1 “Does Moral Philosophy Rest on a Mistake?”, Mind 21 (1912): 36.

2 Ibid., 22.

3 Prichard's complaint appears, not just in the well-known 1912 paper, but in “Duty and Interest” (1928, reprinted in Readings in Ethical Theory, ed. Sellars, Wilfrid and Hospers, John [New York: Appleton-Century-Crofts, 1952], 469-86Google Scholar), where Prichard offers a further argument for the charge that Plato is giving a specious justification for moral obligation. Specifically, Prichard there claims that Plato misconstrues the nature of the challenge posed in Republic by Thrasymachus's Sophism. I am not convinced Prichard has either Plato or Thrasymachus right, but I will leave questions of detailed Republic interpretation aside.

4 “Does Moral Philosophy … ”, 27.

5 Ibid., 28.

6 Ibid., 29.

7 Ibid., 36.

8 Prichard, “Duty and Interest,” 471.Google Scholar

9 The contemporary focus on wrong kind of reason objections stems from work by Rabinowicz, Wlodek and Rlmnow-Rasmussen, Toni raising a problem for certain accounts of the nature of value (''The Sbike of the Demon: On Fitting Pro-Attitudes and Value,” Ethics 114 [2004]: 391–423).CrossRefGoogle Scholar

10 Prichard, “Duty and Interest,” 473, 477.Google Scholar

11 Ibid., 480.

12 Prichard explicitly recognizes this point; he complains of Aristotle that he “really answers two radically different questions as if they were one: (1) ‘What is the happy life?’ (2) ‘What is the virtuous life?'” (“Does Moral Philosophy … ”, 33). Thus he recognizes the strategy but misapprehends its significance.

13 Rawls, JohnA Theory of Justice (Cambridge, MA: Harvard University Press, 1971 I 1999).Google Scholar

14 Ibid., §4.

15 “Wide Reflective Equilibrium and Theory Acceptance in Ethics,” Journal of Philosophy 76 (1979): 256-82. Daniels’ “wide” descriptor is intended to reflect a contrast drawn by Rawls between processes that attempt merely to “smooth out irregularities” in a person's view, and processes that canvass (at least potentially) “all possible philosophically relevant arguments,” and hence may result in a “radical shift” in pre-existing judgments (Rawls, Theory of Justice, 43Google Scholar). See also Daniels, “Reflective Equilibrium,” Stanford Encyclopedia of Philosophy (Summer 2003), ed. Edward N. Zalta; <http:/ /plato.stanford.edu/archives/sum2003/ entries/ reflective-equilibrium/>. Failing to account for this difference in kinds of reflective equilibrium can lead to criticism of the method as shallow and conservative; see, for example, Allen Wood's criticism that the method of reflective equilibrium, as the “standard model of ethical theory” after Sidgwick, “seeks only coherence among commonly held opinions” (Kantian Ethics [Cambridge: Cambridge University Press, 2008], 51)Google Scholar. What Rawls, Daniels, and I are all after is the wide form of reflective equilibrium which, as I hope becomes clear, is simply not open to this objection.

16 Cf. Aristotle, Nicomachean Ethics [hereinafter NE]I.lO.

17 It is here that specific conceptions of virtue and practical rationality are deployed as parts of an idea for human life. For an argument that conceptions of practical rationality are essential for understanding the requirements of virtue, see Russell, DanielPractica/Intel/igence and the Virtues (Oxford: Oxford University Press, in press), chap. 6.Google Scholar

18 On Moral Ends, trans. Raphe! Woolf, ed. Annas, Julia (Cambridge: Cambridge University Press, 2001), vol. 17Google Scholar (emphasis added).

19 Cf. Methods of Ethics (Indianapolis: Hackett, 1907 I 1981), 111.1.4, especially p. 213.

20 In precisely this light John McDowell invites us to consider the Aristotelian nature and work of practical reason as of “second nature” (“Two Sorts of Naturalism,“ in Virtues and Reasons, ed. Hursthouse, RosalindLawrence, Gavin and Quinn, Warren (Oxford: Clarendon, 1995).Google Scholar I take the ancients to be in this way naturalists of a sort, in the way McDowell indicates rather than, e.g., the sort espoused by Philippa Foot (e.g., in her Natural Goodness [Oxford: Clarendon, 2001]).

