A Preliminary Study of Lexical Choices by Academic Disciplines in Student Writing

The tools of statistical natural language processing have the potential to reveal interesting insights on the use of language. Though we often try to be deliberate in our use of language, it is not often clear exactly why it is that we make the linguistic choices that we do. At times, patterns of usage may emerge that could potentially be linked to the language we have been exposed to in our communities. This study attempts to explore this issue by highlighting common and distinctive patterns of usage within academic disciplines characterised by papers written by students in a research university. This paper outlines the corpus and the methods that we employ to tease these structures apart.