¹ Epicurus included, on what seems to me the most reasonable expansion of the meager evidence we have about his views about virtue and the virtues; for discussion, see Mitsis, Phillip, Epicurus' Ethical Theory (Ithaca, NY: Cornell University Press, 1988), pp. 75–76.Google Scholar

² Julia Annas gives a good account of how the unity of virtue might progressively have been arrived at, or defended, in post-Platonic philosophy (Aristotle and the Hellenistic schools down to Antiochus), on the basis of such a bottom-up way of thinking; see Annas, , The Morality of Happiness (New York and Oxford: Oxford University Press, 1993), pp. 73–84.Google Scholar

³ See Plato, , Protagoras, 330a7–b2, 329e2–6.Google Scholar

⁴ See ibid., 331e4–6, 359b6–7, 361a6–b3.

⁵ I.e., the character in Plato's dialogue named after him. Throughout, in referring to Protagoras and discussing his views, as also in what I say about Socrates, I am talking about Plato's characters of those names. I do not intend to be making any claims at all about the historical personages. I leave such historical questions entirely to one side.

⁶ In preparing at 329c2–6 to raise this question, Socrates virtually quotes Protagoras's own remarks at 325a1–2: compare ὡς ἕν τι εἴη συλλήβδην ⋯ρετή at 329c5–6 (justice, etc., “were somehow collectively one thing: virtue”) with κα⋯ συλλήβδην ἓν αὐτ⋯ προσαγορεύω εναι ⋯νδρ⋯ς ⋯ρετήν at 325a1–2 (“what I may collectively term the virtue of a man”). It is Protagoras's own formulations that provoke Socrates' confusion and his subsequent line of questioning.

⁷ Plato, , Protagoras, trans. Lombardo, S. and Bell, K., in Cooper, John M., ed., Plato: Complete Works (Indianapolis and Cambridge, MA: Hackett Publishing Company, 1997).Google Scholar

⁸ For the latter expression, see Vlastos, Gregory, “The Unity of the Virtues in the Protagoras,” in his Platonic Studies (Princeton: Princeton University Press, 1973), pp. 221–65.Google Scholar

⁹ Or instead, if you like, those Platonic Forms.

¹⁰ Thus, Vlastos's effort (in “The Unity of the Virtues,” pp. 234ff.) to reduce Socrates' thesis of the unity of virtue in the Protagoras to a misleading way simply of affirming “biconditionality”—that x is just if and only if x is temperate, and so on—fails. Some psychological and/or metaphysical truth about virtue itself, and the virtues themselves, must underlie and ground any necessary co-instantiation of them. It is this underlying truth that has to be explained and made out in some way, if Socrates' own position in the debate with Protagoras is to be understood.

¹¹ The next argument (Protagoras, 332a4–333b6), concerning temperance and wisdom, does not attempt to establish that these are “such as” one another, but instead that each is directly identical with the other. So it is not a continuation of the line of argument announced at 330b6–7. Having abandoned the first line of argument without completing it, Socrates now shifts to showing, pairwise, as he has now done with justice and piety, that temperance and wisdom, and then justice and temperance (333b7–334c6), and finally courage and wisdom, are related to one another by some transitive relation strong enough to undermine Protagoras's claim of distinctness and difference; when the separate arguments are put together, he will then have established that the same relation relates each one to each of the others.

¹² In formulating this account of what “justice is just” means here, I have been helped by unpublished work by Sean Kelsey on Plato's Phaedo.

¹³ Thus, I do not think that at 331a7–b1 Socrates hoodwinks Protagoras by making him uncomfortable with the idea that justice should be accounted something impious, and scaring him off the true response that justice is not pious (because it is simply neither pious nor impious—not the sort of thing to be either of these things), by illegitimately implying that if it is not pious, then it must also be impious. In fact, if talk of justice's being pious or not is understood in the right way, in accordance with what the context establishes for its interpretation, it is true, as my example shows, that if justice is not pious it must also be impious: if it causes a person to act otherwise than piously in a situation calling for pious action, it causes them also to act impiously.

