Item

Abstract

Humans attribute animacy even to very simple objects displaying self-propelled or goal-directed motion. To test attribution of animacy parametrically using classical psychophysical techniques, we created animations consisting of a single dot that appeared either self-propelled (modelled on the movements of a fly) or moved by an external force (modelled on a leaf drifting in the wind). Both animations were built using the same movement equation and differed in speed and acceleration profiles, allowing parametric morphing from one ‘extreme‘ animation to the other. Low-level stimulus properties (range of screen positions covered, speed or acceleration) did not vary systematically during morphing. 26 naive subjects were asked to rate the ‘extreme‘ animations and 4 intermediate morphs for animacy. Ratings from 19 subjects as well as averages over all subjects could be modelled by a cumulative Gaussian, median PSE was in the middle of the morph range and the median JND was 1.7. These stimuli thus allow parametric testing of animacy perception from single objects with movements modelled on real animate entities.