A Corpus-based Study Of "V N" Colligation Used By Chinese English Majors

EFL learners' constructing grammatically correct but unnatural and unacceptable sentences has gradually shifted scholars and teaching practitioners' sights from traditional grammar to lexico-grammar. The former emphasizes the mastery of syntactic rules, but the latter advocates that form, meaning and function are associated and that lexis and syntax are co-selected. So to approach a word and its use from lexico-grammar perspective, one has to consider the immediate linguistic context, i.e. the grammatical and lexical patterns for the word. In fact, grammatical pattern and lexical pattern roughly correspond to colligation and collocation respectively. It is colligation and collocation that are the meeting-point of lexis and grammar, so many scholars argue for lexis' central place in language learning and regard them as a key parameter for measuring the interlanguage proficiency. Therefore, it is not strange that the study of collocation and colligation has become very heated nowadays. However, researches concerning collocation and colligation would, to a large extent, not be reliable and persuasive without the utilization of corpora because it is corpora that provide lots of natural linguistic contexts for qualitative and quantitative analyses.Up to now, domestic and foreign researchers have made many researches about collocation and colligation; especially after the compilation of learn corpus, comparative studies have sprung up to investigate the learning process of EFL learners. However, the study of Chinese learners' collocational behaviors in "V n" colligation has seldom been touched upon. The present study investigates the distinctive characteristics of "V n" colligation used by Chinese English Majors (CEMs) based on the theory of Interlanguage. Both quantitative approach and qualitative approach are adopted in this paper.In order to observe the overall developmental patterns of this colligation, the author extracts the cc3 performances of three subjects (st2, st5 and st6) from CLEC to make a cross-sectional comparison. Then, the author examines the features of the errors committed by CEMs via error analysis and finds out the underlying factors.Lastly, the author selects several representative verbs and nouns and extracts all their significant collocates from CLEC and BNC respectively with an aim at exploring the panorama of "V n" colligation through Contrastive Learner Corpus Analysis (CLCA). The following are the findings obtained in this study: 1). Generally speaking, CEMs' proficiency is in inverse proportion with committing errors, but the advance is very slow, even there is temporary retrogressive phenomenon. What is more important is that almost all errors committed by three subjects are related to most commonly-used words, especially to 5-, 4- and 3-star words. 2). CEMs' overgeneralization, ignorance of collocational restrictiveness, avoidance, unawareness of semantic prosody or inappropriate mastery of the node's meaning singly or conjunctly results in committing errors, but the major underlying source is learners' and teachers' preference the OCP to the IP. 3). The overuse and underuse phenomena are pervasive in CEMs' "V n" colligation. Neither st5 nor st6 uses various alternative collocates as many as native speakers, which shows that the collocates used by CEMs are apparently insufficient and CEMs have not built an effective collocational network between nouns and verbs. 4) There are similarities and differences between CEMs and native speakers in the use of synonym in "V n" colligation.This paper also delves into the underlying factors behind the misuses, underuse and overuse, and the author does not forget to discuss the implications of the findings for EFL teaching and learning.