Date of Completion

January 1982


Psychology, Experimental




What an environmental object offers or affords for action can be analyzed in terms of the dynamic fit between an organism and its environment. Using methods of intrinsic measurement and similarity analysis, the dissertation examines the consequences of ecosystem dynamics for the visual control of activity. Action boundaries are characterized as 'critical points' in the topology of an ecosystem, and regions of minimum metabolic cost as 'optimal points.' The invariance hypothesis of perception and action proposes that critical points will predict visual category boundaries, and optimal points will predict visually preferred paths of activity.^ Four experiments examine this hypothesis for the activity of bipedal climbing in Short and Tall males, on stairways with varying riser heights. The existence of intrinsic optical information for the ratio of riser height (R) to leg length (L) is demonstrated under natural conditions. In Experiment 1, the perceptual category boundary between 'climbable' and 'unclimbable' stairs is predicted from a geometric analysis of gait. Critical riser height is estimated to be a constant proportion of leg length: R/L = .88. Slides of stairways (R = 20-40 in.) were judged by Short and Tall observers in a categorization task. The perceptual boundary occurred at R/L = .88 for both groups, as predicted.^ In Experiments 2-4, visually preferred riser height is predicted from measurements of metabolic cost during climbing on an adjustable, motor-driven stairmill (R = 5-10 in.; f = 50 steps/min). Minimum energy expenditure per vertical meter (cal/kg-m), indicating optimal riser height, occurred at R/L = .26 for both Short and Tall climbers. In an independent study, slides of stairways were judged in forced choice and rating tasks, resulting in visually preferred riser heights of R/L = .25 and .24 in the two tasks, respectively.^ These results are interpreted as support for the invariance hypothesis. They suggest that visual perception for the control of activity has a basis in the biodynamics of the ecosystem. Implications for stair construction and the design of human-artifact systems are discussed. ^