496

Christian Onof

BERICHTE UND DISKUSSIONEN

Reconstructing the grounding of Kants ethics: a critical assessment

by Christian Onof, London

Abstract: Kants attempts to provide a foundation for morality are examined, with particular focus upon the fact of reason proof in the second Critique. The reconstructions proposed by Allison and Korsgaard are analysed in detail. Although analogous in many ways, they ultimately differ in their understanding of the relation between this proof and that presented in the Groundwork of the Metaphysics of Morals. A synthesis of the two reconstructions is proposed which amounts to combining Korsgaards awareness of the issue of agent-situatedness, with Allisons emphasis upon the pivotal role of the notion of transcendental freedom. The reconstructed proof relies upon a teleological assumption about human agency, and thus does not provide an unconditional grounding for the moral law. After a brief examination of contemporary approaches to the grounding of a universal morality in the broadly Kantian tradition, the paper concludes with a suggestion as to how the value of freedom can form the core of an adequate response to reasons demand for such a ground. Keywords: Kierkegaard, faith, fact of reason, Highest Good. Acknowledgements: I am grateful for comments made on an oral version of part of this paper by Christine Korsgaard and Talbot Brewer, and on a written version of the paper by Sebastian Gardner.

Introduction This paper examines Kants attempts to provide a foundation for his ethics of duty, and focusses particularly upon two contemporary reconstructions of the grounding in the fact of reason. First, the proof from the Groundwork of the Metaphysics of Morals (GMS) is examined briefly. After circumscribing the nature of the fact of reason, Allisons reconstruction of the grounding from the Critique of Practical Reason (KpV) is examined critically and contrasted with Korsgaards. In particular, we show how a valid derivation can be obtained by drawing upon the strengths of these interpretations. That the grounding of morality ultimately fails, is seen to be a consequence of its assumptions. However, we shall show that this result points towards a different type of foundation for an ethics of duty which draws upon the value of freedom.
Kant-Studien 100. Jahrg., S. 496517 DOI 10.1515/KANT.2009.026 Walter de Gruyter 2009 ISSN 0022-8877 Brought to you by | Niedersaechsische Staats- u. Universitaetsbibliothek / SUB Goettingen (Niedersaechsische Staats- u. Authenticated | 172.16.1.226 Download Date | 3/8/12 3:10 AM

Reconstructing the grounding of Kant's ethics: a critical assessment

497

1. Kants grounding of morality in the GMS Kant presents two types of proof of the bindingness of the moral law, which is the requirement to act out of duty, by adopting the moral law as principle of action. According to Allison,1 both use the Reciprocity Thesis2 according to which an agent is free if and only if she stands under the moral law. The first then tries to secure the freedom of the will by appealing to a distinction between two standpoints, the theoretical and the practical, which is seen as underwritten by that between the sensible and intelligible points of view.3 The second adopts a very different strategy, claiming rather that the fact that we stand under the moral law is ensured by a fact of reason.4 Let us briefly examine the Reciprocity Thesis and its role in the deduction presented in the GMS. Allison presents a cogent reconstruction of this thesis based upon the observation that the argument can only work if the notion of agent is not the thin concept of a rational being, or even a rational agent simpliciter, but the rather thick concept of such an agent with a free will (in the transcendental sense),5 i.e. of an agent who is causally and motivationally independent of the natural order. Indeed, if one considers a transcendentally free agent, the ground for the selection of a maxim can never be located in [] anything natural; rather it must always be sought in a higher-order maxim and therefore in an act of freedom.6 The central argument for the necessity of moral bindingness for freedom relies upon the need to find a set of rules which justify the selection of maxims, and this, independently of any particular end. This means that it is based upon looking for principles determining the maxims and justifying the ends chosen by the agent. The requirement that they be rational principles implies that they are universal7 and the requirement that they be independent of particular ends entails their being formal. This means that these are unconditional rules. The conclusion then follows from GMS(I), where the unconditional good is analysed in terms of the will of an agent doing his duty because it is her duty. Therefore, if unconditional rules of action are sought, the agent stands under the requirement of a full justification in accordance with duty.8 Acting morally is therefore required for a transcendentally free agent.
1 2 3 4 5 6 7 8

Allison, Henry: Kants Theory of Freedom. Cambridge 1990, 214 and 239. GMS, III, AA 04: 446463; KpV, 6, AA 05: 2930.1035. GMS, AA 04: 459.1831. KpV, AA 05: 47.1120. Allison, op. cit., 207. Ibid., 207208. In the sense that they would be rational grounds for any rational agent. Allison (op. cit.) notes that some commentators have questioned whether this is meant to amount to the claim that one ought to act only from obligatory maxims. However, a proper understanding of the distinction permissible/mandatory shows that the requirement is to do what is permissible: when a single permissible option is available, this means to do what is obligatory (Beck, Lewis White: A Commentary on Kants Critique of Practical Reason. Chicago 1960, 122; Onof, Christian: A framework for the derivation and reconstruction of the categorical imperative. In: Kant Studien. 1998, 89: 424f).

Brought to you by | Niedersaechsische Staats- u. Universitaetsbibliothek / SUB Goettingen (Niedersaechsische Staats- u. Authenticated | 172.16.1.226 Download Date | 3/8/12 3:10 AM

498

Christian Onof

This purports to establish the necessity of morality for freedom. We shall examine the validity of this proof further and later return to the issue of the sufficiency of morality for freedom. Our immediate concern for the grounding of morality is whether human agents are indeed transcendentally free, in which case the Reciprocity Thesis linking morality and freedom could be the key to a direct argument grounding the moral law. An independent proof of transcendental freedom is therefore needed to ground the moral law. Kant attempts to derive freedom from the conception of an intelligible self one has insofar as one is a rational being. Without going into details, even the most sympathetic commentators judge Kants enterprise of deduction in GMS (III) as a failure.9 The analysis of this failure is often accompanied by a diagnosis of its nature. According to Allison the problem is that Kant is attempting to derive a moral obligation from non-moral premisses. This also follows from Henrichs10 analysis of Kants deduction strategies. This is a standard kind of objection to a way of grounding morality, as Bittner11 observes: many philosophers who tackled the question of why one ought to be moral, viewed the question as meaningless, either because an answer to it would itself have to be moral,12 or because such a question presupposes that moral action is only acceptable if it is viewed as a means to something else.13 Both these views are summarised by Brock,14 in the form of the following dilemma: either the justification for being moral is itself moral, which begs the question because it is morality which is to be grounded, or it is extra-moral, in which case the moral demands could no longer be seen as having the final say in the decision process leading to action. But, as Bittner15 rightly points out, the latter part of this argument relies upon the moral demands sometimes being in conflict with action guided by these extra-moral grounds, for in such cases it would indeed appear that moral demands do not have ultimate authority. This cannot, however, be the case if the deduction of the moral claims from the extra-moral grounds is a successful derivation of the moral requirements from the extra-moral grounds. What the absence of conflict reveals is a connection between morality and the extra-moral grounds. That these extra-moral considerations can be shown to ground morality entails that this connection is not contingent and therefore a priori while the extra-moral character requires it to be synthetic. Although unsatisfactory, one cannot therefore conclude that Kants en9

10 11 12 13

14

15

E.g. Allison, op. cit.; Henrich, Dieter: The Unity of Reason. transl. R.L. Velkley, Cambridge, Mass. 1994. Op. cit. Bittner, Rdiger: What Reason Demands. Cambridge 1989. E.g. Sidgwick, Henry: The Methods of Ethics. London 1922. E.g. Bradley, Francis Herbert: Ethical Studies. Oxford 1927; McDowell, John: Are Moral Requirements Hypothetical Imperatives? In: Proceedings of the Aristotelian Society, Supplementary Volume. 1978, 52:1329. Brock, Dan: The justification of morality. In: American Philosophical Quarterly, 1977, 14:7178. Bittner, op. cit., 18.

