Deontological ethics

7. Kantianism

Ross, then, offers no single principle that can serve to unify morality, nor does he think that morality can be justified from without. But Ross’s view is not the only deontological option: Kant’s categorical imperative (CI), for example, can be seen as an attempt to unify and justify morality (see Kant, I. §§9; Kant 1785).

The CI proposes a test, and actions that fail this test are, Kant claims, wrong. Crucial to the test is the notion of a maxim. We act with certain aims, and these can be specific or general. Maxims are general aims. Thus my maxim may be: make lying promises (i.e. ones I intend not to keep) whenever it benefits me. The CI test asks first on what maxim I propose to act, and then enquires whether this maxim is one that I could will to be a universal law. Here is a rough illustration: the maxim to make lying promises whenever it benefits me cannot be universally willed because its universal adoption would lead to the demise of the very practice on which it relies – namely the practice of promising. Hence making lying promises for my own benefit is wrong.

How exactly the CI test is to be understood and what it would rule out are matters of scholarly dispute. But there is general agreement that Kant’s ethics has a deontological structure. We have just seen, with our example of a lying promise, something of how the Kantian approach might give rise to duties of special relationship. And the test also yields constraints, for agents are forbidden, on an alternative formulation of the test, to treat others merely as a means. Exactly what this entails is again in dispute, but it is intended to rule out such things as lying and killing the innocent even to minimize such behaviour by others. To kill an innocent yourself to prevent other killings, for example, would be to use your victim as a means to minimize victimization. Finally, provided we do not violate the CI, Kant’s system also appears to permit the pursuit of, say, personal projects, since, according to him, we have only a limited duty to help others.

The rationale for Kant’s test lies in a certain conception of rationality. If something is a reason for one agent then it must be capable of being a reason for all. Thus a maxim is not a good reason for action unless it is one on which all agents can act. Any maxim that could not consistently be followed by all, or could not consistently be willed as one that all should follow, is not rationally acceptable – it fails to show respect for the autonomy of all other rational agents. Hence, if the Kantian approach works, morality is justified by appeal to something outside itself, namely, rationality; and we also have, in the CI test, a single unifying criterion of wrongness.