¹⁴ Thus, to say that justice is pious is to say that justice is by its nature a power to cause pious actions, and never a power to cause impious actions. (To say that piety is pious means something connected but slightly different: it is the power that always causes pious actions and outcomes whenever it operates and causes anything at all.)

¹⁵ Plutarch's On Moral Virtue is available in Greek and English in Plutarch's Moralia, vol. 6, ed. Helmbold, W. C. (Cambridge, MA: Loeb Classical Library, Harvard University Press, 1939).Google Scholar Plutarch, a Platonist philosopher, lived from about 50 to about 120 a.d.

¹⁶ The Greek at Protagoras 329d6 is τ⋯ το⋯ χρυσο⋯ μόρια, literally “the parts of gold”—presumably (all) the parts of the totality of gold that there is. This is often or usually translated and understood as several “parts” or “pieces” of gold taken together as a single group, making up a single pile or a single lump of gold. It is possible that the Greek does mean the latter, but the other alternative, I think the more natural way of taking the Greek, should be borne in mind as well; when developed as I think Ariston of Chios developed it (see below), it gives a better account of the unity and correlated plurality of the virtues than the “pieces of gold” reading does.

¹⁷ I agree here up to a point with what Michael Frede says in his introduction to the Hackett edition of the Protagoras (Indianapolis and Cambridge, MA: Hackett Publishing, 1992), pp. xxiii–xxiv: only two real options are considered in the Protagoras, either the identity (in some sense) of all the virtues with wisdom or knowledge of good and bad, or the view that virtue is a whole of disparate parts. But (see below) I do not think the “gold” alternative is in itself vacuous; it looks to me as if Ariston of Chios was trying to make something distinctive of it. Vlastos, Contrast, “The Unity of the Virtues,” p. 225Google Scholar, who takes this option to represent (part of) Socrates' own preference.

¹⁸ Zeno of Citium, the founder of the Stoic school of philosophy, was active in Athens from the last decade of the fourth century to his death in about 262 b.c. Chrysippus was head of the school from about 232 to about 206 b.c.

¹⁹ Laertius, Diogenes, Lives of Eminent Philosophers, VII 161, ed. and trans. Hicks, R. D. (Cambridge, MA: Loeb Classical Library, Harvard University Press, 1925)Google Scholar; Galen, , On the Doctrines of Hippocrates and Plato, ed. and trans. DeLacy, Phillip (Berlin: Akademie-Verlag, 1978), VII 2.Google Scholar Diogenes Laertius is thought to have written in the second or third century a.d.; Galen lived from about 129 to 199 a.d.

²⁰ Perhaps this is why Frede describes it as a “vacuous” option (introduction to Protagoras, p. xxiii).

²¹ On this, see Galen, , On the Doctrines of Hippocrates and PlatoGoogle Scholar, VII 2 (= von Arnim, J., Stoicorum Veterum Fragmenta [Leipzig: Teubner, 1903–1905Google Scholar, with reprints by other publishers], III 373).

²² In translating Plutarch, and in my explication of Ariston's view, I have allowed Ariston to be thinking of the single thing that is virtue as always and continuously entitled to all its more particular names—whether it is currently exercising itself in the specific relation appropriate to that particular virtue-name or not. After all, sight remains light-sight, whether or not it happens at the moment to be grasping light colors, simply because it retains the capacity for, and orientation toward, grasping them, just as sight itself is retained even when one's eyes are closed. And similarly for the knife as salami-cutter and fire as air-heating. However, other sources (Galen, in Stoicorum Veterum Fragmenta, III 374Google Scholar; Clement of Alexandria in ibid., III 376) are explicit in saying that according to Ariston virtue is justice (only) when it “assigns to each his merited share.” (Clement introduces the example of a drachma-piece that gets called “the fare” when given to the shipmaster, but “rent” or “taxes” when used in other ways.) So construed, Ariston is indeed open to the objection which Long and Sedley lodge against him: his view cannot explain “why we should continue to describe a man as courageous even when he is not employing his courage.” See Long, A. A. and Sedley, D. N., The Hellenistic Philosophers (Cambridge: Cambridge University Press, 1987), vol. I, p. 384.Google Scholar But this is so obvious, and so obviously valid, an objection to Ariston's view if so construed that I am very reluctant to think that that was what he meant to say. If it was, then he blundered; what seems clearly to have been his basic idea is perfectly capable of development (as in my text) in such a way as to sidestep any such criticism.