Brought to you by | Niedersaechsische Staats- u. Universitaetsbibliothek / SUB Goettingen (Niedersaechsische Staats- u. Authenticated | 172.16.1.226 Download Date | 3/8/12 3:10 AM

Reconstructing the grounding of Kant's ethics: a critical assessment

499

terprise in GMS (III) is fundamentally wrong-headed. We shall however see that the attempt at a deduction of morality on the basis of a fact of reason is far more appropriate.

2. Fact of reason and grounding of morality The second approach consists in grounding the bindingness of the moral law in a fact of reason. It has been criticised for many reasons,16 but in particular for parachuting a fact to solve a problem in a way which seems to attempt to dispense with a proper argument. What constitutes one strong strand among the objections to it is that, if for Kant the fact of reason forces itself upon us as a synthetic a priori proposition based on no pure or empirical intuition (sich fr sich selbst uns aufdringt als synthetischer Satz a priori, der auf keiner, weder reinen noch empirischen, Anschauung gegrndet ist17), Kants procedure apparently consists in simply cut[ting] off criticism18. Bittner even suggests that it endangers discussion of morality and moral philosophy19. He points out that, on the contrary, the rational cannot be the sort of thing that forces itself upon us20, invoking a broadly Hegelian notion of rationality.21 Below, we shall successively examine the nature of the fact, and how it is supposed to provide a deduction of morality.

What is the fact of reason? Beck22 distinguishes two types of textual accounts of the fact of reason: the first identifies the fact either with the moral law or freedom, and the second with a consciousness of the moral law. Identifying it with the first leads to the problems men16

17 18 19 20 21

22

E.g. Schopenhauer, Arthur: On the Basis of Morality. Indianapolis 1965; Beck, Lewis White: Das Faktum der Vernunft: zur Rechtfertigungsproblematik in der Ethik. In: Kant Studien. 196061, 52:271282; Henrich, Dieter: Der Begriff der sittlichen Einsicht und Kants Lehre vom Faktum der Vernunft. In: Die Gegenwart der Griechen im neueren Denken, Festschrift fr Hans-Georg Gadamer. ed. D. Henrich, W. Schulz, K. H. VolkmannSchluck, J.C.B. Mohr, Tbingen, 1960; Bittner, op.cit.). KpV, AA 05: 31.2728. Bittner, op. cit., 90. Ibid. Ibid. This leads him to shift the focus of attention from the fact of reason and the derivation of the bindingness of the moral law to the weaker recognition of the validity of a principle of autonomy from the notion of ones own yet universal legislation (seiner eigenen und dennoch allgemeinen Gesetzgebung, GMS, AA 04: 432.2930): Bittner does not think that the Categorical Imperative follows from this principle of autonomy (Bittner, op. cit., 7679). Beck, Lewis White: A Commentary on Kants Critique of Practical Reason. Chicago 1960, 166168.

Brought to you by | Niedersaechsische Staats- u. Universitaetsbibliothek / SUB Goettingen (Niedersaechsische Staats- u. Authenticated | 172.16.1.226 Download Date | 3/8/12 3:10 AM

500

Christian Onof

tioned above (i.e. there is arguably no requirement for a deduction, hence no possible discussion) while the second option leaves it open how a mere consciousness can prove moral obligation. Another distinction which Beck makes, is between a fact for and a fact of reason, where the first refers to a value which would be apprehended by an intellectual intuition, and the second to the fact that pure reason is practical (i.e. that it can provide a motive and determine principles for action). He concludes that it is rather the second. Focussing upon the first distinction, those authors who have provided a reconstruction of the proof opt rather for an interpretation of the fact as a consciousness. The choice of the latter interpretation can be justified as follows. (1) If the fact of reason is to be able to ground morality, it must also underpin the motivational power of the moral law.23 (2) The agent must therefore be aware of this fact. (3) Were this awareness not a certainty, it would not be sufficient to ground morality, because of scepticism about its object. (4) The object of this awareness is a purported synthetic a priori truth (5) This synthetic a priori truth is not justifiable using the transcendental apparatus set up in the first Critique. (6) Such certainty therefore clashes with the limits of knowledge identified by the Analytic of the Critique of Pure Reason as an agent is thereby certain of something for which he knows he has no epistemic warrant. This is not a contradiction, but a tension which can be resolved in one of two ways: (6a) the awareness is a given feature of the agents practice, much as space and time are for empirical knowledge, or, (6b) the impossibility of any warrant leads to recasting the object of the awareness as a feature of the agents consciousness, rather than of reality itself. (7a) In case a, the awareness of the fact of reason is actually a consciousness of a practical truth, i.e. of something which, in the agents practice, is taken as if it were true. So the fact of reason is seemingly best interpreted as this practical truth. This accounts for certain statements made by Kant.24 But we note that, insofar as a ground of action for a free agent25 is at stake, it is the agents purported awareness of this fact which will be relevant to the foundational argument. (7b) In case b, the awareness is of a feature of consciousness, so that the fact of reason is itself a consciousness. This interpretation is consistent with many of Kants statements,26 and for Allison,27 is faithful to the bulk of the evidence.

23

24 25 26 27

Since moral action for Kant is determined by the nature of the motive for action, normativity must involve establishing both that the moral law is binding and that it is thereby motivating (Korsgaard, Christine: Creating the Kingdom of Ends. Cambridge 1996, 43). KpV, AA 05: 6.712; KpV, AA 05: 47.1115. The agent must conceive of herself as free (GMS, AA 04: 448.911). KpV, AA 05: 31.24; KpV, AA 05: 42.810. Op. cit. 223.

Brought to you by | Niedersaechsische Staats- u. Universitaetsbibliothek / SUB Goettingen (Niedersaechsische Staats- u. Authenticated | 172.16.1.226 Download Date | 3/8/12 3:10 AM

Reconstructing the grounding of Kant's ethics: a critical assessment

501

(8) From (7a) and (7b), it becomes clear that, whether we interpret Kants use of the term fact of reason as referring to a consciousness or not, it is the consciousness of a certain practical truth which is relevant to the grounding of the moral law, and it makes sense to stipulate that the fact be defined as such a consciousness.28 The next question is: what is this a consciousness of? The texts29 certainly indicate that it is a consciousness of this fundamental law (Bewutsein dieses Grundgesetzes30) which is the moral law. This is the interpretation adopted by Allison,31 specifying however that it is the consciousness of standing under certain constraints which correspond to the moral law in its typified form, i.e. a sort of ordinary language translation of the spirit of Kants categorical imperative. Korsgaard32 concurs in interpreting the fact as a consciousness of the reality of moral obligation33, but adds that it is also a consciousness of the laws capacity to motivate us34. This latter claim has limited textual support,35 for when Kant talks of the consciousness of freedom being identical with the fact of reason, this must mean that originally, one considers them as defined separately. The texts therefore rather point to a consciousness of the constraints of the moral law, thus defining it as a fact of reason but one which lies beyond Becks second dichotomy of fact of/for reason. If we adopt this interpretation how will the deduction function?

