A Corpus-based Account Of The Lexical Use By High School English Learners

The study focuses on researching the English writings written by high school students of China, aiming to offer suggestions to English writing teaching in high school by finding the problems in the writings and analyzing the reasons.To do the research, the author created a writing corpus named English Writing Corpus (EWC). EWC includes four sub-corpora which collect the English writings written by Chinese high school students and some American students based on the same topic. By comparing the writings, the author tries to find the differences in lexical use by Chinese English learners and native English. The author also makes correlation analysis among the factors contributing to the differences including word length, sentence length, and so on.The major findings are: 1) for the Chinese students, the more proficient they are, the closer they are to native speakers of English in the use of English; 2) compared with the writings of native speakers, the writings by Chinese students have a smaller STTR and LD, contain fewer low-frequency words and more shorter sentences, but the average word length of writings by Chinese students is bigger than that of writings by American students; 3) the POS distribution of Chinese students'writings is similar to that of American students writings, but Chinese students can not use function words well, especially conjunctions and prepositions; 4) Chinese students overuse many set phrases, which leads to the overuse of some words, but American students seldom overuse the set phrases; 5) Chinese students are influenced by their mother tongue which can be shown in the use of words, phrases and sentence patterns; 6) Chinese students overuse can and should, and they underuse other modals; 7) in their writings, Chinese students use many lengthy words which are formal as well as many phrases and modals which are informal; 8) correlation analysis shows that the score of Chinese students writing is significantly related to the number of tokens, the number of types, STTR and sentence length, but the correlation between score and word length is not significant.In the end, the author gives some suggestions on English writing teaching based on the major findings of the study.