The Kantian Perspective

For the sentimentalists, those of us who treat moral demands as authoritative do so only because of contingent facts about human nature – most of us happen to be attuned to the sentiments of others.

Kant found this an intolerably shaky foundation for morality, as well as one that does not sit well with our commonsense conviction that the demands of morality obligate everyone. For him, if the will is determined with respect to a contingent goal of an agent, it always aims at some expected pleasure.³⁸ To be sure, the source of that

38 Kant is thus a psychological hedonist when it comes to motives not derived from pure reason. To this extent he is in accord with Hume, according to whom “’Tis from the prospect of pain or pleasure that aversion or propensity arises towards any object” (Treatise 2.3.3, 266), and a long line of others. In the Metaphysics of Morals he emphasizes that pleasure can be

40

pleasure may be the well-being of others, but insofar as it is just a pleasure among others, what he will do is unbearably contingent from a moral perspective. If some other pleasure appears greater or easier or longer-lasting, a benefactor “can even repulse a poor man whom at other times it is a joy for him to benefit because he now has only enough money in his pocket to pay for his admission to the theater”³⁹. In Kantian terms, the Humean view has room only for hypothetical reasons to be moral. Kant himself famously held that as we ordinarily understand it, the demands of morality are both rational and categorical, imperative for anyone regardless of their desires. To vindicate the authority of morality, we must be able to show that even Hume’s ‘sensible knave’ who, lacking sympathy or imagination, desires to take advantage of his fellow-men when able to do so with impunity, has reason to honour the moral law.

Establishing that the knave has reason to act morally will naturally take a lot of argument. Perhaps the cleverest arguments against the Humean view aim to show that instrumental reasoning is only normative or reason-providing if there are categorical reasons to have the desires that serve as its premises. Otherwise it is not the case that one ought to take the means, or has reason to do so. Thus, John Broome has argued that Humeans commit the fallacy of detachment parallel to a similar mistake in modal reasoning: just as it would be a mistake to infer “Necessarily q” from “Necessarily, if p then q” and p, it is a mistake to infer “I ought to φ” from the instrumental principle “I ought to make it the case that if I desire to ψ, I take the best means φ” and desiring to ψ.⁴⁰ In the modal case, for it to be necessarily the case that q, it must also be necessarily, not

associated with desire in two ways, either as cause or effect (Kant 1797/1996, 373–374). In the latter case, the pleasure cannot determine the desire, which must instead be directed by pure practical reason itself. This sort of pleasure is not pathological but moral: “[P]leasure that must be preceded by the law in order to be felt is in the moral order.” (Kant 1797/1996, 511; emphasis in the original)

39 Kant 1788/1996, 157.

40 What I present here is a simplified and modified version of the argument in Broome 1999.

41

merely contingently, the case that p, even if q necessarily follows from p. Similarly, in the case of oughts, for it to be the case that I ought to φ, it must also be the case that I ought to desire to ψ, even if rationality requires φ-ing if one desires to ψ.⁴¹ If there is a sound argument of this type, reasons cannot bottom out in non-rational desires.⁴²

Even if the Humean view of practical reasoning is inadequate, it does not yet follow that a Kantian view must be correct. The distinctively Kantian approach is to argue that categorical reasons derive from the very structure of practical reasoning itself. Like Humeans, Kantians thus understand reasons for action in terms of what it would be rational for an agent to do. For Kant, practical reasoning is a matter of deciding which maxim of action to adopt; a maxim is a subjective principle of action of the type “In circumstances C, I will perform action A in order to achieve end E”⁴³. A maxim is like a policy or law that one gives to oneself. Many contemporary Kantians argue that having such policies, as opposed to merely doing what one most desires, is constitutive of being a person as such. The argument is that being a person is not being at the mercy of passing desires (like Harry Frankfurt’s ‘wantons’) but having a vantage point that transcends particular moments.⁴⁴ Persons are capable of taking a stand on things on the basis of

41 The fallacy can be formulated in terms of the scope of the relevant operator. In the case of necessity, it is a matter of reading □ (p→q) as p→□q, which are clearly not equivalent. Using ‘O’ for ‘ought to’, the practical reasoning case can be formulated as mistaking O (ψ→φ) for ψ→Oφ. The latter principle would obviously allow for detaching the consequent.

42 Christine Korsgaard presents a different argument in the same vein in Korsgaard 1997. It is criticized in Hubin 2001.

43 It is an indication of the contested nature of Kant interpretation that there is no consensus on such a fundamental issue as what a maxim is for him. The formulation I use captures all the elements in at least one of Kant’s own examples, namely “[E] From self-love [A] I make it my principle to shorten my life [C] when its longer duration threatens more troubles that it promises agreeableness.” (Groundwork II, 74)

44 See Korsgaard 1996 and Velleman 2006 for this line of argument.

42

reflection, and taking a stand consists in adopting policies and principles.

Famously, Kant argues that adopting a maxim is rational only if it could at the same time be willed by all as a universal law. What this amounts to and exactly how it is meant to work is the subject of much controversy. A few things should be relatively clear, however.

