as a standard (παραδείγμα) for us to look toward (ἀποβλέπτειν) as we judge whether or not something is virtuous. But this cannot be the whole story; for there is reason to think the present definition of piety (that which all the gods love) satisfies these criteria. It appears to preserve co-extension at any rate, since it is not criticized for being too broad or too narrow.Footnote ²⁶ And yet, despite this, Socrates rejects the definition.

For Socrates, an adequate definition will make clear the essence (οὐσία) of the thing being defined. Unfortunately, Euthyphro’s definition fails to make clear the essence of piety; instead, it merely clarifies how the gods feel about piety. Euthyphro has offered a mere pathos. To make clear the essence of piety, his definition must articulate the unique characteristic responsible for making something an instance of piety. In other words, an adequate definition of F must state the explanatory cause of something’s being F.

Now it might be tempting to conclude, as Ferejohn (Reference Ferejohn1984) does, that the characteristic being sought here is just the capacity (δύναμις) of the virtuous personFootnote ²⁷ so that, when all is said and done, there is only one object being sought after all. In other words, the essence of any individual virtue just is the capacity that underlies all the virtues. This interpretation would unify the primary question and restore the traditional notion that the Laches and Euthyphro are investigating virtue at the same level. But the single capacity (δύναμις) underlying virtue, as an object of inquiry, doesn’t fit the profile we have been given in the Euthyphro. Insofar as the capacity (δύναμις) constitutes the whole of virtue, it won’t provide an adequate model upon which to look (ἀποβλέπειν) when determining whether an act is pious. In other words, the capacity will not distinguish pious acts from courageous acts, nor courageous acts from just acts, and so on. As we have seen (E1), this is an explicit criterion of Socratic definition. It stands to reason that the essence (οὐσία) of piety will be something unique to piety itself, as a distinct part of virtue.Footnote ²⁸

At this point, one might suggest that there are distinct capacities (δυνάμεις) underlying each of the individual virtue-names. But this is also problematic. Socrates denies this position in the Protagoras. Furthermore, we must take into account the extent to which the language of investigation in the Euthyphro differs from the language of investigation in the Laches and the Protagoras. In those dialogues, Socrates openly examined the capacity (δύναμις) of the virtuous person. In the Euthyphro, Socrates makes no mention of capacities at all. Instead, he examines the essence (οὐσία) of piety. Those who take the Euthyphro and Laches to be conducting parallel searches will be hard-pressed to explain the omission of the term δύναμις from the Euthyphro, when this term is used frequently to narrow the search for courage in the Laches. Such commentators will also be hard-pressed to explain the omission of the term οὐσία from the Laches, when this term is used to narrow the search for piety in the Euthyphro. The presence of two distinct levels of inquiry (conceptual and psychological) will provide the best explanation of these facts.Footnote ²⁹

These terminological differences have been a topic of examination for David Wolfsdorf (Reference Wolfsdorf2005), who makes a similar distinction among definitional dialogues.Footnote ³⁰ According to Wolfsdorf, the ‘what is x?’ question is a request for the referent of ‘x.’ As he puts it, the referent being sought in Laches and Protagoras is a psychological capacity (δύναμις), while the referent being sought in Euthyphro and Meno is a metaphysical Form (οὐσία). Unfortunately, if we accept Wolfsdorf’s claim that the search for οὐσία is a metaphysical search for the Form of virtue, it will be difficult to make sense of the part-whole relationship that obtains between them. Wolfsdorf himself never explains this relation. I have suggested instead that we understand the search for essence (οὐσία) as a conceptual search for definition.Footnote ³¹ The passage above (E3) provides important information about the search for an essence. In traditional logic, we distinguish between the extension of a concept and the intension of a concept. The ‘extension’ indicates the set of objects picked out by the concept, whereas the ‘intension’ indicates the internal description under which it picks them out. In requesting the essence (οὐσία) of piety, Socrates is requesting something like the intension of ‘piety.’ According to Socrates, even if all and only pious things are loved by all the gods, this alone would not express the essential nature (οὐσία) of piety, because the definition (λόγος) does not state the explanatory cause, and being loved by all the gods is not intensionally equivalent to ‘piety.’ It may be helpful here to borrow an example from Quine, as some commentators have been inclined to do.Footnote ³² Even if all and only those creatures with a heart are creatures with a kidney, it still does not follow, and is indeed false to suppose that ‘creature with a kidney’ means the same thing as, or defines ‘creature with a heart.’ We learn from (E3) that an adequate definition for Plato requires something like an intensional equivalence between definiendum and definiens. This is an integral part of the process of ‘making clear the essence’ of piety.Footnote ³³ The definition (λόγος)) of F must capture the conceptual essence of F, by expressing its intensional content, or by stating the characteristic that uniquely qualifies something as an instance of F. Footnote ³⁴ A definition of this kind will serve as a standard by which to judge whether or not something is F.

