Abstract
Religion is a ubiquitous cultural phenomenon that ordinarily involves shared belief in supernatural beings. Psychological perspectives on religion have tended to include either motivational biases (e.g., beliefs mitigate anxiety), or cognitive biases (e.g., beliefs depend on how information is processed), neither of which has afforded a complete explanation of religious belief. In this thesis, I propose a cognitive-motivational model that incorporates both perspectives to illustrate a path by which beliefs in supernatural beings can form. Ten studies are reported to support the model. In these, the inherent threat potential of supernatural beings is found to improve memory for them, especially among anxious individuals. Furthermore, believable supernatural beings are found to possess features that differ from their secular counterparts, including greater beneficence, ambiguity, ambivalence, and mind-based abilities. These features are found to facilitate processes of motivated reasoning and emotion regulation (e.g., repression), transforming threatening supernatural beings, that are prolifically remembered by anxious individuals, into more believable beings. Finally, individuals identifying as religious are found to use repression more often than non-religious individuals, particularly to form and maintain positive impressions and beliefs about supernatural beings with the aforementioned features. Taken together, these findings support a model in which cognitive biases for threat detection and motivational biases for emotion regulation can be integrated on a causal pathway from anxious states to religious belief. Although the model focuses on positive beliefs about supernatural beings, many other comforting beliefs that appear in religions (e.g., an afterlife) could be attributed to supernatural beings that are perceived in a positive way. The model answers questions that cognitive and motivational perspectives are unable to answer in isolation, such as why only some supernatural beings are worshiped. This has been called the “Mickey Mouse problem” after the cartoon superhero who has never been worshipped as a god. The model further explains the relevance that a known memory bias for supernatural agents (called the MCI effect) has to religious belief. Most importantly, the model describes the role of anxiety in religion. For centuries, scholars have proposed relationships between anxiety and comforting religious beliefs. To my knowledge, this thesis represents the first detailed analysis of how such a relationship could work. In summary, these findings demonstrate the explanatory power of integrative approaches to the study of religion, and the potential for discovery of other cognitive-motivational pathways that may culminate in religious belief within the proposed model.