Types and Tokens of Lexical Bundles in Civil Engineering Students' Genre-Specific English Articles

The present study sought to find the most frequent lexical bundles, particularly the token of the bundles, in the introduction and discussion sections of civil engineering research articles. it also aimed to examine the forms or types of lexical bundles found in the introduction and discussion sections of the articles. to this end, a quantitative analysis was performed on the use of two-to-five-word lexical bundles, followed by a qualitative analysis of their functions and structures based on the structural taxonomy of lexical bundles proposed by biber, conrad, and cortes (2004). antconc, as a corpus analysis tool, was employed for the extraction of lexical bundles as well as the examination of their types and tokens. likewise, concordancers and word and keyword frequency generators were used for the analysis of word clusters and lexical bundles. a total of 790 lexical bundles in the introduction section and 279 lexical bundles in the discussion section were analyzed based on a set of pre-established criteria (e.g., the frequency cut-off of 20 per million words). the findings revealed that the lexical bundles consisted largely of two-word strings, and the fiveword strings were the fewest. regarding the structural classification of bundles, phrasals were the most frequent ones, whereas clausal ones were the least frequent ones. the study carries important implications, especially for materials developers and course designers who may hopefully benefit from the results in designing eap materials finely tailored to the linguistic needs of civil engineering students.