²³ See also Plutarch, 's On Stoic Self-ContradictionsGoogle Scholar, available in Greek and English in Plutarch's Moralia, vol. 13, part 2, ed. Cherniss, Harold (Cambridge, MA: Loeb Classical Library, Harvard University Press, 1976), ch. 7, 1034cGoogle Scholar, where Plutarch repeats this claim, professing to find in these definitions a contradiction of Zeno's agreement elsewhere with the Platonic position that the virtues do differ from one another by having different natures. Indeed, these definitions would constitute such a contradiction if Zeno's view really were the same as Ariston's.

²⁴ This parenthetical remark reflects the fact that—as one can see already from what Ariston says above about “virtue” (i.e., “knowledge of good and bad”) as against “wisdom”—in Stoic theory phrondsis itself was only one of the particular virtues, not virtue itself: according to Ariston, it is “knowledge” in considering what is to be done and what not done. So later Stoics (perhaps beginning with Chrysippus) who wanted to hold that Zeno's view was not really different from their own insisted that in these definitions he was, nonstandardly, using “wisdom” to mean “knowledge of good and bad.” Socrates in the Protagoras also shows a parallel unclarity in his use of the word sophia. Sometimes the word refers to one among the virtues, sometimes it is used to refer to the comprehensive knowledge that (in some sense) they are all identical with. Perhaps this feature of Zeno's formulations simply reflects his reliance on the Protagoras in formulating his own view.

²⁵ I.e., Ariston said, as we have seen, that justice, e.g., is virtue as “relatively disposed” to busying itself with common enterprises and contracts—i.e., disposed to those things without itself changing internally in any way.

²⁶ Elsewhere Plutarch himself reports that Zeno did in fact maintain that the particular virtues “differ specifically” from one another—but Plutarch claims that this is inconsistent with Zeno's also defining the virtues in the (allegedly “relational”) way he reports Zeno here (in On Moral Virtue) to have done. See Plutarch, , On Stoic Self-Contradictions, ch. 7, 1034c–d.Google Scholar Plutarch reasonably insists on taking the definitions to express Zeno's basic view; it is open to question, however, whether Zeno intended them in the “relational” way that Plutarch also insists on, in order to find a Stoic self-contradiction here.

²⁷ Translation after Long and Sedley, , The Hellenistic Philosophers.Google Scholar Plato uses the expression “a swarm of virtues” in Meno, 72a, where Meno reports Gorgias's idea that there are different and distinct virtues for men and women, children and old people, slaves and free, and for every different sort of action and time of life.

²⁸ Arius Didymus wrote in the early first century a.d. There is as yet no translation in English of the Stoic part of his Epitome of Ethics, but one by Julia Annas is promised in the Clarendon Later Ancient Philosophers series (the other part, on Peripatetic ethics, is due in a second volume in that series, translated by Stephen A. White). Stobaeus compiled his excerpts (Eclogae) in the fifth century a.d.; I refer to them with the pages and lines of the edition of the Greek text by C. Wachsmuth (Berlin: Weidmann, 1884) (as is customary, I indicate this via the “W” appended to each citation). Some relevant passages are translated in Long and Sedley, The Hellenistic Philosophers, chs. 60 and 61; I will provide references to these as appropriate.

²⁹ For a summary account of the Stoic theory of goods and “indifferents,” see Cooper, John M. and Procope, J. F., Seneca: Moral and Political Essays (Cambridge: Cambridge University Press, 1995), pp. xvii–xxiii.Google Scholar

³⁰ Stobaeus, II 5, pp. 58. 9–14W (translated in Long, and Sedley, , The Hellenistic Philosophers, 60K), 62. 15–20W.Google Scholar See also Diogenes Laertius, VII 90 and the first sentence of 91 (where it is necessary to follow the reading of manuscript B as reported in the apparatus criticus of H. S. Long's text [Oxford: Clarendon Press, 1964], and not retain the words κα⋯ περ⋯ φαύλους γίγνονται, if what Diogenes Laertius reports is to make any acceptable sense).

