Abstract
Human observers can identify gender solely from biological motion cues. We asked whether gender classification of biological motion stimuli could be affected by prior gender adaptation. Motion captured male and female actors were used to construct gender-specific prototypes and dots were placed on the major joints producing Point Light Walkers (PLWs). The prototypes were mixed in varying proportions to create a range of test PLWs. On each trial, subjects (n=7) viewed a 100% female, a 50/50 neutral or a 100% male PLW, immediately followed by a brief test PLW. Subjects reported the gender of the test PLW. For each adapting stimulus, a parametric curve was fitted and the critical percentage (Pc), the male/female percentage of the test PLW that was equally likely to be classified as male or female, was calculated. In Exp. 1, adapting stimuli were presented for 12 s. The neutral adapting stimuli had little influence on judgments, with a Pc of 51.3% male. When adapted with a male prototype, the Pc was 58.0% male. In contrast, with female adaptation, the Pc fell to 42.7% male. This result indicates gender-specific adaptation. This finding was replicated in Exp. 2, in which stimuli were sped up by 25% with a proportional decrease in viewing duration. A neutral adapting stimuli yielded a Pc of 49.5% male. When adapted with a male prototype, the Pc was 52.8% male. In contrast, with female prototype adaptation the Pc fell to 44.8% male. Reducing the duration of adaptation reduced the adaptation effect. Our data reveal that adapting to biological motion of one gender increases the likelihood that subsequent test stimuli are judged as the opposite gender. Gender classification based on motion cues is thus pliable and sensitive to recent stimulus experience.
Supported by NIH institutional training grant to HJ and R01EY12872-04 to GRS.