Ecological Resources for Modeling Interactive Behavior and Embedded Cognition

A recent trend in cognitive modeling is to couple cognitive architectures with computer models or simulations of dynamic environments to study interactive behavior and embedded cognition. Progress in this area is made difficult by the fact that cognitive architectures traditionally have been motivated by data from discrete experimental trials using static, noninteractive tasks. As a result, additional theoretical problems must be addressed to bring cognitive architectures to bear on the study of cognition in dynamic and interactive environments. I identify and discuss three such problems dealing with the need to model the sensitivity of behavior to environmental constraints, the need to model context-specific adaptations underlying expertise, and the need for environmental modeling at a functional level. I illustrate these problems and describe how we have addressed them in our research on modeling interactive behavior and embedded cognition.