¹ Cummiskey, David, Kantian Consequentialism, New York and Oxford, 1996CrossRefGoogle Scholar .

² R. M. Hare offers a different line of reasoning for the claim that Kant's ‘formal theory can certainly be interpreted in a way that allows him – perhaps even requires him – to be one kind of utilitarian’. See ‘Could Kant Have Been a Utilitarian?’, Utilitas, v (1993)Google Scholar .

³ There is controversy over what counts as ‘rational nature’ in the humanity formulation, but I do not want that to be the focus of this paper. The details of the definition do not affect the thrust of the arguments in this paper. Cummiskey (p. 85) adopts Christine Korsgaard's definition of ‘rational nature’ (or ‘humanity’) as something like the ‘capacity to set oneself an end’. For various positions on what Kant means by ‘rational nature’, see Hill, Thomas E. Jr, ‘Humanity as an End in Itself’, Dignity and Practical Reason in Kant's Moral Theory, Ithaca and London, 1992, pp. 38–57;Google ScholarKorsgaard, Christine, ‘Kant's Formula of Humanity’, Creating the Kingdom of Ends, Cambridge, 1996, pp. 106–32;CrossRefGoogle ScholarWood, Allen, ‘Humanity as End in Itself’, Proceedings of the Eighth International Kant Congress, vol. I, pt. 1 (1995), andGoogle ScholarDean, Richard, ‘What Should We Treat As An End In Itself?’, Pacific Philosophical Quarterly, Ixxvii (1996)Google Scholar .

⁴ Cummiskey does disagree with some of what Kant says at this level (for instance, Cummiskey, like some other commentators, thinks that the universalizability formulation of the Categorical Imperative does not provide substantial moral guidance by itself), but the main disagreement lies elsewhere.

⁵ As a referee for Utilitas has pointed out to me, there may be room to question whether Cummiskey's Kantian consequentialism is really a form of consequentialism at all. After all, it is unlike most consequentialist theories in several ways. It does not require maximization of the object with the greatest value (rational agency), but rather of the necessary conditions for rational agency, and it requires maximization not just of necessary conditions, but of equal necessary conditions. It also includes asymmetries in one's duties to oneself and others. It may be that the duties Cummiskey describes are not properly called consequentialist, and if Cummiskey only succeeds in showing that Kant's ethics generates particular kinds of non-consequentialist duties, that is not such dramatic news. However, the focus of this paper will be a different objection to Cummiskey – namely that the principles comprising Kantian ‘consequentialism’ (whether that is truly a form of consequentialism or not) do not really follow from the ideas of value in Kant's humanity formulation.

⁶ These categories of duties are not the same as Kant's categories of perfect and imperfect duties to one's self and to others. That is no surprise, since I am reconstructing Cummiskey's argument here and he actually thinks the perfect/imperfect duty distinction is philosophically unjustified for Kant (see pp. 105–23).

⁷ Onora O'Neill is a non-consequentialist Kantian whose views could arguably lead to the conclusion that each of us has a duty to maximally promote the conditions for rational agency. In Constructions of Reason, Cambridge, 1989, pp. 228–33, she argues for an imperfect duty to provide others with the conditions of rational agency. Although she calls the duty imperfect (inasmuch as it is not a duty owed to any specific person), and although O' Neill is not mainly concerned here with exegesis of Kant's texts, the rationale she provides for this duty could be taken as a broadly Kantian rationale for a duty of maximizationGoogle Scholar.

⁸ The duty to see that rational agents are free from destruction might also fall under the category of providing others with as much security as possible.

⁹ Although I think there is a problem with the fundamental strategy of Cummiskey's reconstruction, I do think he is right that some of the details are fairly ‘standard’, inasmuch as Cummiskey adopts roughly the strategy that one prominent commentator, Christine Korsgaard, follows. See Korsgaard, pp. 119–24.

¹⁰ Cummiskey, p. 74. I am simplifying Cummiskey's argument here, leaving out four paragraphs about Thomas Nagel that intervene between the rational egoist response and the quotation just cited. This simplification does not unfairly misrepresent Cum-miskey's main point.

¹¹ Kant, Immanuel, Groundwork of the Metaphysics of Morals, trans. Paton, H. J., New York, 1964, p. 103Google Scholar [Akademie p. 436], and Kant, Immanuel, Critique of Practical Reason, trans. Beck, Lewis White, New York, 1993, p. 60 [Akademie p. 58[, pp. 62 f. [60], and pp. 64–7 [62–4]Google Scholar.

¹² The quotations in this paragraph are from Kant, , Groundwork, p. 96 [428 f.]Google Scholar.

¹³ Kant, , Groundwork, pp. 95–7 [428–30]Google Scholar.

¹⁴ See Kant, Immanuel, Metaphysics of Morals, trans. Gregor, Mary, Cambridge, 1996, pp. 145–232CrossRefGoogle Scholar [379–491], and a Kant, , Groundwork, pp. 97 f. [430]Google Scholar .

¹⁵ This is not the same as saying that she cannot destroy her rational nature for any reason, so it leaves open the possibility that self-sacrifice may be morally permissible.

¹⁶ Like Cummiskey, I am borrowing something like Korsgaard's ‘regress argument’ here. See Korsgaard, pp. 119–24.

¹⁷ This reading of the requirements imposed by the humanity formulation is con-sistent with, and influenced by, Hill, pp. 144 f.

¹⁸ For more on this consequentialist concept of value, see, for example, Kagan, Shelly, The Limits of Morality, Oxford, 1989, pp. 59–62Google Scholar.

¹⁹ This is the title of section III of ch. 5 of the book, pp. 95–7.

²⁰ Kant's, most thorough discussion of relative ends is in Groundwork, pp. 95 f. [27 f.]Google Scholar.