As described by Kant, ethics provides us with the “laws of freedom,” telling us what we ought to do, while physics provides us with the “laws of nature,” merely describing what is. Far less obvious and far more controversial is the distinction that Kant draws within both ethics and physics: between the empirical, or sensible, and the pure, or rational. It is this distinction that is central to Kant’s most important project in ethics: to determine once and for all the exact nature of the moral law.
Section 1 of this chapter argues that there is a single basic procedure employed by Kant, across his writings in ethics, for laying bare the moral law in any of its formulations. I term this Kant’s “Elimination of Sensibility Procedure,” or “ESP.” It is by means of ESP that Kant separates the pure from the empirical in ethics and achieves clarity on the moral law. ESP begins on a reflective, even romantic note, with a moment of wonder. In a world otherwise advancing lock-step according to the laws of nature, we feel wonder at the sheer fact of the existence of morality – at the fact of the existence of moral imperatives and the implied fact of our ability to freely choose out of respect for these imperatives. As Kant most famously describes this wonder at the moral law within us (as well as at the starry heavens above) in his Critique of Practical Reason, it is an “ever new and increasing admiration and awe” (KpV 5:161). But this moment of wonder is just the beginning of Kant’s inquiry concerning the principles of morals, an inquiry that will take him in a markedly different direction than Hume’s before him. The general strategy of what remains of this inquiry is revealed two pages later in the Critique of Practical Reason, in the final paragraph of the book, when Kant explains that we ought to proceed in our study of ethics just as we do in the broader study of nature: We ought to use “a process similar to that of chemistry … [whereby] we … separate the empirical from the rational, exhibit each of them in a pure state, and show what each by itself can accomplish” (KpV 5:163). To achieve this separation, Kant’s ESP highlights a single, crucial feature about the basic fact of morality: its modality. The moral law is unique in commanding with necessity, telling us that we should do something not because it will promote our happiness, but simply because it is the right thing to do, in itself. Not only does the moral law command with necessity, but the will determined by this moral law, in turn, has necessary, unqualified worth. Having identified necessity as a defining characteristic of moral imperatives, Kant’s ESP advances with an eye to Kant’s distinction in kind between reason and sensibility, asking which of our faculties could ground such laws that command “absolutely (not merely hypothetically under the presupposition of other empirical ends), and are necessary in every respect” (KrV A807/B835). ESP, then, systematically rejects our sensibility in part and in whole as a possible source of such a moral law or of a good will determined by this moral law. Having systematically rejected sensibility, ESP turns to reason alone, now clearly distinguished from sensibility, and next specifies how it, in its purity, authors the moral law, and what this moral law is.
The second part of this chapter applies these positive findings regarding ESP to an assessment of recent “constructivist” interpretations of Kant. I point to some problems with a number of constructivist systems and then argue against Korsgaard’s constructivist interpretation of Kant that it, like the other constructivist strategies, rests on precisely the mistake that Kant repeatedly says has haunted the history of moral philosophy, namely, that of conflating sensibility’s pragmatic good and understanding’s moral good. This is the mistake to which the title of the chapter refers. Of course, the title also alludes to H. A. Prichard’s classic 1912 essay, “Does Moral Philosophy Rest on a Mistake?” The mistake to which Prichard drew attention was that of demanding a reason for something that is already self-evident. This mistake also in part defines Korsgaard’s account, as Korsgaard interprets Kant to hold that our sensible incentives as well as our conviction that humanity has value in itself are both in need of further supporting reasons before they can serve as grounds for a choice.
Across Kant’s work in ethics – in his Groundwork, his Critique of Practical Reason, his lectures on ethics, and many other sources – Kant maintains that the single mistake that has most consistently plagued moral philosophers throughout history is a very fundamental one. This is the mistake of not distinguishing properly, i.e., of not distinguishing in kind, between the realm of sensibility and the realm of intellectuality. This is the same mistake that has plagued theoretical philosophy throughout history, with Kant famously observing that it has led some to “intellectualize” sensibility and others to “sensitivize” intellectuality. This failure is just as crucial within moral philosophy because it has brought with it the failure to draw the distinction in kind between two irreducible senses of “good.”
The first kind of good is the pragmatic good, happiness. The kind of desires that aim at this good, according to Kant’s theory of action, are what Kant refers to as our sensible incentives (Triebfedern). Sensible incentives are one of two fundamental kinds of desire, or “conative currency,” as I call it, in Kant’s theory of action.
The second kind of good is the morally good. The morally good has its value in itself, not, say, because of any happiness it may or may not give rise to. Our intellectual incentives, or motives, to choose morally are the second of the two fundamental kinds of desire, or conative currency.
