| Resumo: |
Theories of emotion are currently polarized between categorical models of basic emotions (e.g., fear) and models that propose more fundamental affective dimensions (arousal/valence) that may be combined in different ways to give rise to emotions. Neuroscientific research on Facial Expressions of Emotion (FEE) has typically favoured basic emotions models, but recent evidence suggests that the two views may be fruitfully combined in a hybrid model. In this project, we will investigate the role of affective dimensions in the perception of FEE while retaining emotional categories as relevant levels of analysis. Specifically we will attempt to address the seemingly inconsistent findings reported above by adopting a Predictive Coding (PC) framework. Interestingly, the recent extensions of PC approaches into the affective domain seem to provide a general model that may integrate dimensional and categorical views of emotion and facial expressions. Affective PC suggests that the neural perceptual expectations are intrinsically affect-laden, i.e., they include both predictions about the physical characteristics of the stimulus as well as about its affective value. Phase 1: ERP experiments designed to assess how affective dimensions of FEE relate to other stimulus characteristics such as typicality, basic emotional content, and emotional intensity. Phase 2: cross-sectional ERP study designed to assess developmental changes in FEE processing between the ages of 4 and 15 years. Phase 3: multimodal EEG/fMRI study designed to examine the neural correlates (activation and connectivity) of FEE processing while controlling for affective dimensions and emotional categories. |