Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Neural Decoding of Morphological and Semantic Processing during Chinese Word Recognition across Native Speakers and L2 Learners of Different Proficiency
Poster Session D, Thursday, October 1, 4:30 - 6:30 pm, Wangari Maathai
This poster is part of the Sandbox Series.
Fei Gao1, Fei Gao2; 1Department of Computational Linguistics, University of Zurich, 2Institute of Modern Languages and Linguistics, Fudan University
Morphological and semantic processing are fundamental components of visual word recognition, yet their temporal dynamics in Chinese as a morpho-syllabic language remain incompletely understood, particularly in second-language (L2) learners with different proficiency levels. The present study investigated how morphological and semantic information are processed during Chinese word recognition and how these processes differ between native speakers and L2 learners. Behavioral and EEG data were collected during a visual lexical decision task involving real words, pseudowords, and nonwords. Contrasts between real words and pseudowords were used to index morphological sensitivity, whereas contrasts between lexical items (real words or pseudowords) and nonwords were used to examine semantic processing. Participants included native Chinese speakers (n=22), advanced L2 learners (n=17), and intermediate L2 learners (n=14). Neural representations were examined using time-resolved multivariate decoding with logistic regression. Group-level decoding performance was assessed against chance using temporal cluster permutation tests across the entire epoch, alongside ERP analyses using temporal cluster permutation within predefined regions of interest. Behaviorally, pseudowords elicited lower accuracy and slower reaction times than real words across all groups, with the strongest effects observed in intermediate L2 learners. Mean pseudoword accuracy was 90.2% in native speakers, 73.6% in advanced learners, and 64.2% in intermediate learners, compared with word accuracy of 98.5%, 99.3%, and 94.7%, respectively. Mean pseudoword reaction times were 813 ms, 1216 ms, and 1472 ms across the three groups, compared with 684 ms, 762 ms, and 1017 ms for real words. Time-resolved decoding revealed above-chance word vs. nonword classification beginning at 280 ms in native speakers, 395 ms in advanced learners, and 630 ms in intermediate learners, suggesting progressively delayed semantic processing with decreasing proficiency. In contrast, word vs. pseudoword decoding emerged later, beginning at 415 ms in native speakers and 465 ms in advanced learners, while intermediate learners showed no significant decoding cluster for this contrast. ERP permutation analyses converged with the decoding results. Native speakers showed significant FT7 clusters for word vs. nonword contrasts at 316–464 ms and 570–754 ms, word vs. pseudoword contrasts at 394–532 ms, and nonword vs. pseudoword contrasts at 502–582 ms. Advanced learners showed significant word vs. nonword effects at 340–670 ms and word vs. pseudoword effects at 422–708 ms in a right central/posterior ROI, whereas no reliable nonword vs. pseudoword effect was observed. These findings demonstrate a temporal dissociation between semantic and morphological processing during Chinese word recognition. Semantic processing emerged earlier than morphological sensitivity, whereas morphological decoding was delayed and weakened in L2 learners, particularly at intermediate proficiency levels. The results suggest that morphological processing in Chinese requires more advanced lexical and structural representations than semantic discrimination and highlight proficiency-dependent changes in the neural dynamics of Chinese word recognition.
Topic Areas: Morphology, Language Development/Acquisition