A Corpus-driven Study On Collocation Errors

Collocation acquisition is a difficult problem for Chinese EFL learners. With the development of the computer science and concordance software, corpus linguistics appears in foreign language study and research as a new subject. By using the method of error analysis, the author chooses the EFL learner's corpus—CLEC (Chinese Learner English Corpus) as the research object, and carries out some quantitative and qualitative analyses of the collocation errors committed by the Chinese EFL learners.Because of language transfer, Chinese EFL learners commit a lot of collocation errors in their language output. There are six kinds of collocation errors in CLEC: noun/noun collocation error (ccl), noun/verb collocation error (cc2), verb/noun collocation error (cc3), adjective/noun collocation error (cc4), verb/adverb collocation error(cc5), and adverb/adjective collocation error (cc6). According to annotations of collocation errors in CLEC, the frequencies of different collocation errors are calculated by the concordance software CONCAPP. And the Chi-square tests are carried out to see whether there is a significant difference in all kinds of the collocation errors between Chinese non-English majors (ST3 and ST4) and between English majors (ST5 and ST6).Each kind of collocation error involves two words which belong to different parts of speech except noun/noun collocation, so the two words of different parts of speech with the highest frequencies are analyzed in each kind of collocation. By searching Brown Corpus and consulting Oxford Collocation Dictionary for Students of English, the researcher tries to find out the differences between Chinese English learners and native speakers in the use of collocations and the causes of such differences.Through the above analysis, the following conclusions are drawn:Firstly, four kinds of the collocation errors committed by CET-4 participants (ST3) outnumber those by CET-6 participants (ST4) except noun/noun collocation errors (ccl) and adjective/noun collocation errors (cc4). The total frequency of collocation errors committed by ST3 outnumbers that by ST4, and the collocation errors committed by ST3 are significantly different from those by ST4. Secondly, all kinds of collocation errors committed by juniors and seniors majoring in English (ST6) outnumber those by freshmen and sophomores (ST5), and collocation errors committed by ST5 are significantly different from those by ST6.Thirdly, most of the Chinese EFL learners' collocation errors are caused by language transfer. And the other causes are induced errors, ignorance of collocation restriction, overgeneralization of collocation, redundancy and misunderstanding of the concept.Some suggestions are put forward for collocation teaching and learning according to the above study. Firstly, students should read extensively, especially authentic English materials. Secondly, students should learn to consult the dictionary when necessary. Thirdly, teachers should change the traditional teaching method and pay more attention on lexical teaching rather than grammar teaching. Fourthly, teachers can use the target language corpus in language teaching. Fifthly, teachers can build a learners' corpus to analyze their interlanguage.