Abstract
The categorization of human emotion has remained one of the longest standing debates in the biological and social sciences. Despite the fact that emotions are recognized and expressed universally in six discrete categories (anger, disgust, fear, surprise, happy, and sad), researchers have argued that there might be more basic dimensions or units than the six categories. Here, we employed a perceptual learning procedure to investigate the relationships among the visual processing of the six emotions. Human subjects were trained on an emotion detection task, in which an emotional and a neutral face were presented successively in a random order. Subjects needed to identify the interval in which the emotional face was presented. The faces had a fixed contrast and were embedded in noise with different levels of contrast so that the detection threshold was quantified by the face/noise contrast ratio at which detection performance was 82% correct. Each subject was trained on one emotion for six days. Before and after training, we measured their detection performance on all the six facial expressions. Results showed that training on one facial expression improved performance not only on the trained expression, but also on other untrained expressions. Specifically, training on disgust detection improved the detection on anger, and vice versa; training on fear detection improved the detection on surprise, and vice versa. Such two-way improvements suggest similar neural mechanisms underlying their processing. Furthermore, we found that happiness detection was improved by training on all the six emotions but sadness detection was improved only by training on itself. These findings indicated that happiness might share some components with other emotions, while sadness was separated from other emotions. Therefore, the current study provides new insights to the structure of human emotion, suggesting that the six "basic" emotions might be composed of more basic units.
Meeting abstract presented at VSS 2016