¹ See, for example, Beck, Lewis White, A Commentary on Kant's ‘Critique of Practical Reason’ (Chicago: University of Chicago Press, 1960), pp. 242–4Google Scholar; Garve, Christian, Versuche iiber verschiedene Gegenstande aus der Moral und Literatur (Breslau, 1792), part 1, pp. 111–16Google Scholar; Greene, Theodore, ‘The historical context and religious significance of Kant's Religion', in Religion within the Limits of Reason Alone, Greene, and Hudson, (trans.), (New York: Harper & Brothers, 1960), pp. lxii–lxiiiGoogle Scholar; and Schopenhauer, Arthur, The World as Will and Representation, vol. I, Payne, E. F. J. (trans.) (New York: Dover, 1969), p. 524Google Scholar.

² I will leave translations of ‘Triebfeder’ as ‘incentive’ as they are. However, I will use ‘spring’, ‘spring of action’ (following T. K. Abbott) or ‘drive’ (following L. W. Beck) to refer to what Kant called a Triebfeder. These latter terms better refer to inner sources of action than does ‘incentive’, which is frequently used to refer to external objects that appeal to agents’ desires, and thus has a misleading connotation.

³ Kant provides his most systematic account of the highest good in part one, book two, of KpV; that is the account on which I will primarily draw in this section. For an illuminating account of the development of Kant's notion of the highest good – including the nature of happiness that constitutes the conditioned element, and the role of God in realizing the highest good – see chapter 5 of Forster's, EckartKant's Final Synthesis: An Essay on the Opus Postumum (Cambridge, MA: Harvard University Press, 2000)Google Scholar.

⁴ I use the following abbreviations and translations of Kant's texts. Volume and page numbers refer to the Prussian Academy edition, except for VpR, for which I use only the translation's page numbers, and KrV, for which I use the A/B (first/second edition) format. C: ‘Moral philosophy: Collins's lecture notes’, trans. Heath, P., in Lectures on Ethics, ed. Heath, P. and Schneewind, J. B. (Cambridge: Cambridge University Press, 1977)Google Scholar; Do: ‘What is Orientation in Thinking?’, in Religion within the Boundaries of Mere Reason, trans, and ed. Wood, A. and Giovanni, G. Di, (Cambridge: Cambridge University Press, 1998)Google Scholar; G: Grounding for the Metaphysics of Morals, trans. Ellington, J. W. (Indianapolis: Hackett, 1963)Google Scholar; KpV: Critique of Practical Reason, trans. Beck, L. W. (New York: Macmillan, 1956)Google Scholar; KrV: Critique of Pure Reason, trans. Smith, Norman Kemp (London: Macmillan & Company, 1963)Google Scholar; KU: Critique of Judgment, trans. Pluhar, W. S. (Indianapolis: Hackett, 1987)Google Scholar; MS: Metaphysics of Morals, trans. Gregor, M. J. (Cambridge: Cambridge University Press, 1991)Google Scholar; Rel: Religion within the Boundaries of Mere Reason; TP: ‘On the common saying: this may be correct in theory, but it is of no use in practice’, in Kant: Practical Philosophy, trans, and ed. Gregor, M. J. (Cambridge: Cambridge University Press, 1996)Google Scholar; V: ‘Kant on the metaphysics of morals: Vigilantius's lecture notes’, trans. Heath, P. in Lectures on Ethics; VpR: Lectures on Philosophical Theology, trans. Wood, A. and Clark, G. (Ithaca, New York: Cornell University Press, 1978)Google Scholar.

⁵ I will assume that proportionality is part of Kant's conception of the highest good. For a critical discussion of proportionality, see Reath, Andrews, ‘Two conceptions of the highest good in Kant’, Journal of the History of Philosophy 26 (4) (1988), 593–619CrossRefGoogle Scholar. For a defence of proportionality, see Engstrom, Stephen, ‘The concept of the highest good in Kant's moral theory’, Philosophy and Phenomenological Research SI (4) (1992), 749–80Google Scholar.

⁶ There are a number of ambiguities in Kant's discussion of the highest good and the importance of faith in God for believing that the highest good is realizable. I discuss several of these at length in ‘Kant's criticisms of atheism’, Kant-Studien 94 (2003), 198–219Google Scholar. One of these issues concerns whether we must postulate God's existence, or simply the real possibility of it. Here I shall talk in terms of God's existence (as Kant usually does in KpV. For an argument that we need postulate only the possibility of God's existence, see Ferriera, M. Jamie, ‘Kant's postulate: the possibility or the existence of God?’, Kant-Studien 74 (1983), 75–80Google Scholar.