In notes on Kant’s lectures on ethics from 1774–8 as taken by Johann Friedrich Kaehler, which are reproduced in the better-known 1784–5 Lectures on Ethics Collins, Kant lays out the division between ethics as grounded in either of these two kinds of good, i.e., in either an empirical or intellectual good. He explains that “morality rests either on empirical or intellectual grounds, and must be derived from either empirical or intellectual principles. Empirical grounds are those that are derived from the senses, insofar as our senses are satisfied thereby. Intellectual grounds are those where all morality is derived from the conformity of our action with the laws of reason” (Collins 27:252–3).
Those grounding their ethics on empirical grounds can do so in a number of ways, Kant explains. First, they can ground them on empirical inner grounds. These grounds are inner in that they issue from within our psychological domain, in inner sense, and they are empirical in that they issue from our lower, sensible faculties, which are passive in the sense that they do not presuppose our self-conscious, reflective, active involvement. One type of inner empirical ground is what Kant here calls physical feeling, the feeling of “self love,” which breaks down into vanity and self-interest. This anticipates Kant’s “selfish purpose” of the prudent merchant in Groundwork I (GMS 4:397). Here Kant points us to Epicurus, Helvetius, and Mandeville as philosophers grounding their ethics in these sources. Another empirical inner ground, in addition to physical feeling, is moral feeling, and here Kant points us to Shaftesbury and Hutcheson. This anticipates Kant’s “immediate inclination” of the “friend of mankind” in Groundwork I (GMS 4:398). Kant also identifies empirical external grounds, such as education and government, and here he cites Montaigne and Hobbes as philosophers who ground their moral principles in these external forces.
Time and again, Kant notes the shared modality of all of these empirical grounds for morality: the modality of contingency or conditionality. Each rests on something that is conditional or contingent to us in our status as rational agents, i.e., each rests on something that is not necessary to us simply as rational beings per se; for example, they rest on what happens to provoke a physical feeling or what happens to provoke a moral feeling in us, given our particular sensible nature, or they rest on the particular system of education or government that has shaped us: “However, if the principle of morality rests on self-love, it rests on a contingent ground, for the nature of actions, whereby they bring me pleasure or not, rests on contingent circumstances. If the principle rests on a moral feeling, where the action is judged by the satisfaction or dissatisfaction, by the sensation or in general the feeling of taste, it also rests on a contingent ground… And so it is also with the external grounds of education and government. Under the empirical system, the principle of morality rests on contingent grounds.” (Collins 27:253–4)
But what is so bad about a conditional, contingent foundation for morality? Kant’s response is simple: It is a self-evident fact that moral imperatives simply do command with necessity, unconditionally. And if this is the case, they can only come from a source that holds for all of us necessarily as rational agents, and this source is our purely intellectual faculty of reason. In Kant’s words: “Those principles which are supposed to be everywhere, always and necessarily valid, cannot be derived from experience, but only from pure reason. Yes, the moral law expresses categorical necessity, and not a necessity fashioned from experience,” and that because “all necessary rules must hold good a priori … the principles are intellectual” (Collins 27:254).
So there are two kinds of good, moral and pragmatic. Moral imperatives command categorically; our sensibility cannot ground categorical commands while our reason can; and so the categorical imperative must issue from reason.
Any attempt to determine the morally good is also doomed from the start, Kant argues, insofar as it combines these two senses of good in one, with a single foundation. Here the culprit, as so often for Kant, is the tendency to take reason’s demand for simplicity as constitutive rather than regulative: “Philosophy endeavors to bring about unity in our knowledge, and to reduce it to the fewest principles, [and so] the attempt was made to see if one principle might not be put together out of these two” (Collins 27:249).
Kant believes that the murkiness, across the history of moral philosophy, surrounding the distinction between sensibility and intellectuality, and the failure to isolate the origin of the moral law in intellectuality and specifically pure reason alone, has naturally enough led to a great deal of confusion about the relationship between living morally and living happily. If it is a feeling – whether a moral or physical one – that determines our moral obligations, it begins to appear plausible that moral living has some causal connection with our own pragmatic goal, of (sensible) happiness. Thus throughout history, Kant argues, there have been countless misconstruals of the highest good, or summum bonum, which combines virtue and happiness. Thus some of the mentioned empiricists, Epicurus, Helvetius, and Mandeville, went astray in thinking that the highest good was happiness alone and that virtue would conveniently serve as a means to happiness, while others, the stoics, were equally misguided, reducing the highest good to virtue alone, also believing that virtue would conveniently serve as a means to happiness. Again, Kant’s point is that the moral, intellectual good and the pragmatic, sensible good are distinct in kind, with neither the moral good nor the pragmatic good serving as the necessary or sufficient condition for the other.
