¹ In ethics, then, as elsewhere, we need to distinguish between intuition and belief, since one need not believe one's intuitions. At best, intuition involves something more like a disposition to believe. But of course not all dispositions to believe are intuitions. While it would be useful to have a general characterization of intuitions, this is a complicated subject and I will have to restrict myself to two further remarks. First, intuitions are normally taken to be “immediate” or “spontaneous,” and while this apparently rules out dependence upon current conscious inference or reflection, it seems to leave open the possibility that prior reflection (or current unconscious inference) may have played a role in generating the present “immediate” intuition (and so, among other things, intuitions need not be unlearned). Second, not all “immediate” and “spontaneous” dispositions to believe qualify as intuitions. There is, I think, a further characteristic quality—one that I, at least, find difficult to describe—that is required as well: roughly, its simply “appearing” to one that something is the case. (Thus, although I am immediately disposed to believe that Washington D.C. is the nation's capital, it doesn't seem to me that I have any intuition to this effect.) It is not clear which of these features (or others) are relevant to justifying our reliance upon intuition, in ethics or elsewhere.

² This is similar to an example of Judith Jarvis Thomson's, offered while making a similar point; see Thomson, , Rights, Restitution, and Risk (Cambridge, MA: Harvard University Press, 1986), 257.Google Scholar

³ In point of fact, I don't think it altogether obvious to what extent empirical observations are indeed reliable. Consider shapes and sizes. Many people, I suppose, would be inclined to say, for example, that the stick in the water appears to me to be bent, even if I know better (and so believe it to be straight). But is it also the case—more controversially—that the building in the distance appears small (though I correctly infer that it is large, given its distance and apparent size), or that the penny seen from an angle appears to be oval (though I correctly infer that it is circular)? What, exactly, is it that I observe in such cases? What is it that I “just see,” without further ado? Pursuing these questions would illuminate the nature of empirical observation, and thus might illuminate the nature of moral intuitions as well. But they seem to me quite complex—and they would certainly take us rather far afield—so I am going to put them aside, and assume in what follows that empirical observations are, indeed, generally accurate.

⁴ Not surprisingly, talk of “responding” to the empirical world suggests that the world exists independently of—that is, metaphysically prior to—our empirical sensations and observations. But I take it that even “anti-realists” (who take the world to be somehow metaphysically constituted by our sensations or our reports of their contents) want a way to express the thought that our empirical observations are appropriately connected to the empirical facts, and for present purposes that is the only point at issue. Throughout this essay—both with regard to the empirical world, and the moral domain—I make use of familiar “realist” locutions. But I believe that roughly similar issues arise (concerning the reliability of both moral intuition and empirical observation) for both realists and anti-realists.

⁵ See, especially, Kagan, Shelly, The Limits of Morality (Oxford: Oxford University Press, 1989).Google Scholar

⁶ Strictly, of course, everyone accepts an error theory of at least a rather minimal sort, since no one thinks that moral intuitions can never be mistaken. But I have in mind more ambitious theories of this type.

⁷ In the basic case, a runaway trolley will hit and kill five children, unless you throw a switch which will divert the trolley onto a side track, saving the five, but killing a sixth child trapped on that side track (who would otherwise be safe). A large number of variants of this basic case have been discussed. See, e.g., Thomson, Judith Jarvis, “Killing, Letting Die, and the Trolley Problem,”Google Scholar in Thomson, , Rights, Restitution, and Risk, 78–93Google Scholar; and Kamm, Frances, “Harming Some to Save Others,” Philosophical Studies 57, no. 3 (1989): 227–60.CrossRefGoogle Scholar

⁸ The situation is further complicated by the fact that each side may fail to respond to features that the other side's intuitions mark out as morally significant. Thus, unlike the normal case of color blindness, moral disagreement may actually be closer to a situation in which many groups claim to see one or more colors that some other groups do not, and yet each group still fails to see some of the colors that other groups claim to see.

⁹ See, for example, Singer, Peter, “Sidgwick and Reflective Equilibrium,” The Monist 58, no. 3 (1974): 516.CrossRefGoogle Scholar

¹⁰ See, for example, Peter Unger's discussion of protophysics and psychological grouping principles in his Living High and Letting Die (New York: Oxford University Press, 1996).Google Scholar

¹¹ It might seem that an emotivist or expressivist account of moral claims would have an easy time accommodating these facts, since there is nothing especially surprising in the suggestion that people's emotional (and other) attitudes vary, and that they can be readily generated in response to never before considered cases. But even accounts of this kind, it seems to me, should be troubled by the ease and force with which intuitions can be generated in response to trolley problems (and the like) since it is not at all obvious why these should so readily engage our emotions or other pro-attitudes, nor why minor changes in the cases should elicit such drastically altered reactions.