Deductive strategy Allison and Korsgaard, drawing upon the GMS as well as the doctrine of the fact of reason in the KpV, offer reconstructions of Kants grounding of morality. Allison36 reconstructs the proof approximately as follows: (1) The fact is indeed one of reason since the moral law can have no other origin but a rational one. This is for the following two reasons. First, because it makes universal and necessary claims that are not the product of empirical practical reason (it is rather an ingredient of moral experience, much as space and time are for perceptual experience). Second, because it is a principle of autonomy while empirical principles are heteronomous.

28

29 30 31 32 33 34 35 36

Note that the distinction between fact of and fact for consciousness is hereby blurred; Beck is able to make this distinction because he does not view the fact as a consciousness (Beck, 1960, op. cit.). E.g. KpV, AA 05: 31.24; KpV, AA 05: 91.1924. E.g. KpV, AA 05: 31.24. Allison, op. cit., 233. Korsgaard, op. cit. Ibid., 26. Ibid. KpV, AA 05: 43.1623. Op. cit., 235.

Brought to you by | Niedersaechsische Staats- u. Universitaetsbibliothek / SUB Goettingen (Niedersaechsische Staats- u. Authenticated | 172.16.1.226 Download Date | 3/8/12 3:10 AM

502

Christian Onof

(2) Moreover, insofar as the agent takes this law as governing his will, he has a pure interest in it. (3) The presence of such an interest entails that the law is a possible incentive, which gives a practical grounding to positive freedom, i.e. the capacity to act on the basis of self-imposed principles. (4) To establish freedom practically, it is however also necessary to show independence from causal determination, i.e. negative freedom. This is achieved by observing that the pure interest in the moral law is an interest in a merely intelligible or ideal order of things37 and that the consciousness of the law as a constraint entails a consciousness of the independence of the mechanism of nature (negative freedom). (5) The Reciprocity Thesis is now invoked to show that, insofar as the agent is conscious of freedom, he is conscious of the normative force of the moral law. It is thus not only a possible incentive, but the one we ought to act upon.38 Korsgaards reconstruction39 concurs in effect on points (1) to (3), although it does feature other steps; we shall return to this issue further. Note that Korsgaard does not characterise the moral laws being a possible incentive as positive freedom, insofar as this leaves it open whether the agent is indeed able to act freely in the empirical world. We shall thus refer to Allisons notion as pure positive freedom while Korsgaards notion of positive freedom presupposes negative freedom. Korsgaards argument, which is a reconstruction drawing upon both the GMS and KpV proofs, develops as follows: (4') Our ability to act on the moral law teaches us that we are (negatively) free; (5') Consequently, we are (practically speaking) members of the intelligible world and have a higher vocation than the satisfaction of our own desires40 insofar as we belong to the ground of the world of appearances. (6') We therefore have an incentive to act from duty (positive freedom) which has normative priority: this is the incentive to be moral.

37 38

39 40

Ibid., 242. Note that this reconstruction differs from Allisons earlier views on the deduction (Allison, Henry: Justification and freedom in the Critique of Practical Reason. In: Kants Transcendental Deductions: the three Critiques and the Opus Postumum. Ed. E. Frster, Stanford 1989, 114130). For instance, Allison then took the fact of reason to show that the agent can act out of respect for the moral law. He changed his mind on this partly as a result of Hermans criticisms (Herman, Barbara: Justification and objectivity: comments on Rawls and Allison. In: Kants Transcendental Deductions: the three Critiques and the Opus Postumum. Ed. E. Frster, Stanford 1989, 131136). Op. cit., 170. Ibid.

Brought to you by | Niedersaechsische Staats- u. Universitaetsbibliothek / SUB Goettingen (Niedersaechsische Staats- u. Authenticated | 172.16.1.226 Download Date | 3/8/12 3:10 AM

Reconstructing the grounding of Kant's ethics: a critical assessment

503

Comparison between the reconstructions In comparing the two arguments, we must first note that there is an apparent incompatibility between them. Indeed, although we have noted the differences between the senses Allison and Korsgaard assume for the notion of positive freedom, it may still appear perplexing that Allison derives the possibility of acting from the moral law from a mere interest in it. Korsgaard, on the other hand, requires that we first establish negative freedom (practically) in step (4'). Both argumentative steps are however valid on a proper understanding of the notion of possibility, which is related to the different notions of positive freedom. Korsgaard is more demanding in her use of the word: she requires that it be possible to act on the moral law given ones inclinations, i.e. given the nature of human practice. Such a claim requires that the moral law be shown to sometimes override the value of inclinations. For Allison on the other hand, as long as the agent can consider acting from the moral law, then it is a possibility. In so doing, Allison does not take into account the practical context (inclinations) of the agents action. In other words, it could be the case that any inclination would be deemed more valuable than the moral law. Allison and Korsgaards notions of possibility therefore respectively range over worlds which are possible for agents characterised by their rationality and worlds which are possible for agents who are additionally situated. Korsgaards awareness of the importance of the practical context of action will be seen to be crucial later. We shall distinguish her more stringent notion of possibility by referring to it as situated possibility. There are other differences between the two strategies of reconstruction of Kants derivation. First there is the inversion by which, for Korsgaard, negative freedom entails our membership of the intelligible world, arguably leaving unclear the justification for negative freedom. But the issue in both these reconstructions is to prove from a practical point of view that we have to view ourselves as transcendentally free. Korsgaards richer interpretation of the fact of reason as a consciousness that one could be morally motivated, which we noted above, directly establishes we are negatively free in a practical sense. Second, Korsgaard uses a teleological sense of our membership of the intelligible world, in that she assumes the Highest Good has to be adopted as a rational end. However the nub of the difference between these two strategies, both of which remain close to Kants text, is that each draws upon a different aspect of the Groundwork proof to reconstruct the fact of reason proof, and thereby each resolves the motivational issue differently. Let us examine this point. Korsgaard views the role of our existence in the intelligible world, as presented in KpV, as similar to that in the GMS proof,41 although in KpV, this existence is grounded in the consciousness which is the fact of reason as opposed to our mere
41

Korsgaard, op. cit., 170.

Brought to you by | Niedersaechsische Staats- u. Universitaetsbibliothek / SUB Goettingen (Niedersaechsische Staats- u. Authenticated | 172.16.1.226 Download Date | 3/8/12 3:10 AM

504

Christian Onof

theoretical capacity to formulate theoretical ideas (ibid.). She concludes that the function of the idea of our intelligible existence [] is essentially the same in both books. On the contrary, Allison insists that the line of reasoning is diametrically opposed42 on the grounds that in GMS, positive freedom was deduced from negative freedom, itself derived from the membership of the intelligible world, whereas in the KpV, the deduction moves from pure positive to negative freedom. The reason for Allisons emphasis upon a difference which Korsgaard does not see as essential is that the motivational issue is dealt with by our intelligible existence for Korsgaard and this is the crucial feature which links GMS and KpV, while Allison focusses upon the difference in groundings of freedom in the pure positive sense, respectively in intelligible existence, and in the fact of reason in GMS and KpV. Korsgaard brings in teleological considerations (intelligible existence as a higher vocation) to bear upon her understanding of the motivational aspect. This suggests that, unlike Allison, she does not think that merely considering the rational nature of a free agent, as such delivers the required overriding incentive.

