Causal Proportionality as a Cognitive Budgeting Process

Abstract: Since Yablo (1992), philosophers have been concerned with quantifying the proportionality of cause-effect relationships. Broadly speaking, a cause is proportional to its effect to the extent that the cause is described at a level of detail that “fits” the description of the effect. In this talk, I’ll put forward the view, articulated in Kinney and Lombrozo (2023), that determining the proportional level of description for a particular cause of some effect is a species of a more general task. Namely, we hypothesize that the search for proportional causes can be understood more broadly as a process of balancing a desire for simple, coarse-grained causal representations against a desire for causal representations that preserve decision-relevant information. We hypothesize that each agent’s optimal balance of simplicity and informativeness is determined through cognitive budgeting: agents only expend cognitive resources to form fine-grained representations of their environment when doing so enables them to encode causal information that is relevant to their planning and decision-making. In this talk, I’ll present evidence from experiments in Kinney and Lombrozo (2023; 2024) that confirm this cognitive budgeting account of proportionality. The results of these experiments, I will argue, shed light on why humans adopt particular categorization schemes for representing the causal structure of their natural and social environments.