A Corpus-Based Study Of Delexicalized High-Frequency Verb Collocations Used By Chinese English Majors

The study of collocation has become very heated in recent years. Scholars in and out of China have conducted many researches on this issue from different perspectives and put forward a lot of useful suggestions for English teaching. However, some researches are not based on real life language data, thus the results are not very much reliable and persuasive. The real life language data employed in both quantitative and qualitative analyses can only be obtained from a corpus. It is not an exaggeration to say that corpus linguistics has brought forward a revolution in linguistic research. This thesis, adopting a corpus-based Contrastive Learner Corpus Analysis (CLCA) approach, devotes to the study of the characteristics and overall development patterns of three Delexicalized High-frequency Verbs (DHFVs) collocations, i.e. make, take and give used by Chinese English Majors (CEMs) at three different proficiency levels (beginning level, intermediate level and advanced level).This thesis is a combination of both quantitative and qualitative analyses. Research data are retrieved from St2, St5 and St6, three sub-corpora of the Chinese Learner English Corpus (CLEC). In terms of general English proficiency, St2 is treated as beginning learners (freshmen); St5 is regarded as intermediate learners (juniors); St6, advance learners (seniors). The comparable Louvain Corpus of Native English Essays (LOCNESS) is used as a control corpus, which is composed of essays written by Native American and British students.In this study, we first extract all the noun collocates of the three verbs from St2, St5, St6, and LOCNESS. An Excel calculator is used to calculate Z-score and Mi-score for each extracted noun collocates. Only the noun collocates whose Z-score is higher than 2 and Mi-score is higher than 3 are retained. Then a cross-sectional comparison and a Chinese versus native comparison will be made to reveal the striking patterns of the three DHFVs collocations produced by CEMs, The study shows that, compared to native speakers, CEMs show a strong tendency to overuse the collocations of the three DHFVs, but at the same time, they can not produce various alternative collocates as many as native speakers. There are similarities and differences between CEMs and native speakers. Some collocations produced by CEMs were also shared by native speakers. However, they are different in Z-score and Mi-score, which show the typicality of some collocates is different. The cross-sectional comparisons find the signs of progress in terms of collocation competence of CEMs at three different proficiency levels. However, the advance is not very much significant and even there is a temporary retrogressive phenomenon.This thesis also explains the underlying factors behind the misuses and overuses of DHFVs collocations. L1 interference and ignorance of collocational restriction are vital factors for the misuses and overuses of the collocations of this kind. Strategies like overgeneralization and avoidance are also found to be responsible for the misuse and overuse in the production of DHFVs collocations.Some pedagogical implications are also offered. CEMs' awareness of collocations should be enhanced. Collocations can be better learnt with the reference to Chinese. More attention should be paid to the study of the collocations of common words.