The problems with the deduction A first reason for objecting to the grounding of morality in the fact of reason is, of course, a disagreement with the claim that this is indeed a feature of our moral consciousness. It is difficult to argue either for or against such a claim. One could seek a phenomenological defence of it. Indeed, we for instance find Sartre saying that when I act freely, my act commits me in the face of the whole of humanity.43 This could be taken to substantiate the view that human agents are conscious of the normativity of a requirement of universalisability of their actions. I shall however not pursue this issue which far exceeds the scope of this paper. I shall assume that grounds could perhaps be provided to justify the consciousness that there is a requirement to view ones choice of action as though it were to be applied universally. In the following development, I shall examine Allison and Korsgaards reconstructions in turn, in the light of the notion of a valid grounding of ones action. To characterise the latter, I shall require a concept of purposefully meaningful (p-meaningful) action. To grasp this notion, consider the requirement that an agents action be understood as contributing to at least one of what it understands as being its ends. Action which is shown to contribute to an end, I shall refer to as p-meaningful, and the corresponding requirement as that of p-meaning. I think it is fairly uncontroversial to claim that the requirement of p-meaning is normative for human agency. Moreover, it is clearly implied by many statements in
42

43

Allison, op. cit., 242243. This is a view shared by Ameriks (Ameriks, Karl: Kants Theory of Mind. Oxford 1982, 226). Sartre, Jean-Paul: Existentialism and Humanism. Trans. Philip Mairet, London 1973, 31.

Brought to you by | Niedersaechsische Staats- u. Universitaetsbibliothek / SUB Goettingen (Niedersaechsische Staats- u. Authenticated | 172.16.1.226 Download Date | 3/8/12 3:10 AM

Reconstructing the grounding of Kant's ethics: a critical assessment

505

Kants ethical writings, such as his claim that an action must have an end.44 I shall use this notion to examine whether action from duty is grounded, on Allisons reconstruction of Kants proof from the fact of reason, and then turn to Korsgaards strategy. Allisons argument Allisons argument is a very impressive reconstruction of Kants deduction, but it ultimately relies upon the Reciprocity Thesis (move from (4) to (5)) which we have seen to link a strong conception of freedom, the transcendental one, with morality. As we have seen, this thesis establishes that for an agent whose motivation is his own to choose (pure positive freedom) and who is not causally determined (negative freedom), this agent can, and can only, justify his actions by conforming to the dictates of the moral law (by acting at least permissibly). Let us now examine this claim. Allisons justification in rational terms could be viewed as a special case of justification in terms of p-meaning for a purely rational being. The argument would first consist in observing that, insofar as one shows that a form of action is rational for a purely rational being, one is in effect showing that it serves the purpose of rationality. GMS (I) has shown that the concept of the unconditional good can be analysed as containing that of action from the moral law,45 which entails that a purely rational being acts from the moral law. Such action is p-meaningful for such a being in the sense that it contributes to any end which springs from reason (the only type of end a purely rational being can have). Since it fulfils the requirement of p-meaning, action from the moral law (and only such action) is properly justified for a purely rational being, hence the normativity of the moral law for such a being. With the proposed extension of this justification to human, as opposed to purely rational beings, a problem arises. If Kant had proven there were indeed rational ends which are shared by all human-beings, then an argument on this basis could perhaps be extended to all human beings, a point we shall examine below with Korsgaards argument. But we have not been shown there are such ends on the basis of the sole hypothesis of a free agent. This leaves us with a lack of p-meaningfulness for action grounded in Kants notion of duty alone. This means that, for a human agent, i.e. a non-perfectly rational being, acting from duty is, by this argument, not justified any more than other forms of action. At this point, a proponent of Allisons argument may reply that, in fact his rational justification is of a different type, and it abstracts from any notion of human agents ends. It is the justification that is required of a rational being for her actions insofar as they are free in a transcendental sense. It is indeed clear that, insofar as nothing predetermines such an agents action, any choice she makes has
44 45

MS, AA 06: 385.1. It claims more, namely that this is the whole of the unconditional good, but I would argue that this is not validly argued for.

Brought to you by | Niedersaechsische Staats- u. Universitaetsbibliothek / SUB Goettingen (Niedersaechsische Staats- u. Authenticated | 172.16.1.226 Download Date | 3/8/12 3:10 AM

506

Christian Onof

to be explainable. This amounts to a requirement for justification insofar as any such explanation is submitted to constraints of rationality. But what does this requirement mean? That the statements constitutive of the justification are rationally grounded upon one another, so that if they are of the type I did X because of Y, Y is indeed a reason for X. Note that it is not the stronger requirement that Y should itself spring from reason alone, i.e. that it be an entirely a priori rational ground. The advantage of such a purely rational ground would of course be that it appears to require no further justification. The appeal of such a ground lies therein that it appears complete, thus purportedly showing that the moral law is a sufficient ground for action for a transcendentally free agent. And Allison can argue that, insofar as (4) has established that I have an interest in a merely intelligible order of things, such a justification can indeed be invoked. The problem is that a justification which thus appeals to the moral law can be questioned for the very rationality of the grounding of the action in the moral law. Clearly, for a purely rational agent, it is true that to do X because X is his duty as determined by the moral law, is indeed a rational justification insofar as the moral law determines that which is right for all purely rational agents, as we saw above. In the case of a non-purely rational agent however, this has not been established: we do not know that acting from the universal requirement of duty is rationally justified for an agent qua agent in situation. Is it sufficient for this justification, to observe that X is required by the moral law which applies to all purely rational beings? Not insofar as the agent is in a situation and already acting in the pursuit of particular ends. Sufficient grounds do not appear to be available since the agent, with her particular ends, may perceive the point of view of universal rationality (a point of view she can freely adopt, and from which the moral law is an ultimate justification of action) as remote, even if she has an interest in it.46 The claim that what is sufficient as a justification for all purely rational beings, must therefore also be sufficient for human agents insofar as they are a species of rational beings, is not valid.47 For it requires that it be shown that this justification is sufficient for all rational beings, not merely all purely rational beings. And this has not been demonstrated. The situation is therefore the following. The moral law
46

47

An objector could certainly query the justification that accrues to my action on non-moral motives: such action can be justified in terms of those motives, but the motives themselves have only limited rational warrant. However, the greater (i.e. unconditional) warrant that moral action possesses does not, as such, provide me with a motive for acting morally. The purely rational perspective can only address me as situated agent if it defines a new situation that I have some grounds to move to from the perspective of my current situation, or if it is in some sense constitutive of any human situation. Since neither of these has been shown, the moral motive does not define a possible motive for action. The error involved is akin to that of viewing that, in the Transcendental Deduction of the Categories, once it has been established that a general manifold must conform to the categories for it to be an object of my consciousness (KrV, B 143, AA 03: 115.1418), this must also be the case for the manifold given in space and time. In both cases, the move from genus to species is not warranted.

