Neural circuits of semantic and syntactic processing of Chinese and English

Abstract

There is debate as to whether processing of semantic and syntactic information involves overlapping or separate neural circuits. Whereas some pinpoint Broca’s area (BA 44), as a syntactic processing site, and anterior LIFG as site for semantic processing (1,2), others propose involvement of posterior LIFG in meaning and syntax (4,5,6) or engagement of frontal cortices according to task demands. While in alphabetical languages, syntax and semantics are relatively independent mechanisms, Chinese presents little morphology and high semantic ambiguity and reliance on context, and more interdependence between syntax and semantics.This fmri study explored the neural circuits of semantic and syntactic processing, by asking speakers of Mandarin (who had English as their L2) to judge synctactic or semantic similarities in English and Chinese sentences. For both languages, overall activity for the whole task was shown in areas consistent with previous literature (i.e. left inferior frontal gyrus and posterior cerebellar lobes). Syntactic judgments in both languages were associated with higher activation visual processing areas (cuneus), while semantic judgments for both languages elicited activation in the caudate -a hypothetical centre for language control- (11). However, syntactic judgements in Chinese recruited additional activity in left middle temporal gyrus, consistent with studies showing that disruption of this area involves results in alexia and agraphia for Chinese characters (10). Syntactic judgments in Chinese elicited activation in areas associated with reading unambiguous characters (Left Insula, 12), conflict monitoring (Left cingulate, 13) and memory (parahippocampus, 14). However, syntactic judgments in English elicited activation in auditory/motor speech feedback areas (bilateral postcentral gyrus, 15), the right equivalent to Broca’s area, involved in processing semantic anomalies and in working memory (RMFG, 16) and in cerebellar tonsil, reported in previous studies (8). We argue that overall, syntactic judgements place more demands on higher visual processing areas, whereas semantic judgements might involve control of meaning retrieval. Furthermore, Chinese syntax judgments likely involve control mechanisms of semantic retrieval, whereas in English (the L2), participants are possibly aided by covert speech, and by involvement of part of the network responsible for the processing of Chinese sentences