20th WCP: Why Couldn't Kant Be A Utilitarian?

In his essay "Could Kant Have Been A Utilitarian?", (1) R.M. Hare, analyzing Kant's text, tries to show that Kant's moral theory contains utilitarian elements and it can be properly asked whether Kant could have been a utilitarian though he was in fact not. I take his challenge to the standard view seriously not because it is made by the celebrated moral philosopher but because I find Hare's reading of Kant's text on the whole reasonable enough to lead to a consistent interpretation of Kant's moral philosophy. I agree with Hare on many points which I believe are crucial for discerning a consistent line of thought in Kant's theory of the Categorical Imperative. Nevertheless I hardly believe that it is necessarily concluded from Hare's reading that Kant could have been a utilitarian. This paper aims to show why Kant couldn't be a utilitarian despite the apparently utilitarian elements in his theory. I will seek the answer in Kant's theory itself, not in the biographical matters like Kant's Pietist background, which might be of interest in some other context.

Hare begins his interpretation with one of the formulae of the Categorical Imperative known as the Formula of the End in Itself: "Act in such a way that you always treat humanity, whether in your person or in the person of any other, never simply as a means, but always at the same time as an end." He interprets this as prescribing that one should treat other persons' ends as one's own ends, which is the same as the utilitarian prescription that one should do what will conduce to satisfying people's rational preferences or wills-for-ends. Hare notes that this formula conforms to Bentham's injunction "Everybody to count for one, nobody for more than one." In my view, this is the most reasonable interpretation of Kant's illustrations of the duty to others in Grundlegung.

I put the same point as follows. The status of a person as "an end in itself" consists in being capable of setting and pursuing his/her own ends. (2) This capacity is essential to the concept of the rational agency, as Kant writes: "Rational nature separates itself out from all other things by the fact that it sets itself an end" (GMS 437, cf. MS392). Accordingly, to treat a person as "an end in itself" is to respect a person as a subject who sets and pursues his/her own ends. We should not ignore, much less hinder, others' pursuit of their own ends at our will without their consent.

One of the possible objections to this reading is this: to treat a person as "an end in itself" is to respect him as a being capable of moral action, that is, as a subject of moral autonomy. (3) This sense of "the end in itself" is, to be sure, explicit in Kant's text. But we cannot confine the concept of a person as "an end in itself" to this sense alone. For one thing, such confinement renders the Formula of the End in Itself circular: morality consists only in respecting a person as a moral agent. (4) For another, it makes the illustrations Kant provides for the formula pointless.

Then Hare turns to the Formula of the Universal Law: "Act only on that maxim through which you can at the same time will that it should become a universal law." Hare interprets this as follows. To will the maxim of one's own action to be a universal law means to will it for all situations resembling the present situation, including those in which one occupies all the other possible roles. But in order to will this, one must be willing to undergo not only good but also bad things in all those roles. The result is that one should act on such maxims as do the best, all in all, impartially, for all those affected by my action. This again is a utilitarian principle.

This interpretation is reasonable too. It provides a good interpretation for the concept of "contradiction in the will." Kant illustrates this concept as follows: If I will, for example, the maxim to be a universal law that I neither interfere with the happiness of others nor help others in distress, then it amounts to this: I will a world in which nobody in distress is helped. Such a will contradicts my will to be helped when I am in distress.

Some find this illustration unti-Kantian because it obviously resorts to self-love. But the point here is not that I should help other persons from self-love, that is, in order to be helped when I am in distress, but that a maxim which implicitly presupposes that only I am a rational being that sets and pursues ends, ignoring the fact that all other persons are also such agents, causes contradiction in the will, when universalized. (5) In another word, this thought experiment reveals my attitude to ignore the fact that all others are rational agents who need help when things make it difficult for them to pursue their ends just as I am a rational agent who need help in the like situations. And this interpretation has the advantage of showing clearly the link between the two formulae as Hare summarizes: "If I am to universalize my maxim, it must be consistent with seeking the ends of all the other people on equal terms with my own." This explains Kant's statement well that the principle of universality and that of "end in itself" are two different formulations of one and the same principle. Many critics, failing to see this point, have assumed Kant's moral theory to be an empty formalism.

