In part one of this two-part series I introduced Kant's practical philosophy in relation to practical philosophy in general, and I argued that Kant provides a unique solution to the problem of normativity grounded in the first-personal perspective. In this post I'll try to show how Kant's first-order moral philosophy arises out of this fundamental practical perspective.
For Kant one of the most basic objects studied by moral philosophy is something called a maxim, which is an intention adopted by the will, or roughly speaking a plan of action. There is some controversy about the interpretation of maxims, but for our purposes it will do to think of them as having the form “I will do action A in circumstances C in order to bring about end E.” All actions involve commitment to some maxim or other, though the maxim of an action need not be explicitly and consciously adopted at any time, let alone prior to acting. Maxims form a hierarchical structure, with more general or fundamental maxims rationally constraining more particular ones. For instance, if I adopt the maxim “I will exercise on the weekend to improve my health,” I can't rationally adopt the more particular maxim, “I will watch TV all weekend to catch up with Game of Thrones,” but I can adopt the maxim “I will go for a hike on Saturday to stay in shape.”
Since all actions involve commitment some maxim or other, the will can't be determined to action except by adopting a maxim. In particular, the will can't be determined just by the force of a desire, or what Kant calls an inclination. This is a key difference between Kant's philosophy of action and that of his empiricist predecessors and many present-day Anglophone philosophers. Contemporary philosophers often conceive of action as the result of a belief plus a desire. On this model, an agent combines a belief that the world is a certain way—for instance, “there are no cookies in the cupboard”—with a desire—such as “I want cookies”—and this combination generates an action, such as going to buy cookies. Of course, in most cases people have various conflicting desires. On the belief-desire model, the action that the agent performs depends simply on whichever desire or set of desires is strongest. If the agent in our example desires to eat cookies but desires even more strongly to lose weight, they will not go out to buy cookies.
For Kant, however, action can never be merely the result of forces acting on an agent. Desires may well produce all sorts of effects in a person without any conscious input from that person, but an action is necessarily something that a person can be responsible for, which means that it can only come about through a manifestation of that person's spontaneity. Specifically, desires or inclinations can only contribute to an action to the extent that the agent incorporates them into their maxim.
For example, the person who desires cookies but also desires to get in shape may reflect on each desire and determine that they ought, all things considered, to abstain from eating cookies in order to get in shape. In that case, they incorporate the inclination to get in shape into their will, forming a maxim not to buy cookies. Their decision to incorporate this inclination into their will has nothing to do with the relative strengths of the two inclinations. The inclination to eat cookies may well be stronger in a psychological sense than the inclination to get in shape, but on Kant's model this does not preclude the person's incorporating the latter inclination into their will. The formation of a maxim is an essentially spontaneous or free act that cannot be determined by psychological forces.
This spontaneity of the will is Kant's basic solution to the problem of normativity. It is the core of what I referred to in the previous post as his first-personal voluntarism. As I argued there, from the first-personal practical perspective determinism and naturalism do not have to be theoretically disproven in order to vindicate whatever norms govern the first-personal perspective. (The scope of our knowledge about determinism does have to be shown to be limited to the world of experience, as opposed to the world of things in themselves, but that topic is beyond the scope of this series. If you're interested in this issue, check out my introduction to Kant's theoretical philosophy here.) What emerges from Kant's first-personal reflection on the nature of action is that we are all inescapably committed to the practical reality of normativity, and we reaffirm that commitment every time we act. Because adopting a maxim is essentially an act of spontaneity, it is always something for which the agent who performs it is responsible. It is not something that merely happens to the agent, but rather something that the agent does, thereby affirming their commitment to whatever norms govern the practice of adopting maxims.
Another way of putting this is to say that an agent can only act “the idea of freedom,” as Kant puts it in section 3 of the Groundwork (448). What this means is that in deliberating about what to do I cannot assume the truth of determinism. If I do, I will never act at all, but will simply wait indefinitely to be determined to act. By acting I instead affirm my commitment to my practical freedom, regardless of the theoretical status of determinism. I commit myself to the proposition that I could, in some sense, have acted otherwise, and therefore I am responsible for justifying my decision to act in the way that I did.