Brought to you by | Niedersaechsische Staats- u. Universitaetsbibliothek / SUB Goettingen (Niedersaechsische Staats- u. Authenticated | 172.16.1.226 Download Date | 3/8/12 3:10 AM

Reconstructing the grounding of Kant's ethics: a critical assessment

507

has been shown in GMS (I) to be the rational principle of action which determines the unconditional good. Allison shows that the intelligible point of view from which the moral law overrides any other incentive is one which presents itself to the agent as determining a possible incentive for action. What he does not succeed in establishing is that this point of view itself is one he, as a situated agent with ends, could take as the basis for determining his action. This means that situated possibility has not been demonstrated for the moral incentive. This has two consequences for the Reciprocity Thesis. First, the sufficiency of the moral law as final justification is, contrary to Allisons48 claim, not only not unproblematic, but question begging, since it does not provide a sufficient determination of a situatedly possible incentive. And second, since the possibility of moral action has not been demonstrated for an agent in situation, morality cannot be required of such an agent. For all we can say is that, given that the free agent stands under a requirement of justification, only the moral law could provide such a justification; but since the moral incentive has not been shown to be situatedly possible, this does not translate in terms of a necessary requirement for a situated agent I shall come back to this point later. Morality has thus not been shown to be a necessary constraint for a free agent. But, contrary to standard criticisms of Kants Reciprocity thesis,49 I locate the root of the problem in the argument for sufficiency. Finally, let us note that the problem of a law that is a possible, but not necessarily a situatedly possible incentive, relates to the Hegelian-inspired criticism that the moral law is both external to and imposed upon the agent, as it is only in the idea, in representation50. On the contrary, for Hegel, the moral consciousness is autonomous when its relation to the good is not something foreign to it and also does not lead it into an obstruction or alienation of its freedom51. I shall return to this point further. A way forward from the shortcomings of the Reciprocity Thesis would involve showing that the fact of reason can provide the situated agent with a ground for taking the intelligible point of view, so that the fact of reason and transcendental freedom together would entail that the moral law is sufficient as a final justification, and indeed could be shown also to be necessary. This requires that a bigger argumentative burden be assigned to the fact of reason. Korsgaards argument Some of the spirit of this argument is to be found implemented in Korsgaards reconstruction, in a way which further develops the suggestion made earlier that the moral law must be shown to be p-meaningful. Indeed, Korsgaard is sensitive to the
48 49 50

51

Op. cit., 209. See Allison, op. cit., 209211. Hegel, Georg Wilhelm Friedrich: Werke: Theorie Werkausgabe I: Theologische Jugendschriften 18931800. Frankfurt 1970, 300. Henrich, op. cit., 118; see also Hegel, op. cit.

Brought to you by | Niedersaechsische Staats- u. Universitaetsbibliothek / SUB Goettingen (Niedersaechsische Staats- u. Authenticated | 172.16.1.226 Download Date | 3/8/12 3:10 AM

508

Christian Onof

issue pointed out above, in that she does show an understanding of the issue of situatedness, expressed in terms of our inclinations.52 Indeed, she stresses that what is lacking in the argument up to point (4) is an incentive for us to identify with the free and rational side of our nature53. This agrees with the spirit of the observations I made above about Allisons parallel argument from (1) to (4). With her point (5), she is reconstructing Kant as using our understanding of our higher vocation, as the motivating force that can address the situated agent, and bring him to understand that his moral action can contribute to producing the Highest Good. This injection of a teleological dimension into the argument adds to Allisons proof the required argument for showing that the moral law is sufficient as a final justification. This therefore shows that acting from the moral incentive is a situated possibility. Setting aside the issue of whether the Highest Good is given as a rational end, there however remains a problem for Korsgaards reconstruction of the derivation of the moral law (which does not feature the Reciprocity Thesis). For it is not clear how the motivation of his higher vocation really relates to the agent in situation, and thus whether what is potentially an overriding incentive, will actually be grasped as such in all situations. For the agent may think this higher perspective should prevail in normal circumstances, but when some crisis is to be handled, he may drop this high-minded point of view for the sake of some other more immediate ends and allow himself moral holidays. To this, one could respond that this may well be the case, but that it would not be rational for the agent to do this in view of the fact that the Highest Good is an end, and this is precisely why the moral law represents a universal obligation. As a rejoinder, consider the following. Even if it has been shown that acting by adopting the point of view of our intelligible nature is indeed rational, it is also the case that responding to the situation following certain inclinations is rational, given certain ends, according to our analysis of a rational justification in terms of p-meaning.54 The issue is then whether the consideration of the end which is the Highest Good ought not override that of any other end. It is indeed the case that the Highest Good is presented as overriding in that it is the highest end grounded in reason. But for a non purely rational being, there are other ends, not grounded in reason. From an intelligible perspective, the Highest Good takes precedence, but nothing shows that it ought to from the perspective of a situated agent.55 And, in crisis situations in particular, the pursuit of other more immediate ends may appear to be the more appropriate course of action. So, although Korsgaards reconstruction has dealt with the problem of the sufficiency of the moral law as a final justifi52 53 54 55

Korsgaard, op. cit., 167169. Ibid., 167. We are not seeking full justifications at this point. The claim that we have to think of the Highest Good as an overriding end even from our situated point of view could not be substantiated from a Kantian perspective, unless transcendental freedom were established see below.

Brought to you by | Niedersaechsische Staats- u. Universitaetsbibliothek / SUB Goettingen (Niedersaechsische Staats- u. Authenticated | 172.16.1.226 Download Date | 3/8/12 3:10 AM

Reconstructing the grounding of Kant's ethics: a critical assessment

509

cation, we are still apparently left with the issue of the necessity of morality for a free agent. The moral incentive is here a situatedly possible incentive and we can now understand action from duty as p-meaningful, but the priority of the Highest Good among other ends is only grasped from the intelligible perspective.56 I take this to be closely related to Guyers observation that a free will is not logically or analytically compelled to act only on moral rather than material considerations, but must do so only if it is to maintain or preserve its freedom57. That is, from the perspective of the free will, the value of freedom will require moral action. This appeal to the value of the preservation of freedom from the perspective of the free will is however related to that of the Highest Good from an intelligible perspective in Korsgaards argument, insofar as Kant defines the Summum Bonum as freedom in accordance with a will that is not necessitated to action (die Freyheit nach Willkhr, die nicht neceitirt wird zu handeln58) in his Lectures on Ethics. In both cases, the perspective offered by a higher value would appear to be required for the moral law to be binding. One could object by claiming that this Highest Good is the most valuable end, and this does not depend upon the situation, but is free of perspectival considerations. This amounts to viewing rationality not as a perspective among others, but as a point of view which is beyond any particular perspective. Even if we accept this way of presenting the case for rationality, we note that the fact that we value A more than B does not mean we devote all our time to the pursuit of A and none to that of B. What would be required for this to follow would be for A to override B from any particular perspective, which is not the same as its being more valuable than B independently of particular perspectives. For from the point of view of a particular perspective, values come into play which shed new light upon A and B, values which did not figure in the perspective-free point of view of rationality. As a result, A and B are assessed using considerations which are extra-rational and figure in the background of the point of view in question. And this may lead to B being preferred to A. The problem therefore remains for Korsgaards reconstruction.

