The role of emotions in the maintenance and revision of beliefs
- by Sunny Khemlani
- in News
- posted May 10, 2024
Monica Bucciarelli and Phil Johnson-Laird have two new papers on the role of deontic beliefs and emotions. In a new paper out in Sistemi Intelligenti, they examined the link between the strength of in a particular belief and the strength of the emotions it evokes. Here’s the abstract:
Deontic beliefs related to what is right and what is wrong are among the most divisive in daily life, both between individuals and between cultures. Moral beliefs, which can evoke strong emotions, are a case in point. Previous studies we have conducted have shown that the strength of belief in a deontic assertion about the moral domain or the domain of social conventions is related to the strength of the emotion it evokes and, in particular, to how pleasant or unpleasant the assertion is. However, this effect occurs only with deontic assertions, not with factual assertions of the same content. Mental model theory assumes that emotions play a key role in deontic beliefs because they are not as testable as facts. From the assumptions of the theory derives the prediction that the critical correlation between the strength of belief and the strength of emotion should also occur in deontic beliefs regarding precautionary rules and personal recommendations. The results confirm the existence of a critical correlation between strength of belief and strength of emotion for these types of deontic assertions as well. The relevance of these studies lies in the possibility of deepening the mechanisms underlying the modifiability of beliefs, given that beliefs associated with strong emotions are not readily changed even in the presence of evidence to the contrary.
The paper is available here (paywalled).
Bucciarelli and Johnson-Laird tested some of the consequences of these ideas and described them in a new paper to be presented at CogSci 2024. They find that, in line the mental model theory, assessments of deontic assertions (but not parallel factual assertions) depend on the emotions they evoke. The abstract is here:
An assertion about a fact can in principle be tested in observations. That is impossible for assertions about what is permissible or obligatory, i.e., deontic assertions based on moral principles, conventions, rules, or laws. Many modal logics concern these matters. But an integrated theory of emotions and reasoning predicts that emotional reactions and strength of belief should be correlated for deontic assertions, but not for factual assertions. You can be convinced that it is wrong to take paperclips from the office, and that it is right for society to provide health care for everyone, and your emotional response to these two assertions is likely to correlate with the strength of your beliefs in them. In contrast, you can be convinced both that fresh snow is white and that fossil fuels are making the world hotter, but have an emotional reaction to only the second of these assertions. Grounds for factual assertions are empirical findings. But assessments of deontic assertions depend in part on the emotions that they elicit. Previous studies have corroborated this prediction for moral claims, matters of convention, prudential rules, and personal recommendations. We report two experiments that yield the same interaction for legal pronouncements from the Italian Civil Code compared with parallel factual assertions. People like propositions they believe, and they believe propositions they like. We discuss several remaining unknowns including the potential role of emotions in reasoning about legal and other deontic propositions.