The kind of freedom immediately at issue is what I previously have followed Kant in calling “spontaneity.” Spontaneity is a limited kind of freedom, in that it only involves the capacity to choose between various inclinations. It is more than what Kant characterizes as the “freedom of a turnspit” in his later Critique of Practical Reason (97), meaning it is more than just the “freedom” to be determined in more than one way, as a turnspit can be opened or closed. Rather, it involves a rational choice between multiple inclinations, such as the inclination to eat cookies and the inclination to get in shape. However, it is constrained by the realm of inclination. It does not involve the ability to determine oneself to do something for which one has no inclination at all.
This may sound like hardly any limitation at all, since it is not clear what it would mean for agents to determine themselves independently of inclination entirely. However, Kant claims that this is precisely what they can do. In addition to freedom in the sense of spontaneity, agents also possess freedom in the sense of autonomy, or the capacity to determine their will independently of all inclination.
The argument for this conclusion is similar to the argument for the conclusion that agents can only act “under the idea of freedom.” In deliberating about whether I should act in accordance with one of my inclinations, or in accordance with some requirement of morality which I have no inclination to act on, I cannot assume that it is impossible for me to act on the latter simply because I have no inclination to do so. This is not because it would be wrong to assume so, but simply because from a first-personal practical perspective I encounter no limitation that prevents my will from determining itself independently of inclination.
Kant discusses this issue in the Critique of Practical Reason. He gives the example of a man threatened with execution if he refuses to give false testimony against a man the prince would like to destroy (30). There is no reason to suppose that he has any inclination to give up his life for the sake of the other man, and yet when he reflects from his first-personal perspective on what he ought to do, he can find no grounds for denying that this it is possible for him to do so. He may well intend not to, but there is nothing in the first-personal practical perspective to justify the conclusion that it is impossible, even though it is contrary to all his inclinations. A philosopher committed to the belief-desire model of action would have to reject this possibility and conclude that the man must either secretly desire to give up his life for the sake of duty, or that it is in fact impossible for him to do so. This analysis is forced, and again there are no grounds for it in the first-personal practical perspective. Therefore, the will really is autonomous from a practical perspective. Kant calls this conclusion the “fact of reason.”
So what sort of fundamental maxim should I, a person with an autonomous will, act on? Given that every other person's will is equally autonomous, whatever maxim it is, it must be capable of serving as the fundamental maxim for every person's will. At the same time, of course, it must remain a maxim that I myself can will. Once the problem is posed like this, the solution becomes surprisingly obvious: the fundamental maxim must be, “I will act only on those maxims that I can at the same time will as fundamental maxims for all other persons.” Technically, this is not a maxim in Kant's terminology but rather a practical or universal law, which is like a maxim except that it is capable of determining all persons' wills independently of inclination. This is also what Kant calls a categorical imperative, and the categorical imperative we've just arrived at is what he calls the moral law. Restated in this terminology, we have Kant's famous formula of universal law, the first formula of the categorical imperative: “act only in accordance with that maxim through which you can at the same time will that it become a universal law” (Groundwork 421).
What does it mean in practice to act only on those maxims that one can at the same time will as universal laws? The formula of universal law establishes two tests that a maxim must pass in order to be morally permissible. First, it must pass the contradiction in conception test, which means it must be conceivable as a universal maxim. In order to perform this test, I ask whether I can even conceive of a world in which everyone (including me) acts on the maxim in question.
A classic example of a maxim that fails the contradiction in conception test is a maxim of false promising. If I want to set myself the maxim of making a false promise in order to get some benefit, I must ask myself whether I can even conceive of a world in which everyone (including me) makes false promises in order to get that benefit. As it turns out I cannot, because if everyone made false promises to get the benefit in question, people would no longer take promises seriously (or at least promises concerning the benefit in question), and the institution of promising would effectively cease to exist. Therefore, my maxim could not even conceivably be acted on in a world in which it had become a universal law.