56

57 58

We note that this problem recurs in Korsgaards important attempt to refute scepticism about obligation (Korsgaard, Christine: Sources of Normativity. Cambridge 1996). In her chapter 4, she develops a transcendental argument which shows, successfully I think, that if we value anything, we must value humanity. But, this only establishes the minimal conclusion that we must attach some value to moral considerations, thus refuting complete scepticism about obligation. Her argument is roughly that, if I value X, and X is a partial consequence of Y, then I must value Y. This is true, but I may well attach very little value to Y in comparison with X. I may thus hold that human beings have mostly distinguished themselves in their destructive power and do not, as a whole, have much to commend them. Therefore, she has not established anything like the required overriding obligation of duty. She is, moreover, sceptical about this overriding quality (ibid., 125). Guyer, Paul: Kant on Freedom, Law, & Happiness. Cambridge 2000, 56. MPCollins, AA 27: 344:23.24.

Brought to you by | Niedersaechsische Staats- u. Universitaetsbibliothek / SUB Goettingen (Niedersaechsische Staats- u. Authenticated | 172.16.1.226 Download Date | 3/8/12 3:10 AM

510

Christian Onof

This problem can be seen to revolve around the fact that the agent may not view herself as a unity59 so that she is able to opt for the moral law in some cases and for the principle of self-love in others for instance. In other words, the proof has not succeeded in showing that all the agents ends should be subordinated to a single one such as the Highest Good. The problem is related to Hegels claim that the moral law threatens the individuals sense of self.60

Drawing upon Allison and Korsgaards reconstructions I could at this point be accused of having misrepresented Korsgaards reconstruction of Kants argument. It is true that I have not, as indicated earlier, presented the whole of her argument. Korsgaard61 does in fact provide an argument (the Argument from Spontaneity) to the effect that a purely rational being is required to act morally since the moral law provides the only possible justification. This, together with the fact that as creatures who must act under the idea of freedom, we are bound by the laws of freedom62, amounts to part of Allisons Reciprocity Thesis, namely the necessity of morality as a requirement for a free agent. It might therefore precisely seem to answer the objection I made above, that the claim of necessity remains unresolved for her.63 The problem is that Korsgaard presents this argument before any mention of teleological considerations. This argument therefore shows morality to be necessary for a spontaneous64 purely rational agent. These reconstructions do reflect Kants view in the GMS,65 but in the KpV, he points out that, if the Highest Good were not possible, the constraints of the moral law would be inherently false (mithin an sich falsch sein66). This point is incorporated into Korsgaards argument, but after the claim of necessity is made. This issue might appear to be a point of detail, but it indicates that Kant does, in the KpV, display an awareness of the situation in which the moral agent finds herself, which suggests he is not guilty of developing a moral59

60 61 62 63

64 65 66

This sense in which the Highest Good is one of the agents ends, but only really his as the overriding end of all human action when he adopts the intelligible point of view, is at the core of this lack of unity. Wood, Allen: Hegels ethical thought. Cambridge 1990, 128. Korsgaard, Christine: Creating the Kingdom of Ends. Cambridge 1996, 164. Ibid., 167. Here, I distance myself from Guyers focus upon the value of freedom (e.g. Guyer, Paul: The value of reason and the value of freedom. In: Ethics. 1998, 109(1), 29) as an account of Kants approach to grounding morality, as I think it overlooks Korsgaards analysis of the requirements of practical reason, which I have expressed in terms of p-meaning. Guyers emphasis upon freedom does, however, constitute the key to what I shall propose further as an approach to grounding morality as a regulative constraint on our action. Korsgaard, op. cit., 166. GMS, AA 04: 448.1322. KpV, AA 05: 114.19.

Brought to you by | Niedersaechsische Staats- u. Universitaetsbibliothek / SUB Goettingen (Niedersaechsische Staats- u. Authenticated | 172.16.1.226 Download Date | 3/8/12 3:10 AM

Reconstructing the grounding of Kant's ethics: a critical assessment

511

ity of self-alienation as Hegel claims,67 a point which a sympathetic interpretation of Kants proof should take on board. For, before the issue of the incentive provided by the end of the Highest Good is raised, the constraint represented by the fact of reason could very well appear to laden the agent with guilt as a constitutive feature of his moral consciousness. Indeed, the awareness of the constraint of the moral law would not be sufficient to provide a real incentive, but would generate the kind of cleavage within consciousness which results from the awareness of ones inadequacy in the light of the higher vocation which the moral law represents.68 However, Kant69 insists that the assumption of the Highest Good as an end is required for moral behaviour. And it is to this extent that Kant can be said to pre-empt this form of Hegelian criticism. However, as Ward70 reminds us, Kant also asserts that the moral law is a requirement which exists independently of the consideration of any end. And it is because this view is presented first by Kant, whereupon he then makes the claim about the necessity of postulating we are pursuing the Highest Good, that Korsgaards interprets Kant as she does. But Kants two claims are compatible if the first (requirement to act out of duty) is understood as applying to beings who are purely rational, while the second states the conditions under which this applies to all human beings. On this reading therefore, there is no necessity to act upon the moral law for human agents, prior to having established the necessity to assume the Highest Good as end. As a result, Korsgaards actual reconstruction suffers from dealing with the issue of sufficiency after her version of the argument from spontaneity. The latter can therefore be criticised for the same reason as Allisons Reciprocity Thesis. This problem is solved through a re-ordering of Korsgaards argument. For, whatever the particular situation, given the Highest Good as rational end, the transcendentally free agent must abstract from this situation to ground her action, i.e. for it to be justified. This addresses the worries discussed in the previous sections analysis of Korsgaards argument by drawing upon the valid component of Allisons Reciprocity Thesis (and of Korsgaards Argument from Spontaneity): the requirement of justification for a transcendentally free agent is one which transcends any particular perspective. That is, the perspective-free overriding value of the Highest Good is overriding from any situated perspective for a transcendentally free agent. In other words, the Reciprocity Thesis holds for an agent conditionally upon the Highest Good being a rational end.
67 68

69 70

Wood, op. cit. That such guilt can exist without a situatedly possible incentive is certainly possible: all other inclinations being equal, the agent might act so as to reduce his feeling of guilt, but this would not amount to action from the moral incentive. And this distinction is important to ward off the type of Hegelian inspired criticism which accuses Kant of alienating the agent by making the unity of his self impossible. KU, AA 05: 471.2128. Ward, Keith: Kants teleological ethics. In: Immanuel Kant: Critical Assessments, vol III. Ed. R. Chadwick, London 1992, 251.

