ERP Investigation Into Mental Mechanisms Of Chinese University Students In English Sentence Processing

Researchers generally consider that sentence processing mainly involves syntactic and semantic processing. And the theoretical models concerning the issue come into being: the first is the principle-based model holding the serial or syntax-first model, in which a rigid sequence between syntactic and semantic processing is emphasized. In other words, after the completion of syntactic processing, semantic processing begins and at last the integration takes place. The second is the constraint-based model which suggests that semantic and syntactic processing are relatively independent processes, but syntactic processing is not autonomous. Semantic and syntactic processing interacts with each other from the very beginning of sentence processing, according to specific semantic or syntactic constraints. The third is the concurrent model suggesting that there are no completely independent and rigidly sequential semantic and syntactic processing, semantic processing is concurrent with syntactic processing.To the deficiencies and theoretical discrepancy of behavioral paradigms based on reaction times, since 1980s, researchers have adopted methods of neural cognitive psycholinguistics, among which, event-related brain potentials (ERP) which can make continuous measures on the whole process of language comprehension and connect the research data with the function of the brain, is attached great importance to. Meanwhile, researches make use of the anomaly paradigm which introduces syntactically, semantically anomalous word and both semantically and semantically anomalous word at a certain position of the sentence. As a result, researchers have reached the following conclusions: 1. Semantic and syntactic processing are two different processes. Semantic anomaly produces the N400 effect; syntactic anomaly evokes LAN/ELAN or P600 or the combined LAN and P600 under different conditions. 2. There are three phases in sentence processing: initial syntactic processing, semantic processing and syntactic integration, which is embodied in the neurocognitive model of sentence processing proposed by Freiderici(2002).The results mainly come from the studies run with English and other Indo-European languages as subjects' native language (for instance, German, Dutch, and Spanish), second language, or foreign language. There are few studies which deal with the processing of ESL learners with Chinese as their native language. Therefore, the study on mental mechanisms of English sentence processing in Chinese university students can enrich the relative data of second language processing and comprehension and broaden or modify the related theoretical models. Accordingly, the study systematically explored English sentence processing of Chinese ESL learners, with the purpose of providing some theoretical evidence for English teaching.The study aimed to probe into the following issues by using ERP:1,whether ERP components and time courses of English sentence processing in Chinese university students are different from previous studies; 2,whether ERP components presents differences by employing different research paradigms including the paradigm of sentence-final and sentence-internal anomalies as well as the paradigm of different syntactic structures involving the subject-number agreement and the passive voice of verbs in the sentence. 3,whether the neural theory of three-phase sentence processing can account for mental mechanisms of Chinese university students in English sentence processing.The study randomly selected 20 university students (10 female, 10 male) as subjects, who are right-handed and have passed CET-4. The study included two experiments in which the critical words introducing anomalies were the verbs and the task of subjects was to judge whether the sentences were correct In addition, the indices of the observation were reaction times, the rate of correct responses and ERP indices in the process of the experiments. In experiment 1, experimental materials included 360 simple sentences whose length ranges from 6 to 8 words and were classified into six types (each type includes 60 sentences): control sentences, sentence-internal syntactic anomalies, sentence-internal semantic anomalies, sentence-final syntactic anomalies, sentence-final semantic anomalies and sentence-internal and sentence-final double anomalies involving both semantic and syntactic difficulties. In experiment 2, experimental materials including 240 passive sentences whose length ranges from 6 to 8 words, were classified into four types (each type includes 60 sentences), including control sentences, sentence-internal syntactic anomalies, sentence-internal semantic anomalies and double anomalies involving both semantic and syntactic difficulties. Based on the comparison between simple and complex sentence structures (active sentences-passive sentences), the study explored the characteristics of sentence processing and the interaction between semantic and syntactic processing and further probed into the mental mechanisms of sentence processing.The results indicated:1. Semantic and syntactic anomalies of English sentences elicited the similar ERP components and time courses in sentence processing of Chinese University Students to those of the studies run with English and other Indo-European languages as subjects' native language (for instance, German, Dutch, and Spanish), second language, or foreign language.2. In simple active sentences, under sentence-internal syntactic anomalous condition, the ELAN was visible. Under sentence-final syntactic anomalous condition, the P600 was observed. In relation to semantic anomalies, the sentence-final N400 was significantly larger than the sentence-internal. Under doubly anomalous condition, both the N400 and P600 were detected and the change on amplitude showed an asymmetry compared to single anomalies. To be specific, as for the P600, double anomalies elicited a larger increase in the P600 amplitude. With regard to the N400, double anomalies produced a smaller amplitude. The pattern of results showed the interaction between semantic and syntactic processing.3. In passive sentences, under syntactic anomalous condition, the ELAN and P600 evoked. As for semantic anomalies, the N400 was also visible. Under doubly anomalous condition, both the N400 and P600 were detected and, identical with findings from simple sentences, the change on amplitude showed an asymmetry compared to single anomalies. As to P600, double anomalies evoked smaller P600 amplitude. With regard to the N400, double anomalies produced a smaller amplitude. The findings suggested the significant interaction between semantic and syntactic processing.4. ERPs responses to double anomalies in passive sentences were reliably larger than those to double anomalies in active sentences, which revealed the more significant interaction between semantic and syntactic processing in passive sentences.5. The study found there existed the effect of word position, namely, word position had an impact on ERP components, which is identical with the finding from Hagoort Additionally, under the present experimental conditions, the effect of syntactic structures was detected: in non-marked syntactic structures, semantic processing had a greater impact on syntactic processing. But it was opposite in marked syntactic structures.6. The findings, in part, supported the neurocognitive model of sentence processing proposed by Freiderici.