1 We thank Simon Blackburn, Hallvard Lillehammer, Gopal Sreenivasan, Sergio Tenenbaum, and Ralph Wedgwood for helpful suggestions and are especially grateful to Simon Kirchin for first stimulating this paper and then improving it through a series of immensely helpful oral and written comments.

2 See, e.g., Tappolet, Christine ‘Through Thick and Thin: Good and its Determinates,’ Dialectica 58 (2004) 207–20CrossRefGoogle Scholar.

3 See, e.g., Dancy, Jonathan ‘In Defense of Thick Concepts,’ in French, Peter A. Uehling, Theodore E. Jr., and Wettstein, Howard K. eds., Midwest Studies in Philosophy XX: Moral Concepts (Notre Dame: University of Notre Dame Press 1996), 263–79.Google Scholar

4 McDowell, John ‘Are Moral Requirements Hypothetical Imperatives?’ Proceedings of the Aristotelian Society, Supplementary Volume 52 (1978) 13–29;CrossRefGoogle Scholar ‘Virtue and Reason,’ The Monist 62 (1979) 331-50; and ‘Non-Cognitivism and Rule-Following,’ in Holtzman, S.H. and Leich, C.M. eds., Wittgenstein: To Follow a Rule (London: Routledge & Kegan Paul 1981) 141–62.Google Scholar The acceptance of supervenience and denial of universalizability come in ‘Non-Cognitivism and Rule-Following,’ 144-5.

5 McDowell, ‘Virtue and Reason,’ 332

6 We do not of course claim that adherents of the reductive view typically accept uncodifiability; most reject it. Our claim is just that if uncodifiability applies to thin as well as thick judgments, it cuts no ice against the reductive view.

7 McDowell, ‘Non-Cognitivism and Rule-Following,’ 144Google Scholar

8 Williams, Bernard Ethics and the Limits of Philosophy (Cambridge, MA: Harvard University Press 1985), 130, 141-2.Google Scholar

9 See, e.g., Dancy, ‘In Defense of Thick Concepts,’ 263-4; Hilary, Putnam ‘The Entanglement of Fact and Value,’ in The Collapse of the Fact/Value Dichotomy and Other Essays (Cambridge, MA: Harvard University Press 2002) 28–45,Google Scholar at 34-40; Taylor, Charles ‘Ethics and Ontology,’ Journal of Philosophy 100 (2003) 305–20,CrossRefGoogle Scholar at 305-6; and Tappolet, ‘Through Thick and Thin,’ 213-17.

10 Simon Blackburn has implicitly denied it, when he argues that most proposed thick concepts are purely descriptive, with no specific evaluations built into their content. ‘Frugal,’ for example, while often a term of praise, is the opposite when used of a host dispensing hospitality (‘Through Thick and Thin,’ Proceedings of the Aristotelian Society, Supplementary Volume 66 (1992) 285-99, at 285-7). While this is a possible response to the disentanglement argument, we will follow a different tack and grant both that thick concepts have specific evaluations built into their contents and that their extensions depend on evaluations.

11 Stevenson, C.L. Ethics and Language (New Haven: Yale University Press 1944), 206–7;Google Scholar Hare, R.M. The Language of Morals (Oxford: Oxford University Press 1952), 121;Google Scholar Hare, Freedom and Reason (Oxford: Oxford University Press 1963), 121–9;Google Scholar and Blackburn, Simon Spreading the Word (Oxford: Clarendon Press 1984), 148–9.Google Scholar

12 McDowell and Williams sometimes blur the distinction between the intended and actual extensions of a thick concept, which can make their positions seem relativist. Our formulations are meant to avoid any hint of relativism by clearly separating the claim that a speaker's intended extension for a thick term depends on his evaluative beliefs from the claim that its actual extension depends on evaluative truths. Our defense of the reductive view could be mounted with respect to either claim, but for clarity's sake we keep them separate.

13 See Scheffler, Samuel The Rejection of Consequentialism (Oxford: Clarendon Press 1982).Google Scholar

14 Stephan L. Burton has proposed an analysis of thick concepts that also involves partial indeterminacy but locates it in a different place. He reads ‘x is F’ as ‘x is good for having some instance of properties X, Y, and Z (for specific X, Y, and Z),’ where other instances of the same properties need not make what has them good (‘“Thick” Concepts Revisited,’ Analysis 52 (1992) 28-32). For many terms the specified properties will have to be disjunctive, for example, ‘is an equal distribution, or one proportioned to merit, or …’ But Burton denies that one can specify, even in principle, any narrower property such that all and only the instances of X, Y, and Z that do make things good have that property. Whatever property one selects, even one as specific as equality, some instances of it will be good-making while others are not. Apparently influenced by McDowell's argument about uncodifiability, Burton here denies universalizability; his analyses may therefore appeal to those who likewise reject it. But we wish to retain universalizability, as Sidgwick, Hare, and others did, and therefore take a different line. We do not hold, as Burton does, that the sense of a thick term fully specifies a set of properties; it leaves them partly indeterminate. But there is always some property that fully determines the term's extension, in the sense that all and only instances of it are in the relevant way good-making.

15 Scheffler, ‘Morality Through Thick and Thin: A Critical Notice of Ethics and the Limits of Philosophy, Philosophical Review 96 (1987) 411–34, at 417-18.Google Scholar

16 Gibbard, Allan ‘Thick Concepts and Warrant for Feelings,’ Proceedings of the Aristotelian Society, Supplementary Volume 66 (1992) 267–83, at 280-1CrossRefGoogle Scholar

17 Though Gibbard's analysis of ‘lewd’ makes it involve evaluations of the agent, it does not refer to her motivations but seems to concern only her acts and whether, motivation apart, they exceed some limits. For us this makes Gibbard's ‘lewd’ not a virtue-concept but one suited to the first pattern of analysis and explicable using ‘wrong.’

18 Dancy, ‘In Defense of Thick Concepts,’ 275Google Scholar

19 Putnam, ‘The Entanglement of Fact and Value,’ 38Google Scholar

20 The Oxford English Dictionary defines ‘cruelty’ as involving ‘delight in or indifference to another's pain’; our refinement builds those psychological elements into the term's analysis.

21 McDowell, ‘Non-Cognitivism and Rule-Following,’ 144Google Scholar

22 Williams, Ethics and the Limits of Philosophy, 142Google Scholar

23 See Hurka, Thomas Virtue, Vice, and Value (New York: Oxford University Press 2001)CrossRefGoogle Scholar for such a defense.

24 Sidgwick, Henry The Methods of Ethics, 7th ed. (London: Macmillan 1907), 392–3Google Scholar

25 Rashdall, Hastings The Theory of Good and Evil, 2 vols. (London: Oxford University Press 1907), vol. 1, 59Google Scholar

26 Ross, W.D. The Right and the Good (Oxford: Clarendon Press 1930), 161Google Scholar

27 For an argument of this type, see Ross, The Right and the Good, 135.Google Scholar

28 Some versions of the anti-reductive view may deny that such independent judgments of goodness are possible, since all thin judgments derive ultimately from thick ones. But these are more radical versions of the anti-reductive view than are commonly defended.

29 We owe this second response to Simon Kirchin.

30 Aristotle, Nicomachean Ethics, trans. Ross, D. (Oxford: Oxford University Press 1980), 1096a2Google Scholar