One of the most intriguing arguments of Groundwork III is about a reason to think we have free will—whether or not we really do. The argument is that if we cannot act without thinking of ourselves as free, then we are free, at least in a “practical respect,” and thus we are subject to all of the laws that would apply to us if we actually were free (Kant 1997, 53/4:448). This includes the moral law.
Practical and transcendental freedom. Earlier, in the first Critique, Kant had drawn a distinction between practical freedom and transcendental freedom. Practical freedom can be attributed to rational beings insofar as they are responsive to prudential rules. “Practical freedom can be proved through experience,” Kant wrote. In clarifying this idea he added:
we have a capacity to overcome impressions on our sensory faculty of desire by representations of that which is useful or injurious even in a more remote way; but these considerations about that which in regard to our whole condition is desirable, i.e., good and useful, depend on reason. Hence this also yields laws that are imperatives, i.e., objective laws of freedom, and that say what ought to happen.... (Kant 1998, A802/B830)
Practical freedom, even though it is freedom, remains under the law of sensible nature, because it is a rational being’s freedom to act on considerations of happiness, or on considerations of “that which in regard to our whole condition is desirable.”
Whereas transcendental freedom requires an independence of this reason itself ( ... ) from all determining causes of the world of the senses, and to this extent seems to be contrary to the law of nature.... (A803/B831).
Transcendental freedom would not therefore be knowable through experience. Its reality would require a philosophical proof, and its imperatives would tell us what ought to be done, independently of considerations of happiness.
Freedom in a practical respect. In Groundwork III, Kant’s conception of freedom in a practical respect seems to be different from the practical freedom identified in the first Critique. There, practical freedom could be known by experience, as freedom to act for the sake of more remote sensible goods. But in the Groundwork, the idea of freedom in a practical respect would have to be attributable to every rational being, including even types of rational beings unknown to us. This freedom cannot therefore be proved by experience: “it can be demonstrated only a priori,” Kant wrote (1997, 53/4:448). The freedom he seems to have in mind in the Groundwork is transcendental freedom. There he proceeds under the assumption that all rational beings are free in this sense in idea. By proceeding with this assumption he is not required to prove that any rational beings are transcendentally free in reality, as would have to be proved in theoretical philosophy. Kant says he will operate on the assumption that all rational beings cannot act except under the idea of freedom, which is to say that they cannot act without thinking of themselves as self-determined.
Reason must regard itself as the author of its principles independently of alien influences; consequently, as practical reason or as the will of a rational being it must be regarded of itself as free, that is, the will of such a being cannot be a will of his own except under the idea of freedom, and such a will must in a practical respect thus be attributed to every rational being. (Kant 1997, 54/4:448)
This will suffice for showing that rational beings, just insofar as they are rational, are subject to imperatives besides the hypothetical imperatives directed toward the more remote sensible goods of happiness. For as shown earlier, in Groundwork II, the law of self-determination (formula of autonomy) is expressed in a categorical imperative. The question one wants to ask now, however, is Why does this not prove, in theoretical philosophy, that every rational being is actually free? The answer is that it does prove that rational beings with a will are free—but it does not prove that anyone is a rational being with a will.
A will of your own. The proof that you are a rational being with a will depends on you. You can observe that you have the practical freedom to direct your conduct toward more remote goods. You know you can delay immediate gratification in order, later, to arrive at a far better life condition. Since you have this freedom, at least, you can attribute this to yourself, rather than to a power that the idea of happiness exercises on you. Theoretically, you cannot know that your prudential behavior is not actually determined by the psychological force of the idea of happiness. But if you go on to assume that it is you who directs yourself toward a happier condition later by sacrificing now, then you are acting under what Kant called the idea of freedom. You are therefore acting under the assumption that you have a will of your own; and you presume that you are free in a practical respect. This presumption implies, therefore, that you are subject to the principle of autonomy: act only on maxims that you could at the same time will to hold as universal laws. So, here are your options: assume the direction of your life is managed ultimately by your inclinations and the grip that the idea of your future happiness may have on you, and so give up the idea that you have a will of your own; or, embrace the idea of having a will of your own, and on that basis acknowledge that you, yourself, can direct your life independently of your inclinations and the idea of your happiness—whether or not this is true. Although these are stated in terms of options, it would be a mistake to assume that you have a choice, or that you must decide between the two. The fact is that you really cannot assume the first option. If you could choose between them, then in order to settle on the first option you would actually have to think of yourself as free, and so as already described by the second option.
Transcendental freedom. Remember that the “proof” here is supposed to convince us that we are free in a practical respect, but not as if “transcendental freedom” has been proved in “theoretical philosophy.” The argument only shows us that there is a way we must think about our agency. It does not show us, apart from that, what is true about our agency. It does not show us that we are really rational beings with a will. We must think we are, though we could be deceived.