A Macroscopic Re-examination of Language and Gender: A Corpus-Based Case Study in University Instructor Discourses

The need to re-examine the relationship between language and gender situates in the lack of inconclusive and inconsistent empirical findings over decades. Moreover, several common methodological limitations in the empirical literature of language and gender are noticeable, such as overgeneralization from narrow databases, ignorance of other non-gender factors and limited linguistic features investigated. To counter these limitations and re-examine the relationship between language and gender, this study used a corpus of instructor discourses to quantitatively compare 87 general linguistic features employed by male and female university instructors in their classroom discourses. As gender was defined as a dynamic social construct based on the prevailing dynamic approach in the field, this study examined instructor discourses across different gender groups both as a whole and within specific academic disciplines to evaluate how gender effects might vary under different conditions in the context of university classroom. The corpus of instructor discourses was compiled based on the lecture part of the British Academic Spoken English (BASE) corpus. From this corpus, 98 general linguistic features, informed by previous related studies, were automatically extracted by a Python program created in this study, called Linguistic Feature Extractor (LFE) and these features were filtered to 87 with features of extremely low frequency and high correlation removed. Lastly, given the largely non-normal distributions of the dataset, Mann-Whitney U tests were applied to determine whether there are significant differences in the use of the linguistic features across genders. Cohen’s d values were reported for those statistically significant results. The analysis results indicate that male instructors used more WH relative clauses on object position (WHRO), but fewer desire/intent/decision verbs followed by a “to” clause (TDIDV) and mental verbs (MENV) than female instructors. However, there was no discipline-level gender difference discovered. Overall, the significant differences in the use of the three features suggest that female instructor discourses might be slightly less formal and more engaging than those of their male counterparts. However, given the overall small and limited gender differences in instructor discourses, it was concluded that gender is a non-salient factor for university instructors to deliver lectures, a practice that is highly professionalized