⁷ Of course, it is a mistake to equate teleology with consequentialism. Moreover, commentators have recently challenged the traditional classification and understanding of Kant's theory as deontological and therefore not teleological. See Herman, Barbara, The Practice of Moral Judgment (Cambridge, MA: Harvard University Press, 1993)Google Scholar, chapter 10; Guyer, Paul, Kant on Freedom, Law, and Happiness (Cambridge: Cambridge University Press, 2000)CrossRefGoogle Scholar, chapters 3 and 4; Wood, Allen, Kant's Ethical Thought (Cambridge: Cambridge University Press, 1999), pp. 112–14CrossRefGoogle Scholar; and Auxter, Thomas, Kant's Moral Teleology (Macon, GA: Mercer University Press, 1982)Google Scholar.

⁸ See Engstrom, ‘The concept of the highest good in Kant's moral theory’, sections one and two, for a fuller development of this line of argument.

⁹ Guyer develops and discusses this approach intermittently throughout Kant on Freedom, Law, and Morality, but see especially pp. 339-45 and 386-8. A deep analysis of the differences of these approaches, and an exploration of why Kant sometimes emphasizes one approach and sometimes another, is beyond the scope of this article – although it is clear that he sometimes (for example, in TP) emphasizes the second approach to quell objections regarding the presence of happiness within the highest good.

¹⁰ From what we see around us, nature does not sufficiently cooperate. See, for example, KU 451-3.

¹¹ On the highest good as a social good, see Anderson-Gold, Sharon, Unnecessary Evil: History and Moral Progress in the Philosophy of Immanuel Kant (New York: SUNY Press, 2001)Google Scholar; Yovel, Yirmiyahu, Kant and the Philosophy of History (Princeton: Princeton University Press, 1980)Google Scholar; and Michalson, Gordon, Kant and the Problem of God (Maiden, MA: Blackwell, 1999), chapter 5Google Scholar.

¹² Kant also discusses God's grace as necessary for the full realization of virtue (or for our being given credit for holiness). See Rel 44-53,72-8,190-202. I Also see Hare, John, The Moral Gap (Oxford: Clarendon Press, 1996)Google Scholar.

¹³ In Versuche über verschiedene Gegenstände aus der Moral und Literatur.

¹⁴ Kant's Moral Religion (Ithaca, N.Y.: Cornell University Press, 1970), chapter 2.Google Scholar

¹⁵ Cf. KpV 5: 109-110, where Kant (confusingly) explains how the highest good can be the determining ground or motive for the pure will in a manner ‘consistent with autonomy: because morality itself is the unconditioned,; supreme element within the highest good, no end outside of the will (or prior to or independent of the moral law) serves as the will's determining ground.

¹⁶ By contrast, in the Groundwork, Kant implies that springs [Triebfeder] are all inclination- or need-based, and that the moral law can be practically understood only as a motive [Beweggund, Bewegungsgrund], a reason-based, objective determining ground for an agent's will (G 4: 427,444).

¹⁷ Henry Allison calls this the ‘Incorporation Thesis'; see Kant's Theory of Freedom (New York, Cambridge University Press, 1990), pp. 39–40Google Scholar, 126-7. On how this process works, see Grenberg, Jeanine M., ‘Feeling, desire and interest in Kant's theory of action’, Kant-Studien 92 (2001), 153–79CrossRefGoogle Scholar.

¹⁸ Virtue also involves autocracy, the capacity to master one's inclinations. See , Allison, Kant's Theory of Freedom, pp. 164Google Scholar, 246, 285 n. 35; and Baxley, Anne Margaret, ‘Autocracy and autonomy’, Kant-Studien 94 (2003), 1–23CrossRefGoogle Scholar.

¹⁹ Prima facie, passages from student lecture notes, especially early ones, promise a less reliable representation of Kant's views than passages from his published works. I judge these passages worthy of consideration not only because of their inherent interest, but also because of their number, their similarities and the wide time-span during which they were recorded.

²⁰ Kant does not imply that the prospect of eventual happiness is always sufficient to motivate right action.

²¹ For some clear statements of Kant's distinguishing between judging some-thing to be right and being motivated by that, see VpR 141; C 27: 274-5 [1428].

²² One can find concerns about passages like (la)–(3b) in a number of well-known criticisms of Kant's highest good, such as Greene's ‘The historical context and religious significance of Kant's Religion’;. I suspect, however, that because some of these critics also make the mistake of being disturbed by the presence of happiness in the highest good, or by not paying enough attention to the end/drive distinction, these critics were not able to focus on the troubling claims I have listed.

