A Comparative Study of Lexical Bundles in the Abstracts of Chinese Master’s Theses and Research Articles by English Expert Writers

This study compares the use of lexical bundles in the abstracts of Chinese master’s theses and those of research articles written by English expert writers. Two small corpora were constructed: a corpus of master’s thesis abstracts and a corpus of expert research article abstracts. Based on Hyland’s (2008b) functional classification framework, four-word lexical bundles were extracted and functionally annotated using AntConc. The findings show that: (1) the normalized frequency of lexical bundles in the expert corpus is significantly higher than that in the master’s corpus, indicating a greater degree of formulaic density among expert writers; (2) in terms of functional distribution, text-oriented bundles account for the highest proportion in the expert corpus (45%), reflecting stronger discourse organizational awareness, while research-oriented bundles account for the highest proportion in the master’s corpus (42%), reflecting a focus on research background and methods; (3) master’s thesis abstracts exhibit certain weaknesses, including a preference for verb-based bundles and vague quantifiers, overuse of certainty verbs, avoidance of first-person pronouns, and formulaic use of bundle patterns. This study provides empirical evidence for the teaching of academic English writing.