Beyond Valence and Arousal: Distributed Facial Patterns Linked to Specific Emotions Cannot be Reduced to Core Affect

Emotions are central to human life and, as such, are of primary interest in basic psychological research. There is widespread agreement that emotions involve subjective experiences that can be described with discrete natural language labels and often involve changes in bodily states, but there is ongoing debate about how specific and differentiated these bodily states are, and how they relate to emotional labels. Recent work showed that objective measures derived from the face can be used to accurately classify discrete emotional labels. However, it remains possible that the facial patterns associated with discrete emotions merely convey valence and arousal (i.e., "core affect") and that this information can be utilized to deduce discrete emotion categories. In light of this, in the current work, we examined whether the facial patterns that reliably distinguish between emotional states are reducible to valence and arousal. Our findings support the position that the human face contains rich information that can be used to predict people's emotional states and that this information is not reducible to core affect. We discuss the implications of this work to the debate concerning the nature of emotions.