A Corpus-based Study Of Chinese EFL Learners' Verb+Noun Collocational Errors In English Writing

This descriptive study is designed to explore the characteristics of Chinese EFL learners'V+N collocational errors based on CLEC, to investigate the verbs and nouns that cause most problems to the learners'V+N collocations, and to attempt to provide some tentative explanations for these errors. A total of 1,566 V+N collocational errors in CLEC are examined and descriptive classification of these errors is made by adopting a combined taxonomy which divides these errors into 14 subcategories, followed by a diagnosis-based classification into 13 subcategories. The number of each subcategory of V+N collocational errors is counted and normalized. Analyses are carried out in relation to the learners'L2 proficiency and writing condition differences. Comparison is employed as the principal method of data analysis and the data of both descriptive and diagnosis-based categorizations are first analyzed with regard to L2 proficiency differences, followed by the analysis of the data with regard to writing condition differences. After that, close examinations on the wrongly used verbs and nouns in the learners'V+N collocations are conducted. Through analyzing the results, we have the following findings:1) No specific pattern emerges in Chinese EFL learners'V+N collocational errors. However, all learners show a heavy reliance on their mother tongue.2) Learners'V+N collocational errors display different characteristics with regard to their L2 proficiency differences. The result of ST2, ST5 and ST6 learners'V+N collocational errors turns out to be somewhat surprising: most subcategories of these errors in ST6 outnumber those in ST5 and some even outnumber those in ST2. In ST3 and ST4, however, the number of each subtype of these errors is negatively correlated with the learners'L2 proficiency levels in most cases.3) Different characteristics of learners'V+N collocational errors arise due to the writing condition differences. The result of both descriptive and diagnosis-based classification shows that most subcategories in examination writing outnumber those in free writing. Learners in examination writing commit more Misanalysis, Incomplete/Misunderstanding of the Concept errors and tend to rely more heavily on their mother tongue.