| Summary: | Although facial expressions are one of the most important ways of communication in human society, most studies in the field focus only on the emotional aspect of facial expressions. The communicative/conversational aspects of expressions remain largely neglected. In addition, whereas it is known that there are culturally universal emotional expressions, less is known about how conversational expressions are perceived across cultures. Here, we investigate the underlying dimensions of the complex space of emotional and conversational expressions in a cross-cultural context. For the experiments, we used 620 video sequences of the KU facial expression database (62 expressions of 10 Korean actors), and 540 video sequences of the MPI facial expression database (54 expressions of 10 German actors). Four groups of native German and Korean participants were asked to group the sequences of the German or Korean databases into clusters based on similarity, yielding a fully crossed design across cultural contexts and databases. The confusion matrices created from the grouping data showed similar structure for both databases, but also yielded significantly less confusion for own-culture judgments. Interestingly, multidimensional scaling of the confusion matrices showed that for all four participant groups, two dimensions explained the data sufficiently. Most importantly, post-hoc analyses identified these two dimensions as valence and arousal, respectively, for all cultural contexts and databases. We conclude that although expressions from a familiar background are more effectively grouped, the evaluative dimensions for both German and Korean cultural contexts are exactly the same, showing that cultural universals exist even in this complex space.
|
|---|
