| Through the way of genre analysis, this paper analyses the generic structures and language features of research article (RA) English abstracts. In order to obtain a convincing conclusion, the author builds a corpus which is a random sample of50RA abstracts in linguistics of Chinese writers from CNKI and of another50abstracts in linguistics of native English writers from SSCI. Using Bhatia’s IMRC model and Swales’CARS model, the author analyses the generic structures and hedging, tense, voice and first person pronoun between Chinese and English scholars, analyzing their similarities and differences and finding the reasons of these differences.The results show that the generic structures of native English speakers’abstracts are basically accordance with IMRC model, while Chinese learners’are often IR model. In the two corpuses, the frequencies of "introduction" and "result" are nearly100%, so the author takes them as obligatory moves. The great difference is the lack of "method" and "conclusion" in the abstracts of Chinese scholars’papers. In language features, the use of hedging, tense and voice did not make many differences between Chinese and English scholars. But the frequency of first person pronouns on the two corpora is low, and relative to the Chinese scholars, native English scholars use it more frequently.Based on the results and discussion, the present study make a tentative attempt to offer a pattern for the generic structure of English thesis abstracts in linguistics for reference.The author hopes to help Chinese RA writers and students better understand abstract structures, language features and function and gain better skills about reading and writing RA abstracts. |