Jumping ahead to 1793 for a moment, to Kant’s “On the Common Saying: That May Be True in Theory, but It Is of No Use in Practice,” we get a taste for how consistent Kant was, across his writings, in pinning failures in ethics on the problem of the conflation of the moral and the pragmatic in ethics. Here Kant summarizes Garve’s complaint about Kant as follows: that all action must be directed to what we prefer, that what we prefer must be something we take to be “good,” that what we take to be good is what is described by the word happiness, and that, accordingly, the motive for acting on the moral law, too, must be our pursuit of happiness (TP 8:281). In response, Kant predictably explains that “this argument is nothing but a play upon the ambiguity of the word the good; for this [can be taken to mean] either what is good in itself and unconditionally as opposed to what is evil in itself, or else what is only conditionally good as compared with what is a lesser or greater good…” and that what we have here is “a difference in kind” (TP 8:282).
Likewise, in his 1795 “Perpetual Peace” Kant considers the objection that morality demands of us things that are impossible in politics. His response is that morality never commands what we cannot do, and that what must really be meant here with the objection is that morality demands of us things that are impossible to do in a manner consistent with our happiness. But if this is the tension, then acting on the moral law is only impossible if we misconstrue the moral law itself as a “general doctrine of prudence.” And to misconstrue the moral law as a general doctrine of prudence is, of course, in effect, to “deny that there is a [doctrine of morals] at all” (ZeF 8:370). Indeed, as Kant would put it two years later, in his 1797 Metaphysics of Morals, these philosophers who conflate the pragmatic and moral good are guilty of “the euthanasia (easy death) of morals” (MS 6:378).
What is important to note about Kant’s discussions about the persistent confusion around the roles of sensibility and intellectuality in morality – both for the purpose of understanding Kant’s criticisms of rationalism and empiricism in ethics and, in turn, for recognizing constructivists’ misunderstanding of Kant’s criticism of rationalism in ethics, as we will see in our next section – is that they do not only implicate empiricists, but also rationalists. For while rationalists would seem to be on the same page as Kant in their reliance on reason for their findings in moral philosophy (and all else), Kant sees significant daylight between his and their positions. In Kant’s view, they simply are not in a position to do a good job of relying on reason: Their failure is due not to their reliance on reason’s insights per se but to their failure to distinguish in kind between reason and sensibility, as they accordingly fail to separate out reason’s insights from sensibility’s. Thus, just as Kant, in the Critique’s chapter on the Amphiboly, sees both empiricists and rationalists guilty of the conflation of empirical and intellectual concepts, so, too, does he here see both empiricists and rationalists guilty of this conflation in the moral realm, with the result that neither is able to isolate reason and identify the moral law, including the rationalists.
Thus in his lectures on ethics, Kant attacks the perfectionism of Alexander Baumgarten, a Wolffian. Baumgarten’s principle of morality, he argues, demands that we do good and abstain from evil. This principle, however, suffers the same fate for Kant as Garve’s and others’ conceptions of morality, insofar as it fails to distinguish the pragmatically good from the morally good. What Baumgarten offers is accordingly not merely a tautology, insofar as it tells us simply to do the good thing. It is also, and even worse, in Kant’s view, a “principium vagum” (Collins 27:264), because it does not even manage to distinguish as desiderata definitions of a sensible and a moral good before failing to define the moral good.
In failing to distinguish between a moral and a pragmatic principle, Baumgarten has also failed to identify a principle of obligation, Kant argues, because only a moral principle obligates, while a sensible principle, of skill or prudence, is a mere problematic imperative. To the extent that Baumgarten also enjoins us to seek perfection, he likewise is confused, Kant argues, because “perfection and moral goodness are different” (Collins 27:265): Perfection, Kant argues, concerns the generic “completeness of man in regard to his powers” (Collins 27:265), while goodness “consists in the perfection of the will, not the capacities,” so that Baumgarten is again looking in the wrong place, in addition to never specifying what the moral law is that would perfect our will, and in the end gives us only a principle that is tautologous and vague in an amphibolous manner.
Our review of these sources from before and after the Groundwork positions us to consider Kant’s rejection of rationalist perfectionism in the preface to his 1785 Groundwork. Here Kant rejects rationalist perfectionism again, this time as presented by Wolff himself. And again, his criticism homes in on the conflation of empirical and intellectual imperatives in this rationalist perfectionism. Thus, Kant strikes preemptively against those who might see him peddling stale goods already supplied by Wolff, by making clear how Wolff has fallen short of what Kant will attempt to do. How so? Again, Wolff has merely “considered volition in general” (GMS 4:390). Wolff has, that is, failed to isolate the rational, or moral volition, because he has failed to strip it of all that is empirical. In other words, Wolff has failed to recognize, as described in the Critique’s Amphiboly, the distinction in kind between the two basic sources of determination of actions: rational ones, deriving from reason, and empirical ones, deriving from our inclinations. Here Kant summarizes these points about how Wolff and other rationalist perfectionists fail, saying that they
do not distinguish motives that, as such, are represented completely a priori by reason alone and are properly moral from empirical motives, which the understanding raises to universal concepts merely by comparing experiences; instead they consider motives only in terms of the greater or smaller amount of them, without paying attention to the difference of their sources (since all of them are regarded as of the same kind); and this is how they form their conception of obligation, which is anything but moral, athough the way it is constituted is all that can be desired in a philosophy that does not judge at all about the origin of all possible practical concepts, whether they occur only a posteriori or a priori as well.
