Précis to a practical unified theory of cognition and action: Some lessons from EPIC computational models of human multiple-task performance

Abstract : Experimental psychology, cognitive science, and human factors engineering have progressed sufficiently far that a practical unified theory of cognition and action is now foreseeable. Such a theory soon may yield useful quantitative predictions about rapid human multiple task performance in applied settings. Toward this end, an Executive-Process/Interactive-Control (EPIC) architecture has been formulated with components whose assumed properties emulate fundamental perceptual, cognitive, and motor processes. On the basis of EPIC, a theorist may construct detailed computational models that characterize multiple task performance under both laboratory and real world conditions. For example, EPIC computational models provide good accounts of response latencies and accuracies from the psychological refractory period procedure, aircraft cockpit operation, and human computer interaction. As a result, major commonalities in performance across various task domains have been discovered, and efficacious principles for designing person machine interfaces have been identified. The substantive and methodological lessons learned from these advances constitute an instructive precis to further utilitarian theoretical unification.