³¹ The passage is printed in von Arnim, Stoicorum Veterum Fragmenta, as III 280, and translated in Long, and Sedley, , The Hellenistic Philosophers, as 61D.Google Scholar

³² At 62. 7W, Anus had said: “Of all these virtues [viz., the ones that are knowledges] the end is to live following nature; but each of them makes a person achieve this through its own particular means.”

³³ It seems to me unacceptable for Long and Sedley to say (in The Hellenistic Philosophers, p. 384) that it is true, according to Chrysippus and the other Stoics, that each virtuous action is done from all the virtues only to the extent that when an action is done from temperance, for example, it must avoid being done simultaneously with cowardice, injustice, or folly (and that this can be avoided because the temperate man knows, in a secondary way, all the “theorems” of courage, justice, and practical wisdom). It is perfectly clear that Arius Didy-mus and Plutarch (see n. 34) understand the thesis in a much stronger sense than that, and if Long and Sedley's interpretation of Chrysippus can only support such a weak conclusion, that just shows a defect in their interpretation. See further below.

³⁴ Plutarch confirms this point in On Stoic Self-Contradictions, ch. 27, 1046e–f: “They say that the virtues imply one another not only in the sense that anyone who has one has them all but also in the sense that anyone who does anything on the basis of one of them does it on the basis of all. For they say that neither is a man who does not have all the virtues perfect, nor is any action perfect that is not done on the basis of all the virtues.” Although in citing the Stoic doctrine of the mutual implication of the virtues (VII 125) Diogenes Laertius does not mention that it includes the claim that each action is done on the basis of all the virtues, his way of explicating it (VII 126) actually does imply it, as we will see below.

³⁵ Long, and Sedley, , The Hellenistic Philosophers, p. 384.Google Scholar

³⁶ This is presumably the correct response to Plutarch's attempt in On Stoic Self-contradictions, ch. 27, 1046f–1047a, to find Chrysippus contradicting his theory that each action is done on the basis of all the virtues, when he claims that the good man is not always acting courageously (⋯νδρίζεσθαι) because, for that, certain particular impressions (phanta-siai)—presumably ones of harm about to happen to one, etc.—have to have been received and appropriately responded to. It is in such a situation—where specific sorts of responses are needed to specific sorts of impressions—that the “higher reaches” of courage's knowledge that I have spoken of are brought into play. For that reason, these actions, but not others where courage nonetheless does have work to do, can count as “acts of courage.” (Another way to put the point is that, as Chrysippus says, every action of the virtuous person is done κατά or “on the basis of” each and every one of the virtues; but it does not follow from the fact that an act is done “on the basis of” courage (⋯νδρεία) that it is a case of ⋯νδρίζεσθαι, “acting courageously.”)

³⁷ Diogenes Laertius, VII 126: τ⋯ δ⋯ ποιητέα κα⋯ αἱρετέα ⋯στ⋯ κα⋯ ὑπομενητέα κα⋯ ⋯μμενητέα κα⋯ ⋯πονεμητέα.

³⁸ In support, Chrysippus could point to the fact that, in his own formulations, essentially the same language Zeno had used recurs: of justice dealing with “distributions,” temperance with “choices,” courage with “endurances,” and so on. It must be admitted that Zeno's quite brief formulations do not actually give any detail that clearly betrays (as Ariston's, for example, do) the differentiation of the virtues by way of specific, mutually disjoint areas of life as their province. So far as the language goes, they could be interpreted as making reference instead to different “areas” in the constitution of a single virtuous action.

