Rhythmic Similarity hypothesis for cross-cultural vocal emotion recognition

This study examines the newly formulated Rhythmic Similarity hypothesis, which proposes that cross-cultural vocal emotion recognition is more accurate between users of languages with similar rhythmic structures. To disentangle the close relationship between culture and language, this study tested how well American English and French listeners recognized emotions in recordings produced by native speakers of Dutch and Korean. Since French and Korean share similar rhythmic structures, the prediction was that French listeners would outperform American English listeners in recognizing emotions in Korean. However, this prediction was not supported by our data, disconfirming the Rhythmic Similarity hypothesis. Furthermore, emotion recognition accuracy for both listener groups (American English and French) was higher in Dutch than in Korean, supporting the Cultural Proximity and Language Distance hypotheses. Additionally, recognition accuracy was above-chance for all emotions and was affected by arousal, valence, and basicness in ways consistent with previous findings.