7. Virtues as definitionally distinct

Socrates occasionally describes each virtue as a distinct part (μέρος) of a single whole. This has puzzled commentators who aim to explain the unity thesis. I suggest that the part-whole relation makes the most sense within the two-level account, where it fits nicely into the framework of the conceptual search. On my interpretation, Socrates considers the virtues (wisdom, justice, courage, temperance, and piety) to be conceptually, or definitionally distinct. The part-whole relation simply refers to a definitional overlap – the meaning of ‘virtue’ will be broad enough to encompass the various meanings of the individual virtue-names. And so, the fact that Socrates now suggests that piety is a ‘part of justice’ should not come as an unpleasant surprise.

Socrates is proposing a part-whole relation between justice and piety, and the suggestion occurs at the conceptual level. In the Euthyphro, there is no counteracting effort to unify the virtues into a single thing (δύναμις), as there was in the Laches. Piety and justice are conceptually distinct. Yet, it is clear that the definition of justice will somehow encompass the definition of piety, just as the definition of virtue will encompass the definitions of each of the individual virtues. And so, it is suggested here that ‘justice’ casts a wider net than ‘piety, ’ just as it was suggested previously (L1) that ‘virtue’ casts a wider net than ‘courage.’

Of course, the commentary above may seem unsophisticated in light of Socrates’ analogy at 12d-e, where Socrates mentions ‘number’ and ‘odd’ to illustrate the overlapping relation between justice and piety. We should notice that ‘number’ and ‘odd, ’ in addition to having different intensions, also have different extensions (which means that some instances of justice are not instances of piety). This may appear to threaten the biconditional view, as we can no longer contend that something is just if and only if it is pious. Yet, the example is not as threatening as it appears. As Brickhouse and Smith have observed, a perfectly good explanation is made available by the text.Footnote ³⁵ Socrates initially requested the quality (εἶδος) distinctive of both pious acts and pious persons (7a), so that the definition might serve as a model by which to identify pious things in general. When it comes to persons, the virtue-names have the same extension (δύναμις), which is all biconditionality requires. Yet, when it comes to actions, they certainly do not. The virtue-names manage to pick out different actions by means of their different intensions (so that some just acts are not pious). If we suppose that ‘justice’ connotes something which expresses the right relation to others, and we suppose that ‘piety’ connotes something which expresses the right relation to the gods, then we can see how this part-whole relation might hold at the conceptual level, insofar as the ‘gods’ are considered a subset of ‘others.’Footnote ³⁶ But the mere fact that the meanings of these terms are related in this way does not imply that their psychological states must also be distinct. The virtue-terms may still refer to one and the same psychological state (δύναμις), the expert knowledge of good and bad. This is where the biconditional interpretation proves correct, since the conceptually distinct virtues are inseparable within the agent.

But this gives rise to another question. If indeed the virtues are identical with respect to the underlying psychological state, what explanation is there for their definitional variance? How exactly do the definitions vary?Footnote ³⁷ A plausible suggestion comes from Brickhouse and Smith (Reference Brickhouse and Smith1997, Reference Brickhouse and Smith2010). They explain that a single expertise can have different applications, and that these applications can be distinguished conceptually by their various results (ἔργα). In the early dialogues, when a virtue has been identified as a kind of scientific knowledge (ἐπιστήμη), Socrates will typically inquire into the result (ἔργον), as he does here (12e): ‘What result (τινος ἔργου) would the science [piety], which provides service (θεραπίαν) to the gods, ultimately accomplish.’ Sciences are distinguished from one another by reference to an object (the subject matter) and by reference to a result (the product). Medicine, for instance, concerns the body and produces health; carpentry concerns building-materials, and produces buildings. According to Brickhouse and Smith, the knowledge of good and bad (qua psychological state) will have different applications. These applications are differentiated conceptually by their results (ἔργα). On this interpretation, virtue as a whole produces benefit. Justice produces benefit to others, piety produces benefit to the gods, and so on. Unfortunately, this interpretation runs into problems. There is no indication that a single science will produce multifarious results. On the contrary, there is one clear indication that piety will not produce benefit to the gods. At 13a-b, the suggestion is abandoned under Socrates’ influence.Footnote ³⁸ Therefore, although I am in agreement with Brickhouse and Smith that the virtues are unified by a single psychological state (δύναμις), I cannot accept the claim that the virtues are distinguished conceptually by their individual results (ἔργα).

