A Corpus-based Study Of Verb-noun Collocation Errors In Chinese College English Learners' Test Writings

Collocation is a universal linguistic phenomenon and it is very important for the achievement of native-like competence. It has been recognized as one of the ways that differentiate native speakers from foreign language learners. Previous studies indicate that collocation errors have been one major type of lexical errors. Error Analysis is one of the primary methods used to investigate learner language, i.e. interlanguage. Presently with the development of computer technology and the emergence of enormous corpora and well-designed concordance softwares, corpus-based error analysis has become an important approach to second language acquisition research within applied linguistics.In 2003, the largest learner corpus in mainland China, Chinese Learner English Corpus (CLEC) was published, which is of great significance to corpus-based linguistic research in China. Based on this error-tagged corpus and the theories of Interlanguage and Error Analysis, the present study focuses on investigating the verb-noun collocation errors in the compositions of Chinese College English Learners under examination writing condition. It aims to find the problems and difficulties of Chinese learners in the process of learning English verb-noun collocations.Based on the observation, analysis and classification of verb-noun collocation errors retrieved from test writings of two sub-corpora of ST3 and ST4 in CLEC, eight categories of verb-noun collocation errors are explored in this study with the help of concordance software, MicroConcord. These eight categories of verb-noun collocation errors are defined as: non-matching of the verb and the second noun in the coordinated phrases, overgeneralization and coinage according to previous knowledge, extraordinary matching of the verb and noun, arbitrary selection of synonymous verbs, overuse of delexical verbs, literal translation, arbitrary selection of synonymous nouns and others. From the result of independent sample t-test, it is indicated that there is no statistically significant difference in the competence of verb-noun collocation between learners with different English proficiency. While the results of the frequency distribution show that learners often make the following three typical categories of errors in their writings: arbitrary selection of synonymous verbs, overuse of delexical verbs and literal translation. Furthermore, there are some changes over the distribution of these three typical categories of verb-noun collocation errors between the learners with different English proficiency. Then independent sample t-test of these three typical errors is carried out. The result shows that significant differences exist in the categories of arbitrary selection of synonymous verbs and literal translation, but no significant difference in overuse of delexical verbs. Finally the underlying causes have been explored from different perspectives.Based on the investigation, this study provides some implications for English language teaching and learning in China, including emphasis on the quantity and quality of the input of authentic language, the introduction of learners corpora and native speakers corpora into language classroom, the rejection of electronic dictionary and so on.