Abstract
Previous research has established that emotional information conveyed through body and facial movement is systematically and reliably identified (Bassili, 1979, Dittrich et al., 1996, Atkinson et al., 2004). Furthermore, the neural circuitry involved in such emotion perception (e.g., Heberlein et al., 2004) overlaps with the neural areas involved in the visual analysis of human motion per se (e.g., Puce & Perrett, 2003). This raises the question of whether emotion detection processes and action detection processes interact. Specifically, do emotion related processes contribute to the detection of human movement? Or, is human action first detected and then subsequently interpreted by emotional processes? We examined the relationships between emotion and action detection with psychophysical discriminations of point-light walkers of various emotional states. Emotional gaits were performed by professional actors and were captured with an optical motion tracking system. Pilot studies were used to identify gaits that conveyed anger, sadness, happiness, or emotional neutrality equally well. Point-light walkers were then placed within masks of positionally scrambled but otherwise identically moving points. On half the trials, a point-light walker was present within the mask. On the remaining trials, the walker was scrambled and thus absent. Observers reported whether or not the human form was present in each movie. Discrimination performance was emotion dependent. Furthermore, performance depended upon the observer-relative direction of gait as discrimination was best when point-light actors walked toward observers. Importantly, sensitivity analyses indicated that detection accuracy was modulated by emotion and direction interactions. These results suggest that the visual analysis of human motion cannot be understood independently from the visual analysis of emotion.
Acknowledgement: Supported by NIH EY12300