Figure 1. Illustration of a bistable point-light walker.

Illustration of a single frame of the bistable point-light walker (without perspective cues), and the two veridical interpretations with the point-lights superimposed. Whilst both veridical interpretations are equally plausible, observers usually prefer the FV interpretation. Thus, the objectively bistable point-light walker does not correspond to the subjectively bistable one. To obtain subjectively bistable stimuli, we performed a preliminary adjustment task in which perspective cues carrying the information concerning the veridical orientation of the point-light figure were gradually manipulated. On each trial, participants were asked to indicate whether the visually presented stimulus was perceived as FV or FAV. Individual perspective levels were determined by fitting a cumulative Gaussian function to the proportion of FV responses in the different perspective conditions, and selecting the perspective manipulation corresponding to the 30%, 50%, and 70% FV thresholds.