Graduate School of Informatics, Department of Informatics Associate Professor
Cognitive Science from an Ecological Rationality Perspective
Human perception and decision-making can be modeled as the brain’s readout of high-dimensional sensory representations. The structure of these representations is constrained by environmental and biological regularities, such as inhomogeneities in the external world, the spatial distribution of retinal cells, and the stochastic nature of neural firing. Under such restrictions, psychological phenomena—including perceptual illusions and metacognitive biases—can emerge as natural outcomes of constrained optimization rather than as failures of the system.
We argue that identifying the structure of sensory representations and explaining human behavior as their necessary consequences offers a powerful meta-theory for cognitive science. Adopting this perspective allows us to move beyond merely reproducing isolated effects and toward constructing unifying frameworks that connect diverse phenomena across multiple cognitive domains.