Exploring word associations in academic engineering texts

Abstract

Given the importance of lexis in language description, this study attempts to integrate the lexical approach to describe a specialised language for teaching and learning. In addition, this paper demonstrates the use of the correspondence analysis (CA), one of the multivariate techniques, as a useful tool to describe a language. As such, this is a corpus-based study of verbs among academic engineering text types. A larger engineering corpus (E2C) was constructed by combining two specialised corpora, consisting of two text types, namely reference books (RBC) and journal articles (EJC). The Wordsmith 6 program was used to extract 30 key-key-verbs from E2C. The British National Corpus (BNC) was used as the reference corpus. The CA was conducted with these key-key-verbs by computing the frequency values of the verbs generated for each corpus: E2C, RBC, EJC and BNC. The findings include the visual display of the complex inter-relationship of the verbs among the corpora, thus, demonstrating the potential use of the CA as a tool for specialised language description. The empirical observations of the verbs may lead to significant findings on the features of the academic engineering texts types; thus, this study promises more well-informed future investigations into other linguistic features, rhetorical functions, and pedagogical implications involving the academic engineering texts.