I was thinking of Quine's "change the logic, change the subject," saying, and thought over "change the deontic logic, change the deontic subject," and so then I wondered if deontic logics without ought-implies-can must be about something other than those logics with it. Neither would be the "true" deontic logic, mayhap, and so there would be no absolute definition of morality/normativity behind them and to which they were supposed to answer; no reason to judge the acceptant or rejectant of "ought-implies-can" as in a more essentially well-grounded debate position than the other, and so no sufficient grounds for rejecting the implication as having counter-deterministic consequences.

On the other hand, without the rather contrived denial of agglomeration, a pair of conflicting obligations does become a command to enact a contradiction and then "ought-implies-can" would mean either that contradictions can be enacted (which few would claim) or that the conflict between the obligations has to be resolvable somehow (which is quite disputed). If we accept that ought-implies-can is sometimes materially true (see below), do we have to think that it is always materially true, though?

Material implication: sometimes we start from knowledge of possibility to then reason by modus tollens that something is not really obligated. For example, if Sam can't turn into a giraffe by talking to a full moon, then he's not obligated to do so, and we can know the antecedent before recognizing the consequent. However, the other main use of "ought-implies-can" is to establish far-reaching metaphysical claims such as the logical possibility of a joint created-uncreated personage (Anselm) or the metaphysical possibility of having done otherwise than we actually did (Kant). It is the latter kind of reasoning that I mean by "material implication" modulo "ought-implies-can" and so my query is specifically whether the Kantian inference need be made universally (i.e. whether every obligation can a priori establish our metaphysical capacity to act otherwise or whether it's logically possible for some-but-not-all obligations to do so).

  • 1
    I do not think it "has" to be so taken. If one motivates OIC by "morality is a guide to action", or some such pragmatics, then dropping OIC does "change the subject". However, Kant was not so pragmatic, and there are many examples where people are not physically able to do X, while acknowledging that they still have a duty to do it on something like Kant's terms, see e.g. Mizrahi. So, arguably, OIC may be dropped without changing the subject, perhaps, it even "has" to be dropped as absolute. Kant's own formulation of OIC has a ceteris paribus, btw.
    – Conifold
    Aug 3 at 16:58


You must log in to answer this question.