ABSTRACT
There are two distinct ways one can be indecisive about a question: one can be ambivalent or one can be uncertain. We present linguistic and normative reasons in support of this view, including evidence about the inappropriateness of using epistemic modals when one is ambivalent about whether something is the case despite having complete evidence about it. The evidence suggests that when we are ambivalent about whether something is the case, the question is treated as neither open nor settled. We characterize this aspect of ambivalence as a kind of attitude toward penumbral cases of vague predicates.
