A Corpus-assisted Multimodal Discourse Analysis of Malaysian Sports News Discourse: Exploring the Representation of Female and Male Athletes

Abstract

Cases of gender bias which occur in sports discourse have been numerous, as reported by past studies, primarily in the context of ‘Western’ sports news discourse. However, most of this research has been non-linguistic or has paid relatively little attention to sports reporting in other regions. The aim of this thesis is to analyse sports news discourse from Malaysia, with special emphasis on ‘dominant’ discourses, which involve instances of objectifying, trivialising, and stereotyping female athletes. The investigation focuses on the types of bias indicated by differences in terms of the number of sports news stories written about female and male athletes, and also in terms of portrayal – such as instances that emphasize gender-stereotypical images of female and male athletes. The thesis compares news reports about female and male athletes, based on a new 3-million-word corpus of English-language Malaysian online sports news discourse. The corpus was gathered from 7640 articles collected from the sports news section in six English-language Malaysian newspapers. Drawing on a previous study by Caple (2013a), the thesis focuses on analysing three main categories of representation: i) gender and sex, ii) activity and passivity and iii) emotivity, emotion and volition. In order to consider both verbal and visual data, i.e. text and images, the thesis combines linguistic and semiotic analysis, utilising a framework known as corpus-assisted multimodal discourse analysis (CAMDA), which has not yet been fully developed or applied in a large-scale empirical study (Bednarek & Caple, 2014). In so doing, I combine both quantitative and qualitative approaches to analyse the compiled corpus, making use of different corpus linguistic techniques (especially keywords analysis) as well as using various discourse analytical frameworks. The thesis demonstrates how linguistic analysis of a large corpus can be combined with analysis of both words and images of a smaller sub-set of the data.