Mechanistic explanation and dynamical cognitive science

Cognitive phenomena are now thought to arise from exceedingly complex, flexible, and large-scale mechanisms distributed across brain, body, and the environment. This has elicited doubts about the future viability of mechanistic explanation: the idea that (cognitive) phenomena can be adequately explained by describing the organized systems of structures and processes from which they arise. By considering contemporary examples from dynamicist cognitive psychology, evolutionary robotics, systems neuroscience and other branches of cognitive science, I argue that these doubts are unfounded. Specifically, I show that sophisticated methods of description and analysis such as dynamical modeling and graph-theoretic analysis can be used to describe cognitive mechanisms of ever-increasing complexity, flexibility, and size. Moreover, I argue that powerful computer simulations allow researchers to discover and reason about mechanisms that cannot be easily identified in nature. Rather than abandon mechanistic explanation, therefore, researchers in cognitive science invoke novel methods and concepts to discover and describe increasingly complex, flexible and large-scale cognitive mechanisms. By putting a spotlight on these methods and concepts, I seek to provide both an accurate understanding of cognitive scientific practice as well as an improved philosophical conception of mechanistic explanation.