Abstract
The horizontal-vertical illusion has often been interpreted in a three dimensional context, namely that the vertical bar in a “T” figure is perceived as longer because it refers to a line slanted in depth. Therefore, this illusion might be useful to investigate perceived depth in brain damaged patients, who might be specifically impaired in processing 3D scenes. Before using this illusion as a probe in to 3D perception, we were interested in measuring the 3D component in this illusion. For this purpose, we asked normal observers to compare the vertical and horizontal lengths of three figures (“T”, “L”, “+”) at different orientations (0°, 90°, 180°, 270°). We used the method of constant stimuli where we manipulated the aspect ratio of the figure. Participants had to judge which line (horizontal or vertical) was the longest and we estimated the point of subjective equality from the psychometric function. We found that two parameters - a vertical/horizontal parameter (p) and a bisecting parameter (q) - could explain all conditions presented. Parameter p refers to an overestimation of the vertical segment relative to the horizontal one, and parameter q to the tendency to overestimate the dividing line relative to the divided line, irrespective of their orientation. Participants showed consistent behaviors through all conditions. On average, parameter p corresponded to 5% of overestimation and parameter q to 17% of overestimation. If parameter p is related to a depth illusion, it would correspond to a figure presented in a plan slanted at about 18°. This dissociation between two independent parameters supports the use of the horizontal-vertical illusion to study more accurately the mechanisms involved in neglect patients and suggests it could be a useful tool to diagnose this population of patients.