¹ Sidgwick, Henry, The Methods of Ethics, 7th ed. (Chicago: University of Chicago Press, 1907)Google Scholar; this work will be referred to hereafter as ME, and page references will be given parenthetically in the text.

² Sidgwick's utilitarian reliance on such “secondary principles” closely resembles Mill's view; see Brink, David O., “Mill's Deliberative Utilitarianism,” Philosophy & Public Affairs, vol. 21 (1992), pp. 92–93.Google Scholar Moreover, it should be noted, this reliance on moral rules and motives is compatible with act utilitarianism. Acting on the best rules or motives will result in some wrong acts; but because these acts will be part of an optimal pattern of behavior, an act utilitarian can represent them as cases of blameless wrongdoing. If so, Sidgwick's reliance on rules and motives does not imply rule utilitarianism or motive utilitarianism

³ Cf. Simmons, A. John, “Utilitarianism and Unconscious Utilitarianism,” in The Limits of Utilitarianism, ed. Miller, H. and Williams, W. (Minneapolis: University of Minnesota Press, 1982).Google Scholar

⁴ Compare Rawls's description of the process of “reflective equilibrium”; see Rawls, John, A Theory of Justice (Cambridge, MA: Harvard University Press, 1971), pp. 19–21, 46–53, 579–81.Google Scholar Rawls claims a Sidgwickian pedigree for his view in A Theory of Justice, p. 51n.Google Scholar

⁵ Elsewhere, I have argued that the egoist rationale is metaphysically robust; see Brink, David O., “Sidgwick and the Rationale for Rational Egoism,” in Essays on Henry Sidgwick, ed. Schultz, B. (New York: Cambridge University Press, 1991).Google Scholar

⁶ Sidgwick thinks that there is something right or defensible about each of his three methods. Intuitionism is defensible when it is understood as a foundationalist epistemological view, whereas egoism and utilitarianism are substantive normative views. Sidgwick concludes that egoist and utilitarian first principles each rest on a fundamental intuition; it is this fact that constitutes the dualism of practical reason.

⁷ See also Sidgwick, Henry, “The Establishment of Ethical First Principles,” Mind, vol. 4 (1879), pp. 107–8Google Scholar; the essay will be referred to hereafter as “First Principles,” and page references will be given parenthetically in the text.

⁸ This metaphysical dependence of moral truths about action tokens on truths about right-making properties asserts of each token that it has the moral properties it does by virtue of certain properties, which that token possesses, being right-making properties. As such, this kind of metaphysical dependence is compatible with a wide variety of metaphysical views about the nature of properties and their relations to particulars.

⁹ Someone might wonder whether this metaphysical dependence of particular truths on general truths requires the epistemic dependence of particular beliefs on general beliefs. There are properties on which the gender of particular chicks depends and in virtue of which chicken-sexers sort particular chicks by sex. But we may nonetheless be inclined to ascribe knowledge to the reliable but inarticulate chicken-sexer who cannot cite these properties that justify his practice. This externalist view about knowledge may seem to undermine the inference from metaphysical dependence to epistemic dependence. But I am not sure about this. General beliefs, like any others, need not be articulate to play a role in justification or knowledge. We ascribe tacit or implicit beliefs when they are needed to explain regularities in a person's behavior. If the chicken-sexer's reliability is counterfactually stable in the relevant way, then there is a regularity whose explanation requires ascribing to the chicken-sexer implicit recognition of those properties that make the sex of a chick detectable. If the chicken-sexer's reliability is not counterfactually stable in the relevant way, then we are less inclined to ascribe knowledge to him, though we may attribute a certain knack to him.

A different worry is that the relata in metaphysical dependence are different from those in epistemic dependence. With metaphysical dependence, the fact that x is F depends upon the fact that x is G and on the fact that G is an F-making feature. But with epistemic dependence, it may seem, the justification of my particular belief that x is F depends upon some general belief—not about G being an F-making feature, but about the reliability of the cognitive process by which I formed my belief that x is F, for example, the belief that I am a reliable detector of F. However, it seems reasonable to suppose that the justification for this belief rests upon my being able to detect G and my belief that G is an F-making feature. If so, metaphysical dependence must eventually be represented in the content of one's justified beliefs.

¹⁰ There is an interesting difference in the apparent plausibility of locating the foundations at the level of general principle in the moral domain (philosophical intuitionism) and the nonmoral domain (philosophical foundationalism). Whereas Sidgwick argues that philosophical intuitionism is the most plausible form of intuitionism, philosophical foundationalism has not seemed especially plausible; instead, it has seemed more plausible to locate the foundations at the level of particular beliefs in the nonmoral domain (particular foundationalism).

This difference can, I think, be explained. Classical foundationalism treats as self-evident beliefs reporting (occurrent) sense data (e.g., the belief that I now see a reddish patch on my visual field). These beliefs concern private objects and, as a result, might seem to be infallible; they are beliefs whose private content is such that it is not clear how one could mistakenly hold a belief with that content. By contrast, particular moral beliefs (e.g., the belief that this act is wrong) concern common objects; as such, they seem eminently fallible.

