¹ As will transpire, a closer-fitting title would be ‘How to comprehend, confirm, and criticise evaluations’. I take my epigraphs from Frost, ‘The Black Cottage’, and McDowell, , Mind, Value, and Reality (Cambridge, Mass: Harvard University Press, 1998), p. 114Google Scholar. Jennifer Hornsby and David Wiggins warned me of a few errors.

² Not for the first time, I borrow the terminology of ‘lumpers’ and ‘splitters’, in affectionate remembrance, from sermons on the Trinity given at Balliol by the Revd Frank McCarthy-Willis-Bund.

³ Wiggins, David, ‘Ayer's Ethical Theory: Emotivism or Subjectivism?’, in A. J. Ayer: Memorial Essays, ed. Griffiths, A. Phillips (Cambridge University Press, 1991), p. 180Google Scholar.

⁴ On Austen's use of ‘elegant’, cf. Phillipps, K. C., Jane Austen's English (London: Andre Deutsch, 1970), pp. 51–3Google Scholar.

⁵ Hacker, Peter, Appearance and Reality (Oxford: Basil Blackwell, 1987), p. 204.Google Scholar

⁶ We may say that, in a minimal sense definable as one-way necessary covariance, the pot's elegance ‘supervenes’ upon its precise proportions in that any variation in the former requires a variation in the latter, though not vice versa. However, it would be odd to say that the elegance obtains ‘in virtue of’ the precise proportions (which would be a relation not just of co-variance but of dependence) when these can vary minutely without affecting the manner of the elegance, and detectably without affecting the fact of the elegance. It might be better to say that the manner of the pot's elegance obtains in virtue of an approximate shape, and the fact of its elegance in virtue of a very approximate shape — supposing that there are such things as more or less approximate shapes; but we would have to bear in mind that the boundaries of those would only be definable by reference to the elegance (as they are recognisable only by experience of it).

⁷ Some may still urge a metaphysical reduction that is neither epistemic nor definitional: the sense of a value-term might be a route not to a sui generis value-property, but to an essentially open-ended disjunction of properties (where each property was distinguishable non-evaluatively, though the disjunction was only determinable evaluatively). This piece of ontological cleansing invites an objection, and a caveat. The objection is that value-properties commonly come in degrees, but the disjunction does not. The caveat is that, even if every one of the disjuncts is a natural property in the sense of pulling its weight within an empirical science, the disjunction may not be a natural property in that sense. (On ‘natural’, cf. D. Wiggins, ‘A Neglected Position’, in Haldane, J. and Wright, C. (eds), Reality, Representation, and Projection (Oxford University Press, 1993), pp. 330–1.Google Scholar) Indeed, the disjunction is idle even as an explanation of particular evaluations (it would be crazy to attempt to spell it out in explaining any of them). One might call it the reflection that the value sheds upon reality neutrally described, and compare the shadow cast by a cloud upon a contoured landscape: the shape of the shadow is fully determined by the occlusion shape of the cloud, the laws of optics, and the lie of the land; a creature may intentionally move into sun or shadow; but the cloud does not map any natural boundaries that could explain either the shadow or anything else.

⁸ Thus it is not the case that a value-property is either reducible to natural properties, or never inferrable from them.

⁹ No doubt aptly, talk of ‘realism’ tends to invoke or provoke other ‘ism's, notably scientism. Even Paul Valéry can write: ‘L’homme froid est par là le mieux adapté à la réalité, laquelle est indifférente. - Les choses n'avancent ni ne retardent, ne regrettent ni n'esperènt’; Tel Quel (Paris: Gallimard, 1941), p. 120Google Scholar. It is not clear why, or how, ‘things’ are to determine their own categories.