In response, Brickhouse and Smith may contend that the search for piety breaks down (13e 12) precisely where Euthyphro fails to identify the result of piety. In their most recent work (Reference Brickhouse and Smith2010, 159–167), they suggest that Socrates actually leads Euthyphro into thinking about piety as a part of justice in terms of ‘some distinctive result that [piety] produces.’ For Brickhouse and Smith, piety will be distinguished from justice by means of its distinctive result (ἔργον). In my estimation, this is where Brickhouse and Smith have gone wrong. It is one thing to distinguish the virtues by means of identifying different applications of a single knowledge; it is quite another thing to distinguish the individual virtues by their distinctive results. A close examination of the Euthyphro will show that, although the investigation does break down where Euthyphro fails to identify the result of piety, this ‘break down’ is not a failure to distinguish piety as a part of justice. Euthyphro has already given Socrates a satisfactory answer about piety as a part of justice. At 12e 5, Euthyphro answers that piety is ‘the part of justice concerned specifically with the service to the gods [rather than human beings].’ Socrates applauds Euthyphro’s answer (at 12e 8), saying the answer is ‘very well-put!’ From this moment forward, Socrates no longer questions Euthyphro about piety as a part of justice. This confirms that Socrates is satisfied with the answer Euthyphro has given.

Of course, Socrates does request one more small thing from Euthyphro. This ‘small thing’ concerns the ‘all-noble result’ of piety (πάγκαλον ἔργον).Footnote ³⁹ But since the all-noble result is never discovered, there is nothing in the text that rules out the possibility that piety and justice will produce the same result. What we discover is that piety will not produce a benefit to the gods (13b). Thus, we are led to consider who receives the benefit; and the most plausible answer is that piety (like justice) results in a benefit to human beings.Footnote ⁴⁰ It not unreasonable to suppose that piety and justice produce the very same result.

Even if piety and justice do produce the same result, they will be distinguishable already, simply in virtue of the relational category Euthyphro has outlined in his response at 12e 5. A unique result (for each of the putative virtues) is not exactly needed to fulfill the role of distinguishing them. In fact, we may have good reason to be skeptical of the idea that the virtues produce distinctive results (ἔργα). In Book 5 of the Republic, Plato offers an account of how different capacities (δυνάμεις) should be distinguished. At 477c1-d5, Socrates says ‘that which relates to a different thing and produces a different thing, I call a different δύναμις.’ If this is how Socrates distinguishes between different capacities (δυνάμεις), we have stumbled upon something important. For Brickhouse and Smith describe each virtue as having a distinct range of application together with a distinctive result. By doing so, I believe they are in danger of describing distinct capacities. This would undermine their own account of what unifies the virtues. We must take care, it seems, not to do too much to distinguish the virtues, lest we obliterate the thing that unifies them.

