¹ This paper was first given at a conference on moral psychology at Monash, August 1991, and has since been read at the University of Queensland, the Australian National University, and the University of Delhi. I am indebted to those present on all these occasions for stimulating and searching comments. I am especially grateful to Philip Pettit and Richard Holton for helpful discussion, and to Margaret Wilson and Christine Korsgaard for written comments on an earlier draft.

² According to Zweig, Arnulf in his introduction to Kant: Philosophical Correspondence, 1759–1799 (University of Chicago Press, 1967), 24.Google Scholar

³ Letter to Kant from Borowski, Ludwig Ernst, probably 08 1791Google Scholar. The correspondence between Kant and Maria von Herbert, and the related letters, are in Volume XI of the Prussian Academy of Sciences edition of Kant's works (de Gruyter, Walter, 1922)Google Scholar. The English translations given in this paper are closely based on those of Arnulf Zweig, partly revised in the light of the Academy edition, and very much abridged. See Zweig, Arnulf, Kant: Philosophical Correspondence, 1759–99Google Scholar. ©1967 by The University of Chicago. All Rights Reserved. I make use of the translations with the kind permission of Professor Zweig and the University of Chicago Press. Readers who would like to see fuller versions of the letters than those given here should consult the Academy edition, or the Zweig translations.

⁴ Kant, Immanuel, The Doctrine of Virtue, Part II of The Metaphysic of Morals, Gregor, Mary (trans.) (Harper and Row, 1964)Google Scholar. One wonders whether these parts of The Doctrine of Virtue may have been influenced by Kant's thoughts about Herbert's predicament. An alternative explanation might be that The Doctrine of Virtue and Kant's letter to Herbert are both drawing on Kant's lecture notes.

⁵ Justin Oakley has commented on what he calls the ‘moral ineptitude’ of people who act from the motive of duty alone: an ineptitude that prevents them from achieving the very goal duty aims at, for example, that of comforting a friend. See ‘A Critique of Kantian Arguments against Emotions as Moral Motives’, History of Philosophy Quarterly 7, No. 4 (10 1990), 441–59.Google Scholar

⁶ First published in 1797; an English translation is included as an addendun to Beck, L. W.'s translation of The Critique of Practical Reason (University of Chicago Press, 1949).Google Scholar

⁷ As she says on page 327 of ‘The Right to Lie: Kant on Dealing with Evil’, Philosophy and Public Affairs 15, No. 4 (1986), 325–49.Google Scholar

⁸ Korsgaard's views on Kant and lying are developed in ‘The Right to Lie’ (ibid.). Her views on Kant and friendship are developed in ‘Creating the Kingdom of Ends: Responsibility and Reciprocity in Personal Relations’, forthcoming in Philosophical Perspectives 6: Ethics, Tomberlin, James (ed.) (Atascadero, California: The Ridgeview Publishing Company, 1992)Google Scholar. As will become evident, I owe a very great debt to Korsgaard's approach in both papers.

⁹ Kant is wrong about this. We often value particular items in such a way that they aren't replaceable by a duplicate: it is this very teacup that I value, this very house, this very painting.

¹⁰ ‘Freedom and Resentment’, in Freedom and Resentment (Methuen, 1974), 1–25.Google Scholar

¹¹ Ibid., 9.

¹² Op. cit., note 8.

¹³ Groundwork of the Metaphysics of Morals, Paton, (trans.) (Harper and Row, 1964), 435.Google Scholar

¹⁴ Strawson, , op. cit., note 10, 9.Google Scholar

¹⁵ The phrase is Korsgaard, 's, op. cit., note 7, 335.Google Scholar

¹⁶ Op. cit., note 13, 429. Kant has many formulations of the categorical imperative. I restrict myself in this paper to the formula of humanity, partly because it best captures the ‘attractive ideal’ Korsgaard and I are hoping to find.

¹⁷ Ibid., 430. This characterization draws heavily on Korsgaard.