¹⁰ James Griffin attempts an anodyne usage when he characterises a moral intuition as ‘merely a moral sentiment or belief that persons have independently of any moral theory or philosophy that they might adopt’; Value Judgement (Oxford: Clarendon Press, 1996), p. 3Google Scholar. However, when he then expresses scepticism about ‘piecemeal appeal to intuition’, citing R. M. Hare, and writes, ‘Intuitions … are just beliefs’, he is surely lapsing into a sense of an epistemic gap to be bridged, apparently by ‘theory’. It is true to say that he offers no word of explanation as to why he wants to call nontheoretic moral beliefs ‘intuitions’ when, apparently, he would not apply the term to non-theoretic identifications of blends of colours or commonplace causes. It is presumably a lesson from Davidson that beliefs of a conceptual category to be ‘just beliefs’ (until rescued by a theorizing deus ex machina) would not even be beliefs.

¹¹ Think also of explanations that cite dispositions. Dispositions are not parts of mechanisms (even when they are grounded in mechanism); and yet they can surely be explanatory in their own way.

¹² Cf. Wiggins, D, Needs, Values, Truth: Essays in the Philosophy of Value, 3rd edition (Oxford: Clarendon Press, 1998), p. 153Google Scholar, note 17.

¹³ One may be tempted to say that there is external causation where there is no mental reflection. But intelligence can work swiftly and implicitly; and thought may open the door to a causal input.

¹⁴ Arrington, Robert L., Rationalism, Realism, and Relativism (Ithaca, NY: Cornell University Press, 1989), p. 175Google Scholar.

¹⁵ Note also that the presence of a matt black surface may explain a perception, causally or otherwise, even though it reflects no light. (I owe this point to John Hyman.)

¹⁶ Williams, Bernard, Ethics and the Limits of Philosophy (London: Fontana, 1985), p. 129Google Scholar.

¹⁷ Though I find this helpful, its mention of ‘commendation’ should be accepted as a decent piece of lexicography rather than as an implicit contribution to philosophy; cf. Ziff, Paul, Semantic Analysis (Ithaca, NY: Cornell University Press, 1960), pp. 228–32Google Scholar. Ziff is also salutary on ‘approval’; ibid., pp. 223–7.

¹⁸ Cf. Urmson, J. O., The Emotive Theory of Ethics (London: Hutchinson, 1968)Google Scholar, chapter 9. Does this introduce an element of relativism whereby an ascription of goodness may come out as true or false from different points of view? No, for the relevant point of view, as indicated by the context, already fixes the content of what is being said. A judgement inviting conflicting verdicts from different points of view is to that extent indeterminate — which is not to say that how things show up from a given point of view is always indisputable.

¹⁹ Gallie, W. B., ‘Essentially Contested Concepts’, Proceedings of the Aristotelian Society, 56 (1955/6)Google Scholar; Plato, Euthyphro 7b6–8a2. Cf. Myles Burnyeat: ‘It is typical of a virtue concept that its range should be liable to controversial extension or modification. For to state and defend criteria for collecting manifestations of a virtue is to articulate a way of grouping certain phenomena which exposes something of one's outlook on life in general’ ‘Virtues in Action’, in Vlastos, G. (ed.) The Philosophy of Socrates (New York: Doubleday, 1971), p. 212.CrossRefGoogle Scholar

²⁰ , Williams, Ethics and the Limit of Philosophy, p. 145.Google Scholar

²¹ Let us concede that a lack of commitment can carry a cognitive cost; for it may generally be the case that someone to whom the culture is alien is not good at discriminating or resolving or detecting borderline or debatable or novel instantiations of a value. Yet this is less plausible of someone fully acculturated but also ambivalent. No one may have a keener nose for class differences than an inverted snob.

²² Newman, J. H., The Idea of a University, ed. Ker, I. T. (Oxford: Clarendon Press, 1976)Google Scholar; Raverat, G., Period Piece (London: Faber and Faber, 1952).Google Scholar

²³ Raverat, , Period Piece, p. 217Google Scholar.

²⁴ French Writers (Library of America, 1984), p. 162Google Scholar.

