Cristina Ballarini

I argue that credences do not participate in an epistemically central kind of mental process—active (i.e. deliberate, person-level) reasoning. My argument hinges on the empirical finding that human thinkers tend to “neglect alternatives” when deliberately reasoning with uncertainty: in cases where thinkers recognize that their uncertainty is distributed over various possibilities, they tend to engage in downstream reasoning that attends to just one possibility at a time. A model on which thinkers reason with beliefs about probabilities better accounts for the empirical results. I conclude that beliefs play an epistemically important role in human minds that credences do not. Besides motivating a new kind of dualism about belief and credence, my argument additionally suggests a way of characterizing the functional difference between credences and beliefs about probability.

The New Evil Demon Problem presents a serious challenge to externalist theories of epistemic justification. In recent years, externalists have developed a number of strategies for responding to the problem. A popular line of response involves distinguishing between a belief’s being epistemically justified and a subject’s being epistemically blameless for holding it. The apparently problematic intuitions the New Evil Demon Problem elicits, proponents of this response claim, track the fact that the deceived subject is epistemically blameless for believing as she does, not that she is justified for so believing. This general strategy—which I call the “unjustified-but-blameless maneuver”—is motivated, in part, by the assumption that the distinction between epistemic justification and blamelessness is merely an extension of the familiar distinction between moral justification and blamelessness. In this paper, I consider three ways of drawing the distinction between justification and blamelessness familiar from the moral domain: the first in terms of a connection with reactive attitudes, the second in terms of the distinction between wrongness and wronging, and the third in terms of reasons-responsiveness. All three ways of drawing the distinction, I argue, make it difficult to see how an analogous distinction in the epistemic domain could help externalists explain away the intuitions which underwrite the New Evil Demon Problem. Motivating the unjustified-but-blameless maneuver, I conclude, is a much less straightforward task than its proponents tend to assume.