Why do coherent belief systems tolerate incoherence? Belief update failures in society

Our research agenda is centered around a novel conceptual framework of belief formation and belief updating that integrates coherence and incoherence, describes how they interact, and specifies which parts of a person’s
belief system are (in)coherent. The idea of our framework is to distinguish core variables, about which the person forms beliefs that are characterized by a strong emotional valence and that are less bound by coherence;
and inferred variables, about which the person forms beliefs to maximize coherence with observed variables (or “evidence”) and their core beliefs. Core variables might include a person’s ability in a domain important
to them; and inferred variables might include whether employers or society in general reward these abilities. In interpreting evidence and consequently in updating their beliefs about inferred variables, the person uses as
guidance a mental model of “how the world works,” i.e., how core, inferred, and observed variables relate to each other. This gives rise to a complex system of core and inferred beliefs. Our conceptual framework builds on the idea that emotions and evidence have distinct direct effects on belief formation: emotions primarily affect core beliefs, while evidence predominantly shapes inferred beliefs. However, the interaction between core and inferred beliefs, driven by the need for coherence, introduces complex indirect and feedback effects. Specifically, changes in emotions can affect core beliefs, and thereby
indirectly alter inferred beliefs. Conversely, new evidence can influence a person’s emotional state, thereby affecting core beliefs. This is resolved in a (sometimes fragile) equilibrium of core and inferred beliefs that
features both coherence and incoherence. Given its multifaceted nature, our framework will allow us to organize our thinking about beliefs in diverse scenarios. We develop and apply this framework through six closely connected subprojects. In doing so, we follow a micro-to-macro-to-micro approach: on the one hand, we start from psychological principles and build towards social implications; on the other hand, new insights on aggregate phenomena will feed back into a better understanding of the underlying psychological mechanisms. In the first set of projects, we investigate core beliefs. We first formulate formal theories of core belief formation, assuming a strong relation to the person’s self concept. Motivated by the hypotheses developed here, we turn to study experimentally how emotional and cognitive dynamics shape core beliefs through attention and arousal. We then dig deeper into the formation of
core beliefs by focusing on their emergence in the developmental process. In the second set of projects, we investigate inferred beliefs and how they interact with core beliefs. We first study how to influence inferred beliefs via language and content. We then broaden the scope to study interactions between individuals who hold false beliefs. And finally, leveraging the results from the previous subprojects, we investigate economic and political implications. Based on our novel insights on these macro-level questions, we will go back to the micro level and refine our theories of belief formation and updating. In combining the expertise of psychologists, linguists, and economists, we believe that this approach can result in a deeper understanding of belief formation and belief updating across domains.