Abstract
Our concept bald displays all of the features paradigmatic of vagueness. For instance, it is sorites-susceptible. Moreover, the concept allows for cases of a distinctive kind in the middling part of the sorites series -- cases for which it seems ''there is just no fact of the matter'' whether the member is bald. Our concept ought displays some of these same features. For instance, it too is sorites-susceptible. However, there is a difference. The middling cases in an ought sorites series lack the ''no fact of the matter'' phenomenology. In this paper, I ask: why do we have the no fact of the matter phenomenology for descriptive concepts? And what's special about our concept of ought that it lacks that phenomenology? I start by giving a rigorous characterization of the ''no fact of the matter'' phenomenology in terms of questioning attitudes. When it comes to middling cases of our concept bald, it is confused to continue to ask whether the person is bald. This isn't so for middling cases of our concept ought: we can reasonably wonder what one ought to do in the middling cases. I then offer an explanation for this zetetic difference. I argue that, asking questions about middling cases in a sorites is typically costly. But for certain normative concepts, the questions are inescapable, and so it's a cost we must pay.