| This thesis questions the use of the theory of affordance, a psychological theory of perception, as the basis for a symbolic design methodology in human-computer interaction. By comparing the psychological writings on affordance of James J. Gibson, James E. Cutting, and Edward Reed, with the design-oriented writings of Donald Norman, William W. Gaver, and Paul Dourish, the thesis identifies key differences in the writers' underlying assumptions. While psychological writings seek a theoretical mechanism to explain perception and its relationship to behavior, design-oriented writings attempt to associate perception and behavior through a formalization that can be leveraged by designers. Despite the distinction in their goals, these ideas about affordance are often used interchangeably, allowing designers to justify formalized decisions that remain unsupported by the psychological discourse.;In an effort to highlight the disagreement between the two definitions of affordance, this thesis traces each definition back to a set of underlying epistemological and ontological assumptions. While existing literature has established a connection between the James J. Gibson's writings and the philosophical tradition of phenomenology, little work has been done to uncover the assumptions of Donald Norman's definition of affordance. By associating his symbolic and formalized conceptualizations with those of early artificial intelligence work, the thesis relates Norman's definition with neo-Kantian ideas that fundamentally oppose the phenomenological ideas that form the basis for human-computer interaction. Based on these findings, this thesis questions the validity human-computer interaction's current symbolic design methodology. |
