Causal discovery and generalization

The fundamental problem of how causal relationships can be induced from noncausal observations has been pondered by philosophers for centuries, is at the heart of scientific inquiry, and is an intense focus of research in statistics, artificial intelligence and psychology. In particular, the past couple of decades have yielded a surge of psychological research on this subject – primarily by animal learning theorists and cognitive scientists, but also in developmental psychology and cognitive neuroscience. Central topics include the assumptions underlying definitions of causal invariance, reasoning from intervention versus observation, structure discovery and strength estimation, the distinction between causal perception and causal inference, and the relationship between probabilistic and connectionist accounts of causal learning. The objective of this forum is to integrate empirical and theoretical findings across areas of psychology, with an emphasis on how proximal input (i.e., energies impinging on sensory systems) can be translated into reliable and accurate mental models of distal causal connections. Our main focus will be on three broad questions of (i) how humans, and other animals, discover the causal structure of their environment, (ii) how such inference is generalized across spatial and temporal contexts, and (iii) how it shapes cognition, perception and action. Although this dialogue is mainly devoted to the psychological study of how humans acquire and generalize causal knowledge, a deeply related question, and one that has profoundly impacted developments in cognitive science, is how such processes can or should be computed by intelligent systems. Consequently, we strongly encourage contributions, while maintaining a primary focus on psychology, to draw heavily on statistics, philosophy, computer science, and any other discipline grappling with the constraints and opportunities afforded by a furthered understanding of causal representation.