Can fictions sometimes reveal important information about what beliefs audience members hold? I argue that a case can be made that emotional responses to some horror fictions can reveal that audiences harbor beliefs in the supernatural, beliefs that audience members might otherwise deny holding. To clarify the terms of the discussion, I begin with an overview of two leading theories of belief: the representational and dispositional accounts. I explore the role of belief in the production of emotional responses by posing (...) a hard question that none of the leading theories answers directly: Why are some fictional scenarios and events so much more effective than others? I argue that the answer has to do with belief, that is, the beliefs about the world that audiences bring to fictions. After laying the groundwork, I argue that cultural differences in audience responses to some horror fictions might be best explained by what supernatural beliefs they hold. After developing the case, I offer several reasons to be skeptical of this conclusion. (shrink)

Pragmatists have argued that doxastic or epistemic norms do not apply to beliefs, but to changes of beliefs; thus not to the holding or not-holding, but to the acquisition or removal of beliefs. Doxastic voluntarism generally claims that humans acquire beliefs in a deliberate and controlled way. This paper introduces Negative Doxastic Voluntarism according to which there is a fundamental asymmetry in belief change: humans tend to acquire beliefs more or less automatically and unreflectively, but they tend to withdraw beliefs (...) in a controlled and deliberate way. I first present a variety of philosophical, empirical and logical arguments for Negative Doxastic Voluntarism. Then I raise two objections against it. First, the apparent asymmetry may result from a confusion of belief with other doxastic attitudes like assumption, supposition, hypothesis or opinion. Second, the apparent asymmetry seems to vanish if we focus on doxastic states rather than just beliefs. Some rejoinders and their consequences for the vague concept of belief are sketched. (shrink)