On Supererogation: One Should Go When Going Is Good Enough and Not Going Is Not

Abstract

Major existing approaches to prioritizing modality (which subsumes deontic and bouletic modality) suffer from the problem of supererogation (‘giving too much’), vali dating the inference from the premises (i) that p should be the case and (ii) that ‘p and q’ is better than ‘p (and not q)’ to the conclusion that q should be the case. I propose a novel account of should and related prioritizing-modal expressions, under which should(p) essentially means that making p happen is, or constitutes part of, the sole way (among the contextually prominent alternative options) to make the situation ‘good enough’.