21 Euthydemus, 278e-282a. This is also much of the picture of eros in Plato's Symposium, and the assumption that everyone seeks happiness is common in other dialogues as well; besides Republic, Apology, Crito, Gorgias, and others all feature argument set in a framework built on this psychological assumption.

22 Kant's version of this point is most clearly formulated in “Religion within the Boundaries of Mere Reason,” in Religion and Rational Theology, trans. Wood, Allen W. and George di Giovanni (Cambridge: Cambridge University Press, 1996),Google Scholar Ak. p. 6:24. See also the Metaphysics of Morals, trans. Gregor, Mary (Cambridge: Cambridge University press, 1991),Google Scholar Ak. p. 6:385. Allison, Henry calls this Kant's “Incorporation Thesis,” and claims that it “underlies virtually everything that Kant has to say about rational agency” (Allison, Kant's Theory of Freedom [Cambridge: Cambridge University Press, 1990], 40).CrossRefGoogle Scholar

23 This is a major conclusion of Socrates’ argument against Callicles in Plato's Gorgias. As to Aristotle, see John McDowell, “Two Sorts of Naturalism,” in Hursthouse, Lawrence, and Quinn, §10. A similar but more complex picture is given by the Stoics, on whose view the life of virtue and practical wisdom is thought to be equivalent to living “according to nature” (cf. Diogenes Laertius 7.85, 94, 108, in Hellenistic Philosophy, ed.Inwood, Brad and Gerson, L.P. (Indianapolis: Hackett Publishing 1997), II-94;Google Scholar see also Striker, Gisela “Following nature: a study in Stoic ethics,” in her Essays in Hellenistic Epistemology and Ethics [Cambridge:, Cambridge University Press, 1996]).CrossRefGoogle Scholar More generally the relation between “impulse” (horme) and reason is complex in ways that complicate the picture I suggest here. However, even here, Diogenes Laertius glosses the ideas of Zeno, Cleanthes, and other Stoics as claiming that “reason supervenes as the craftsman of impulse” (7.86, Long and Sedley, ed., The Hellenistic Philosophers [Cambridge: Cambridge University Press, 1987], 57a, p. 346).

24 See the introduction to Political Liberalism (New York: Columbia University Press, 1993); see also Daniels, “Reflective Equilibrium,”.

25 Philebus, lld. All quotations for Philebus are from Dorothea Frede's translation (Indianapolis: Hackett, 1993) and will be cited by Stephanus page number.

26 Ibid., 65a.

27 Ibid., 20d-21a.

28 Ibid., 21c.

29 Ibid., 60c. Here Frede translates hikanon as “self-sufficiency,” but I retain the “sufficiency” of the earlier passage.

30 Plato doesn't acknowledge this distinction here in Philebus, but we can see an analog at work in the discussion of the ideal polis in Republic. There, Adeimantus and Socrates agree that the polis is needed because none of us individually is “self-sufficient” (autarkes). The polis is the solution to that problem, but there too there is a distinction between senses in which the polis can be self-sufficient. The analog to the first (weak) alternative is the mere viability of a polis- which of course is better than a polis that is incapable of self-perpetuation (although Socrates says even such a polis has “grown to completeness” [te/ca]-Rep. 37le [Grube/Reeve translation]). However, it is evident that a polis might be sufficient in that weak sense while lacking things that might contribute to its value as a polis, hence fail to be sufficient in the second and stronger of the two senses above. The distinction between these senses is, in fact, brought out by Glaucon's objection to the basic polis Socrates sketches initially (Rep. 372c-e ). Socrates’ merely self-sufficient (“healthy“-Rep. 372e) city nevertheless imposes “hardships” on its citizens, Glaucon complains, and Socrates agrees to consider the origin of a “luxurious” city, which does not lack the furnishings and niceties that the merely self-sufficient one does. So Plato can recognize just the sort of distinction that the appeal to the “lacking in nothing” condition wants, but doesn't explicate it in defending his conception of the best life.

