语法语境下汉语名动分离的ERP研究

Neural Distinction between Chinese Nouns and Verbs in the Grammatical Context: An ERP Study

运用ERP技术，从语法角度，通过词语搭配判断任务，考察汉语名词和动词加工的脑神经机制。实验结果显示，在适合的语法语境中，名词、动词和动名兼类词所诱发出ERP差异主要反应在P200、N400和P600三个ERP成分上。在正确的语境中，名词诱发出更大的P200，而动词则诱发出比名词更大的 N400和减小的P600；当动名兼类词分别用作名词和动词时，虽然二者的N400没有显著差异，但前者诱发出一个增大的P600。根据实验结果认为：汉语名词和动词具有不同的神经表征和加工机制，名词和动词的语法功能在汉语名动分离中起了重要的作用。

Many current researches in neurolinguistics focus on the different possible neural representations of nouns and verbs. Evidence from neuropsychological and neuroimaging studies suggests different neural representations of nouns and verbs in Indo-European languages. Specifically， verbs are represented in the frontal region， while nouns are represented in more posterior brain areas. This is consistent with the results of electrophysiological studies. In contrast to other Indo-European languages， Chinese is characterized by the lack of inflectional morphology. Some investigators argued that nouns and verbs in Chinese are not dissociable with regard to grammatical classes but rather are differentiated by their semantic properties. The current study aims to investigate whether noninflectional nouns and verbs in Chinese could be dissociated at the neural level and whether the dissociation reflects their grammatical roles.All the nouns and verbs used in the experiments were all words that were not characterized by inflectional morphology. There were three sets of stimuli used in the experiments: (1) disyllabic unambiguous nouns， (2) unambiguous verbs， and (3) noun-verb ambiguous words. These stimuli were combined with a priming word to construct two contrastive phrase contexts， namely， noun-predicting (e.g.， “yi Q + _”) and verb-predicting (e.g.， “bu M + _”). The subjects were asked to decide whether the presented phrase represented a legal or illegal phrase. ERPs in response to the three sets of stimuli in different contexts were recorded. The ERPs were then analyzed using a repeated-measures ANOVA.We observed significant ERP differences between nouns and verbs in intervals of 145~225 ms， 250~400 ms， and 450~650 ms. Unambiguous nouns elicited a larger P200 than did unambiguous verbs. When the context completely matched their role， unambiguous verbs elicited a larger N400 and a smaller P600 than unambiguous nouns. Similarly， ambiguous words used as nouns elicited a larger P600 than did ambiguous words used as verbs.Our results do not support the claim that nouns and verbs in Chinese are not dissociable in terms of grammatical classes. On the contrary， the results demonstrated clear ERP responses that reflect a grammatical process. We argue that although the Chinese language lacks inflectional morphology， grammatical features are embedded in nouns and verbs and function regardless of whether there are explicit grammatical signs in these words. Grammatical features play an important role in the neural processing of Chinese nouns and verbs.