Hare admits that his interpretation does not seem to explain what Kant calls duties to oneself, such as the prohibition of suicide and indolence. And he points out that Kant's own arguments for such duties depend on some different senses of "treat as an end", behind which Hare smells out heteronomous principles such as God's will or Nature's purpose. But he shows that the prohibition of indolence can be dealt with in the utilitarian interpretation of Kant's theory, with the prohibition of suicide left inexplicable.

Hare's interpretation, as he rightly admits, fits merely what Kant calls duties to others. This is not however, a sign of defect on the part of Hare's reading, for Kant himself never succeeded in explaining duties to oneself on the ground of the Categorical Imperative. Kant's arguments for duties to oneself are confused in both Grundlegung and Metaphysik der Sitten. In Grundlegung, Kant resorts to some other principles such as the purpose of nature or the nature of rational beings. And in Metaphysik der Sitten, Kant introduces a new concept, the concept of "the end that is also a duty," to explain duties to oneself and imperfect duties. I think this introduction of the new concepts suggests that Kant himself is aware that the explanation in Grundlegung is not satisfactory. However, Kant's new argument for such duties in Metaphysik der Sitten is no more convincing than that in Grundlegung, and even more clearly shows Kant's confusion, as follows:

For suppose there were not such duties: Then there would be no duties whatsoever, and so no external duties either. For I can recognize that I am under obligation to others only insofar as I at the same time put myself under obligation, since the law by virtue of which I regard myself under obligation proceeds in every case from my own practical reason; and in being constrained by my own reason, I am also the one constraining myself. (MS 417-18).

This passage properly represents Kant's consistent claim that one constrains oneself in moral action, that morality consists in acting in conformity with duty (pflichtmaessig) from duty (aus Pflicht). But this can never be a ground for the existence of duties to oneself, for acting in conformity with duty from duty is an element common to all duties, not only duties to oneself, as Kant writes elsewhere.

"...to every ethical obligation there corresponds the concept of virtue, but not all ethical duties are thereby duties of virtue. Those duties that have to do with what is formal in the moral determination of the will (e.g. that an action in conformity with duty must also be done from duty) are not duties of virtue. Only an end that is also a duty can be called a duty of virtue. For this reason there are several duties of virtue (and also various virtues), whereas for the first kind of duty only one (virtuous disposition) is thought, which, however, holds for all actions." (MS383)

If the self-constraint were a ground for duties to oneself, then all duties performed without external constraint would be duties to oneself, even the duty to help others in distress, for example, when performed "from duty". This is inconsistent with Kant's original claim.

I propose the following solution: it is duties to others that the Categorical Imperative grounds, and duties to others consist of those concerning others and those concerning oneself. Suicide and indolence are against duties concerning oneself, not to oneself. And the ground for such duties lies not in a person's relation to the self but to others: suicide and indolence most of the time result from or result in negligence of responsibility to others. (6)

It is certain that Hare stands quite close to this interpretation when he turns to Metaphysik der Sitten to resolve the tension between utilitarian and non-utilitarian elements in Kant's thought. Kant explains that "the end that is also a duty" is "one's own perfection" and "happiness of others." Hare asserts that the perfection that Kant is after is moral perfection, which consists in the acquisition of virtue. But virtue, in turn, consists in nothing else but the disposition to fulfill the duties to others, otherwise the concept of moral perfection would go circular. Moral perfection must be of form, duties to others making its content.

From the above, not only I conclude that Hare's reading is consistent with Kant's argument, but I count it as a promising step to the systematic interpretation of Kant's moral philosophy. For all that, I cannot answer his question in the affirmative whether Kant could have been a utilitarian. For Kant never recognized happiness to be "an end in itself", an end which is intrinsically valuable.

Not that Kant was indifferent to human happiness. Such mis-understanding, as many would admit, belongs to the past. It is true that he rejected happiness as the ultimate moral principle, and yet he maintained that happiness was the only end that we can presuppose "a priori" and "with certainty" to be present in all human beings, and he defined morality as worthiness for happiness, seeing the highest good in the synthetic connection of morality with happiness. No doubt he believed in the essential value of human happiness. Nonetheless, Kant denied that happiness in itself was the ultimate end. Why? Several types of arguments appear in Kant's text to disqualify happiness for a moral principle. Let us take up just two of them.

