Discovering Structure in the Moral Domain

Abstract

Early moral psychologists identified the moral domain with a class of actions that negatively impacted the wellbeing of others or violated their rights. However, anthropological work suggested that this view failed to capture the full extent of the moral domain, which can include victimless actions (e.g., food taboos), especially among socially conservative or non-Western individuals. Which kinds of acts are included in the moral domain? Along which dimensions do the acts differ from one another? Paper 1 utilizes a data-driven approach to mapping the moral domain, revealing a simple two-factor structure that captures variance in moral judgments across individuals, as well as reliable cross-voxel pattern information within individual brains. The remaining papers investigate judgments of agents who perform “harmful” acts (e.g., assault) versus “impure” acts (e.g., incest), which are each representative of the separate factors discovered in Paper 1. In Paper 2, we see an asymmetry in people’s causal attributions for the actions of harmful versus impure agents: impure acts are judged as more internally generated, and less due to the situation, compared to harmful acts. This asymmetry is due to differences in abnormality, a key dimension along which the moral domain may be organized. Paper 3 probes agent evaluations: how are harmful and impure agents expected to act in other contexts? People expect harmful agents to be harmful but not impure. In contrast, people expect impure agents to be both impure and harmful. This effect is connected to a model of the moral domain with a conceptual “core” of dyadic harm, surrounded by a periphery of victimless moral violations. Together, this work highlights a simple structure in the moral domain that can explain moral judgments, causal attributions, action predictions, as well as patterns of activity in the cortex.