And so, let us now explore a related solution. In human life, there are indeed a number of relational categories toward which the knowledge of good and bad (qua psychological state) might be applied. The definitions of the individual virtue-names may simply outline these relational categories or domains. Justice, for instance, does involve one’s relationship to others, as temperance involves one’s relationship to self, and piety does involve one’s relationship to the gods. Courage, it seems, involves one’s relationship to an uncertain future. That being said, we should pay special attention to courage in this context, since the only definition privileged enough to bear the ‘Socratic’ label in the early dialogues happens to be the definition of courage as ‘knowledge of what is to be feared and dared.’ This definition surfaces both in the Laches (195a) and in the Protagoras (360c-d). Of course, the capacity (δύναμις) of the virtuous person is being sought in both dialogues. For this reason, the ‘knowledge of what is to be feared and dared’ is immediately equated with the knowledge of good and bad (the whole of virtue). Had Socrates been investigating the essence (οὐσία) of courage instead, the definition may have proven acceptable.Footnote ⁴¹ At any rate, the definition of courage constitutes the most promising definition of any virtue to be found in the early dialogues. If we follow this model faithfully, the definitions of the individual virtues may diverge simply because they outline different relational categories to which the knowledge of good and bad can be applied.Footnote ⁴² Since there is no independent result or accomplishment (ἔργον) mentioned within the context of this definition, and since no distinctive result is ever discovered for any of the putative virtues, I suggest that the virtues all share a single result (namely, εὐδαιμονία). This seems to be the implication of certain passages in the Euthydemus and Charmides. Footnote ⁴³ The knowledge of good and bad, as a single capacity in the soul, is said to result in the production of good and happy human lives. I suggest, therefore, that eudaimonia is the shared result of the virtues, and that the ‘all-noble result’ of piety (as with any of the virtues) is the production of good and happy human lives.

But regardless of how we individuate the virtues, in the end, we learn that there is potential for reconciling the claim that the virtues are definitionally distinct parts (the biconditional view) with the claim that they are all one and the same psychological thing (the identity view). In the next section, we will turn to the Meno. The Meno offers compelling evidence that our two levels of inquiry are in fact distinct.

8. Meno and the two distinct levels

In the opening lines of the Meno, Socrates is asked whether virtue can be taught:

Meno’s question is connected to the psychological search – it is the same question that generated the unity thesis in the Protagoras. Footnote ⁴⁴ The Laches too arose from the question of whether virtue can be taught.Footnote ⁴⁵ In these dialogues, the question was a request for the psychological state of the virtuous person. The same holds for (M1), where Meno wants to discover how the psychological state of virtue comes to be present. Yet, Socrates’ response (71a-b) is important:

Socrates is distinguishing between two separate questions (again), and his response implies that one must be answered before the other.Footnote ⁴⁶ This has come to be known as the Priority of Definition: Socrates is insisting that the first question what virtue is (τί) is epistemically prior to the second question the sort of thing virtue is (ὁποῖος). I will argue that the second question (ὁποῖος) concerns the psychological state (δύναμις) of the virtuous person, as evidenced by (M1), where Meno sought to know how the psychological state arises. The prior question (τί) concerns the definition (οὐσία) of virtue.Footnote ⁴⁷ For, despite (M1), the initial investigation of the Meno is heavily decorated with the vocabulary of the conceptual search. According to Socrates, an answer to the first question (τί) must pick out all and only those things that are virtuous (73d); it must make clear (δηλῶσαι) the essence of virtue (72b-c); and it must serve as a standard (παραδείγμα) by which to judge whether an action is virtuous (72c). The object of inquiry is finally stated at 72a-c.

Commentators agree that the first section (70–77) of the Meno resembles a dialogue of definition.Footnote ⁴⁸ The vocabulary echoes the search for piety in the Euthyphro. That being said, Meno has a very different agenda, as his opening lines indicate. He intends to inquire at the psychological level, to develop a psychological theory.

Socrates is perfectly aware of Meno’s agenda. When Meno fails to offer an adequate definition, Socrates issues three types of answers that could be given in response to the ‘what is x?’ question. The first two examples answer the question ‘what is shape?’ while third example answers the question ‘what is color?’

• (1) 75b10: ‘shape is that which, alone of the things that are, always accompanies color.’

• (2) 76a: ‘shape is the limit of a solid.’

• (3) 76d: ‘color is an effluence from shapes which is commensurate with sight and perceptible.’

The examples appear to offer rare insight into the type of answer Socrates wants from the primary question. Unfortunately, the relations between these examples, and their precise Socratic rankings, have been the subject of controversy. For this reason, interpreters have suggested that Socrates (and perhaps Plato too) had no clear idea of the type of answer the primary question required.Footnote ⁴⁹ I think this suggestion is incorrect. My own opinion is that the examples above provide valuable insight into the relation between our two distinct levels of inquiry. They speak to the epistemic priority of the conceptual level over the psychological. In order to support this reading, I will examine the examples one-by-one, and provide an account of their precise Socratic rankings.