Whereas classical foundationalism has seemed the most plausible form of foundationalism for the nonmoral domain, for familiar reasons, I do not think that it is. Classical foundationalism can hope to rest our knowledge on infallible foundations only by restricting these foundational beliefs to private objects. But this threatens to commit us to skepticism, because it seems we cannot move from beliefs exclusively about private objects to beliefs about common objects by deductive or abductive inference. Moreover, it is doubtful that infallible beliefs are self-justifying. Their justification would seem to depend upon recognition of their infallible character, and this will involve recognition of how beliefs with a certain kind of content could not be held mistakenly. But then a particular infallible belief will depend for its justification on general beliefs whose truth explains why the particular belief is true.

¹¹ On a familiar internalist reading of Sidgwick's views about the relationship between morality and rationality, he faces a problem reconciling the “dualism of practical reason” with his own philosophical intuitionism. For the dualism of practical reason consists in the fact that egoism and utilitarianism each rest on a fundamental intuition. But, on this internalist reading, egoism and utilitarianism are competing theories about morality and rational conduct. Yet if, as Sidgwick claims, fundamental intuitions must be mutually consistent (condition [c]), then it is not true that both egoism and utilitarianism can rest on fundamental intuitions. I argue that this is one reason, among others, to favor an externalist reading of the dualism of practical reason, according to which egoism is the correct theory of rationality, whereas utilitarianism is the correct theory of morality; see my “Sidgwick and the Rationale for Rational Egoism” (supra n. 5), section 2.

¹² It is sometimes held that dialectical reasoning has probative value only if one's initial beliefs are credible independently of their inferential connections. Cf. Lewis, C. I., An Analysis of Knowledge and Valuation (La Salle: Open Court, 1946), p. 339.Google Scholar But this is not the view that the present interpretation ascribes to Sidgwick. For the starting points of dialectical method are common moral beliefs; what Sidgwick (in some moods) takes to be self-evident are first principles, not common moral beliefs.

¹³ This passage is a problem for philosophical intuitionism as long as “proof” is understood as justification. The passage presents no problem for intuitionism, under this strategy for reconciling intuitionism and dialectical inquiry, if “proof” is understood to refer only to discursive demonstration or reasoning. Sidgwick clearly denies that “proof” here refers to discursive demonstration (ME, 419–20). At one point in this context, he seems to equate proof with dialectical reasoning (ME, 420), But then the passage would make the puzzling assertion that the dialectical support for utilitarianism makes the dialectical justification of utilitarianism as complete as possible. So I think it is unclear exactly how Sidgwick does want to understand proof in this passage. For this reason, I would not want to resist the first reconciliation strategy solely on the basis of this passage. The difficulty in squaring this reconciliation strategy with Sidgwick's reliance on the regress argument seems to be sufficient reason to resist this strategy.

¹⁴ Cf. Singer, Peter, “Sidgwick and Reflective Equilibrium,” The Monist, vol. 57 (1974), pp. 490–517CrossRefGoogle Scholar; Schneewind, Jerome, Sidgwick's Ethics and Victorian Moral Philosophy (Oxford: Clarendon Press, 1977), chs. 9–10Google Scholar; and Irwin, T. H., “Aristotle's Methods of Ethics,” in Studies in Aristotle, ed. O'Meara, D. (Washington, DC: Catholic University of America Press, 1981), p. 203.Google Scholar

¹⁵ “First Principles” (supra n. 7) was published two years after the second edition of The Methods of Ethics.

¹⁶ See Aristotle, , Posterior Analytics, 72al–5Google Scholar, and Metaphysics, 1029b3–12.Google Scholar

¹⁷ And this is not simply a development of contemporary moral and political theory. Consider, for example, Aristotle's dialectical reasoning from common or respected opinions at various points in the Nicomacltean Ethics or John Stuart Mill's defense of utilitarianism's implications in chs. 2 and 5 of Utilitarianism.

¹⁸ So, unlike Singer, Schneewind, and Irwin, I do not think one can provide consistent intuitionist readings of Sidgwick's various claims.

¹⁹ We can explain how it is that we are generally reliable only by assuming (defeasibly) that our general moral outlook is roughly right. For according to dialectical methods, the credentials of a particular belief or theory are normally assessed in the context of other beliefs that are not themselves in question. If we demand that all of these beliefs be justified at once, then according to dialectical methods, their justification can only be a matter of their mutual support. See Brink, David O., Moral Realism and the foundations of Ethics (New York: Cambridge University Press, 1989), pp. 122–25.CrossRefGoogle Scholar

²⁰ Protagoras offers just such a story to explain to Socrates why moral knowledge is distributed more widely than other forms of craft knowledge; see Plato, 's Protagoras, 323a, 324d–325a.Google Scholar

²¹ Whereas some may wish to derive the content of morality from strategic egoism or evolution, I have no such reductive ambitions. My claim is that, if the moral facts are roughly what common-sense morality says they are, then strategic and evolutionary considerations help explain why these facts are ones that we reliably recognize and attach importance to.

²² It is a defect of my discussion in Moral Realism and the Foundations of Ethics, ch. 5, that it omits this explanation of the reliability of considered moral judgments. Whereas that discussion aims primarily to explain why our considered moral beliefs are likely to be true, this discussion aims primarily to explain why true moral beliefs are likely to be commonly held.

²³ A more satisfactory treatment can be found in my Moral Realism and the Foundations of Ethics, pp. 197–209.Google Scholar