Kratzer semantics for modals and conditionals generates the prediction that sentences of the form if p, ought p are trivially true. As Frank and Zvolenszky show, for certain flavors of modality, like deontic modality, this prediction is false. I explain some conservative solutions to the problem, and then argue that they are inadequate to account for puzzle cases involving self-frustrating oughts. These cases illustrate a general problem: there are two forms of information-sensitivity in deontic modals. Even generalizations of Kratzer semantics (...) that predict these two roles for information, e.g. Kolodny and MacFarlane, predict that they vary together. I propose a generalization of Kratzer semantics that allows the two roles for information to vary independently of each other. (shrink)

Metaethical contextualism is a form of moderate contextualism according to which the truth-conditions of normative utterances are sensitive to some salient standard, norm, or theory that is determined by the context of utterance. Metaethical contextualism implies that two speakers might utter grammatically contradictory normative sentences without expressing contradictory propositions, leaving the view vulnerable to the 'problem of lost disagreement'. The problem of lost disagreement occurs when two parties to a dispute disagree, but their utterances don't express exclusionary truth-conditional content: they (...) might both be right. While metaethical contextualists have proposed plausible solutions to the problem of lost disagreement, these solutions are ultimately unsatisfying, at least in contexts where one would expect exclusionary truth-conditional content. In this paper, I develop a form of metaethical contextualism that predicts exclusionary truth-conditional content in some contexts of disagreement, thereby assuaging any concerns that metaethical contextualism fails to adequately account for normative disagreement. (shrink)