By contrast, Kant explains that his own moral philosophy will present a foundation, or groundwork, of a metaphysics of morals, precisely by isolating our intellectual faculty of reason and its a priori insights. He tells us that “moral philosophy rests entirely on its pure part” (GMS 4:389) and that “a metaphysics of morals is thus indispensably necessary” (GMS 4:389) among other reasons because otherwise “morals themselves are liable to all kinds of corruption” (GMS 4:390).
We have now seen how Kant views his own moral philosophy departing most fundamentally from previous moral philosophies by virtue of its distinction in kind between sensibility and intellectuality. And we have seen how these previous accounts include even rationalist, perfectionist ones. Turning to Groundwork I and II we can also use this background regarding this fundamental distinction in kind to expose a common strategy in the first and second sections of the Groundwork, which are otherwise so different. This is the mentioned Elimination of Sensibility Procedure, or ESP. I will review this process only briefly before turning to the topic of constructivism.
Kant begins Groundwork I with a simple assertion: We all recognize, as a fact, that a good will is good without qualification. Even where such a good will is useless, he writes, it will nonetheless “like a jewel, still shine by its own light” (GMS 4:394). Likewise, Groundwork II properly gets started when Kant simply asserts the following: “Finally, there is one imperative which immediately commands a certain conduct without having as its condition any other purpose to be attained by it. This imperative is categorical… This imperative may be called that of morality” (GMS 4:416). As in his lectures on ethics, in Groundwork I Kant takes it for granted that we all recognize the unqualified and necessary value of a good will. Likewise, in Groundwork II Kant argues that we all recognize that there is a moral law and that it commands with necessity. These starting points are moral realist starting points, asserting the value in itself of a good will and the existence of a moral law that commands categorically.
And this is what gets the ball rolling for Kant. It is ESP that takes us to the categorical imperative, which, in turn, is what supplies us with a method for testing the moral status of particular maxims, a method referred to by Rawls as the “CI-procedure.” These starting points are what constructivists neglect, not recognizing that they are what serve as the crucial foundation for Kant’s eventual embrace of the categorical imperative, which supplies the constructive procedures that they so value, in the various formulations of the categorical imperative. In other words, Kant does not think that the categorical imperative and its procedures come from nowhere or are themselves constructed without an underlying moral realist foundation. Instead, it is because we first recognize that wills determined by the moral law are good without qualification and that there are moral imperatives that command with necessity (and that determine a good will), that we can subsequently recognize the various formulations of the categorical imperatives as formulations of the moral law. As Kant puts this point – about the need to begin our exploration into the nature of the moral from our recognition of the fact of its necessity – in the Critique of Practical Reason, “But how is the consciousness of the moral law possible? We can come to know pure practical laws in the same way we know pure theoretical principles, by attending to the necessity with which reason prescribes them to us and to the elimination from them of all empirical conditions, which reason directs” (KpV 5:30).
It is this “elimination … of all empirical conditions” to which we next turn. In Groundwork I, on the basis of the conclusion of the necessary worth of a good will, Kant is able to systematically eliminate all empirical, or sensible, determinations of the will, as such determinations will only always yield a will with conditional value. Having eliminated all the offerings of sensibility, Kant asks what else could possibly determine the will, and now, having eliminated sensibility, we arrive at the Formula of Universal Law (FUL), or at least an early version of it. In Groundwork II, the structure of the argument is the same, with Kant first arguing that moral laws are by their nature categorical, eliminating as possible moral laws all hypothetical imperatives of sensibility. The question is then what the moral law could be, given a lack of empirical grounds, and the answer is again FUL (and then, after switching to consideration not of the form but of the matter of such a necessary moral law, the Formula of Humanity as an End in Itself, etc.). Kant thus recognizes a distinction in kind between sensibility and intellectuality; sees other accounts failing to recognize this distinction and therefore arriving at confused accounts of morality; and then himself sets things right by carefully navigating the waters of this distinction, eliminating the offerings of sensibility, and isolating the teachings of pure reason.
