Relativity in the perception of emotion across cultures

Abstract

A central question in the study of human behavior is whether or not certain categories of emotion, such as anger, fear and sadness (termed "discrete emotions"), are universally recognized in the nonverbal behaviors of others (termed the "universality of attribution hypothesis"). In this dissertation, the universality of attribution hypothesis was revisited in order to examine whether individuals from remote cultural contexts perceive the same mental states in nonverbal cues as individuals from a Western cultural context. The studies described in this dissertation removed certain features of prior universality studies that served to obscure the underlying nature of cross-cultural perceptions. In study 1, perception of posed emotional vocalizations by individuals from a US cultural context were compared to those of individuals from the Himba ethnic group, who reside in remote regions of Namibia and have limited contact with individuals outside their community. In contrast to recent data claiming to support the universality hypothesis, we did not find evidence that emotions were universally perceived when participants were asked to freely label the emotion they perceived in vocalizations. In contrast, our findings did support the hypothesis that affective dimensions of valence and arousal are perceived across cultural contexts. In the second study, emotion perceptions based on facial expressions were compared between participants from US and Himba cultural contexts. Consistent with the results of Study 1, Himba individuals did not perceive the Western discrete emotion categories that their US counterparts did. Our data did support the hypothesis that Himba participants were routinely engaging in action perception, rather than mental state inference. Across both cultural contexts, when conceptual knowledge about emotions was made more accessible by presenting emotion words as part of the task, perception was impacted. In US participants, perceptions conformed even more strongly with the previously assumed "universal" model. Himba participants appeared to rely more on mental state categories when exposed to concepts, but a substantial amount of cultural variation was still observed. Finally, in Study 3, perceptions of emotion were examined in a US cultural context after the focus of participants was manipulated, either onto mental states (broadly), emotions or behaviors. Perceptions of emotion did not differ substantially across these three conditions, indicating that within a US cultural context the tendency to infer mental states from facial expressions is somewhat inflexible. Overall, the findings of this dissertation indicate that emotion perception is both culturally and linguistically relative and that attempts to apply the Western cultural model for emotions as a universal one obscures important cultural variation.