1. Happiness gives a general rule but never a universal law, even if universal happiness is considered to be the object, because its content is empirical and indeterminate. (KPV 25-26, 36 cf. GMS418)

2. Happiness is a natural, not moral end of human beings, except when happiness of others is sought as "an end that is also a duty".(MS387-88)

Argument 1 may surely be a good characterization of happiness, which is common to one's own happiness and happiness of all, but it is not adequate for disqualifying happiness for a moral principle. For we can still consider happiness of all to be a formal principle as long as happiness can be presupposed "a priori" to be an end of all human beings, and judge whether or not a certain action is morally permissible by examining whether or not its maxim meets the necessary conditions of happiness, for example, safety, freedom, health, etc.

Such judgement, to be sure, requires empirical information, but can still be a moral judgement in Kantian sense. We have to keep in mind that Kant's insistence on moral laws being a priori does not mean that no empirical information is relevant to moral judgement. The principle expressed in the formulae of the Categorical Imperative is supposed to be a priori given, but in judging whether a maxim is in accordance with the principle a priori, it is necessary to take empirical information into consideration. Otherwise how could we tell whether a maxim which is by definition material is in accordance with the formal principle? Happiness may well be a formal principle in this sense. So if Kant's criticism of the principle of happiness were nothing more than the argument 1, then Hare's question could be answered in the affirmative.

It is argument 2 that is crucial. One's own happiness is a natural end which human beings naturally pursue, while happiness of others can be a moral end, an end the pursuit of which is a duty. This means that the principle that makes happiness a moral end is not happiness itself but something else. What is it? It is the very principle of a person "as an end in itself". And to treat a person as an end in itself is, as I have discussed above, to respect a person as a rational being that sets and pursues his/her own ends. This can be understood as promoting his/her happiness, which is not wrong but misleading. It is more accurate to understand it as respecting a person's freedom to pursue his/her ends, that is, freedom to be happy his/her own way.

To set and pursue an end is an essential element of free action, as Kant writes:

An end is an object of free choice [Willkuer], the representation of which determines it to an action... Every action, therefore, has its end; and since no one can have an end without himself making the object of his choice into an end, to have any end of action whatsoever is an act of freedom on the part of the acting subject, not an effect of nature.(MS384-385)

Therefore, respecting a person as a subject of ends includes respecting the person's freedom. Happiness is not valuable in itself; it is valuable because it is set and pursued as an end by the a rational being. In Kant's words, the value of happiness is relative; that of a free rational agent absolute. Thus the most fundamental principle of Kant's moral philosophy may well be to respect the freedom of every person. (7)

Note here that Kant's concept of freedom in practical sense has two aspects: freedom of choice (Willkuer) and freedom of will (Wille) or autonomy of will. The former is freedom to choose whatever end or maxim one would prefer; The latter is freedom to act morally, that is, to act in conformity with duty from duty. These aspects together constitute the freedom of a person as "an end in itself". Any interpretation that ignores either aspect leads to inconsistency. The prevalent interpretation that nothing but moral autonomy of will is the real freedom and as such worthy of respect, for example, is obviously inconsistent with Kant's theory of imputability of actions. If one's freedom of choice were not real, how actions could be imputed to him/her?

I believe we are almost ready to conclude: Kant could not be a utilitarian even if the principle that we should care about other person's ends, the sum of which makes happiness, lies at the heart of his moral theory. For Kant grounds this principle not on the idea that happiness in itself is the ultimate end, but on the principle that every rational being is "an end in itself" and his/her freedom should be respected as such.

I said "almost" because why we can say that every rational being is valuable in itself is a question which cannot be handled within the scope of this paper. If this question proves to be answered on the principle of happiness, we have to suspend our conclusion. But this is a question not only on Kant's part but also on the utilitarians' part: it has been questioned whether Bentham's injunction "Everybody to count for one, nobody for more than one," which is, as Hare points out, the utilitarian counterpart of the Formula of the End in Itself, has any ground in classical utilitarianism. (8) As long as this question is open, we may well ask in turn, for example, "Could J.S. Mill have been a Kantian?"