Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Distinct neural responses to words and pseudowords in emotional prosody perception between native and non-native speakers
Poster Session F, Friday, October 2, 2:45 - 4:45 pm, Wangari Maathai
Cheng Xiao1, Jiang Liu1, Rutvik H. Desai1, Ma Feilong1; 1University of South Carolina, USA
Introduction. Understanding how native speakers and L2 learners differ is critical for understanding language acquisition, comprehension, and usage. The In-Group Advantage (IGA) hypothesis (Elfenbein & Ambady, 2002) posits that native speakers have an advantage over non-native speakers in recognizing emotional prosody. However, current empirical evidence evaluating the IGA hypothesis remains mixed. This discrepancy can be partly explained by the choice of stimuli: some research uses real words, whereas other studies use pseudowords. This distinction is particularly crucial for tonal languages, where the same acoustic feature (i.e., f0) encodes both word meaning and emotional prosody. In this study, we examine how native speakers and L2 learners differ in their neural responses by analyzing fMRI responses to real and pseudo- words in a tonal language, Mandarin Chinese. Methods. Participants. Thirty right-handed, healthy young adults (18 females; mean age: 22.9 years) participated in this study, including 15 L1-Chinese L2-English speakers and 15 L1-English L2-Chinese speakers. Stimuli. We used Chinese real words and pseudo-words spoken with positive and negative prosody, recorded from one female and one male Chinese speaker. All stimuli were controlled for syllable length and lexical tone. Real words were additionally matched for frequency, valence, arousal, concreteness, and age of acquisition, and L2 Chinese learners passed a vocabulary screening for the real words to ensure comparable semantic knowledge. Pseudowords were newly created using lexical gaps so that they remained entirely unknown to both groups. The perceived valence and arousal of both emotional prosodies were validated by an independent norming study (N = 42). fMRI acquisition and analysis. For each participant, we recorded their neural responses during four functional runs of 11 minutes each. Each run comprised 24 randomized 26-second blocks, and each block was followed by an emotional judgment task. We preprocessed the fMRI data using fMRIPrep, hyperaligned them across all participants, and compared the neural responses to real words and pseudowords between native speakers and L2 learners. Results. The native group recognized emotional prosody more accurately than the learner group in both real words (B = 1.671, p = 0.003) and pseudowords (B = 1.321, p = 0.008), supporting the IGA hypothesis. Furthermore, native speakers and L2 learners showed distinct neural responses. There were significant interactions between group and stimuli (t = 6.041, p < 0.001). Specifically, while processing real words, native speakers showed increased activation in the angular gyrus (AG) and decreased activation in the middle frontal gyrus (MFG) compared to L2 learners. While processing pseudowords, native speakers showed decreased activation in the AG and increased activation in the MFG compared to L2 learners. We repeated these comparisons using only positive prosody or only negative prosody and obtained consistent results. Conclusion. We found that native speakers and L2 learners showed distinct neural responses during the emotional judgment task, and these differences were modulated by the stimulus category. Specifically, the differences observed while perceiving real words and pseudowords were opposite in direction. This interaction effect may help resolve the discrepancies in the existing literature regarding the IGA hypothesis.
Topic Areas: Multilingualism, Prosody