Walkable slants of simulated plane ground surfaces yield lower discrimination thresholds than non-walkable slants.

Abstract

Seen in central perspective, sets of parallel lines yield gradients of convergence when aligned with the observer's direction of gaze, or foreshortening when orthogonal to it. A squares tiling and a random distribution of lines combine both features. Such stimuli, simulating plane, slanted ground surfaces, were shown to naïve observers who, in a two-alternative forced-choice task, had to decide in which of the two pictures of a trial the simulated slant was steeper. Difference thresholds were smaller for walkable slants (23-33°) as opposed to non-walkable ones (40-55° and 59-79°), but, for the walkable slants, they did not differ significantly between convergence and foreshortening stimuli (1.9° vs. 2.0°), nor did they do so between a squares tiling and a random distribution of lines (1.5° vs. 1.3°). For intermediate slants, thresholds were smaller for foreshortening stimuli and a random distribution of lines. The prospects of future studies, using different axes of slant, juxtaposing different textures within single trials, tilting surfaces, or adding gradients of light and shade, are outlined.