Because the maxim fails the contradiction in conception test, there is a perfect duty not to act on it, which means that I must never act on it under any circumstance. This can also be called a negative duty, because it forbids certain maxims without prescribing any particular other ones. (It emerges in Kant's 1797 book The Metaphysics of Morals that the perfect/imperfect distinction between duties may not be precisely equivalent to the negative/positive distinction, but we don't need to go into that here.)
Even if a maxim passes the contradiction in conception test, it may still fail the second test of the formula of universal law, the test of contradiction in volition. This test asks whether the maxim could not only be conceived as a universal law, but whether I could also will that it serve as a universal law. A classic example of a maxim that passes the contradiction in conception test but fails the contradiction in volition test is a maxim of universal nonbeneficence, or a maxim of never helping others. I can conceive of a world in which nobody ever helps anyone else, but I cannot will that such a world come to be, because like everyone else I too will need help at some point in my life. Generally, one cannot will that a maxim become universal law if doing so would require willing something contrary to a practical interest that belongs to all human beings as such, such as receiving help from others. Another maxim that fails this test is a maxim of never improving one's own skills, or the capacity of one's will to set and pursue ends (including ends required by morality). Some skill is by definition necessary to achieve any end at all, so a maxim of entirely neglecting one's skills is contrary to a necessary interest of the will, and therefore fails the test of contradiction in volition.
If a maxim fails the contradiction in volition test, there is an imperfect duty not to act on it. This means that one must not adopt the maxim as it stands, but one may adopt a maxim which is something less than the negation of the original maxim. For instance, although I must not refuse to help others all of the time, I needn't agree to help them all of the time either. Instead, I can adopt a maxim of sometimes helping others. Imperfect duties can also be called positive duties, because unlike negative duties which merely forbid maxims, positive duties also require the adoption of certain other maxims.
As we've just noted, however, the maxims required by imperfect duties do not require that we take any particular actions in order to remain committed to them. For instance, I can remain committed to a maxim of sometimes helping others even if I decline to help them on an indeterminate number of possible occasions. Of course, if I pass up every opportunity to help, it becomes less and less plausible to claim that I am still committed to a maxim of sometimes helping. The question of which kinds of practices satisfy and do not satisfy imperfect duties, like the question of which maxims a person is committed to in general, is plausibly a matter of interpretation.
That, then, is the formula of universal law. As we've seen, it specifies two tests for the moral permissibility of maxims, resulting in perfect and imperfect duties not to act on maxims that fail those tests. Now, you may have noticed that the formula of universal law itself doesn't look exactly like a maxim. It specifies a general act to be performed in all circumstances—only to act on those maxims that can at the same time be willed as universal laws—but it doesn't specify an end for that act. Although, as I noted before, the categorical imperative is not technically a maxim, it does still have the form of one, which means that it too has an end.
What, then, can that end be? As with the “act” portion of the categorical imperative specified in the formula of universal law, whatever the end is, it must be one that the will can legislate to itself autonomously, that is, independently of inclination. As with the formula of universal law, once the question is posed in these terms the answer becomes surprisingly obvious: the end of the categorical imperative must be the will itself. There is no other end that the will can set itself that is independent of inclination.
What does it mean for the will to set itself as its own end? Clearly it can't be exactly the same thing as setting other sorts of ends. In most cases we set ourselves as ends things that we don't already have, and by setting them as ends we commit to bringing them about. This can't be what it means to set the will itself as an end, because the will of course already exists. Nor can it mean committing to bring about more wills, i.e., more persons, as a utilitarian would interpret the imperative.
What it does mean becomes clear if we reflect again on the significance of the autonomy of the will. As we saw before, the actions of the will are essentially spontaneous, meaning that inclinations cannot determine the will except by being incorporated into its maxims. This also means that they cannot be normative or reason giving independently of the will. Only by being incorporated into an agent's maxim does an end become normative for the agent who incorporates it. This does not, of course, mean that there are no ends that are universally normative—as we've just seen, the will itself is just such an end. But it does mean that the will only becomes normative through its own autonomous action. And when the will sets itself any other end, the normativity of that end is conditional on its being set by the will. Another way of putting this point is to say that all ends other than the will itself are conditional ends. That is, their normativity is conditional on their being set by the will. The will itself, by contrast, is an unconditional end, because it is the source of its own normativity. Because of its autonomous nature, it is rationally committed to setting itself as its own end and treating itself as the source of the normativity of all its other ends.
