Right and Coercion: Can Kant’s Conception of Right be Derived from his Moral Theory?

Abstract Recently, there has been some discussion about the relationship between Kant's conception of right (the sphere of juridical rights and duties) and his moral theory (with the Categorical Imperative as its fundamental norm). In section 1, I briefly survey some recent contributions to this debate and distinguish between two different questions. First, does Kant's moral theory (as developed in the Groundwork and the Critique of Practical Reason) imply, or validate, a Kantian conception of right (as developed in the first part of the Metaphysics of Morals, the Doctrine of Right)? In other words, is the Categorical Imperative sufficient to show that the fundamental principles of right are normatively valid? Second, does Kant's conception of right presuppose his moral theory? In other words, is the Categorical Imperative necessary to show that the basic principles of right are normatively valid? In this paper, I will be primarily concerned with defending a negative answer to the first of these questions. In section 2, I will discuss Paul Guyer's attempt to vindicate a positive answer to the same question. In section 3, reasons will be given why any attempt to derive Kant's conception of right from the Categorical Imperative must fail because of the analytic connection between right and coercion. Keywords: Kantrightcoercionjuridical dutiesCategorical ImperativeMetaphysics of Morals Notes 1 I will point to a difficulty with this reading below (cf. p. 96). 2 This assumption is challenged in Pogge, Citation2002. 3 This is the way Kant introduces the distinction between morality and legality in the Critique of Practical Reason (5: 81); it is echoed in the MM at 6:219. As I have argued elsewhere (Willaschek, Citation1997), Kant's 'official' introduction of the distinction in the MM (6: 214) differs from the one in the CPR in locating the difference between morality and legality within the laws. This new way of drawing the distinction between legality and morality tells against the traditional reading according to which the universal law of right is just a limited application of the Categorical Imperative. 4 In addition to these considerations, a number of other passages in the Metaphysics of Morals speak in favour of the traditional interpretation, too; see Wood, Citation2002: p. 6. 5 Cf. the similar claim in Willaschek, Citation1997: pp. 221ff. 6 What I call the 'traditional interpretation' here was called the 'official view' there. 7 Wood allows that juridical duties, considered as ethical duties, "can be brought under the principles of ethics." (Wood Citation2002, 9) He continues: "To this extent it may be correctly said that Kant's theory of right falls under or can be derived from the principle of morality. That is, this may be said in so far as juridical duties are regarded not merely as juridical but also as ethical duties. Considered simply as juridicl duties, however, they belong to a branch of the metaphysics of morals that is entirely independent of ethics and also of its supreme principle." (ibid.). I take this to mean that the normative validity of juridical duties may yet be derived from the moral law, but not their validity qua juridical duties, but only qua ethical duties. The specific way in which legal norms bind us cannot be derived from, or reduced to, the character of a moral obligation. 8 'Kant's idea is that we do not have to appeal to anything other than the idea of freedom itself in order to justify the innate right' (Guyer, Citation2002: p. 46). 9 It still seems to me that when Kant calls the 'universal law of right' a 'postulate that is incapable of further proof', he suggests that this law cannot be derived from any higher or more general principle. Against this, both Guyer and Pippin have argued that Kant, by saying that the postulate is incapable of further proof, admits that it is capable of some proof. But this is not so clear in the German original, which says: 'gar keines Beweises weiter fähig', which might be translated more literally as 'not further capable of any proof at all' or perhaps 'for which no proof at all can further be given'. The precise meaning of the German original is not entirely clear to me, since the phrase 'weiter fähig' ('further capable') is quite uncommon and occurs in Kant only this one time. Because of the strong 'gar keines' ('none at all'), I am inclined to disregard the term 'further' altogether and to read Kant simply as saying that no proof of the postulate is possible. Of course, this is also unsatisfactory in that it leaves unexplained why Kant puts in the qualifier 'weiter' ('further'). In any case, the situation is not as clear as Gregor's translation suggests. 