I now turn to constructivism. As is well known, John Rawls revitalized normative ethics in the 1960s and 1970s, most famously with his 1971 A Theory of Justice. Rawls was writing not only at a time when normative ethics was out of favor, but also at a time when, in the area of metaethics, there was a great deal of hostility toward moral realism. There was, likewise, a great deal of hostility toward rational intuitionism, which claims that we can arrive at insights regarding objective moral truths by means of rational intuitions that are self-evident. That is, these self-evident truths do not have a binary justification structure, where the proposition demands a separate justification, but instead have a unary justification structure, where the proposition is self-evident. Rawls and the later Kantian constructivists do not challenge what had been (at the time of Rawls’s writing) this dominant hostility toward moral realism and yet they hope to secure an objectivist ethics, not a mere relativism. They hope to square this circle by recourse to a new “constructivist” strategy, according to which they will offer an underlying procedure instead of some alleged insight of some alleged moral fact. This procedure will then be used to construct the content of the moral doctrine. But while this approach certainly has us investing a procedure with a great deal of responsibility for determining the content of the doctrine, this does not necessarily eliminate a place for moral realism or even a dependence on one. For the obvious question that still remains is Why should we defer to this procedure and not some other one, if any at all? In the words of Onora O’Neill, discussing the procedures of the various formulations of the Categorical Imperative in relation to Rawls, “The more demanding question is whether they [these procedures] can themselves be justified without reintroducing some form of moral realism by the back door” (O’Neill 2002:355).
In an attempt to answer this question about the justification of Rawls’s “CI-Procedure” itself, and to do so without recourse to moral realism, Rawls, after asking the question “Is the CI-Procedure itself constructed?” answers “No, it is not. Rather, it is simply laid out. Kant believes that our everyday human understanding is implicitly aware of the requirements of practical reason, both pure and empirical” (Rawls 1989:99) – a characterization of the starting point for justification of the CI Procedure that sounds like rational intuitionism and in turn moral realism. Rawls also says that this procedure is justified on a coherentist basis, as it is a procedure that squares with our conception of free and equal persons as rational and reasonable. But, again, why defer to our conception of a person as reasonable, and how is this different from having a rational intuition of what it is to be reasonable, or moral? Moving to Rawls’s political theory, we find a parallel in his account of the Original Position. How is this procedure justified? Rawls again offers a coherentist justification, whereby we look at the principles dictated by the Original Position and see whether they conform to our “considered judgments.” Again, however, whose considered moral judgments is Rawls referring to, and why should they or we defer to them, if not because they or we think that these considered judgments grasp moral truths, and how would that avoid moral intuitionism and moral realism? As is well known, Rawls eventually backs away from these views, or at least becomes more explicit about the limited scope of his claims. That is, finding his objectivist account vulnerable to the charge of moral realism, Rawls scales back his ambitions, limiting himself to more modest, political, relativist conclusions, that find their grounding in local, non-universalizable sources.
Before moving on to discussion of the next attempt at a Kantian constructivism, in the work of Onora O’Neill, it is worth pausing to ask why Rawls might take such pains to avoid ascribing to Kant a moral realism or (rationalist) intuitionism that could ground the constructive procedures spelled out by the categorical imperative. After all, we saw Rawls conceding that for Kant the categorical imperative itself “is simply laid out” by our “everyday human understanding” (Rawls 1989: 99), and he also acknowledges that “Kant says that the moral law can be given no deduction, that is, no justification of its objective and universal validity, but rests on the fact of reason. This fact (as I understand it) is the fact that in our common moral consciousness we recognize and acknowledge the moral law as supremely authoritative and immediately directive for us” (Rawls 1989: 102). Moreover, Kant consistently employs other examples of moral realist language, as when he asserts that rational nature is “something whose existence has in itself an absolute worth” and that “rational beings are called persons inasmuch as their nature already marks them out as ends in themselves” (GMS 4:428); that the moral law “rests on the inner nature of the action, so far as we apprehend it through the understanding” (Collins 27:254); and that “the moral imperative expresses the goodness of the action in and for itself, so that moral necessitation is categorical and not hypothetical” (Collins 27:255–6).
Did Rawls’s own philosophical discomfort with intuitionism incline him in this direction? Did he feel pressure to exhaust alternatives to an intuitionist interpretation of Kant, given that, as J. L. Mackie put it about the intellectual climate in which Rawls was writing (in Mackie’s 1977 Ethics: Inventing Right and Wrong), “Intuitionism has long been out of favor, and it is indeed easy to point out its implausibilities” (Mackie 1977: 38). Regardless, the main reason that Rawls cites for rejecting a moral realist interpretation of Kant’s ethics, which reason is then echoed in the work of O’Neill (O’Neill: 355), is Kant’s rejection of a Leibnizian/Wolffian/Baumgartian perfectionism (Rawls 1989: section 3). But as we have seen, while Kant does indeed reject these rationalists’ views, the charge he levels against them is not one of moral realism. Far from it, Kant pinpoints their failure to adequately distinguish, i.e., to distinguish in-kind, between reason and sensibility; and as a result of this, and with a mere distinction-in-degree in place between reason and sensibility, they bungle the distinction between moral good and pragmatic good, in the end having only heteronomous goods to show for it. While Kant acknowledges that their error is an extremely difficult (and transcendental) one to identify, Kant believes that once he has done so (at no small cost in effort), reason can now be exercised to yield insights into facts about moral right and moral wrong. Ironically, having understood Kant’s criticisms as, instead, targeting the rationalists’ use of reason in a moral realist manner, the Kantian constructivists go on to not only reject a moral realist interpretation of Kant but also embrace an alternative interpretation of Kant that imputes to him a constructivism that itself falls prey to precisely the timeless problem that Kant identified in the rationalists’ ethics, of conflating reason and sensibility.
