128(2). 203-235), as both groups of listeners performed similarly on average in both recordings. Likewise, the Prosodic Proximity hypothesis was not supported by our data, since French listeners did not outperform American English listeners in recognizing emotions in Korean, although French and Korean obviously share basic prosodic structures. Instead, the effect of culture and language was emotion-specific, varying across Speaker Language and/or Listener Language. Additionally, despite the significant impact of the emotional dimensions of arousal, valence, and basicness on recognition accuracy, these dimensions did not consistently predict accuracy, since individual emotions deviated from the general patterns of accuracy due to emotion-specific features. Together, these findings provide a more nuanced perspective on the relative contributions of universality, culture, language, and emotional dimensions to cross-cultural emotion recognition, especially for listeners unfamiliar with the target language.
Liang et al. (Mon,) studied this question.