‘Discourse reading’ predication and gender: putting together and discourse analysis

Abstract

This paper proposes a methodological synergy between corpus linguistics and discourse studies. We use a literary corpus composed of more than 200 books of Brazilian literature (about 5 million words) and present a quantitative and qualitative analysis that takes the gender element as an analytical operator, selecting one of its dimensions: characterizations attributed to female and male characters. Searching for linguistic structures indicative of predication, we distinguished four axes of analysis (appearance, emotion, character e social role), expanding a type of research that is normally restricted to case studies. Using computational tools, we switch the glasses that bring the text to a close or a distant reading, making the best of each of these perspectives. The results reinforce the idea of female characterization centered on appearance (specifically on beauty), but also signal the importance of taking into account not only the high frequency elements of the corpus.

Keywords:
Digital Humanities; Corpus linguistics; Text mining; Discourse analysis; Gender studies