Rejecting Rawls’s particular version of a constructivist interpretation of Kant’s ethics, Onora O’Neill incorporates into her own version an alternative move that we earlier saw Rawls quickly reject, namely, a constructivist justification of the constructivist principles themselves. To the question, Why this procedure? Or Why should I follow it? Onora O’Neill offers a constructivist response: Kant’s morality “begins simply with the thought that a plurality of agents lacks antecedent principles of coordination, and aims to build an account of reason, of ethics, and specifically of justice on this basis. He thinks of human beings as doers before they become reasoners or citizens” (O’Neill 2002:362). O’Neill addresses the question of whether the mere need to coordinate a plurality of agents’ actions is a sufficiently restrictive one to yield rich moral content, and she asserts that it is, albeit under the assumption that “coordination” rules out coercion and deception. But it seems that this assumption only pushes back the question, to why this sort of coordination should be chosen. Here O’Neill appeals to the notion of the “arbitrary,” noting that some factors are arbitrary and should not be taken into account. Indeed, Kant does reject appeal to “arbitrary” considerations, when he rejects merely subjectively valid considerations. But this rejection is not dictated by a morally neutral concept. Instead, Kant rejects merely subjectively valid considerations insofar as they conflict with an independent moral criterion, of objectivity and universality. For subjectively valid considerations may well be far from arbitrary in another sense, according to Kant: They may be those considerations crucial to our attainment of our pragmatic good, happiness. If we adjust O’Neill’s account to Kant’s usage, we would accordingly have to disambiguate the term “arbitrary” and instead use the term “morally arbitrary” here, but in this case we would still need to explain the basis for this moral judgment.
We turn, finally, to the constructivist view of Christine Korsgaard. Korsgaard’s account is the most ambitious of the constructivisms in an important sense, and in a sense that also makes it the most alluring of the constructivist options – if not the most successful. We saw that, without a moral realist foundation, Rawls’s constructivism opts for a relativist anchoring, while O’Neill aims higher, hoping to achieve objectivity through an anchoring in the fact of our predicament as a plurality of agents without antecedent coordination, though in the end she is unable to wring such a rich morality from such a normatively spare set of facts. By contrast, Korsgaard wants to anchor morality in something even more basic to our human condition and even more universal, arguing that absolutely everyone, regardless of who they are or what they think or do, is, and must be, committed to the categorical imperative, simply by virtue of making self-conscious choices: “When you will a maxim you must take it to be universal. If you do not, you are not operating as a self-conscious cause, and then you are not willing. To put the point in familiar Kantian terms, we can only attach the ‘I will’ to our choices if we will our maxims as universal laws” (Korsgaard 1999:27). By arguing that morality is the condition for any self-conscious choice, or, elsewhere, any rational choice or any voluntary choice whatsoever,1 and that there is no coherent alternative to moral choice, her account is the most ambitious in its claims of objectivity and universality.
Here Korsgaard’s interpretation consistently centers on what is commonly referred to as her “regress argument interpretation” of a passage leading up to Kant’s presentation of the Formula of Humanity formulation of the categorical imperative in the Groundwork. Here is the familiar passage:
Now I say that man, and in general every rational being, exists as an end in himself and not merely as a means to be arbitrarily used by this or that will. He must in all his actions, whether directed to himself or to other rational beings, always be regarded at the same time as an end. [1] All the objects of inclinations have only a conditioned value; for if there were not these inclinations and the needs founded on them, then their object would be without value. [2] But the inclinations themselves, being sources of needs, are so far from having an absolute value such as to render them desirable for their own sake that the universal wish of every rational being must be, rather, to be wholly free from them. Accordingly, the value of any object obtainable by our action is always conditioned. [3] Beings whose existence depends not on our will but on nature have, nevertheless, if they are not rational beings, only a relative value as means and are therefore called things. [4] On the other hand, rational beings are called persons inasmuch as their nature already marks them out as ends in themselves, i.e., as something which is not to be used merely as means and hence there is imposed thereby a limit on all arbitrary use of such beings, which are thus objects of respect.
