ABSTRACT
Based on long-standing, widely accepted notions of a “universal language of emotion”, much of research has used “universal” FACS-coded facial expressions to inform broad fields. However, as described, FACS-coding largely represents Western facial expression signals, thereby limiting and biasing knowledge of the true complexities of facial expressions across cultures and affecting related fields (e.g., social neuroscience). In order to redress the balance and expand scientific knowledge beyond the boundaries of Western, educated, industrialized, rich and democratic (WEIRD) populations (Henrich, Heine, & Norenzayan, 2010), it is imperative to identify facial expression signals that are representative of emotion communication in different cultures. With the advent of new data-driven methods and technology-e.g., the Generative Face Grammar (GFG; Yu et al., 2012)—it is now possible to more comprehensively explore the specific face signals (and their combinations) associated with the perception of different emotion categories across diverse cultures (see also Gill, Garrod, Jack, & Schyns, 2012, for perceptual modelling of social traits). As a result, use of bottom-up (i.e., unbiased) data-driven methods will highlight rather than erase genuine cultural differences and similarities in facial expression signals and their corresponding perceptual categories (i.e., decoding).
