Kant on Freedom and Evil

An argument still frequently raised against Kant’s theory of freedom goes as follows: Kant seems to have two conceptions of freedom, which may be termed, following Sidgwick, the rational and the neutral conception. According to the rational conception, freedom involves acting on the moral law. It is only when we act on the moral law, the idea goes, that we are actually free of the determination of our will by natural desires. Rational freedom, then, seems necessary in order to account for our ability to act morally. The neutral conception, on the other hand, involves our choosing between good and evil; that is, knowing the law, we choose to act on it or to willingly violate it. Without this conception of freedom it becomes impossible to blame agents for their immoral actions.

The standard criticism (formulated in some way or other by Reinhold, Sidgwick, and Prauss) is that these two conceptions clearly can’t work together: either we are only free when we follow the moral law, in which case we cannot freely choose to violate it and are never responsible for immoral actions, or we are always free, in which case there seems to be no internal connection between freedom and the moral law. Nelson Potter and, more recently, Henry Allison, among others, have pointed out in response that Kant actually defines freedom as the ability to follow the moral law, which means that one can be free without ever actually acting morally, so long as one is capable of acting morally. But the misunderstanding (and criticism) keep coming back, and to address them I want to look at a passage in the Metaphysics of Morals, where Kant deals explicitly with the problem (or at least as explicitly as he ever deals with the major problems) and suggest two important consequences of this passage.

The sentence in question reads:

die Freiheit nimmermehr darin gesetzt werden kann, daß das vernünftige Subjekt auch eine wider seine (gesetzgebende) Vernunft streitende Wahl treffen kann (6:226)

Roughly:

freedom can never be located in this, that the rational subject can also make a contentious choice against his (lawgiving) reason

If we keep in mind the context—Kant is talking about the definition of freedom—this isn’t a particularly strange claim. In fact, I want to insist that it is much closer to plain common sense than to the wacky metaphysics of which Kant is sometimes accused.

Here is the first point: Kant is talking about the rational subject acting against his reason. It makes sense that freedom—as an ability of the subject—cannot be defined in this way. Why? Consider an ordinary object, like a chair. We may roughly define a chair as an object for sitting on. The definition is perhaps incomplete, but it would be a very strange mistake to complete the definition by adding that it is also possible not to sit in a chair. It is of course true that there are chairs in the world in which no one is sitting at the moment, but that is not a good reason to define a chair as an object for sitting on or not sitting on. The first part belongs to the definition of a chair, while the second part belongs to a description of how chairs are sometimes actually used in practice and has nothing to do with the definition. By adding the second clause we do not simply confuse the definition of a chair, but we in fact ensure that we are not giving a definition of it at all. “An object for sitting on or not sitting on,” you see, would be a description that does not apply in any essential way to chairs: it would also cover tables, beds, floors, elephants, other people, etc. Similarly, it would be very odd to define the freedom of a rational subject as the subject’s ability to act in accord with or in opposition to its reason. To be rational means that one’s freedom lies in the exercise of that rationality, not its abandonment.

But now let’s look at the second point: In parenthesis, Kant puts the important word “lawgiving” before the word “reason.” Why is this important? Because it seems to refute the claim by some prominent analytic Kantians (see, for example, Korsgaard’s argument in “Skepticism About Practical Reason”) that immorality is just a species of irrationality. It is not. There is, of course, something irrational in violating one’s reason. But it is not as obvious that there is something similarly irrational in violating one’s lawgiving reason. Kant is, in other words, tipping his hat to Hume’s famous quip:

‘Tis not contrary to reason to prefer the destruction of the whole world to the scratching of my little finger. (Treatise of Human Nature, Bk. 2, 3, iii)

There is nothing contrary to reason in this. There is, however, something contrary to lawgiving reason. But this sort of contradiction is not a logical one: it is a contradiction between the standards set by reason, on one hand, and evil on the other.

Kant never denies that we can freely choose to act contrary to the moral law, and he does not claim that it is irrational to do so. This is implied by his account: if evil were nothing more than a species of irrationality, then we would not be responsible for our evil actions any more than we are responsible for an innocent mistake in multiplication. This raises a new difficulty, of course: how can we understand the freedom of a rational being as allowing for action contrary to lawgiving reason? I will not address the problem here. My goal has, rather, been to deal with two common misreadings by pointing out, first, that Kant’s definition of freedom in terms of the moral law does not, in fact, rule out the possibility of freely acting contrary to the law and, second, that action contrary to the law is nevertheless freely chosen and not simply an error of reasoning. The problems that remain with the conception of freedom, I believe, must thus be addressed through the problem of evil.