Kant developed three principal formulas of the categorical imperative. The first and most well known of these is the so-called “formula of universal law.” The second is the “formula of humanity.” This one has been almost as interesting to commentators and critics as the first formula. To some, in fact, the formula of humanity is an even more interesting expression of Kant’s basic insights in ethics.

Means, ends and transitivity. A “means” is something whose value depends upon its use. The value of its use depends in turn upon the value of what can bring about, and so upon the value of an “end.” But an end can often have a use also, serving as a means to some further end. A can be a means to B, which can be a means to C, which can be a means to D, and so on. An “end in itself,” sometimes called a “final end,” is the termination of such a chain. It is not a means to anything further; there is nothing it can be used for. It is what everything else in the series is ultimately used for. Happiness is usually thought of as an end in itself. Just about everything useful is useful ultimately for happiness; but happiness, despite its value, is useless. The value of use is transitive. In the chain of uses above, B’s value depends on C’s, which depends in turn on D’s. So it would be foolish mistake to use D as a means of bringing about B, or to trade D for B. That would amount to valuing the means more highly than its end. It would be a colossal mistake, therefore, to value any means more highly than an end in itself. Imagine that the devil offered you any new car you may desire in exchange for your losing all prospect of future happiness. You would be foolish to accept his offer.

The value of humanity. What Kant wants to convey with the formula of humanity, likewise, is that it is at least as foolish to treat humanity as a means for attaining any other end. When he refers to humanity as an end in itself, he means that every other end, like A, B, and C above, is valuable because of its relation to humanity. This includes happiness. Consequently, it is foolish, or unreasonable, to treat humanity as a mere means to anything else, including happiness. But more than that, it is morally wrong. If the validity of the categorical imperative depends on humanity’s existing as an end in itself, as it does according to Kant’s moral realism, then the categorical imperative should forbid any action that treats humanity merely as a means. By this line of reasoning Kant sees that an alternative version of the categorical imperative can now be expressed as: “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 ” (Kant 1997, 38/4:429).

The four examples, again. Following that statement of the new version of the categorical imperative, known as the “formula of humanity” (or of “humanity as an end”), Kant returned to the four examples of perfect and imperfect duties to oneself and others, showing how these same duties can be derived from the idea of treating humanity as an end. (1) Here the suicide example works as follows: the man who would commit suicide in order to avoid continued misery would treating his humanity (“humanity in his own person”) merely as a means. He would be, essentially, trading his humanity for happiness. (2) The example of the lying promise shows that to swindle someone is to treat his humanity merely as a means, since “he whom I want to use for my purposes by such a promise cannot possibly agree to my way of behaving toward him” (Kant 1997, 38/4:429-30). Scholars have generally supposed that by this Kant meant something like the following: the idea of humanity, sometimes referred to also as “rational nature,” includes the cognitive powers that aim essentially at an accurate representation of reality. The power of belief is a good example. Since here the promisor cannot succeed without giving the promisee a false belief, he is using the promisee’s power of belief merely as a means. We treat the humanity of another person as an end, therefore, when in our dealings with them we avoid intentionally inducing false beliefs. For the imperfect duties: (3) Kant’s example of neglecting natural talents presents a man whose action would not, he writes, “harmonize” with the idea of humanity as an end in itself (Kant 1997, 39/4:430). The idea seems to be that although humanity already exists as an end in itself, there are still not-yet-existing ends for human beings to bring about: one of them, or one set of them, relates to our latent natural talents. To choose not to develop these talents, for the sake of personal happiness, is a way of using one’s own humanity merely as a means. (4) Another end of every human being is his or her own happiness. To harmonize one’s conduct with this end of humanity is to contribute toward realizing the happiness of others: “the ends of a subject who is an end in itself must as far as possible be also my ends” (Kant 1997, 39/4:430). This is not an injunction to work full time for others’ happiness, however. It would be wrong to treat humanity in oneself merely as a means to the ends of others. But how much this duty may require of us remains here an open question. Kant will have something more to say about it a dozen years later, in The Metaphysics of Morals. His view expressed there will be that the moral law cannot say how far we must go, or what personal sacrifices we ought to may, for the happiness of others. This is a consideration of virtue (Kant 1991, 94-5/6:390-1).

Derivation of the formula of humanity. The argument by which Kant derived the formula of humanity has attracted scholars’ attention recently. Some claim that he supports that formula with reasoning along the following lines: in order to value anything at all, it is necessary to regard oneself, or humanity in oneself, as the source of all value, and so as absolutely good. That is why the value of humanity must always be respected—because it is the basis for the value of anything else (see Hill 1991, 103; Korsgaard 1996, 119-24; Wood 1999, 124-7). This interpretation of Kant’s argument has not held up well to criticism, however. One clever objection is that, by following that line of argument, in order to regard anything at all as evil it must be necessary to regard oneself as evil (Kerstein 2002, 59). Another complaint is that even if it may be necessary for us to value our own humanity in order to value anything else, it does not follow that we ought to value humanity in any other person (Darwall 2009, 149-50). Kant’s argument in deriving the formula of humanity seems to be slightly different from the one represented and criticized. He writes something to this effect: (1) Subjectively, every human being regards him- or herself as an end in itself; (2) Every other rational being does so as well, and for the same reason; (3) Therefore, objectively, rational beings exist as ends in themselves, and are not to be used merely as means to any other ends. (Kant 1997, 37/4:429). This argument, admittedly, is not very illuminating. It is not especially clear how the step from (2) to (3) is supposed to be understood: from every rational being’s subjectively thinking something about him- or herself, to this thought’s being objectively true about them all. Although Kant does claim, in a footnote, that (2) is only a “postulate” that he will prove later in the Groundwork, he does not seem to have made good on that promise.

The formulas of humanity and of universal law. Although Kant claimed that all versions of the categorical imperative he derived are equivalent (Kant 1997, 43/4:436), some scholars doubt that they are. Korsgaard finds several differences between the formula of universal law and the formula of humanity, including that, as she claims, the argument of the suicide example fails under the formula of universal law, but succeeds under the formula of humanity (Korsgaard 1996, 151-2; 143-4). But she may not be correct in this, owing to the way she interprets the “contradiction” expected for maxims failing the universalization test. Wood, on the other hand, contends that the two formulas may often appear to give different results because the formula of universal law is only a rough preliminary to the formula of humanity. He claims that “Kant’s formula of choice for applying the moral law is not the formula of universal law but the formula of humanity” (Wood 1999, 110). But this claim is hard to believe, considering what Kant wrote in the Groundwork: “one does better always to proceed in moral appraisal by the strict method and put at its basis the universal [law] formula of the categorical imperative” (Kant 1997, 44/4:436).