Emotion in Action: Correlates of Emotion Recognition from Body Movements

Abstract

Human emotions can be viewed as a principal currency for relationships as they are often the underlying motivational force for human behavior. Thus, the ability to perceive and correctly interpret the emotions of our conspecifics constitutes a core competence and is of great adaptive value, as it allows us to navigate our behavior towards an adequate response. The human body has been identified as a potent source to quickly retrieve a wealth of information about affective states. In this regard, previous research has discussed factors that modulate emotion recognition, including kinematic and postural properties of the movement, the presence (or absence) of an interacting partner as well as individual characteristics of the observer. The major aim of this dissertation is to establish further factors that modulate the perception of emotions from body movements from interactions. To tease these factors apart, three experiments were carried out. The first experiment investigates which parts of the body convey the most salient information that ultimately allows for an identification of the respective affective state. Moreover, it is addressed how the emotional expressivity of the observer influences emotion recognition. The second experiment assesses how the spatiotemporal coupling of the agent8s actions influence emotion recognition. The third experiment extends previous research designs by implementing a novel virtual reality apporach to investigate whether the recognition of emotion is influence by the perceived (co)presence with the interacting agents. Moreover, gender differences are assessed using a subjective valence rating scale. The results indicate that the information that is retrieved by the observer to recognize an emotion depends on the type of emotion that is displayed. Whereas the emotions anger and happiness can be recognised more reliably from arm movements, sadness seems to be perceived mainly via the cues from the head and torso. Moreover, our results suggest that the recognition of anger and affection depends on a spatiotemporal coupling of the agent9s actions while the recognition of happiness and sadness does not. With respect to interindividual differences, we found that more emotionally expressive individuals display higher emotion recognition accuracies than individuals with lower subjective emotional expressivity. Lastly, our results do not support the notion that there are gender differences regarding the valence perception of emotional interactions.