Abstract
Models of spatial perception typically assume no contribution of vestibular information to the perception of distance despite the importance of updating object distances during motion and expected variations in the distance to the ground plane during head tilt. To investigate the possible contribution we investigated depth perception with different body postures but with visual context cues constant. Experiments were conducted in the York University Tumbled Room, a realistically decorated room (8’ x 8’ x12’ ) oriented orthogonal to gravity, and in a normally oriented control room. The length of a visual line, projected onto the far wall by a laser, was compared to a tactile reference rod. An adaptive psychometric procedure was used to control the length. Ten observers were tested in an upright posture with respect to each room binocularly and monocularly with the room visible, and binocularly with only the line visible. The line length judged as equal to the reference rod was significantly longer (p=0.015) in the tumbled room by 108.8% ± 5.7% than in the control room. The effect was larger during monocular viewing (p<0.01, 113.6% ± 9.6%) but went away in the dark. The longer line length required to appear equal to the reference rod when subjects were on their backs suggests that, as in the moon illusion, it appeared closer (up to 74% of the actual distance ). By keeping the visual context constant, this is the first demonstration that posture has a direct effect on perceived size and, by implication, distance. That the effect was largest with monocular viewing suggests a moderating effect of stereopsis (as in the moon illusion). The fact that the effect disappears in the dark suggests that it is due to a conflict between visually and gravitationally defined reference frames rather than directly due to posture relative to gravity.
Meeting abstract presented at VSS 2013