A Corpus-based Error Analysis On Nonnative English Majors' Oral Test Production

Error analysis (EA) as a branch of Applied Linguistics undoubtedly has made much contribution to the study on second language and foreign language learning. It not only throws some light on the process of EFL teaching and learning, but also helps us gain some insights into the strategies the learners employ in the learning process. A careful study of the errors usually reveals the common characteristics shared by the language learners, which is believed to be of great help to both the language teachers and the language learners. The domestic error analyses of spoken English mainly concentrate on middle school students' spoken English and non-English major's spoken English. Little has been done to the English major's spoken English, especially to expression produced in TEM-8 oral tests. The author thinks that students who take part in spoken test of TEM-8 have a comparatively high language proficiency level in China, and therefore, errors in the production of their speech can represent the characteristics of speaking proficiency of Chinese advanced learners. Therefore, it is necessary to do a systematic study on their oral English.With the help of Error Analysis and Spoken and Written English Corpus of Chinese Learners 2.0 (SWECCL 2.0) which was designed by professor Wen Qiufang, the author tries to find out the main error types of Chinese advanced English majors' spoken English. The procedures are as follows:First, randomly choose 8 records of high level students,8 records of low level students and 24 records of middle level students from each year (2003-2007). Totally 200 test-takers' audio records and their corresponding transcribed texts are chosen as the sample.; Second, find out all the errors on the basis of international phonetic alphabet, grammatical rules of English and A Dictionary of Common Errors in English. Third, classify all the errors according to its linguistic category. Fourth, use AntConc to count the total number of each error type and then use SPSS to do a comparative analysis between error types and students' scores.After above analysis, the author reaches the conclusion that the errors of speech production of TEM-8 oral test takers mainly occur at phonological level, grammatical level, lexical level and discourse level. Among these errors, phonological errors are the most distinctive, and second is the grammatical errors, then discourse errors and last lexical errors. The comparative analysis of students with high oral English scores and students with low oral English scores shows that errors made by excellent students in spoken English are significantly different form those of weak students, particularly in the phonological errors, lexical errors and discourse errors.