Kant and Relevant Descriptions

Filed under: , by: Kevin

To follow in a serious of unfortunately serious posts, I thought I would note some of my worries about Kant's ethics, although I am by no means an ethicist (to quote Jerry Seinfeld: "not that there's anything wrong with that").

Consider the second formulation of the categorical imperative:

Act only according to that maxim whereby you can at the same time will that it should become a universal law.

But how do we set limits on the relevant descriptions that are appropriate in individuating an act and the maxim that is to be applied to it? If I am performing two morally salient actions (stealing and providing for my family), which aspect will be primary in the description? The question of which description will be given priority is somewhat minor, and there are ways around it. But I think a more difficult question is at what level of specificity the description should be applied. First, it may be so general as to be the conjunction of both of these acts, or so specific that moral considerations fade into microphysical considerations. Second, what criteria do we have for bracketing context out of our descriptions of actions and maxims? Perhaps I might not universalize "Do not steal," but perhaps I would universalize "Steal to provide for your family" or some other concatenation of specificities. Yet there is a very strong possibility that without limiting bounds on what descriptions will count as relevant, we can generate descriptions that will render any act moral in virtue of some conjunction of aspects that contribute to the individuation of the act and its related maxim.

This is a separate question from the more traditional difficulty of how to decide between two conflicting imperatives. I take it that this latter problem can be resolved--I know of one colleague is actively working on the issue. My question precedes that one, however, for it challenges how we come to identify an act and a maxim in the first place. What are the appropriate, non-arbitrary grounds of individuation? What are the constraints on description?