Abstract
Evidence has suggested that differences in performance exist when comparing tasks involving visually directed locomotion to those involving visual perception of space. Such evidence indicates that systematic errors apparent in visual space perception do not influence the motoric responding associated with such visual exposure. For instance, when subjects are required to verbally match distance intervals oriented in a frontal-parallel plane to those oriented in-depth when provided with static visual cues alone, a robust foreshortening effect is observed for in-depth estimates — an effect that disappears when responding via blindfolded locomotion. Our intention was to replicate this finding using a slightly different task and to examine additional conditions in which dynamic visual information and non-visual information are concurrently available. The current experiment was conducted in a large, open, outdoor environment. Subjects were required to view a target distance oriented in-depth and match that distance, first in an in-depth plane and subsequently in a frontal-parallel plane (forming a 90 deg angle). Responding occurred in one of three ways: 1) verbally positioning a target from the original viewing position using static visual information, 2) blindfolded locomotion, 3) locomotion with vision. Results demonstrated that visually perceived estimates varied as a function of orientation, with distance estimates produced in the frontal-parallel plane resulting in a substantially smaller extent than those oriented in-depth. Such a result is in agreement with other findings demonstrating a dissociation between perception and action in locomotion. For visually directed action, a strong effect of optic flow was also found. When subjects walked without vision, a consistent undershoot of the target distance was observed, whereas when subjects walked with vision, a consistent overshoot of the target distance was observed.
Supported by an NSERC and a CFI grant to HJS.