As Korsgaard characterizes it, Kant’s search here for something with categorical value is not an attempt to find something that will render coherent a specifically moral choice by supplying this sort of choice with a grounding in something with unconditional value, even though Kant has said he is searching for “something whose existence has in itself an absolute worth, something which as an end in itself could be … the ground of a possible categorical imperative” (GMS 4:428). Instead, and in what should already give us pause, in Korsgaard’s view Kant is searching for something with categorical value to serve as the ground for any rational, sufficient, or coherent choice whatsoever (Korsgaard 1996a:162, 120). So from the start Korsgaard takes what is actually Kant’s specifically moral inquiry, into whether there is something with absolute worth that could render coherent our moral choices, for a mere generic inquiry, into whether there is something with absolute worth that could render coherent any rational choice whatsoever. Moreover, the search as she characterizes it is not in the form of Kant’s rejection of so many candidates in a list of possible categorically valuable ends. The items are instead said by Korsgaard to be connected in a different way, as a single value regress, where Kant would have each item in the list grounded in the value of the next item (Korsgaard 1996a:120).
Korsgaard accordingly argues first, and in my view correctly, that [1] Kant rejects objects of inclinations as having absolute value, asserting that their value is conditional, not categorical (Korsgaard 1996a:121) and that he next [2] identifies inclinations as the source of the value of these objects but likewise rejects them (Korsgaard 1996a:122). She argues that Kant then looks for the source of value of these inclinations, that he then fails to find anything of value at all, including rational nature, and that he then concludes that we are therefore choosing even though we fail to locate anything of absolute value at all that could serve as the basis for our choice, and that we must therefore, in spite of ourselves … be implicitly valuing our own humanity… and that we must therefore, in spite of ourselves, be valuing the humanity of all.
Before moving on, I will now pause to address the title of this chapter, “Does Constructivism Rest on a Mistake?” and tie in some points covered in Section 1. The title is an allusion, of course, to H. A. Prichard’s famous 1912 essay “Does Moral Philosophy Rest on a Mistake?” Prichard’s main point was that we naturally know right and wrong, that it is self-evident (providing a unary justification procedure rather than a binary one resting on some further proposition), but that, in response to the challenge to morality posed by our inclinations (like Kant’s “natural dialectic,” at GMS 4:405), we mistakenly disregard this self-evident, foundational grasp of value, looking elsewhere for a proof of value.
Next, we saw in Section 1 of this chapter that Kant consistently charges others in the history of philosophy with a failure to recognize that sensibility and intellectuality are distinct in kind, a recognition at the heart of Kant’s critical turn in philosophy, and that for lack of this distinction in kind, these philosophers committed the “euthanasia of morals” (TL 6:378) or, in Kant’s words in the Groundwork, “substitute[ed] for morality some bastard patched up from limbs of varied ancestry” (GMS 4:426). Here Kant repeatedly judges these ethicists with having failed to recognize that there are two basic and irreducible conative currencies: our sensible inclinations and our moral motives. We also saw that Kant, in his attempt to isolate the moral law of our intellect, accordingly eliminates, time and again in his treatment of the topic, the offerings of sensibility before turning to intellectuality.
Returning now to Korsgaard’s interpretation, we can see that what Korsgaard has in effect done, in shooting for the stars with a radical argument for why every self-conscious choice commits us to the moral law, is to commit precisely the error that Kant sees plaguing the history of moral philosophy, and most importantly his rationalist predecessors and contemporaries, who see all choice as directed toward what is best, where they draw no distinction in kind between what is best sensibly and intellectually. She has not recognized the distinction in kind between sensibility and understanding and between inclinations and moral motives. She has instead argued that any and all choices are ultimately grounded in one kind of value, namely a commitment to the moral law, or humanity. Moreover, this failure to recognize this distinction in kind manifests itself in the form, twice, of the sort of mistake to which Prichard drew our attention: the mistake of not recognizing when we have reached rock bottom in a self-evident fact. Kant does not think inclinations have moral value, but he also does not think that they need have moral value in order to be an irreducible kind of conative currency. Kant recognizes that the burden is not on us to provide a reason for why an inclination should incline; inclinations per se do already incline us, even when we are choosing from a reflective distance, which is why they are states of the faculty of desire, or the Begehrungsvermögen (it is accordingly worth noting that Korsgaard misidentifies Kant’s incentives as pleasures rather than as desires, because that latter but not the former already have us desiring, inclining, or being motivated toward the object in question, whereas pleasure per se does not2). Korsgaard’s lack of recognition of the fact that for Kant inclinations can provide grounds for action independent of their moral status is thus the moment at which she fails to recognize the irreducible distinction in kind between the sensible and intellectual realms, and so it is the amphibolous moment in her interpretation that I refer to as its “intellectualization moment.”
