Hostname: page-component-594f858ff7-pr6g6 Total loading time: 0 Render date: 2023-06-07T18:32:31.330Z Has data issue: false Feature Flags: { "corePageComponentGetUserInfoFromSharedSession": true, "coreDisableEcommerce": false, "corePageComponentUseShareaholicInsteadOfAddThis": true, "coreDisableSocialShare": false, "useRatesEcommerce": true } hasContentIssue false

On the Possibility of Kantian Retributivism

Published online by Cambridge University Press:  01 September 2009

New York


One of the most potent motivations for retributivist approaches to punishment has been their apparent connection to an ethical background shaped by the Kantian notion of morally autonomous and rational human agency. The present article challenges the plausibility of this connection. I argue that retributivism subverts, rather than embodies, the normative consequences of moral autonomy, justifying a social practice that conflicts with the considered judgments that the proper recognition of moral autonomy would authorize. The core of my case is the analysis of whether a punishment should be understood as a restriction of a criminal's freedom properly understood. I argue that the affirmative view faces serious difficulties that have not been, and are not likely to be, resolved by retributivist justifications that draw their support from Kantian moral theory.

Research Article
Utilitas , Volume 21 , Issue 3 , September 2009 , pp. 276 - 296
Copyright © Cambridge University Press 2009

Access options

Get access to the full version of this content by using one of the access options below. (Log in options will check for institutional or personal access. Content may require purchase if you do not have access.)


1 Cottingham, John, ‘Varieties of Retribution’, Philosophical Quarterly 29 (1979), pp. 238–46CrossRefGoogle Scholar, describes no less than nine ‘varieties of retributivism’, and his list is, surely, far from complete to date.

2 Both the Oxford English Dictionary and Cottingham's taxonomy of usage in justificatory theories of punishment (Cottingham, ‘Varieties of Retribution’) identify this to be ‘the basic or fundamental sense of “retribution”’.

3 Kant, Groundwork of the Metaphysics of Morals, trans. H. J. Paton (New York, 1964), p. 114.

4 Kant, Groundwork, p. 108.

5 In one of his most strongly worded statements about their relationship, Kant says that they ‘reciprocally imply each other’ (‘Critique of Practical Reason’, Critique of Practical Reason and Other Writings in Moral Philosophy, trans. L. W. Beck (Chicago, 1949), p. 140).

6 See Thomas E. Hill, ‘Humanity as an End in Itself’, Dignity and Practical Reason in Kant's Moral Theory (Ithaca, 1992), pp. 38–57, for an interpretation of the notion of an ‘end in itself’ that emphasizes its dependence on the capacity for a self-chosen autonomous will.

7 Kant, Groundwork, p. 96.

8 Kant, Metaphysics of Morals, trans. Mary Gregor (Cambridge, 1996), p. 25.

9 Taken to the extreme, this is tantamount to saying that desert does not automatically imply a moral permission, let alone a moral duty, to punish (cf. David Dolinko, ‘Some Thoughts about Retributivism’, Ethics 101 (1991), pp. 537–59). Indeed, as I argue below, even the assertion of the freedom-consistency of a punishment that prevents the commission of a crime by the punished subject is not unproblematic.

10 This practice of (possibly coercive) preemption bears close resemblance to what Sorell, Tom, ‘Punishment in a Kantian Framework’, Punishment and Political Theory, ed. Matravers, Matt (Oxford, 1999), pp. 1027Google Scholar, calls ‘paternalistic coercion’ – a coercive practice to which a state might resort to bring citizens’ behavior in line with the behavior they would engage in had they been autonomous – including, in particular, behavior that is consistent with the laws of a just state. A key assumption that underlies both of them is the error theory of choice that Sorell argues ‘provides the background in Kant for the pursuit of autonomy’ (p. 27) but is often missed by Kant's modern interpreters. The error theory provides a natural background for the claim, which figures prominently in the discussion below, that a criminal does not lose his or her humanity, and so his or her claim to autonomy, upon committing a crime.

11 Sarah Holtman, ‘A Kantian Approach to Prison Reform’, Jahrbuch für Recht und Ethik (1997), pp. 315–31, at p. 6.

12 Byrd, B. Sharon, ‘Kant's Theory of Punishment: Deterrence in Its Threat, Retribution in its Execution’, Law and Philosophy 8 (1989), pp. 198CrossRefGoogle Scholar.