31 Frede's explanation (Pizilebus, xxxi) is that this is elenchtic argumentation. Socrates has as his goal securing the consent of Protarchus to his conclusion about the human good, and his consent ipso facto imposes these conditions on it. Even if this is right, however, it does nothing to justify Protarchus’ concurrence. Why must the good be perfect or sufficient to be desired as such?

Gosling presses the problem of the lack of justification for these formal conditions even more forcefully (Plato: Pl!ilebus [Oxford: Clarendon, 1975]). Suppose, he argues, that if faced with the prospect of a life of pleasure without knowledge of any kind, Protarchus had responded that he would gladly choose such a life. What sort of response could Socrates give? He obviously could not say that Protarchus would be irrational in virtue of such a choice; as Gosling puts it, in this context such a response is “not just going round in a circle, it is pirouetting“ (182). Gosling believes that the best prospects for a justification of the conditions may be the idea that they are required for a life to be a human life, but that Plato leaves this idea undeveloped.

32 Frede comments that “Perfection and sufficiency might look like the same thing, but perfection stresses that no further additions are possible, sufficiency that nothing is lacking” (PI!ilebus, 14, n. 2). Yet Socrates immediately goes on to conclude of pleasure and knowledge that “if either of the two is the good, then it must have no need of anything in addition. But if one or the other should tum out to be lacking anything, then this can definitely no longer be the real good we are looking for” (20e). The operative sense of lacking in nothing here appears to be precisely the same as incapable of addition, so the two conditions collapse into one.

33 I thank Joshua Gert for helpful discussion of this point.

34 NE I.4. The debate over Aristotle's conception-whether it is a life of practical reason plus external goods, or contemplation-is long and deep, and I have little to add, though I take the former to be his settled view. I take what I say here to be neutral between these two positions.

35 NE I.7:1097a25-36 (Ross/Urmson translation).

36 NE 1.7:1097b8-20.

37 NE 1.5:1095b25. The same idea is in the Eudemian Ethics:” … if it depends on the individual and his personal acts being of a certain character, then the supreme good would be both more general and more divine, more general because more would be able to possess it, more divine because happiness would then be the prize offered to those who make themselves and their acts of a certain character“ (EE 1.3: 1215a15-19). The same idea also seems at work in Aristotle's suggestion that “security” is part of the self-sufficiency that is a condition of happiness, at Rhetoric 1360b29.

38 NE VIII.l:115Sa2-4.

39 It is clear that, though the former is used to make the strong claim about friendship (a life lacking it is not simply not the best life, it is not a life worth living at all), it is the latter which Aristotle has in mind in the main, as a criterion for a conception of eudaimonia. This stronger condition of lacking in nothing is one Aristotle refers to as completeness (teleion) in the Metaphysics. There the first definition of that which is complete is that outside of which it is not possible to find any of the parts proper to it (V.16: 102lb12). Things which are complete in this sense, he says, “lack nothing in respect of goodness and cannot be excelled and no part proper to them can be found outside” (1021b32). Notably, his definition of completeness in NE 1.7 makes no reference to lacking in nothing.

40 Socrates says “it's almost impossible to establish a city in a place where nothing has to be imported” (Rep. 370e-37la), and this is true even of the merely complete, not “luxurious,” polis.

41 NE IX.9: 1169b3-8.

42 NE IX.9: 1169b9-21.

43 Should we take this to mean that happiness is within our power tout court, or merely that it is not within the power of others to deprive us of it? Some ancient texts (e.g., Socrates’ defence in Apology) suggest the latter, but I believe Aristotle's discussions of the idea suggest the stronger. Virtue might not be the only way to secure myself against the depredations of others, but it plausibly is a defence against challenges to happiness that may occur otherwise. I thank Daniel Russell for discussion of this point.

