More on Epistemic Modals

The questions are obviously related, but they should not be conflated, at least not well in advance of theory. Very roughly, the first is a question of semantics, the second of pragmatics (or of the semantics– pragmatics interface). The puzzle about supposition and epistemic modality I raised in section 1 of Yalcin 2007 was primarily used to motivate a specific answer to the first question — the semantic question. As for the second question, I said little about it until later in the paper. When I finally came to defend a view about it, it was on the basis of more abstract considerations: certain views about consequence (Sect. 4), judgements about the truth values of certain unembedded epistemic modal sentences, and the idea, already suggested by others, that a plausible formal pragmatics for unembedded epistemic modal talk could be given without having to assume that these sentences have an informational content in any usual sense. But my treatment of the second question was tentative. This is because my view was (is) that answers to it turn on considerations of a more global character, and are, unfortunately, not easy to settle with a quick puzzle. With this background, let me turn to Barnett (this volume). As Barnett understands me, I use the puzzle that appears in section 1