According to the first example (75b10), ‘shape is that which alone, of the things that are, always accompanies color.’ Socrates says he would be pleased if Meno could describe virtue in this manner. But Meno rejects the first definition because it uses a term that has not yet been discussed (‘color’). Moments later, Socrates indicates (79d) that the criticism is mutual, saying ‘if you remember [Meno] … we rejected the sort of answer that tries to give an answer in terms of things that are still being searched for, and have not yet been agreed upon.’ But the first example is open to another (specifically Socratic) criticism. According to the example, color will always be accompanied by shape. And while this appears to leave open the possibility of having shape without color, we know that Socrates would not offer a definition that was too broad or too narrow. We must recognize that the example was meant to capture all and only instances of ‘shape, ’ so that the terms are co-extensive.Footnote ⁵⁰ Yet, if definiendum and definiens are merely co-extensive, the definition will fail to capture the essence (οὐσία) of shape. Recall the Euthyphro. Socrates rejected the definition of piety (‘what all the gods love’) because it failed to capture the essence of piety. Euthyphro had offered a mere pathos. The same criticism applies to our first example, which fails to capture the essence of shape. Despite preserving co-extension, it fails to state the characteristic that uniquely qualifies something as an instance of shape.Footnote ⁵¹

According to the second example (76a), ‘shape is the limit of a solid.’ Socrates carefully secures the meaning of the relevant terms (limit, plane, solid) before offering this overtly geometrical example, which follows the standard pattern of a definition by genus and differentia, and is explicitly endorsed by Socrates, who later indicates (at 76e) that it is somehow better than the third and final example. In fact, the second example appears to satisfy Socrates’ criterion for definition. There is no reason to think that it fails to capture the conceptual essence (οὐσία) of ‘shape.’ In contrast to the first example, Meno offers no criticism. He appears to accept the definition along with Socrates. Yet, he also appears curiously underwhelmed by the geometrical definition. He openly gravitates toward the third example instead. A careful reader will wonder why Meno favors the third example over the second, despite lacking a criticism of the second.Footnote ⁵² It is here that the ‘two-levels’ interpretation yields a most compelling explanation.

Prior to the third example, Meno confronts Socrates. He demands yet another type of answer to the ‘what is x?’ question. We have observed that Meno, from the very beginning, has expressed more of an interest in a psychological theory of virtue than a definition. Socrates apparently recognizes this, and responds appropriately (76a-b).

Socrates is aware that Meno wants an account of the natural thing (πρᾶγμα) of virtue. Thus, at 76d, Socrates reluctantly agrees to indulge Meno ‘in the manner of Gorgias, ’ giving his interlocutor the kind of answer he wants.

According to the third and final example, ‘color is an effluence from shapes which is commensurate with sight and perceptible.’ Needless to say, Meno finds the answer completely satisfactory (76d), while Socrates has reservations. We should notice that the answer is based on Empedocles’ theory of vision: Meno’s preferred answer fits within the framework of a scientific theory related to a particular capacity (δύναμις), the capacity of the human eye. Socrates does not reject the third example. But when Meno exclaims that it is ‘excellently put!’ Socrates says the second example is better. According to Socrates, Meno would prefer the second example too, if he would remain long enough to hear about it. But Meno won’t be converted, as he makes perfectly clear (77a) ‘I would stay, Socrates, if you would give me many such answers.’

These examples (2 and 3) therefore expose a methodological disagreement between Socrates and Meno concerning the examination of virtue. Socrates’ model-answer (example 2) is framed in accordance with the conceptual level of inquiry, and is meant to capture the essence (οὐσία) of shape; Meno’s model-answer (example 3) is framed in accordance with the psychological level, and is meant to offer a scientific explanation of the natural phenomenon (πρᾶγμα) color, which involves a certain capacity (δύναμις). By my lights, Socrates endorses the second example for one reason: the conceptual search is epistemically prior to the psychological search.

As we shall see, this reading is supported by the dialogue’s second portion (77–100) no less than by its first (70–77). As Meno persists, a visible shift occurs, from the vocabulary of the conceptual search to the vocabulary of the psychological search. At 77b, Socrates attempts to revive the conceptual search for definition, and instructs his interlocutor to trace the boundaries of virtue ‘in the pattern … just now received.’ Meno’s answer is both defiant and informative.

