Object-extracted Preference In The Processing Of Chinese Relative Clauses:ERP Evidence

Through an event-related potentials (ERPs) study, we examined the processing of different types of Chinese (Mandarin) relative clauses, i.e., subject-extracted relative clause modifying the subject of the main clause (SSR) versus object-extracted relative clause modifying the subject of the main clause (SOR), subject-extracted relative clause modifying the object of the main clause (OSR) versus object-extracted relative clause modifying the object of the main clause (OOR), to test the universality and language specificity of sentence comprehension processes. The results of this study support object-extracted preference in both subject relative clause and object relative clause, i.e., SORs and OORs. In particular, ERP results in relative clause region showed stronger P600effect in SSRs than in SORs, reflecting a canonical strategy preference, and stronger N400effect in verbs than in nouns, reflecting easier understanding in nouns. ERP results in main clause object and the word after it showed stronger P600effect in SSRs than in SORs, suggesting thematic structure effect on syntactic construction and the processing preference of the whole sentence. ERP results in the third word showed stronger N400effect in OSRs than in OORs, reflecting the influence of meaning interference, and stronger P600effect in OSRs than in OORs, reflecting the canonical strategy preference. ERP results in the fourth word showed stronger N400effect in OSRs than in OORs, suggesting that current meaning processing can be overridden by previous meaning processing. ERP results in relative head and the word after it showed stronger P600effect in OSRs than in OORs, suggesting LPC effect, the effects of parallel function preference and perspective shifting, and the processing preference of the whole sentence.