Shadows as sources of cues for distance of shadow-casting objects.

Abstract

Shadows are neglected sources of information about shadow-casting objects' distance (location). In an experiment using real shadows, participants judged the distance of two rods, either with shadows (illumination from the left) or without shadows. Without shadows, a thin rod raised slightly above the surface was incorrectly judged to be more distant than a thick rod, but with shadows the thin rod was accurately judged to be closer. For judgments to be correct, shadows had to overwhelm competing height and stimulus-size cues. Exp. II involved nonoverlapping shadows produced by a light placed in front of the rods. Again, without shadows the thin-elevated rod was incorrectly judged to be more distant, but with shadows it was correctly judged to be closer. Exp. III involved two sets of thick and thin elevated rods and frontal lighting. Judgments were accurate when shadows were present but inaccurate when shadows were absent. Experiments conducted through the American Psychological Society Web-research site replicated results of Exp. I, II, and III. A final Web experiment replicated results of the others, but with shadows produced by illumination from the right. Three properties of shadows--angles, interposition, and positioning--are possible sources of distance information associated with cast shadows.