Next, notice the subsequent item in the list, [3]: animals. The ESP reading of this part of Kant’s argument for the Formula of Humanity formulation of the categorical imperative has Kant here eliminating, as a candidate for something with unconditioned value, not merely our inclinations in particular but our sensibility as a whole. And how might Kant represent our sensibility as a whole, independent of our reason? As an animal, because animals are just that: sensibility without an accompanying intellectuality. And so the question for Kant is whether animals could be an end with categorical value, and he of course, given his views on animals, answers No. But, turning back to Korsgaard’s interpretation, though Korsgaard claims that a regress to value is ongoing, she never mentions the place of animals on this list, even though her regress interpretation would imply that animals are the basis of the value of our inclinations (Korsgaard 1996a: 122).
We next turn to our intellectual, or rational nature, in [4]. Here, Korsgaard does not have Kant recognizing that it is the nature of rational beings as rational that makes them ends in themselves, with Würde, or dignity. This is the second instance, in Korsgaard’s interpretation of Kant’s argument for the Formula of Humanity, of the sort of move that Prichard flags, and I refer to it as Korsgaard’s “moral anti-realist moment.” Rather than understanding Kant to be helping us to conceptually isolate rational nature from sensibility so that we can now recognize that it is what has value in itself, Korsgaard would have Kant turning to rational nature and not recognizing its value in itself, and thus have Kant coming up empty-handed in an attempt to identify something, anything, with unconditional, absolute value.
Where we saw Kant repeatedly recognizing that there are two distinct kinds of irreducible conative currency, moral and pragmatic, Korsgaard thus has Kant recognizing none. Korsgaard instead has Kant leading us, by regress, to a place where there is no ground for any choice (Korsgaard 1996a:122). I accordingly refer to this as the “conative vacuum moment” of her interpretation.
As Korsgaard describes it, Kant believes we are confronted with a need to choose despite lacking any conative currencies that incline or impel us to make a choice. Faced with this conative vacuum, we make choices anyway, Korsgaard asserts, and this fact of choice in the face of a conative vacuum tells us that … we must implicitly be committed to the value of not only our own humanity but also the humanity of everyone else (Korsgaard 1996a:122).3 Given the vast riches that emerge from this conative vacuum, I call this part of Korsgaard’s argument her “moment of magic.” And because this commitment supposedly grounds every choice, even immoral ones, it follows that any choice at odds with the value of humanity is at odds with its own motivational underpinning. Accordingly, choice at odds with the value of humanity is both unmotivated and also incoherent, given that is contradicts its real ground, commitment to the value of all humanity. As a result, on this interpretation there is no basis for immoral choices, and, what’s more, any choice that we might commonly refer to as “immoral” is actually not only unmotivated but also incoherent. Having thus equated coherent choice with moral choice, Korsgaard merges the pragmatic and the moral, as throughout pre-Kantian history (Korsgaard 1996a:123).4
The first big problem with this constructivist attempt to shift the burden of normativity onto some implicit commitment is that normativity does not work that way for Kant. Kant would reject this as one version of the Euthyphro problem: The moral law is not the moral law because we are committed to it; instead we are committed to it because it is the moral law. And we must think for ourselves and use our reason to ascertain that it is the moral law, and only then will we respect it and be committed to it. That we would take pleasure in something or even be implicitly committed to it still leaves what both G. E. Moore and Kant would see as the wide open question, of why it is morally good. This is the open question that would still face Korsgaard’s “implicit commitment” at the end of the day, just as it does O’Neill’s plurality of uncoordinated agents. In Kant’s view, morality is too important a thing to delegate to an implicit commitment. Moreover, if it were thus delegated, the result would not be an ethics of autonomy, as we would here not be submitting to the authority of our own reason, but instead to something outside it.
Moreover, Kant simply rejects the view that the moral choice is the only coherent choice. Unfortunately, given that we are embodied beings with a faculty of sensibility, we do have irreducibly sensible inclinations, and in Kant’s view these inclinations are not grounded in an implicit commitment to all of humanity but instead in our sensible nature, and they can provide their own all-too-coherent basis for choice that is nonetheless all-too-human, and immoral. As Kant asserts against his predecessors and contemporaries alike, it is this recognition that allows us to not only clarify the true meaning of the moral law, but also properly gauge the nature of the challenge of moral living and respond in a clear-eyed manner. It is Kant’s recognition of the fact that inclinations are not ultimately grounded in a commitment to moral law, that choices can indeed be coherent even if immoral, and that moral living and happiness can sadly go their own separate, coherent ways, that allows Kant to recognize that one of the most important tasks of moral living is to do our part to cultivate our inclinations and develop our character so that, at least to the extent possible, we can bring happiness and morality into harmony in this life.