44 Laertius, Diogenes (DL) 10.128, “Letter to Menoeceus,” in Hellenistic Philosophy, 30.Google Scholar

45 DL 10.131, in ibid., 30-1.

46 Cicero reports Epicurus as claiming that “there is no need of reason or debate about why pleasure is to be pursued and pain to be avoided,” that we “perceive“ this, just as we perceive “that fire is hot, that snow is white, that honey is sweet“ (De Finibus 1.30, in Hellenistic Pl!ilosopl!y, 57).lt seems equally apparent, however, that the sophisticated form this aversion takes in the mature theory-ataraxiaisn't simply animal attractions and aversions. This connection needs argument, and I take that argument to consist in equilibrium between considerations at a variety of levels.

47 DL 10.122, in ibid., 28.

48 DL 10.128, ibid., 30.

49 DL 10.130, in ibid.

50 DL 10.131, in ibid.

51 DL, Principle Doctrine XXI, in ibid., 33-34.

52 Cf. Principle Doctrines VII, XIV, in ibid., 32.

53 VC 77, in Hellenistic Pllilosoplly, 40. The “freedom” Epicurus is speaking of here is not metaphysical freedom (on which his views are vexed), but freedom from pain and the “disturbance” that is the product of unsatisfied desire. Cf. Cicero, De Finibus 1.37.

54 VC 67, in Hellenistic Philosophy, 39.

55 For an argument that none of these components should be seen as foundational, but all provide mutual support, see Annas, Julia “Ethics In Stoic Philosophy,“ Phronesis 52 (2007): 58–87CrossRefGoogle Scholar.

56 Cicero, De Finibus, trans. Rackham (Cambridge, MA: Harvard University Press, 1914), IV.53.Google Scholar

57 Stobaeus, Anthology 2, 6e, in Hellenistic Pl!ilosopl!y, 212.

58 Stobaeus, Anthology 2, llh, in ibid., 223.

59 Stobaeus, Anthology 2, llg, in ibid., 223.

60 Of the many Stoic texts making this point, perhaps the clearest is Stobaeus, Anthology 2, 7, in ibid., 213.

61 Cf. Cicero, De Finibus 5.83: “For since the Good consists solely in virtue and in actual moral worth, and neither virtue nor moral worth, as they hold, admits of increase, and since that alone is good which necessarily makes its possessor happy, when that which alone constitutes happiness does not allow of increase, how can anyone possibly be happier than anyone else?“

62 Diogenes Laertius VII.93, in Hellenistic Philosophy, 193.

63 Diogenes Laertius VII.l27-8, in ibid., 201-2. The same point, in the same language, is found in Stobaeus (5b2, in Hellenistic Philosophy, 204), who adds that the virtuous man “has the good and divine virtue within himself, which is why he escapes all vice and harm” (11m, in Hellenistic Philosophy, 229). Cicero's formulation of Stoicism agrees that “magnanimous and strong-hearted men are able to despise and ignore everything which fortune can bring to bear against man” (De Finibus 111.29, in Hellenistic Philosophy, 239).

64 Cf. Epictetus, Discourses, III.7.5. As with Epicurean theory, this doctrine raises problems for including friendship as an essential part of the best human life which it is not clear the Stoics can adequately address, though I cannot take up here the notion of friendship in Stoic ethics.

65 For that matter, it affects judgments on Level I as to what lives are good (the wise man on the rack can be happy), and interacts with other Level IV views (e.g., Stoic theories of emotion, action, human development, etc.). See below.

66 De Finibus 111.24, in Hellenistic Philosophy, 238.

67 Stobaeus llg, lls, in Hellenistic Philosophy, 222, 231.This summary greatly compacts important features of Stoic philosophy of mind and action, which give special technical senses to ‘impulse’ and ‘striving'; cf. Stobaeus 9ff., in ibid., 230ff.