10 Ripstein, Citation2004 also argues against the widespread identification of coercion with sanctions. 11 Kant: 'wo kein Zwang besorgt werden darf'. Mary Gregor translates 'where no coercion may be applied'; but besorgen here rather means 'to worry about', and dürfen in Kant often means 'must' rather than 'may'. 12 I develop this theme in more detail in Willaschek, Citation2002 and Willaschek, Citation2005. 13 I do not want to claim that this will do. Kant's official reason why moral duties may not be coercively enforced is that their fulfilment requires that they be obeyed for their own sake (e.g. 6: 219). But this is not convincing. First, the Categorical Imperative requires us to act on maxims that can hold as universal laws; it does not require us to act from the motive of duty. Even though, according to Kant, the moral value of our actions depends on the motive from which we act, this is not part of the content of the Categorical Imperative, but merely an implication of its categorical character. Hence doing one's duty, as specified by the Categorical Imperative, can be coercively enforced. (For instance, it is possible to force someone to keep a promise.) Second, being coerced into doing one's duty does not exclude that one acts from duty if acting from duty is understood counterfactually: someone S does F from duty just in case doing F is in accordance with duty and S would not have done F if doing it had not been in accordance with duty (see also Willaschek, Citation1997: pp. 214ff.). The real reason why moral duties and obligations, as opposed to juridical ones, cannot be legitimately enforced by coercion seems to me to be that to moral duties there correspond no claims on the one who has the duty in question. This is obvious in the case of ethical or wide duties: if I have a wide duty to help the needy, no one in particular has a claim to my help. Something similar seems to be true in the case of non‐juridical promises (a narrow duty): even though my friend is under a narrow duty to keep his promise, I do not have a claim on him – I can, as Kant puts it, 'remind him' of his promise, but I cannot require him to keep it. 14 Note that the impossibility of conflicting rights could be guaranteed by 'uniting' the spheres of external freedom completely ad hoc; that this uniting must happen 'in accordance with a universal law of freedom' is an additional requirement. 15 This idea that right is inherently social in a way that the moral sphere is not is meant as an alternative to Arthur Ripstein's ingenious suggestion that we can arrive at Kant's conception of right by adding to his moral theory the postulate 'that rational beings occupy space' (Ripstein, forthcoming b: p. 000). Even though I agree with Ripstein's important observation that Kant models juridical relations on special ones, I think that spatiality, for Kant, is neither necessary nor sufficient for juridical relations. First, according to Kant, we occupy space not as rational, but as sensible beings. Second, assuming with Kant that we can understand the idea of non‐sensible rational beings at all, there is no reason why they should not stand in juridical relations, for instance, by entering into contracts. (Maybe the idea of coercing non‐sensible beings into fulfilling their legal obligations is difficult to grasp, but so is the idea of communication between such beings, and many other ideas about them. The reason is the difficulty of the notion of a non‐sensible being, not that of its being subject to coercion.) To put the same point differently, many legal relations simply do not presuppose that we are 'embodied rational beings' (ibid., p. 000). Finally, the idea that space introduces 'a new type of incompatibility relations' (ibid., p. 000) would require that two acts that are morally indistinguishable (do not differ in moral worth) can be distinguished juridically (differ with respect to their legality). But since all illegal acts, according to Kant, are also morally prohibited, this is not Kant's view. Rather, as Ripstein himself points out (ibid., p. 000), the converse is true: legally indistinguishable actions can differ in moral worth, but not vice versa. I think that Ripstein is correct in claiming that the universal principle of right requires 'a new type of incompatibility relations'. However, as I see it, the new type of incompatibility relations cannot be merely spatial, but has to be distinctively social in the sense implied by the impossibility of conflicting rights. 16 This I take to be Arthur Ripstein's strategy in Ripstein, forthcoming b. 17 I argue for this option in Willaschek, Citation1997. 18 On the translation, see fn. 8 above.