13 For example, Holtman, ‘A Kantian Approach’; Byrd, ‘Kant's Theory’; Murphy, Jeffrie G., ‘Does Kant Have a Theory of Punishment?’, Columbia Law Review 87 (1987), pp. 509–32CrossRefGoogle Scholar; Scheid, Don E., ‘Kant's Retributivism’, Ethics 93 (1983), pp. 262–82CrossRefGoogle Scholar. Murphy, ‘Does Kant’, provides evidence of considerable revisions and ambiguities in Kant's own statements on punishment from his earlier and later works.

14 For example, Mundle, C. W. K., ‘Punishment and Desert’, Philosophical Quarterly 4 (1954), pp. 216–28CrossRefGoogle Scholar; Hart, H. L. A., ‘Prolegomenon to the Principles of Punishment’, Punishment and Responsibility (Oxford, 1968), pp. 127Google Scholar; Byrd, ‘Kant's Theory’.

15 Burgh, Richard W., ‘Do the Guilty Deserve Punishment?The Journal of Philosophy 79 (1982), p. 195CrossRefGoogle Scholar, emphasis added.

16 This echoes Sorell's argument that ‘paternalistic coercion’, e.g. in taking cigarettes away from a smoker, is consistent with Kant's Principle of Humanity because the smoker could reasonably share in the end of that action (‘Punishment’, pp. 12–14).

17 Among many points of textual support for this claim, see Kant's characteristic argument about the inappropriateness of rape as a punishment for a rapist, because that would effectively imply the lack of respect for the rapist's humanity (The Metaphysical Elements of Justice, trans. John Ladd (Indianapolis, 1965), p. 333). This is so because, as Kant insists in the Groundwork, freedom is a property of all rational beings ‘endowed with a will’ (Groundwork, p. 115). If it could be shown that the rapist were incapable of will (i.e. incapable of practical reason), respect for humanity would cease to be a constraint on his punishment. Of the more recent Kantian treatments of punishment, Morris, Herbert, ‘Persons and Punishment’, The Monist 52 (1968), pp. 475501CrossRefGoogle Scholar, and Holtman, ‘A Kantian Approach’ and ‘Toward Social Reform: Kant's Penal Theory Reinterpreted’, Utilitas 9 (1997), pp. 3–21, develop arguments emphasizing the importance of punitive practices’ respecting the personhood of criminals.

18 See the explicit statements by Kant, Metaphysics, p. 25. Morris, ‘Persons and Punishment’, p. 497, maintains that a criminal, in choosing a crime, necessarily wills her punishment; withholding it, therefore, constitutes a denial of her freedom and of her claim to a rational nature. Similarly, Jeffrie G. Murphy, Kant: the Philosophy of Right (Macon, 1994; 1st edn., 1970), p. 121, argues that with regard to punishment, ‘the criminal himself has no complaint, because he has rationally willed or consented to his own punishment’.

19 Kant, Groundwork, p. 97.

20 Kant, Metaphysics, p. 108.

21 Cf. Murphy, Kant.

22 Fleischacker, Samuel, ‘Kant's Theory of Punishment’, Essays on Kant's Political Philosophy, ed. Williams, H. L. (Chicago, 1992), pp. 191212Google Scholar, presents the most sophisticated recent defense of this claim, based in part on his explicit rejection of the freely willed punishment criterion. As I argue above, however, that criterion is called for both as the most consistent reading of the exegetical evidence and as an inference of the overall analytical framework of Kant's moral theory.

23 Rather than, say, its consequences, which are strictly outside the purview of this argument.

24 Morris, ‘Persons and Punishment’.

25 Since moral law is the ratio cognoscendi of autonomy that is equivalent to human dignity (Kant, ‘Critique’, p. 119fn).

26 Rawls, John, ‘Legal Obligation and the Duty of Fair Play’, Collected Papers, ed. Freeman, S. (Cambridge, 1999; 1st edn. 1964), pp. 117–29Google Scholar; Morris ‘Persons and Punishment’; Murphy, Kant. The criticisms presented below could be seen as applying also to Nino's ‘consensual theory of punishment’, which shares some of its features with the revealed preference argument as well (C. S. Nino, ‘A Consensual Theory of Punishment’, Philosophy and Public Affairs 12 (1983), pp. 289–306).