68 Tusculan Disputations, 5.40-1, in Long and Sedley, ed., The Hellerzistic Plzilosophers (Cambridge: Cambridge University Press, 1987), (63L), 397. Seneca remarks that “At all events the mind must be withdrawn from all externals into itself” as a condition of “making ourselves adaptable” to whatever external conditions come to bear on our plans and ends (On Peace of Mind, 14.1-2, in Hellenistic Plrilosophy, 243).

69 The Stoics actually have two such doctrines, one personal, one social. Here I refer to the doctrine of personal oikeiosis, but both figure in as Level IV contributions to overall moral theory. Cf. Cicero, De Finibus III.16, III.23. For discussion of the doctrines, see Striker, Gisela “The Role of Oikeiosis in Stoic ethics,” in Oxford Studies in Ancient Philosophy 1 (1983): 145-67.Google Scholar

70 Discourses III.24.56, trans. Matheson, in The Stoic and Epicurean Pl1ilosophers (New York: Random House, 1940), 395.Google Scholar See also Stobaeus, Anthology 2.11m (in Hellenistic Philosophy, 11-95).

71 A bellwether of this insight is Kraut's, Richard “Two Conceptions of Happiness,” Philosophical Review 88 (1979): 167-97;CrossRefGoogle Scholar there are few more comprehensive discussions than Annas, Julia’ TllC Morality of Happiness (New York: Oxford, 1993).Google Scholar

72 Prichard, “Duty and Interest,” 473.Google Scholar

73 Aristotle is better known for the doctrine that our habituation shapes the way we see things (NE Vl.2: 1139a33-b5), but Plato makes essentially the same case in Laws, Book II.

74 John McDowell has made essentially this point on more than one occasion. He criticizes the idea of taking prudential notions to explicate moral ones, and vice versa: “What makes it the case that the conception of excellence is a unity, insofar as it is one, is not that prima facie conflicts are resolved by asking what will maximize some independently recognizable goods but that the results of those efforts at discernment tend to hang together, in the way that particular facts hang together to constitute a world” (“The Role of Eudaimonia in Aristotle's Ethics,” in Essays on Aristotle's Ethics, ed. Amelie O. Rorty [Berkeley: University of California Press,1980). 367-69); see also his appeal to the model of Neurath's boat in “Eudaimonism and Realism in Aristotle's Ethics,” in Aristotle and Moral Realism, ed. Robert A. Heinaman (Boulder, CO: Westview Press, 1995), 213-14. There he argues that (in our terms) Aristotle himself seeks only narrow equilibrium in his own views, but we can understand that equilibrium to be broader in the ways we have explored here.

75 See, for example, Scanlon's, T.M.Wllat We Owe to Eacll Other (Cambridge, MA: Harvard University Press, 1999)Google Scholar and Darwall's, StephenTile Second-Person Standpoint (Cambridge, MA: Harvard University Press, 2006).Google Scholar

76 For example, he takes Plato's project to be to vindicate our “ordinary moral thoughts or convictions” -unqualified -as against the Sophists’ charge that these thoughts are false (Prichard, “Duty and Interest,” 472).

77 I defend the claim that it can in “Virtue Ethics and Deontic Constraints“ (manuscript).

78 I hasten to add that I do not suggest that this is what the ancients saw themselves as doing. There is no evidence that they were disturbed by questions to which what I suggest here might represent an answer. The Stoics did indeed see their theory as tightly integrated and in effect subject to a sort of coherentist justification, as we have observed. My suggestion is that we construe the argument between ancient accounts -and in particular the dispute between conceptions of the good -as a search for a reflective equilibrium.

79 Isaiah Berlin suggests this very concern as a virtue of pluralism, value in “Two Concepts of Liberty,” Four Essays on Liberty (Oxford: Oxford University Press, 1969), §8.Google Scholar

80 I am most grateful to the Earhart Foundation for a grant enabling research for this paper. I also thank Julia Annas, Joshua Gert, Nathaniel Goldberg, James Lesher, and Daniel Russell, and the editors of this volume and their students, for helpful comments on earlier drafts.