Evidence regarding cross-cultural regularities in emotional experience is central to the science of emotion. How many varieties of emotion are preserved across cultures in the emotions evoked in distinct situations? Are emotion categories or affective features (e.g., valence, dominance) more culturally universal? Do emotional experiences fall into discrete clusters or span continuous gradients? We provide answers to these questions by examining high-dimensional cross-cultural regularities in the emotions evoked by music using large scale data collection and analysis methods. We analyze hundreds of thousands of judgments from participants in the US (N = 1,011) and China (N = 895) of the emotions and affective appraisals evoked by 1841 music samples. We uncover 13 distinct dimensions, or varieties, of emotional experience that are preserved across cultures. Cross-cultural regularities in emotional response are represented by emotion categories such as “awe”, which drive culture-specific predictions of affective features such as valence and arousal. However, the emotion categories evoked by music are not discrete—they can be blended together in many ways. Our findings, visualized within an interactive map (https://s3.amazonaws.com/musicemo/map.html), reveal a complex, high-dimensional space of emotional states evoked cross-culturally by music.
The culture in which people live plays an important role in shaping their sense of self (Ryder & al., 2000) and emotional patterns (De Leersnyder, 2014). At the same time, previous research showed that individuals who live in between languages and cultures report an intriguing emotional hybridity (Pavlenko, 2005) and consider linguistic socialisation as an intense process of personal transformation (Panicacci & Dewaele, 2017). This research adopts an innovative approach to the topic by placing biographical and linguistic factors side by side. Statistical results from 468 migrants in English-speaking countries, supported by 5 follow-up interviews, revealed that the age of migration, the length of stay and the status in the host country were unrelated to participants’ perceptions of the heritage (L1) and host (LX) culture. Conversely, migrants’ language use, especially for expressing emotions, as well as L1/LX self-perceived dominance and emotionality linked with their attachment to the culture that produced that language. In other words, the cognitive and emotional embracement of the language contributed to enforce participants’ sense of belonging to the culture, explaining a variance of respectively 12.2% and 13.5% in their L1 and LX acculturation levels. Findings thus highlighted the crucial role of language use and affective socialisation in shaping individuals’ cultural belonging.
Despite the broad consensus about the necessity to study emotion in non-industrialized societies, emotion in many cultural niches remains underexplored. The present investigation focuses on hunter-gatherer populations—a rare type of society today, but a dominant one throughout human history. We examined two hunter-gatherer groups—the Maniq and the Mlabri—focusing in particular on emotion terminology in their respective languages. The two groups are in many ways similar: they are small traditionally hunter-gatherer societies of 300-400 individuals each, they live in Thailand, and speak related Austroasiatic languages. They differ, however, in that the Mlabri were recently forced to settle in permanent villages and turned to waged labor, while the Maniq are still predominantly nomadic and continue to forage for subsistence. We explored the meaning space of emotion using emotionally evocative stimuli: the Amsterdam Dynamic Facial Expression Set (ADFES) (Van der Schalk et al. 2011) and emotional scenarios adjusted to both cultural settings (cf. Boster 2005). We found that Maniq and Mlabri were similar in that they lacked dedicated terms for certain emotion concepts, e.g., disgust, and participants in both groups often relied on bodily or situational descriptors when responding to facial expressions, e.g., nose-scrunching, stink, etc. The two languages nevertheless differed in their inventory of emotion terms, use of metaphor, and their speakers’ appraisals of some emotional scenarios. We discuss these differences in the light of the ethnographic background and recent changes in livelihood, and consider the possible role of sociocultural factors in shaping the domain of emotion.
Leading theories of emotion recognition tend to be perceptual and context invariant. This is untenable. Here I argue that we need a theory of emotion recognition, and hence emotion concepts, that allows for context sensitivity. I argue that this is best explained by proposing that emotions are recognized and conceptualized using scripts. According to the dominant view, emotion recognition happens automatically, fast, and reliably (Ekman, 1969, 1971, 1999). Ekman’s claims seem to entail that people, independent of culture and context, have a universal capacity to recognize specific facial muscle movements as emotional expressions. But studies on emotional expressions show that social context has a lasting effect on how emotional expressions are interpreted (Hess, 2009, Crozier & de Jong, 2012, Gendron & Feldman Barrett, 2017). Some theories of emotions allow for context sensitivity. This is a key feature of theories that identify emotions with scripts (Russell, 1991, 2003). As a theory of emotions, this is controversial, but my proposal here is to extend Russell’s approach to emotion recognition. That is, in this paper I argue that emotion recognition, hence concepts, works via scripts. Scripts ground our understanding of emotion terms and allow us to categorize emotions in real world situations. Scripts also build in socially inculcated assumptions about group differences. In this way our emotion concepts also become loci of cultural knowledge and biases.