The activation and selection of Chinese verbal classifiers: an ERP study

This study investigated the activation and selection of Chinese verbal classifiers during speech production to evaluate how lexico-syntactic features are retrieved, and how these processes are modulated by specific production demands. In two picture-word interference experiments (bare verb naming and verbal classifier phrase naming), we manipulated classifier congruency and semantic relatedness. Behaviourally, both tasks elicited robust semantic interference effects in naming latencies. Crucially, a classifier congruency effect in accuracy emerged only in verbal classifier phrase production. ERP results revealed classifier effects during the N400 time window in both experiments, and these effects were modulated by semantic relatedness. Ultimately, our results support the WEAVER++ model, aligning with previous evidence from nominal classifiers. These findings suggest that verbal classifier features are automatically activated during lemma retrieval, whereas competitive selection emerges only when structurally required by an agreement context.