Brought to you by | Niedersaechsische Staats- u. Universitaetsbibliothek / SUB Goettingen (Niedersaechsische Staats- u. Authenticated | 172.16.1.226 Download Date | 3/8/12 3:10 AM

512

Christian Onof

This can be summarised in the following hybrid reconstruction of the proof from the fact of reason: (1) The fact is one of reason since the moral law can only have a rational origin. (2) Since the agent takes the law as governing his will, he has a pure interest in it. (3) The presence of such an interest entails that the law is a possible incentive, which grounds pure positive freedom practically. (4) Negative freedom is established by observing that consciousness of the law as constraint entails consciousness of ones independence of natural causality. (4") From (3) and (4), we must practically view ourselves as transcendentally free (positive freedom). (5") From (4"), we must view ourselves as having an intelligible dimension, and assuming the Highest Good as a common end for beings with such a dimension, this defines a higher vocation than following our inclinations. (6') From (5"), we have an incentive to act from duty. (7) The Reciprocity Thesis holds for an agent conditionally upon the Highest Good being a rational end: the sufficiency of the moral law is a consequence of (6') and that morality is required for the transcendentally free agent is a result of the need for a final justification. (8) From (4") and (7), we are required to act from this incentive. The completeness of the grounding of the moral law is thus achieved if one is able to show that the pursuit of the Highest Good is an end given through the fact of reason (or some other feature of the faculty of reason). The validity of this claim about the Highest Good is too large a topic to do proper justice to in this paper, but I could refer to an abundant literature to suggest that it is widely thought that the pursuit of Highest Good has not been shown by Kant to be derivable from the mere consideration of rationality.71 This should come as no surprise given the central role it assigns to the concept of happiness. What this suggests is that Korsgaards introduction of the Highest Good into the argument is too thick a notion, and that a thinner concept would be needed to fill the gap identified in Allisons reconstruction of the Reciprocity Thesis. But that is where the problem lies: apparently no such concept is available to Kant.72 In line with the observations made earlier, we could look for such a concept beyond morality itself. The problem for Kantian ethics would be that if one goes beyond mere rationality, there is nothing else available since Kants ethical principles are constructed from pure rationality itself. This might suggest some revision of the content of the ethical principles is required, and for these new principles, minimal grounding assumptions (thicker then mere rationality) could be sought.
71 72

Beck, 1960, op. cit., 244255. The other natural candidate would be the Kingdom of Ends. But to show that this is an end of reason is tantamount to grounding the moral law. This can therefore not constitute a starting point in a foundational argument.

Brought to you by | Niedersaechsische Staats- u. Universitaetsbibliothek / SUB Goettingen (Niedersaechsische Staats- u. Authenticated | 172.16.1.226 Download Date | 3/8/12 3:10 AM

Reconstructing the grounding of Kant's ethics: a critical assessment

513

One can view important ethical theories developed in the 20th century which are direct inheritors of the spirit of Kants ethical programme as attempts in this direction. Thus Rawls on the one hand, and Apel and Habermas on the other, have sought to construct a framework for universal normative practical principles. What do they draw upon to ground their frameworks? Rawls73 requires that we consider an a-historical ideal situation, the original situation, from which we can decide which kind of social contract would produce a fair and just society. Apel74 and Habermas75 require that an ideal discussion between all members of a society lead to a decision upon what rules the participants will follow. The question is whether the starting points of these approaches are uncontroversial and can therefore ground ethical claims. For Rawls, the question is about the distribution which is to be agreed upon in the ideal state. This is not unproblematic, for, as Nozick76 argues, if we are to distribute benefits and burdens among persons, this assumes clear boundaries between persons and assets. But these are not given, and are potentially difficult to draw since real individuals are always already found with assets. In other words, the apparently neutral idea of a just decision in an ideal situation is not a conception which springs from mere rationality, and any such decision will involve social conceptions (persons, assets, ) that do not meet with general assent, hence a problem arises for such an approach to normative grounding which again has its roots in an issue of situatedness. Apel and Habermas try to minimise the assumptions their discourse ethics rely upon by presenting them as derivable from rationality together with basic claims about communication. Apel thus talks of a transcendental grounding for his ethics:77 the conditions of the ideal discussion are presupposed by everyday argumentative discourse. This does not, however, ensure that some consensus will be obtained. In other words, it could be argued that the thin concept of the transcendental conditions of everyday argumentative discourse is not sufficient to ensure that any ethical principle receives general assent as a result of a process of discussion. The problem could be understood as resulting from the absence of any requirement of shared purposes. If no assumption such as that of the Highest Good as a common end is available, there is no guarantee of convergence. Without going further into the discussion of these ethical theories, we can conclude that the problem of grounding Kants moral law does not obviously disappear by considering alterations of the content of morality because of worries that are similar to those raised in our examination of Kants argument.
73 74

75

76 77

Rawls, John: A Theory of Justice. Oxford 1971. Apel, Karl-Otto: Diskursethik als Verantwortungsethik eine postmetaphysische Transformation der Ethik Kants. In: Kant in der Diskussion der Moderne. Ed. J. Schnrich and Y. Kato, Frankfurt am Main 1996, 326359. Habermas, Jrgen: Diskursethik Notizen zu einem Begrndungsprogramm. In: Moralbewutsein und kommunikatives Handeln, Frankfurt am Main 1983, 53126. Nozick, Robert: Anarchy, State and Utopia. New York 1974. Op. cit., 332.

Brought to you by | Niedersaechsische Staats- u. Universitaetsbibliothek / SUB Goettingen (Niedersaechsische Staats- u. Authenticated | 172.16.1.226 Download Date | 3/8/12 3:10 AM

514

Christian Onof

4. Beyond the fact of reason What the above analysis has shown is that reason arguably requires a grounding of our actions in the moral law, but that human agentss situation is such that they cannot have access to such a ground. This is a state of affairs closely related to that of the faculty of reason in the first Critique. For, although reason seeks the unconditioned, the limits of our epistemic powers imply that it cannot access anything unconditional through knowledge. The second Critique was meant to compensate for our epistemic failings by giving us access to an unconditional principle, the moral law. Instead of knowledge of the unconditioned, reason was to provide a binding moral law and a correlative notion of rational faith. It now seems that the practical domain is plagued by a similar frustration of the faculty of reason as in the epistemological domain, and on analogous grounds. In the practical domain, the search for the conditions for any given conditioned78 is replaced by the demand for a justification for any action or adoption of ends. This means that, rather than climbing towards an unconditioned through a series of conditions, we advance towards a ground through a series of p-meaningful justifications. However, our attempt to reach the ground is thwarted, not, as in the theoretical case, by our epistemic limitations, but rather by our limitations as situated agents. Both however, are connected with our faculty of sensibility, respectively in its providing intuitions that are required to give content to conceptual representations, and in its being the source of inclinations and ends. I would like to suggest, on the one hand, that if there is indeed such a similarity, reason may play a role in regulating agency as it plays a role in regulating knowledge; and on the other, that such similarity hides a subsisting lack of symmetry between the two domains. Let us first clarify the similarities between the theoretical and practical domains. Recalling that in the theoretical domain, Ideas of Reason represent ways in which reason could attain the unconditioned if this were a possible object of knowledge, we see clear parallels emerging. In the practical domain, the moral law would provide the ground of action which reason requires for perfectly rational beings, i.e. beings who are not situated in a sensible world. In the theoretical domain, the ideas of God, the soul, would fulfil the rational requirement for the unconditioned, for beings whose epistemic capacities are not limited by the faculty of sensibility (i.e. beings with an intellectual intuition). In the practical domain, the moral law stands as the ultimate principle of action for purely rational beings. But, further, for beings like us whose epistemic capacities are limited by our sensible form of knowledge, the Ideas of Pure Reason provide a horizon which guides the unification of knowledge. That is, what we know can be organised by reference to God, the soul, The Ideas of Pure Reason enable us to view our objective knowledge as if it were of a Nature created by God, of a soul, These ideas are
78

KrV, A 336/B 393, AA 04: 213.35/ AA 03: 259.1517.