²³ For an argument against, and an alternative to, seeing Kantian autonomy as all or nothing, see Engstrom, Stephen, ‘Conditioned autonomy’, Philosophy and Phenomenological Research 48 (3) (1988), 435–53CrossRefGoogle Scholar.

²⁴ Wood, Kant's Moral Religion, chapter 4.

²⁵ ‘Disbelief is ambiguous. Kant's primary target is dogmatic atheism (which denies the existence, and real possibility, of God), but he is also critical of sceptical atheism (which accepts God's real possibility but does not affirm his existence). Also ambiguous is the problem disbelief presents for respect. Some of Kant's accounts suggest a logical or conceptual problem (that if one does not believe in God's existence, or at least real possibility, one cannot believe that the highest good is a real possibility, and thus the moral law will strike one as ‘inherently false’ (KpV 5:114)); others suggest a more practical, perhaps even psychological, problem (that if one does not believe the highest good is a real possibility, one's respect for the moral law will become ‘weakened’ and one's ‘moral attitude’ ‘impair[ed]’ (KU 5: 452-3). For more on both these points, see my ‘Kant's criticisms of atheism’.

²⁶ Passage (2e), which includes, ‘If I have so conducted myself as to be worthy of happiness, I can also hope to enjoy the latter, and such are the springs of morality’ (C 27: 304), is perhaps the passage least amenable to the previous interpretation, for it – much more than (2a)–(2d) – suggests that an agent's hopes for her ow n happiness themselves constitute springs of moral action.

²⁷ For the sake of brevity, I will not here explore whether the fact that it is hope for our deserved happiness rather than happiness simpliciter that reduces the tension between the claim of passages (2a)–(2e) and Kant's notion of autonomy. I consider this in the discussion of claim (1).

²⁸ What is unexpected is not Kant's claim that the representation of duty alone cannot move a human agent to action, but his suggestion that something other than respect for the moral law (which one would expect to accompany the representation of duty) is an appropriate spring of action to attach to that representation.

²⁹ Kant's justification for ascribing radical evil to the human race is itself the topic of ongoing debate. See Wood, Allen, Kant's Ethical Thought, pp. 283–90Google Scholar; and Allison, Henry, ‘Ethics, evil, and anthropology in Kant: remarks o n Allen Wood's Kant's Ethical Thought’, Ethics 111 (3) (2001), 594–613CrossRefGoogle Scholar, especially 605-10.

³⁰ Marcia Baron considers this in ‘Freedom, frailty, and impurity’, Inquiry 36 (1993), 431–41CrossRefGoogle Scholar.

³¹ For another treatment of Kant o n moral motivation, religion and emotion, see Ameriks, Karl, Kant and the Fate of Autonomy (New York: Cambridge University Press, 2000), pp. 317–31CrossRefGoogle Scholar.

³² Of course, from a practical point of view, we and he must see him as able to act rightly for its own sake.

³³ See, for example, Baron, Marcia, Kantian Ethics Almost Without Apology (Ithaca, N.Y.: Cornell University Press, 1995), pp. 152–6Google Scholar.

³⁴ See, for example, Baker, Judith, ‘Do one's motives have to be pure?’, in Philosophical Grounds of Rationality, Grandy, Richard and Warner, Richard (eds) (London: Oxford University Press, 1986), pp. 457–73Google Scholar.

³⁵ Also see MS 6: 406 for Kant's comments on using ‘feelings that accompany the constraining power of the moral law (e.g., disgust, horror, etc…)’ which ‘make its efficacy felt, in order to get the better of merely sensible incitements’.

³⁶ I am not thinking about how the feeling of sympathy or the desire for happiness can motivate the agent (alone or in combination with morality). Rather, I am thinking about how the agent's association of beneficent, right, or good action with satisfaction and future happiness alters or shapes the way the agent views her options: how her positive associations with morally good behaviour make wrong action look less tempting from the point of view of self-love than they would were the agent not sympathetic or lacked hope in the highest good. More subtly, because of the universal nature of the highest good and the necessity of working together with others to bring it about, faith in its possibility removes the common association of promoting others’ well-being with sacrificing one's own, and encourages the association between others’ happiness and one's own. So it removes some of the dazzle from temptations to short-change others or ignore their happiness, and puts helping others in a more attractive light.

³⁷ Again, even if a few of these passages from Kant's early writings or his students’ lecture notes are inconsistent with Kantian autonomy, this would not undermine our sense of Kant's considered view: that human agents are autonomous. But such passages could increase our understanding of the development of Kant's views of motivation, and of his moral objections to atheism.