¹ First person reasons are sometimes called “subjective” reasons, and third person reasons “objective.” However, because the terms “subjective” and “objective” have so many uses (including a different use later in this paper), I have adopted the “first person”/“third person” terminology to avoid confusion.

² For this view of morality, see Baier, Kurt, The Moral Point of View (Ithaca, NY: Cornell University Press, 1958), 189–91, 308–15Google Scholar; Frankena, William, Ethics (Englewood Cliffs: Prentice-Hall, 1963), 18Google Scholar; Cahn, Steven M., Puzzles and Perplexities (Lanham, MD: Rowman and Littlefield, 2002), 68–69Google Scholar.

What is meant by an action's having a “favorable relation” to one's interests? Roughly, that the action would make one better off, or one has justification for believing that it would do so, or one's expected welfare conditional on one's performance of the action is greater than one's unconditional expected welfare, … or something in this neighborhood. I leave the phrase “favorable relation” in the text to avoid unnecessary debates about exactly what the relationship between an action and one's interests must be for one to have prudential reason to perform the action.

³ For this view of ethics, see Hobbes, Thomas, Leviathan (1651; Cambridge: Cambridge University Press, 1996)Google Scholar; Rand, Ayn, The Virtue of Selfishness (New York: Signet, 1964)Google Scholar. Platocan also be read this way inThe Republic (Indianapolis: Hackett, 1974)Google Scholar.

⁴ For this view of moral reasons, see Kant, Immanuel, Foundations of the Metaphysics of Morals, trans. Beck, Lewis White (Indianapolis, IN: Bobbs-Merrill, 1959), 31–34Google Scholar; Joyce, Richard, The Myth of Morality (Cambridge: Cambridge University Press, 2001), 42–45CrossRefGoogle Scholar. On the notion of the “favorable relation” to desire-satisfaction, see note 2 above.

⁵ On this extra content of “morality,” see my “Values and Morals: Outline of a Skeptical Realism,” Philosophical Issues 19 (2009): 113–30CrossRefGoogle Scholar.

⁶ For a more precise account of objectivity, see my Ethical Intuitionism (New York: Palgrave Macmillan, 2005), 2–4Google Scholar.

⁷ An exception is the case in which a non-absurd proposition is one of an infinite class of similar alternatives. For instance, if one knows only that a given variable must take on exactly one value in a certain continuous range, then, in the standard treatment, each of the continuum many possibilities will be assigned probability zero. However, these cases are not relevant to the argument of this paper.

⁸ The Probabilistic Reasons Principle is an indicative conditional, but the nested conditionals (a) and (b) are subjunctive.

⁹ The need for the “thereby” qualifier is created by the tendency of philosophers to misinterpret “if-then” statements as material conditionals, that is, to interpret “If P then Q” as meaning “Either P is false or Q is true,” such that there is no requirement that P and Q should have any connection to each other. My use of “thereby” blocks this interpretation.

¹⁰ I thank Loren Lomasky for pointing this out, using a different example.

¹¹ I thank Justin Weinberg for something vaguely like this objection.

¹² I thank Sarah Raskoff for discussion of this issue. See also Kolodny, Niko, “Why Be Rational?” Mind 114 (2005): 509–63CrossRefGoogle Scholar.

¹³ The sentence “S has the goal of doing what S has most reason to do” is ambiguous. One reading, which I call the de dicto reading (after a distinction in the philosophy of language literature), is the following: “S wants the following to be the case: [for all x, if S has most reason to do x, then S does x].” The other reading, which I call the de re reading, is this: “For some x, S has most reason to do x, and S wants the following to be the case: [S does x].” The de dicto reading is the correct one.

¹⁴ For discussion of virtuous and benign infinite series, see my “Virtue and Vice Among the Infinite” in Ad Infinitum: New Essays on Epistemological Infinitism, ed. John Turri and Peter Klein (New York: Oxford University Press, forthcoming). Note that the practical-reasons regress is clearly benign on the account defended in that article, since the infinite series of practical reasons does not require anything to instantiate an infinite intensive magnitude.

¹⁵ See David Hume, A Treatise of Human Nature (Buffalo, NY: Prometheus), 413–18, 455–70; Smith, Michael, The Moral Problem (Oxford: Basil Blackwell, 1994), 12Google Scholar; Mackie, J. L., Ethics: Inventing Right and Wrong (New York: Penguin, 1977), 40–41Google Scholar; Gibbard, Alan, Wise Choices, Apt Feelings (Cambridge, MA: Harvard University Press), 33Google Scholar; Hare, R. M., Objective Prescriptions and Other Essays (Oxford: Clarendon, 1999), 1Google Scholar; Nowell-Smith, P. H., Ethics (Baltimore, MD: Penguin, 1954), 39–42Google Scholar; Joyce, Richard, The Myth of Morality (Cambridge: Cambridge University Press, 2001)CrossRefGoogle Scholar; and Fumerton, Richard, Reason and Morality: A Defense of the Egocentric Perspective (Ithaca, NY: Cornell University Press, 1990), chap. 3Google Scholar.

¹⁶ See Hume, Treatise of Human Nature, book II, part III, section iii.