Brought to you by | Niedersaechsische Staats- u. Universitaetsbibliothek / SUB Goettingen (Niedersaechsische Staats- u. Authenticated | 172.16.1.226 Download Date | 3/8/12 3:10 AM

Reconstructing the grounding of Kant's ethics: a critical assessment

515

regulative for our knowledge. In the practical domain, we might similarly want to define some regulatory role for the moral law. But how could this be specified? Translating directly from the theoretical case, this might be insofar as one would organise ones agency as if moral constraints were binding. But moral constraints do not operate in this way, i.e. by providing a way of bringing unity to pre-existing actions. Also, it is not by characterising our agency as if a moral law were binding for it, that sufficient grounds will ever be available to override inclinations. That is, if reason regulates human agency, it must do so in a stronger sense than in the theoretical domain, which reintroduces a dissymetry between the practical and theoretical domains along the lines of Kants original intentions. In what sense can one go further than a mere as if conception of the regulatory? The key must lie in the fact that, as we have seen in this paper, we have a case where reason requires of non-purely rational agents that they act morally, namely the case of beings who share the Highest Good as an end. How could this result be used? On the one hand, what ends we adopt is up to us. Ultimately, we are therefore free to become the kind of beings we choose to become, within constraints defined by our circumstances. On the other hand, however, this cannot mean that we should be compelled to adopt the Highest Good as an end, as there would be no further grounds for so doing except morality itself, which leads to a vicious circle. To avoid this circle however, we need only note that however we choose our ends, the same freedom of the will is an underlying pre-condition. The value of this very freedom is thereby recognised in all such choices. Moreover, as we noted earlier, the value of freedom and its preservation could arguably be substituted for the fact of reason and the Highest Good in defining a moral end in the foundational argument we reconstructed above. I am only sketching an argument here, but the key must lie in noting that valuing freedom leads to a respect of human agency in any of its instantiations, and a rejection of any form of surrendering of ones freedom to serve other values.79 Additionally, taking the value of freedom of the will as the key to the normativity of the moral law is clearly in the spirit of Kants writings, as Guyer80 convincingly demonstrates. Guyer sees this strand of Kants moral thinking as, moreover, Kants most compelling line of thought81, in particular insofar as the capacity to rule ourselves by our own principles [] possesses and offers us an extraordinary sort of
79

80 81

To properly understand how this could help define an incentive to act morally, we need to consider that in placing ones faith in the value of freedom, one will simultaneously make certain individual choices. This means that the content of morality would feature universal features such as the moral law as well as individuated features. The latter would address the issue of the indeterminacy of Kants ethics that leaves room for further specification of action which is arguably morally relevant (Onof, op. cit., 1998). As a result, the individual agents motivation would be grounded in this individuated involvement in constructing moral requirements: this constructivist ethics would provide the agent with a motivation fulfilling the role of the Highest Good in providing her with an incentive to act morally. Guyer, 1998, op.cit., 2235. Ibid., 28.

Brought to you by | Niedersaechsische Staats- u. Universitaetsbibliothek / SUB Goettingen (Niedersaechsische Staats- u. Authenticated | 172.16.1.226 Download Date | 3/8/12 3:10 AM

516

Christian Onof

dignity (ibid.), one which is readily graspable even without philosophical reflection. It is this very value of freedom that I suggest should be at the core of a justification of morality. But grounding the overriding role of such a value could not be achieved any more than grounding the fact of reason.82 In the light of the above, it would therefore seem that espousing this as the highest value is what reason requires, although reason itself cannot guide us further along this path. That is, reason points to the need to adopt this as the core value definitive of practical normativity, but reason cannot ground it. We therefore have to go beyond rationality, it would seem, to grasp this value. This is the kind of commitment which is required of a practical agent, and it could be characterised by a leap of faith, one which lies between the Kantian and Kierkegaardian notions of faith. The Kantian notion of rational faith is circumscribed within the bounds of reason. Practical rationality leads to certain conclusions (e.g. the existence of God or the immortality of the soul), which cannot be objects of knowledge because of the theoretical limitations of the faculty of reason.83 The Kierkegaardian notion of faith is one which leaves reason behind, to the extent that it requires the acceptance of the paradox that is involved in encountering the absolute.84 A practical grounding of the moral law in the value of freedom finds roots in the dignity that characterises free will for any practical agent, and meets the requirement of practical reason for p-meaning, i.e. for a final ground of action. It is therefore in tune with the needs of practical rationality. But here, it is not reason in its theoretical employment, but also practical reason which falls short of the task of securing such a foundation. Practical reason can therefore only call for such a leap of faith beyond reason so as to secure a proper grounding for human agency.85 Far from amounting to a rejection of rationality as in Kierkegaards religious phase, such a leap secures the possibility of a regulation of human agency by the dictates of pure practical reason.86
82

83 84

85

86

Assigning a central role to the value of freedom is, of course, a typical characteristic of existential ethical claims. Sartre, in particular, in his struggle to put together an ethics, focussed upon freedom as providing the ground (Sartre, Jean-Paul: Being and Nothingness. London 1958, 627). The commitment to freedom involves choosing to live ones life authentically (Heidegger, Martin: Being and Time. Oxford 1962, 267). This existentialist notion provides a useful characterisation of the sense in which it is by choosing what kind of being one is, that the grounding of a morality becomes a possibility. KpV, AA 05: 133.2332. Kierkegaard, S.: Philosophical Fragments. Trans. H. V. Hong and E. H. Hong, Princeton 1985, 51. That the value of freedom is the only possible object of such a faith cannot be shown within the context of this paper. In the theoretical domain, the pretensions of reason are curbed by the faculty of sensibility through the faculty of understanding which defines the limits of possible knowledge through a set of principles. Reason can then organise that knowledge according to its regulative principles. In the practical domain, the regulative role of reason is key to the rationality of human agency. Insofar as the content of the moral law is derivable from pure practical reason, reason does find its true calling in the practical domain. But sensibility remains a limiting condition in the sense that it defines a situation which lies beyond the context of pure

Brought to you by | Niedersaechsische Staats- u. Universitaetsbibliothek / SUB Goettingen (Niedersaechsische Staats- u. Authenticated | 172.16.1.226 Download Date | 3/8/12 3:10 AM

Reconstructing the grounding of Kant's ethics: a critical assessment

517

As such, it may be described as meta-rational faith. If the above is correct, it provides a grounding for morality that is valid for any individual who commits herself to it.

Conclusion An examination of two contemporary reconstructions of Kants derivations of the moral law has been carried out. It shows that this law can be properly grounded on the assumption that the Highest Good is an end adopted by all rational beings. This result is important even though it does not provide an indisputable grounding for Kants ethics. Having noted the problems encountered by other attempts to ground universal ethical principles in the broadly Kantian tradition, the need for a leap of faith guided by reason, and involving a commitment to the value of freedom, is suggested as a way of providing an ultimate justification of action.

rationality. The difference with Kants original foundational programme is that sensibility cannot simply be ignored, but reason needs to work with it to define an incentive for acting morally which overrides other incentives. It is in putting ones faith in the value of our capacity for free choice that this can be achieved.

Brought to you by | Niedersaechsische Staats- u. Universitaetsbibliothek / SUB Goettingen (Niedersaechsische Staats- u. Authenticated | 172.16.1.226 Download Date | 3/8/12 3:10 AM