Motion- and stationarity-based figure/ground segregation: The effects of temporal asymmetry

This study investigated figure/ground segregation based on both common movement ("common fate" for the Gestaltists) and perceived stationarity (nonmovement), i.e., a figure presented in two different locations and embedded in noise either appeared moving or stationary when it was perceptually segregated. The effect of differences in the duration of a figure (a three-dot line) in each location was studied in Experiment 1. A figure was presented for the same duration (duty cycle = 0.50) or for different durations (duty cycles of 0.25 and 0.75) in each location (the Temporal Symmetry and Asymmetry conditions, respectively). The figure always appeared to move when it was segregated from background noise (motion-based figure/ground segregation) for Temporal Symmetry. In contrast, figure/ground segregation could be based on the perception of stationarity as well as motion for Temporal Asymmetry. Thresholds (50%) were measured as the number of added noise dots that resulted in perceiving motion of the figure and for perceiving the figure itself. Additional results demonstrated that the figure was perceived in the location having the.75 duty cycle. It is proposed that higher energy in the 0-Hz frequency component in this location results in stationarity-based figure/ground segregation, whereas higher energy in time-varying frequency components underlies motion-based figure/ground segregation. In Experiment 2, hysteresis was observed in the Temporal Asymmetry condition in the transitions between perceived motion and stationarity and between perceived stationarity and noise. In the Temporal Symmetry condition, hysteresis was observed in the transition between perceived motion and noise. The presence of hysteresis indicates that perceived stationarity in the Temporal Asymmetry condition is not due to insufficient activation of motion detectors, but is a stable perceptual state based on mechanisms that are sensitive to 0-Hz energy and compete with motion-detecting mechanisms sensitive to energy in time-varying frequency components. Judgments of motion direction in Experiment 3 provided tentative evidence that both duration of the briefer frame and duty cycle influence perceived motion.