Deliberative Modality under Epistemic Uncertainty

It is standard practice in linguistic semantics to treat a modal like ought as expressing quantification over possible worlds that are selected and ranked according to certain contextual parameters (Kratzer, 1977, 1981, 1991, 2012). On this view, different modal flavors correspond to different values for these contextual parameters. In the following, we will call the particular modal flavor at play in practical deliberations like (1) deliberative modality,1 and we will assume that, together with at least deontic (in the narrow sense, that is, related to rules and laws) and bouletic (related to wishes) it belongs to the so-called prioritizing modal flavors (in contrast to epistemic and dynamic ones, cf. Portner 2009).2 The ability of this approach to provide an adequate account of inferences involving sentences that express deliberative modality has been questioned. One recent such challenge is raised by the discussion of the miners puzzle in Kolodny and MacFarlane (2010):3