Corpus-based study on syntactic complexity of texts by L1 and L2 learners

Based on 90 argumentative texts from the corpora of LOCNESS and WECCL, paper attempts to make a study syntactic complexity those texts. All data are processed online by Web-based L2 Syntactic Complexity Analyser, then with corpus tool AntConc IBM SPSS Statistics 20. The findings include: for one thing, L1 learners use more complex sentences verb phrases, embedded clauses, in their writings, compared learners. For another, writing may dependent clauses longer size than despite tend often. Finally, some pedagogical suggestions made how improve course teaching.