¹⁸ To say this is not yet to address Kant's own characterization of what's wrong with lying in his letter, namely that it is a failure of duty to the self. That is because I think that what he says here is less plausible. He says, in the Doctrine of Virtue, that to lie is to violate humanity in one's own person. It is to use one's natural being, one's power of speaking, as a means to an end that is nothing to do with ‘the intrinsic end’ of speech, which is the communication of thought. In this respect, lying is an unnatural use of one's natural self, like masturbation (op. cit., note 4, 93, 94). Herbert's view, that honesty is some thing one owes to a friend, owes to others, has far more going for it, and finds independent support in Kant.

¹⁹ Ibid., 430. Kant is alluding to John 8:44.

²⁰ Ibid., 466.

²¹ Op. cit., note 13, 429.

²² See also Zweig, , op. cit., note 2, 24.Google Scholar

²³ Op. cit., note 8, 4.

²⁴ Op cit., note 4, 469.

²⁵ Ibid., 468, 469.

²⁶ Ibid., 471, my italics. This is a remarkable metaphor for a philosopher who finds in the autonomous human self, and its self-legislating activity, the only source of intrinsic value.

²⁷ Ibid., 471.

²⁸ Ibid., 471.

²⁹ Op. cit., note 8, 8.

³⁰ Op. cit., note 13, 397.

³¹ Ibid., 398.

³² For a defence of the latter reading, see Herman, Barbara, ‘On the Value of Acting from the Motive of Duty’, Philosophical Review 90, No. 3 (1981), 359–82.CrossRefGoogle Scholar

³³ Kant, Immanuel, Critique of Practical Reason, Beck, L. W. (trans.) (Macmillan, 1956), 97, 101.Google Scholar

³⁴ Ibid., 119.

³⁵ Ibid., 80.

³⁶ Ibid., 117.

³⁷ Op. cit., note 13, 428.

³⁸ Op. cit., note 4, 407.

³⁹ Op. cit., note 33, 118.

⁴⁰ Ibid.

⁴¹ See Wolf, Susan, ‘Moral Saints’, The Journal of Philosophy 79 (1982), 419–39CrossRefGoogle Scholar, on the perils of sainthood.

⁴² See for example op. cit., note 4, 456.

⁴³ Op. cit., note 33, 109.

⁴⁴ Letters 1, 3 and 4 above. Elisabeth Motherby was the daughter of Kant's friend Robert Motherby, an English merchant in Königsberg.

⁴⁵ Op. cit., note 4, 462, my italics.

⁴⁶ In the non-ideal case, she says, one's actions may be guided by a more instrumental style of reasoning than in ideal theory. But Korsgaard also wants to say that non-ideal theory is not a form of consequentialism. The goal set by the ideal is not one of good consequences, but of a just state of affairs (op. cit., note 7, 343). I am unhappy with this defence: her view does apparently endorse a kind of consequentialism. It permits actions that are likely to have a certain good consequence, namely a just state of affairs. (Of course this is not consequentialism of a utilitarian variety.)

⁴⁷ Ibid. There are constraints on what we are permitted to do for the sake of this end. Briefly: on Korsgaard's view, the formula of universal law provides constraints on what we are permitted to do in our attempts to make the Kingdom a reality, when faced with evil. In the case of lying to the murderer, she believes that the formula of universal law, correctly applied, will yield a permissive conclusion. Korsgaard's reconstruction is subtle and complex, drawing on a Rawlsian distinction between ideal and non-ideal theory. I have not done Korsgaard justice here, and readers are referred to her article for the fuller picture. I should also say that my final conclusion, namely that in this case Herbert may have had a duty to lie, is not endorsed by Korsgaard herself.

⁴⁸ Letter 3, my italics.

⁴⁹ Op. cit., note 4, 471. Kant's ignorance of Antipodean bird life is (just) forgivable.

⁵⁰ Ibid., 471.

⁵¹ Ibid., 434, 435.

⁵² There is one final letter from her on the record, dated early 1794, in which she expresses again a wish to visit Kant, and reflects upon her own desire for death.

⁵³ Ibid., 424.