This answer is conveniently summarized (78c) as ‘the ability (δύναμις) to procure good things.’ In this way, Meno conforms his answer to his own preference (example 3), Footnote ⁵³ rather than following the pattern of Socrates’ model-answer (example 2). It is noteworthy that Meno’s answer resembles the Socratic account of quickness from Laches, where Socrates openly engaged the psychological search.Footnote ⁵⁴ But Socrates is not seeking this type of answer in the Meno. He is seeking an answer to the conceptual search, which requires a definition. For this reason, Socrates reconstructs Meno’s answer (78c5), omitting the psychological term so that virtue is ‘being of the sort to acquire good things.’ Once Meno’s list of good things is exposed (health, wealth, status) it becomes clear that such things can be acquired either justly or unjustly. Meno is forced to agree that the just acquisition of good things is required, which allows Socrates to reject the definition on grounds that it defines the whole of virtue, using a part of virtue.

As Socrates pushes the conceptual search, it is suitable that he should exploit the part-whole relation. For, as it turns out, Meno cannot find his way through the conceptual landscape. At 79b, the primary question (what is virtue?) is posed once again. Meno must start the conceptual search over ‘from the top.’

Perhaps out of frustration, Meno responds with his famous paradox. He asks how Socrates can even look for virtue, when he does not know at all what virtue is. (If Socrates already knows what virtue is, he need not look for it. But if he does not know what virtue is, he won’t recognize a virtue once he encounters it). Socrates’ answer involves the theory of recollection, which I do not have space to discuss.Footnote ⁵⁵ I will confine myself to the discussion that follows the demonstration (82b-85b) using the slave boy, a demonstration that can be summarized with one key insight: knowledge of what something is (τί) can be attained, like geometrical knowledge, by searching one’s own inner resources.Footnote ⁵⁶

Having once labored to convince Meno that it is worthwhile to continue the search for definition (85b- 86d), Socrates starts the conceptual search again, but only to find more resistance:

Predictably, the Priority of Definition is presented again. Socrates informs Meno that he really ought to inquire into the essence of virtue before considering whether virtue is the sort of thing to be taught. The persistence of Meno prevails.

Now, in order to examine the sort of thing virtue is, they must bypass the prerequisite definition, and proceed by use of a hypothesis (ὐπόθεσις). From this moment, the language of investigation shifts entirely from the language of the dialogue’s first section (70–75) which featured the vocabulary of the conceptual search, to that of the dialogue’s second section (75–100) which features the vocabulary of the psychological search. Commentators have failed altogether to notice this shift. Even Wolfsdorf (Reference Wolfsdorf2005), who examines the subtle difference in terminology among definitional dialogues, fails to notice the shift in terminology here in the Meno.Footnote ⁵⁷ The shift begins with Socrates’ three examples, which exposed a methodological disagreement between Socrates and Meno. Prior to these examples, the terms πρᾶγμα and δύναμις simply do not occur. After the examples, the terms εἰδός and οὐσία simply do not occur.Footnote ⁵⁸ A few passages will further illustrate this point.

In this passage, the second question of the Meno is permanently affixed to the psychological search. The new investigation will concern a psychological reality (περὶ τὴν ψυχὴν ὄντων). Moments later (89d), Socrates ties this reality to the technical term, πρᾶγμα, and proceeds to solidify the connection (96a) through repetition.Footnote ⁵⁹

This confirms that when Socrates previously agreed to switch from the prior question what virtue is (τί) to the secondary question the sort of thing virtue is (οποῖον), he thereby moved from the conceptual search for essence (οὐσία) to the search for a psychological state (πρᾶγμα). By keeping essences and psychological states distinct, and by emphasizing the priority of the conceptual level over the psychological level, the Meno vividly reconfirms the presence of two distinct levels of inquiry, and provides important information about the relationship between them. Finally, the investigation in Euthyphro is clearly concerned with the essence (οὐσία) of piety, while the investigations of the Laches and Protagoras are concerned with the psychological state (δύναμις) of the virtuous person. Thus, there is clear evidence that these two levels have been kept distinct throughout the dialogues of definition. Understanding the primary question as a springboard for two distinct searches manages to relieve a great deal of tension in the early dialogues, and it introduces a compelling new approach to Socratic doctrine. I have argued that this interpretation is essential to understanding the unity thesis.