Abstract
This thesis addresses a range of interdependent themes centred on how emotions are picked up from the body movements of actors engaged in natural communication. There were 4 main aims. The first considered the importance of the social context. The second was to further our understanding of the mechanisms responsible for the perception of affect from biological motion. The third considered how the skill to read affect from body movements develops in children. The final aim was to further elucidate what is known about affect. A corpus of point-light displays were created from the body movements of pairs of female and male actors engaged in interlocution whilst emoting. A MacReflex motion analysis system was used to record their movements. Observers viewed either the original and/or various permutations of the original displays and made judgements about the portrayed emotions. The results of the first set of experiments, which tested whether subjects could recognise the emotions and also explored the effects of inversion, were in line with previous findings which showed that emotions can be recognised from biological motion and that inversion impairs recognition. However, this latter effect varied, depending on the emotion shown. Of central importance, the following two experiments considered the social context of the displays. For most of the emotions tested, it was found that seeing a natural interaction made a statistically significant enhancement to how well many emotions were recognised. This finding highlights the need for natural stimuli and the need to consider emotions at a social, as well as an intrapersonal, level. The mechanisms were further investigated in the next set of experiments. It was shown that biological motion of affect is processed differently to biological motion of locomotion. The final investigation focused on how children develop this ability: Children as young as five are able to identify some of the emotions shown and by age nine development is almost at adult levels. In summary, the findings reported in this thesis show that the social context is important in emotion and future studies should incorporate the social context in the stimuli. They also lend support for a) categorical theories of emotion and b) the existence of distinct modules in the mind.