The unconditional/conditional distinction is one of two basic distinctions that can be drawn among ends. As we've seen, it is just the distinction between an end whose normativity has no condition or is its own condition (the will), and an end whose normativity is conditional on the existence or activity of something else, as all ends other than the will are with respect to the activity of the will. The other basic distinction is between ends that are intrinsically normative and ends that are merely instrumentally normative. An intrinsically normative end is normative for its own sake, not merely as a means to some other end. For instance, I may take my friendships to be normative for their own sake, so that I would continue to find them worthwhile even if they didn't produce any side benefits for me. An instrumentally normative end, by contrast, is normative only as a means to some other end. For example, I may eat quinoa because it's good for me, even though I don't really like the taste. If it weren't good for me I would no longer eat it, because it's not intrinsically normative for me.
As you may have noticed, the unconditional/conditional distinction and the intrinsically normative/merely instrumentally normative distinction are somewhat independent of one another. Both friendship and quinoa are merely conditional ends, of course, because the will is the only end that is unconditionally normative. In other words, conditional ends can be either intrinsically or merely instrumentally normative. Indeed, the same end can be both intrinsically and instrumentally normative, though of course it can't be both intrinsically and merely instrumentally normative.
The unconditional end that is the will, however, cannot be merely instrumentally normative. As the source of all normativity, if it were not normative in itself then nothing else would be normative either and so there would be nothing for it to serve as a means towards. Like other intrinsically normative ends, it may of course also be instrumentally normative—the will helps us achieve all sorts of other ends—but it cannot be merely instrumentally normative. Therefore, the will is an unconditional and intrinsically normative end.
Kant's technical term for the will is “humanity.” This term has no necessary relation to the biological species homo sapiens. Rather, it applies in principle to the end-setting capacity of any rational being which has that capacity. As a matter of fact Kant does not believe that animals are rational in the requisite sense, though that doesn't mean that he denies that they are rational in any sense at all. If Kant turned out to be wrong about animals, they too could be said to have “humanity.” Using this new term, we can now formulate the conclusion of the previous discussion, which turns out to be the second formula of the categorical imperative, or the formula of humanity: “So act that you use humanity, whether in your own person or in the person of any other, always at the same time as an end, never merely as a means” (Groundwork 429). In other words, treat humanity or the will like the unconditional, intrinsically normative end that it is, not like a merely conditional or instrumentally valuable end.
What this means in practice is a matter of some dispute, but like the formula of universal law it has a negative and a positive component. Indeed, Kant claims that all of the formulas of the categorical imperative are just different aspects of one and the same moral law, so they should produce the same duties. One plausible interpretation of the negative component of the formula of humanity is that it requires us not to treat any person (including ourselves) in a way to which that person could not consent. The modality (“could”) is crucial here. The issue is not whether the person actually does consent, or even whether they would consent if given the opportunity. Rather, it is whether they could consent given their status as autonomous rational beings. Things to which persons as such could not consent include coercion, whether violent or not; manipulation, including being lied to; and having their rational capacity destroyed, whether by injury or death (including, Kant thinks, by suicide). All of these cases involve the circumvention, manipulation, or destruction of a person's rational capacity.
To return to the lying promise example from the discussion of the formula of universal law, if I want to extract some benefit from you, I have two basic options: I can either try to persuade you to grant me the benefit, in which case I recognize that the decision whether or not to grant it ultimately lies with your own will; or I can attempt to circumvent your will, for instance by lying to you. In the latter case you could not consent to my maxim of making a lying promise in order to extract the benefit, because if you were aware that it was my maxim it would no longer be possible to act on it, for reasons similar to those discussed with regard to the contradiction in conception test of the formula of universal law. If you knew that my maxim was to lie to you, it would be conceptually impossible for me to act on it, since lying requires at least possible ignorance on the part of the person being lied to. As with the formula of universal law, when a maxim violates the negative requirements of the formula of humanity—when it subjects a person to some treatment to which they could not consent—there is a negative, perfect duty not to act on that maxim.