First, this formulation of the Categorical Imperative⁴⁵ is meant to be a formal criterion, a filter through which any maxim must pass to be rational. The criterion must be formal, because reasons are essentially the sort of considerations that have authority for any rational agent. Rationality cannot depend on the material of the maxims, contingent desires and goals, since people have, or at any rate could have, all sorts of desires or goals.⁴⁶ By the same token, the maxim must be universalizable – if it is binding on me qua rational agent, it is binding on everyone qua rational agent. As many have recently argued, for a policy to be rational, it must be justifiable to any rational agent, suitably qualified – for Habermas, to any agent willing and able to enter into an uncoerced discourse⁴⁷, for Scanlon, to any agent who is motivated “to find principles for the general regulation of behavior that others, similarly motivated, could not reasonably reject”⁴⁸.

Second, here as elsewhere, the fundamental formal principle of rationality is that of avoiding contradiction. Kant offers two different tests for this. The contradiction in conception test asks whether the

45 I follow the current convention and capitalize the Categorical Imperative when talking about the test that maxims must pass in order to determine which are “categorical imperatives” in the plural, i.e.

prescriptions that obligate unconditionally. I will here discuss only the Formula of Universal Law version and leave aside the issue of whether the other formulations are equivalent, as Kant claims.

46 “And how should laws of the determination of our will be taken as laws of the determination of the will of rational beings as such, and for ours only as rational beings, if they were merely empirical and did not have their origin completely a priori in pure but practical reason?” (Groundwork II, 63)

47 See Habermas 1983, Habermas 1996.

48 Scanlon 1998, 6.

43

maxim is even conceivable as a universal law; if not, it will not be justifiable to any rational agent. Kant’s most famous and persuasive example of this is the case of deceitful promises. Suppose I am considering the possible maxim “I will make a deceitful promise in order to advance my interests”. If it were a universal law, it would be something like “Everyone will make a deceitful promise in order to advance his interests.” But if everyone made deceitful promises when it suited them, the whole institution of promising would collapse: no one would take anyone’s word for anything.⁴⁹ Thus, the maxim cannot be coherently universalized. A maxim whose contrary fails the contradiction in conception test identifies a perfect duty or unconditional obligation. Since failing to keep promises fails the test, keeping promises is such a duty – for Kant, it is never acceptable to break a promise. The other test Kant offers for identifying rational maxims is contradiction in will. According to him, there are maxims that can be conceived as universal laws but cannot be coherently willed as such. One of his examples is failing to develop one’s talents.

The idea seems to be that given some basic facts about human beings, they cannot achieve any goals without either themselves or others having the skills they require. If nobody had any skills, no goals would be reached. Assuming that as agents we are always seeking one goal or another, we cannot coherently will a world in which no goals are reached.⁵⁰ If this sort of admittedly contrived reasoning holds water, we have an imperfect duty to develop our talents. Imperfect duties obligate only sometimes and to some extent – we need not spend all our time developing our talents.

Suppose that we do, in fact, get determinate content for the moral law by way of the categorical imperative test, and so as a result of pure reasoning. (Since I am not here concerned with

49 As Kant puts it, “the universality of a law that everyone, when he believes himself to be in need, could promise whatever he pleases with the intention of not keeping it would make the promise and the end one might have in it itself impossible, since no one would believe what was promised him but would laugh at all such expressions as vain pretenses.” (Groundwork II, 74)

50 Compare Kant 1797/1996, 522–523.

44

arguing for any particular position, I will adopt an agnostic stance on whether or not Kant’s tests yield determinate results.) Why should anyone care about this – that is, why should we adopt only rationally acceptable policies? Well, if, as it seems, the question about why one should care about something is a question about what reasons we have to care about something, it is already answered. In David Velleman’s words, there is “something self-defeating about asking for a reason to act for reasons”⁵¹. Similarly, when we ask if something really has authority, we seem to be asking for reasons to obey its commands, and we already know that reason commands us to obey the commands of reason. So while we can, as a matter of fact, act contrary to morality, as we all too often do, its normative authority over us is inescapable and the demands it makes upon us genuine. When we experience something as having this sort of authority, we have the subjective feeling of respect (Achtung) for it:

“What I cognize immediately as a law for me, I cognize with respect, which signifies merely consciousness of the subordination of my will to a law without the mediation of other influences on my sense.”⁵² Kant hastens to add that respect is not a cause of the moral law (that is, the will is not presented with a hypothetical imperative “If you respect the moral law, do x”) but an effect of the law. Kant does not present this as an empirical hypothesis; rather, consciousness of a law one has given oneself is what respect is.⁵³ Thus, insofar as moral

51 Velleman 2006, 19.

52 Groundwork I, 56n. This kind of respect combines elements of fear and inclination: the law gives rise to something like fear insofar as it may command us to act against our own happiness, and something like inclination, since it is something that we ourselves will as rational agents, not something external.

53 Thus, to say that one acts out of respect for the law, as Kant sometimes does, is just to say one’s action is determined by consciousness of the law, and since the latter results from an exercise of pure reason, reason itself can be practical. I emphasize this, because otherwise there is a temptation to read Kant as going back on his claim that reason itself can be practical when he says things like “neither fear nor inclination, but solely respect for the law, is the incentive which can give an action moral worth” (Groundwork II).

45

demands are reason’s own laws, both their experienced and normative authority is explained.