³⁹ In his account of these virtues, Aristotle accepts—with a vengeance, one is tempted to say—the view that makes them each keyed to a single area of human life and conduct: temperance for how to behave in earing, drinking, and sex; courage for how to behave in war and related dangers; “good temper” for how to react to insults and in general how to understand and relate to one's own personal dignity; and so on. Plato's account irt the Republic is markedly different. I return to this difference below.

⁴⁰ That this qualification is intended is plain from the immediately preceding context, where the contrast between the “full” virtues and the so-called “natural” ones is a contrast within the scope of ethical virtue, only. Aristotle does not maintain here, or clearly anywhere else, that there is reciprocity between the possession of theoretical-intellectual virtue and ethical virtue. (Since Plato regards sophia as, in Aristotle's terms, simultaneously both theoretical and practical, for him this question hardly arises in a clear way.)

⁴¹ See Irwin, T. H., “Disunity in the Aristotelian Virtues,” Oxford Studies in Ancient Philosophy, supp. vol. 1988, pp. 61–78 (see esp. pp. 66–72)Google Scholar; and Annas, Julia, The Morality of Happiness, pp. 73–79.Google Scholar

⁴² See Aristotle, , Nicomachean Ethics, VI 5, 1140b11–21.Google Scholar

⁴³ In On Stoic Self-Contradictions, chs. 15–16, Plutarch reports a whole series of passages from a work of Chrysippus, “Against Plato on Justice,” that object to things Plato says in the Republic about justice as a virtue of individual persons.

⁴⁴ For this account of the different virtues, see Plato, , Republic, IV, 441c–443b.Google Scholar

⁴⁵ It is easier to see what this last, negative clause means for spirit and appetite than for reason. However, one should note that a reason that does its natural job must not only work hard at knowing the truth about human values, across the board, and reach decisions on that basis, it must in doing so not put the pursuit of its own pleasure ahead of anything else (and thus must not take on the functions of appetite, the pleasure-seeker); likewise, reason must not form such a strong impression of its own excellence and preeminence (i.e., the excellence and preeminence of the person whose reason it is) that it acts with intellectual arrogance and self-promotion (thus usurping some of spirit's functions).

⁴⁶ It is not often noted that it is in such terms as these that Socrates first introduces his account of the virtues (of the individual person) in Book IV of the Republic. Having worked out his account of the virtues of the city (428a–433e) in the order wisdom-courage-temperance-justice, he begins his account of the individual's virtues (after brief and purely formal references to wisdom and courage, 441c9–d4) with his target, justice itself (441d5–442b10), going on after that to give his definitions of courage (442b11–c4), wisdom (442c5–9), and temperance (442c10–d3). And in stating what justice or each-part's-doing-its-own-job entails, he makes a point of saying that, for reason, that includes its being wise, for spirit, its being courageous (441e4–5, 442b8).

⁴⁷ It seems that Plato has good reasons for dismissing as impossible the contrary position, where reason has its full virtue of wisdom, while the other virtues of justice, courage, and temperance are nevertheless wanting because the other parts of the soul are unruly. It is implicit in the whole scheme of education for the Republic's guardians that people whose appetites and spirited desires are not reined in and shaped and controlled through early moral training will not have the capacity to think correctly and learn what needs to be learned about true values in order to follow the course of intellectual training needed to develop their minds to the point where they achieve this knowledge. Of course, that early training does not establish full courage, justice, and temperance, but it does establish the first stages in their full development. And one might be willing to grant Plato that anything short of that full development (whatever exactly that might consist in) would prevent the actual establishment in one's mind of the full knowledge of good and bad that is required for wisdom—given, as I noted above, that that means the sort of permanent and unwavering grasp that could not be guaranteed if wayward desires or impulses of spirit, of any sort, remained.

⁴⁸ This is not the only point at which a view like Chrysippus's, otherwise similar to Plato's, comes off stronger: because Chrysippus does not recognize separate soul-powers independent of reason, he is in a better position to argue that with the one perfection of reason (and only with it) comes the perfect condition also of those aspects of an action that show its courageousness, temperance, and justice.

⁴⁹ See Plato, , Statesman, 306a–bGoogle Scholar, and, for the quoted words, 308b and 310a.