The basic positive duties required by the formula of humanity are the same as those required by the formula of universal law: to improve one's own skills and to adopt some of the ends of others when they are morally permissible (Groundwork 430). These duties can be derived independently of the formula of universal law by reflecting on the nature of ends in general. As we saw in introducing the formula of humanity, an end doesn't have to be something nonexistent that we aim to bring about. However, it does have to be something we can aim at and realize in some sense or other. Therefore, if humanity is to serve as the end of the moral law, it must not be a merely negative end (one that generates negative duties forbidding the adoption of maxims), but also a positive end (one that generates positive duties requiring the adoption of maxims).
In other words, we must aim at the fulfillment of humanity in some sense. There are two basic ways in which we can do this: by working to fulfill the ends set by a person, or by working to improve the skills which a person can use to achieve their own ends. With regard to ourselves, the former cannot be a duty, because the fulfillment of one's own ends—what Kant calls happiness—is by definition what one aims at in setting any ends at all. Since it's not even conceptually possible to fail to aim at our own happiness, it cannot be a duty to do so. We can, however, fail to improve our skills, so we have an imperfect duty to see to it that we do improve them. The converse is true with regard to duties to others. We can help others achieve their ends, so we have an imperfect duty to do so. We cannot, however, directly help them improve their skills, as skills in Kant's sense are a matter of the autonomous activity of the will, which no force outside of the will (including another person's will) can effectively determine. Therefore, the duty to fulfill the end of humanity entails the imperfect duties of improving one's own skills and helping others achieve their morally permissible ends.
There is one more feature of the categorical imperative related to but distinct from its end. You'll recall that in introducing maxims I mentioned that they are organized hierarchically, with the more general or fundamental ones rationally constraining the more applied, particular ones. As we've seen, the categorical imperative ought to be one's most fundamental maxim, or, properly speaking, it ought to constrain all of one's maxims as a practical law.
This is more or less what is meant by Kant's famous (to some, infamous) claim in the opening sentence of the Groundwork that nothing “could be considered good without limitation except a good will” (393), or as he later puts it, that “an action from duty has its moral worth not in the purpose to be attained by it but in the maxim in accordance with which it is decided upon” (399). To have a good will is (or at least centrally involves) making the categorical imperative or the moral law one's fundamental maxim (properly speaking, practical law) and setting one's other maxims in accordance with it. An action that exemplifies a good will is morally worthy, and its moral worth derives not from the end of the maxim but rather from the maxim itself's being chosen in a way that is compatible with the moral law being one's fundamental practical principle.
These claims are relatively straightforward, but they have probably resulted in more misunderstanding and misrepresentation than any other single feature of Kant's philosophy. As with most such cases, one could wish that Kant had expressed himself more clearly, but the bulk of the blame must lie with interpreters who have paid insufficient attention to what he actually wrote. The main problems with the understanding of the notion of moral worth have concerned its relation to inclination, which as we saw before is roughly equivalent to desire. Many Anglophone interpreters, especially before the renaissance in English-language Kant studies in the 1960s and '70s, read Kant as claiming that moral worth is opposed to inclination, so that an action could only be morally worthy if it was done without desire or even in a manner contrary to all desire. These interpreters paint a picture of Kant as a stuffy old Prussian moralist, admonishing us to do our duty by submitting to an arbitrary and legalistic moral system, according to which our actions only have moral worth to the extent that performing them makes us miserable.
As with most misinterpretations, there are some quotations that, taken out of context, can seem to support this reading. These include Kant's claims that an action that is motivated solely by inclination is merely “in conformity with duty” as opposed to “from duty” and therefore has “no inner worth” (Groundwork 398-9), and that if I am to perform an action “from duty” I must “put aside entirely the influence of inclination” and be motivated by the moral law “even if it infringes upon all my inclinations” (400-1).
