Causal reasoning without mechanism

Area of Science:

• Cognitive Psychology
• Causal Inference
• Philosophy of Science

Background:

• Understanding unobservable mechanisms is key to inferring causal structures from observable events.
• Current models often rely on mechanistic or probabilistic knowledge, which may not always be available.
• A gap exists in explaining human causal reasoning when detailed mechanistic information is absent.

Purpose of the Study:

• To introduce and validate the domain-matching heuristic as a model for human causal reasoning.
• To investigate how individuals infer causal relationships without explicit mechanistic understanding.
• To identify the specific domains people utilize in this heuristic reasoning process.

Main Methods:

• Participants were asked to cluster artifacts to identify commonly used mechanism domains.
• The domain-matching heuristic was tested by examining causal attribution, prediction, and judgment in adults and children.
• Experimental tasks assessed subjective understanding of causal links based on domain congruence.

Main Results:

• Analysis of artifact clustering revealed three primary mechanism domains: mechanical, chemical, and electromagnetic.
• Participants' causal inferences, predictions, and judgments aligned with the domain-matching principle.
• Both adults and children demonstrated reliance on this heuristic across various causal reasoning tasks.

Conclusions:

• The domain-matching heuristic provides a robust explanation for how humans perform causal reasoning with limited mechanistic knowledge.
• This heuristic simplifies the inference of cause-effect relationships by focusing on domain coherence.
• Findings suggest a fundamental cognitive strategy for navigating causal complexity beyond explicit mechanistic or probabilistic models.