²⁵ Platts, M., Ways of Meaning (London: Routledge & Kegan Paul, 1979), p. 249Google Scholar. If this punctures the pretensions of the ‘realism’ of which Platts takes semantic depth to be a sign, he might prefer to say that conceptions of the gentlemanly can always become more refined, but without any increase in understanding. Then the thesis of realism will transcend that of cognizability in applying only to values that we privilege in some way still to be explained.

²⁶ Newman, , The Idea of a University, p. 179Google Scholar.

²⁷ Gibbard, A., ‘Morality and Thick Concepts’, Proceedings of the Aristotelian Society, Supplementary Volume 66 (1992), pp. 267–8CrossRefGoogle Scholar.

²⁸ Burnyeat, M., ‘Virtues in Action’, p. 216Google Scholar.

²⁹ Cf. Wiggins, , Needs, Values, Truth, pp. 95–6Google Scholar.

³⁰ Thus, on this view, I may detect a value as showing up from some evaluative (that is, ‘sentiment-involving’) point of view that I can occupy imaginatively, and yet recognise no related practical reasons. Yet we might say that I value a thing only if it speaks positively to a sentiment that I take to heart. This would seem a matter for stipulation.

³¹ Wiggins, , Needs, Values, Truth, p. 370Google Scholar.

³² ibid., p. 158.

³³ Cf. Plutarch, , Moralia 208cGoogle Scholar: ‘When some criminal was submitting calmly to torture, Agesilaus remarked, “What an thoroughly wicked man he is to apply endurance and fortitude to evil and shameful ends!’” What makes this citable as a‘Spartan Saying’ is that it turns the screw: the fortitude of a villain is not only not a virtue — it is a further vice.

³⁴ The problem may be briefly stated as follows. ‘This act is courageous’ may be true, on my view, in virtue of the sense of ‘courageous’; yet ‘This act is to be done’ can hardly be made true by the sense of ‘to be done’. Of course, one can specify, say, ‘morally to be done’ (if that is sufficiently determinate). But then the judgement is no longer inherently practical: if ‘morally’ is a qualifier and not an intensifier, ‘This act is morally to be done, but I will not do it’ is neither admirable nor necessarily acratic.

³⁵ Cf. Wiggins, D., ‘Reply to Roger Crisp’, in Lovibond, S. and Williams, S. G. (eds) Essays for David Wiggins: Identity, Truth and Value (Oxford: Blackwell, 1996), pp. 263–4Google Scholar.

³⁶ Le Roman de Tristan et Iseut dans I'Oeuvre de Thomas Malory (Paris: Librairie Ancienne Honore Champion, 1925), p. 137Google Scholar.

³⁷ ‘Je suis un chevalier errant qui chascun jor voiz aventures querant et le sens du monde, mes point n'en puis trouver’; The Works of Sir Thomas Malory, ed. Vinaver, E., 2nd edition (3 vols.; Oxford: Clarendon Press, 1967), p. 1494Google Scholar.

³⁸ ibid., p. 1491.

³⁹ Stendhal, , Romans et Nouvelles (2 vols.; Paris: Gallimard, 1952), vol. I, p. 1200Google Scholar.

⁴⁰ ibid., vol. II, p. 707.

⁴¹ Stendhal's habitual irony makes him ambivalent: where his Italian original describes the crowd's sudden emotion when they realised that Giacomo's younger brother Bernardo was reprieved, he noted, ‘On voit bien comment chez un peuple esclave de la sensation présente, la pitié serait pour le coupable qui va souffrir’; ibid., vol. II, p. 1456.

⁴² Cf. Williams, Bernard, ‘Internal and External Reasons’, in Moral Luck (Cambridge University Press, 1981)CrossRefGoogle Scholar.

⁴³ It turns my earlier quarrel with him into a quibble that neither of these points is an objection to the role that Griffin still hopes to find for theory.