These quotes do make it sound as though Kant sees action from duty, and therefore moral worth, as entirely incompatible with inclination. However, his talk of the “influence of inclination” can be misleading. All of these quotations are from section 1 of the Groundwork, where Kant is seeking to elucidate “common moral cognition,” with the aim of discovering the content of the categorical imperative and determining how it could determine the will. As we've seen, the autonomous nature of the will means that the categorical imperative cannot itself be grounded in inclination. However, that does not mean that action from duty or from the categorical imperative must exclude maxims of inclination entirely. It only means that such maxims cannot serve as the categorical imperative, or the fundamental principle of the will.
The real contrast Kant is drawing with the distinction between actions from duty and actions merely in accordance with duty is between the actions of a will whose fundamental principle is the categorical imperative, and the actions of a will whose fundamental principle is a maxim of inclination (what Kant calls a maxim of self-love), but which nevertheless sometimes sets itself derivative maxims that happen to agree with the maxims that would be set by a good will in the circumstances. These maxims are morally right—which just means that they are in accordance with the requirements of morality—but they do not have moral worth, because they are derived from a fundamental principle of self-love rather than morality. They are merely in conformity with duty rather than from duty.
Such actions are possible precisely because the requirements of duty do in fact sometimes coincide with the inclinations of self-love, contrary to what the caricatures of Kant would lead one to believe. For example, Kant claims that “there are many souls so sympathetically attuned that, without any other motive of vanity or self-interest they find an inner satisfaction in spreading joy around them and can take delight in the satisfaction of others so far as it is their own work” (Groundwork 398). The fundamental maxim of such a person might be, “I will help others whenever possible, because it pleases me.” Kant describes the actions of such people as “amiable” and “honorable,” and even deserving of “praise and encouragement,” but not of “esteem” (398). This is because the fundamental principle of these actions is that of self-love rather than that of morality. The actions of “sympathetic souls” may be in conformity with what the moral law requires, but that is only because such souls have been lucky enough never to find themselves in a situation in which the requirements of the moral law are in tension with those of self-love.
If such souls were to fall on hard times, Kant suggests, their inclinations might shift so as to no longer be in agreement with the requirements of morality. In that case, if they continued acting on the fundamental principle of self-love, their actions would no longer even be in conformity with duty. If, however, one came across a person afflicted by suffering and without the benefit of a natural love for humanity, who nevertheless helped others just as reliably as the sympathetic soul during good times, then one could be sure that one was dealing with a person acting from duty rather than merely in conformity with duty. Because such a person would by definition have no inclination to act as morality requires, one could be sure that any of their actions in comformity with duty would also be from duty, that is, motivated by a fundamental commitment to the moral law.
This by no means entails that it is impossible for the actions of a sympathetic soul to also have moral worth. They perfectly well can have moral worth, as long as the sympathetic soul makes the moral law rather than the principle of self-love their fundamental principle. Again, Kant's purpose in section 1 of the Groundwork is to elucidate common moral cognition, so he has chosen examples that he believes make particularly stark the contrast between actions from duty and actions merely in accordance with duty. He does not mean to claim that inclination is excluded from morally worthy action altogether, only that it must be subordinated to the moral law. Otherwise the will would not really be acting autonomously at all, but would at best give the appearance of so acting. The requirement to act from duty as opposed to merely in conformity with duty, then, derives from the same ground as the categorical imperative itself: the autonomy of the will.
There is far more to say about Kant's moral philosophy, but if you've followed along until now, you'll have a solid understanding of his fundamental ideas. There are two (some people count three) other formulas of the categorical imperative introduced in the Groundwork, which deal with aspects of Kant's larger practical philosophy, which I discussed in the first part of this series. Specifically, they deal with the relation between his practical and his theoretical philosophy, and with what might be called his social philosophy, which is discussed in his 1793 book Religion within the Boundaries of Mere Reason. Like the formula of universal law and the formula of humanity, these formulas are meant to reveal different formal aspects of the moral law while generating the same content or duties as the other formulas. There are also many other features of Kant's moral psychology and first-order moral philosophy discussed in the Religion book as well as in the Metaphysics of Morals which the Groundwork was meant to prepare for. These issues are beyond the scope of this series, but feel free to let me know in the comments if there's any particular issue you'd like me to write more about.