Corpus-based comparison of lexical complexity in L1 and L2 postgraduate academic writing

This study adopted a corpus-based, contrastive approach to lexical complexity in the academic writing of first language (L1) and second language (L2) postgraduates. Lexical complexity scores were extracted using the Lexical Complexity Analyzer from the Corpus of Arab Proficient Users of English (CAPUE), consisting of Saudi academics’ dissertations in applied linguistics. To investigate the potential differences between this corpus and native speakers’ corpus, the lexical complexity of writing material from the CAPUE and Corpus of English Native Speakers (CENS) were compared. The computational system employed 25 lexical complexity measures to investigate differences in the two groups’ lexical density, sophistication, and variation. The results revealed similar lexical density in the writing of both groups; however, the texts by L1 researchers were more lexically complex for most measures of sophistication and variation. The results have implications for teaching English for academic purposes and highlights areas with inappropriate lexical choices. These findings call for the design of pedagogical interventions to enhance the lexical complexity development of L2 postgraduates.