Talk by Ed Baggs, June 13th

On Thursday, June 13th, Ed Baggs (blog, twitter, scholar) will deliver a talk that should be of broad interest to cognitive science students, psychologists, and beyond. Ed provides a thought-provoking perspective on ecological psychology, based on an analysis of the relation of an organism to its surround. The talk will take place at 3 p.m. in B2.24, the boardroom on the second floor of the Computer Science Building.

Title: The limits of explanation in embodied cognitive science

Embodied cognitive science seems to offer a radically different type of explanation from that found in computational cognitive science. It promises to explain particular behaviours in terms of perceptual or mechanical coupling between an actor and some structure in the actor’s surroundings. Perceiving and acting are held to be a matter of ‘maximizing or minimizing something visual [or perceptual], or centering or symmetricalizing something, not merely a matter of reacting to a stimulus’ (Gibson 1972, unpublished note quoted in Reed 1988, p. 77). The resulting empirical programme has proved successful and productive. But the methodology appears only to be suitable for the study of behaviours where there is a clearly definable goal that can be described in terms of an optimal solution. Balancing on a beam or swinging for a home run may be describable in such terms; constructing a sentence or a termite mound, maybe less so. For this reason, it is sometimes argued that embodied explanations are inherently limited and cannot “scale up”. I propose that this scaling-up metaphor is misleading. It implies that a single type of explanation should exhaustively capture all types of behaviour. In fact, the optimization mode of explanation does not even capture what is perhaps the most fundamental form of behaviour, namely exploration. I further suggest that “distinctively human” forms of behaviour arise through a dialectical tension between optimization and exploration, and such behaviours require a qualitatively different type of explanation.