27 The Metaphysics of Morals, p. 108. Of course, ‘dualism’ is a radical simplification, since there is likely to be a continuum of morally more and less acceptable motivations, but it captures the sense of multiplicity relevant for the discussion that follows.

28 Wolff, R. P., The Autonomy of Reason: A Commentary on Kant's ‘Groundwork of the Metaphysics of Morals’ (New York, 1973)Google Scholar; Singer, Marcus G., ‘Reconstructing the Groundwork’, Ethics 93 (1983), pp. 566–78CrossRefGoogle Scholar.

29 Even if it can bind it to will a carefully circumscribed restitution. It bears emphasizing that the argument considered here is not whether the contractarian argument is, in principle, morally binding, but whether it is morally binding just because it is contractarian.

30 Meyer, Michael J., ‘Stoics, Rights and Autonomy’, American Philosophical Quarterly 24 (1987), pp. 267–8Google Scholar.

31 Recall that the maintained assumption is that the alternatives to punishment include less punitive preemptive measures, not merely therapy. This makes implausible the claim that the rational moral self may be under an obligation to will punishment as a way of constraining the immoral self: in the presence of effective preemptive measures, there would, presumably, be no criminal actions left to constrain.

32 The implications of the distinct conceptualizations of the ‘split-level self’ for the meaningfulness of the notion of autonomy are the subject of a substantial volume of articles, catalyzed, in particular, by Frankfurt, Harry, ‘Freedom of the Will and the Concept of a Person’, Journal of Philosophy 68 (1971), pp. 520CrossRefGoogle Scholar. Although controversies on the precise nature of the division of the self and the relationship between its components persist, there seems to be a general agreement that a robust notion of autonomy requires the recognition of some such division (Christman, John, ‘Constructing the Inner Citadel: Recent Work on the Concept of Autonomy’, Ethics 99 (1988), pp. 109–24CrossRefGoogle Scholar).

33 Andrew Oldenquist, ‘An Explanation of Retribution’, The Journal of Philosophy (1988), p. 469.

34 Hill, Thomas E., ‘Kant on Punishment: A Coherent Mix of Deterrence and Retribution?’, Respect, Pluralism, and Justice: Kantian Perspectives (Oxford, 2000), p. 198CrossRefGoogle Scholar.

35 For example, Goldman, Alan, ‘The Paradox of Punishment’, Philosophy and Public Affairs 9 (1979), pp. 4258Google Scholar; Quinn, Warren, ‘The Right to Threaten and the Right to Punish’, Philosophy and Public Affairs 14 (1985), pp. 327–73Google Scholar.

36 Deigh, John, ‘On the Right to Be Punished: Some Doubts’, Ethics 94 (1984), p. 197CrossRefGoogle Scholar, makes a parallel point in his criticism of Morris’ argument: ‘To put the problem succinctly, to hold, as Morris does, that pardoning a criminal against his wishes constitutes a wrong done to him is to deny that the authority of those responsible for meting out the punishment includes discretion to forgo the sanction. And this should strike us as rather odd.’

37 For example, Quinn, ‘The Right’.

38 The necessity of punitive measures is, of course, deduced not from the Categorical Imperative alone, but from its conjunction with (a) empirical facts about the actual intentions, moral psychology, and physical capability of the criminal; and (b) the extent of the deprivation of freedom required by the mechanism that is expected to prevent the crime. I explore some of these factors below.

39 Cf. Sorell, ‘Punishment’.

40 In particular, the present argument neither addresses nor implies a criticism of the distinctly non-Kantian, naturalist versions of retributivism, such as that of Moore, Michael S., ‘The Moral Worth of Retribution’, Responsibility, Character, and the Emotions: New Essays in Moral Psychology, ed. Schoeman, F. (Cambridge, 1987), pp. 179219Google Scholar, or of the theories like Mabbott's (‘Punishment’, Mind, New Series 48 (1939), pp. 152–67), which make no explicit connections between retributivism and morality.

41 In T. M. Scanlon's felicitous phrase, ‘even the most familiar of moral principles are not rules which can be easily applied without appeals to judgment’ (What We Owe to Each Other (Cambridge, 1998), p. 199).

42 I thank Sanford Gordon, Catherine Hafer, Sarah Holtman, Paul Kelly, Diana Marian, and the anonymous reviewer for their helpful comments on various drafts of this article.