Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Comparing the neural correlates and effective connectivity of different types of syntactic movements in Mandarin Chinese: An fMRI study.
Poster Session C, Friday, October 25, 4:30 - 6:00 pm, Great Hall 3 and 4
Tingwu Lee1, Shiaohui Chan1; 1National Taiwan Normal University
Filler-gap dependencies, or syntactic movements, involve a relation between a moved filler and the gap that it left. Previous neuroimaging studies generally assume that the left inferior frontal gyrus (L-IFG) and the left posterior superior/middle temporal gyrus (L-pS/MTG) play crucial roles in processing syntactic movements. However, whether the brain tells the subtle differences of movement types and distinguishes them remains unclear. For instance, Europa et al. (2019) adopted the fMRI technique and found that object clefts and passives in English only differ in the intensity of the L-IFG activity, while both structures share the same neural connectivity network. This result suggests that the human brain may not distinguish movement types when they are processed. Several fMRI studies on Chinese passives also have similar results supporting the view of the movement of the Patient object to the subject position (Feng et al., 2015; Want et al., 2021). However, the dominant formal syntactic account for deriving Chinese passives is the null operator movement analysis, involving a base-generated subject with the Patient role, a null operator moved sentence-internally, and a post-verbal gap left by the operator (Huang, 1999). The operator (OP) serves a deictic sign bridging the Patient subject and the post-verbal gap ([e]), as in nanhai bei nuhai [OP] pian-le [e] ‘The boy was deceived by the girl.’ The theoretical account is incompatible with the movement hypothesis from the fMRI studies, but it raises a new question about the neural foundation of null operators. Thus, the aim of this fMRI study is twofold. First, I compared the neural correlates of two movement types in Mandarin Chinese, particularly the passive and topic constructions. The latter was chosen because it represented a typical case of object movement to the sentence-initial position. Second, I attempted to draw inferences from the neural activity results to examine the validity of the operator movement account. By adopting the whole-brain analysis and the region of interest analysis, it was found that both passive and topic constructions triggered the L-IFG and the L-pMTG equally intensively. However, the dynamic causal modeling analysis showed that the Chinese passive construction failed to trigger the modulatory effect from the L-IFG to the L-MTG as the Topic construction successfully did. Such a modulatory effect is argued as crucial evidence reflecting the processing of filler-gap dependency: the L-IFG is activated to maintain and retrieve the information of the filler and then pass the information to the posterior temporal region for establishing filler-gap integration. As a result, the object movement analysis of Chinese passives is called into question, but there is no direct evidence for the neural correlates of operator movement found in this study. We assume that the strong activations of the L-IFG and the L-MTG are to deal with the structural complexity and the non-canonical thematic role order. The result also suggests that the brain does not process all structures with non-canonical word orders equally.
Topic Areas: Syntax and Combinatorial Semantics,