Auditory Emotional Prosody Perception Using Pseudo-Speech Stimuli in Native and Non-Native Listeners

Objective

To assess emotional recognition ability using corpora of emotions conveyed through nonlinguistic pseudo-sentences in native and non-native listeners (Turkish and English).

Methods

A cross-sectional design was employed, including a total of 60 young adults (aged 18–25 years). Of these, 30 were American English-speaking participants with no knowledge of Turkish, while the remaining 30 were age-, sex-, and education-matched Turkish participants. Emotional recognition scores and reaction times were assessed following audiological measurements using a one-interval, five-alternative forced choice method. A hundred stimuli recorded by Turkish speakers were presented, including 5 emotions × 2 speakers × 10 pseudo-sentences. The emotions tested were “neutral,” “happy,” “angry,” “surprised,” and “panicked.”

Results

No statistically significant differences exist between the groups' recognition of “neutral” and “angry” emotions. However, significant differences were observed in the recognition of happy, surprised, panicked and the mean scores of the emotions. Reaction times showed that nonmeaningful pseudo-sentences elicited similar listening efforts between native and non-native listeners.

Conclusion

Overall results suggest that while there may be recognizable vocal cues irrespective of languages for expressing angry and neutral emotions, this does not apply to all emotions. These results point to the fact that the type of test materials may play an important role when measuring emotional recognition among different cultures using auditory stimuli. In terms of reaction time results, pseudo-sentences could be used for cross-language auditory emotion recognition, however with certain emotions.