The perception and cognition of emotion from motion
Emotional expression has been intensively researched in the past, however, this research was normally conducted on facial expressions and only seldomly on dynamic stimuli. We have been interested in better understanding the perception and cognition of emotion from human motion. To this end 11 experiments were conducted that spanned the perception and representation of emotion, the role spatial and temporal cues played in the perception of emotions and finally high level cognitive features in the categorisation of emotion. The stimuli we employed were point-light displays of human arm movements recorded as actors portrayed ordinary actions with emotion. To create them we used motion capture technology and computer animation techniques. Results from the first two experiments showed basic human competence in recognition of emotion and that the representation of emotions is along two dimensions. These dimensions resembled arousal and valence, and the psychological space resembled that found for both facial expression and experienced affect. In a search for possible stimulus properties that would act as correlates for the dimensions, it emerged that arousal could be accounted for by movement speed while valence was related to phase relations between joints in the displays. In the third experiment we manipulated the dimension of arousal and showed that through a modulation of duration, perception of angry, sad and neutral movements could be modulated. In experiments 4-7 the contribution of spatial cues to the perception of emotion was explored and in the final set of experiments (8-11) perception of emotion was examined from a cognitive perspective. Through the course of the research a number of interesting findings emerged that suggested three primary directions for future research: the possible relationship between attributions of animacy and emotion to animate and inanimate non-humans. The phase or timing relationships between elements in a display as a categorical cue to valence and finally the unexplored relationship between cues to emotion from movements and faces.