UTILISING CORPUS STYLISTICS TO FACILITATE LITERARY ANALYSIS: AN ASSESSMENT OF THE EFFECTIVENESS OF SEMANTIC DOMAINS IN IDENTIFYING MAJOR LITERARY THEMES IN A SELECTION OF CHARLES DICKENS’ NOVELS

Abstract

Though still in an early stage of development, corpus-assisted literary analysis is becoming increasingly popular as having the full potential of corpus linguistics methodology for literary stylistics. This paper argues that corpus linguistic procedures can be considered an addition to the analytical inventory of traditional stylistics. It aims to explore how corpus linguistic procedures, particularly semantic domains, can be effective in detecting major literary themes in fiction. In order to do so, five corpora were compiled: a corpus for each of the four novels of Charles Dickens’ selected (i.e., Oliver Twist, David Copperfield, Great Expectations and Our Mutual Friend) and a compiled corpus combining all four novels. Wmatrix 5, with the BNC Sampler-Written as a reference corpus, was used to extract the key semantic domains in each corpus respectively. The literature on the selected novels was consulted to identify the major themes. Then, it was verified whether these themes were reflected in the corpus analysis, and, finally, the extent to which the procedure was effective in reflecting the major literary themes was also explored. The findings confirmed the effectiveness of the procedure of analysing semantic domains in studying literary texts, particularly in relation to their themes.