ABSTRACT
Support for the universality hypothesis primarily cites universal facial expression recognition (i.e., decoding, e.g., Ekman et al., 1969). Yet, only a subset of the six “universal” facial expressions is universally recognized (i.e., namely happy, surprise, anger, and sad) suggesting a universal language of fewer than six facial expression signals (see also Sauter, Eisner, Ekman, & Scott, 2010, for evidence of cultural specificity and universality in emotion vocalizations). Interestingly, this subset does not include the facial expression signals of fear and disgust, although both are largely considered to be “primitive” on the basis of their biological origins (e.g., Susskind et al., 2008) and association with rapid deep brain activity (e.g., Whalen et al., 2004, but see also Pessoa & Adolphs, 2010 for a discussion). Instead, low (but significantly above chance) recognition of “universal” facial expressions suggests that these signals contain elements of universal signals of emotion. For example, consider a facial expression of the Oryia women in Bhubaneswar, India where “the tongue extends out and downward and is bitten between the teeth, the eyebrows rise, and the eyes widen, bulge, and cross” (Shweder, 1991, p. 246). Based on the available information, we would either conclude that the emotion is fear (or surprise) on account of the widened eyes, or disgust based on the protruding tongue. Given that the facial expression represents surprise/embarrassment/fear, accuracy would exceed chance (but not be high enough to support accurate communication). Thus, by extracting the universal elements (e.g., eye widening, eyebrow raising), while ignoring unfamiliar signals (i.e., crossing of the eyes), a basic level of communication could be achieved.
