The Detective and Sensation Fiction of Wilkie Collins: A Computational Lexical-Semantic Analysis
Keywords: computational lexical-semantics; detective fiction; genre classification; sensation fiction; theme analysis; vector space clustering (VSC); Wilkie Collins
DOI: fiction; theme analysis; vector space clustering (VSC); Wilkie Collins
Theme and genre classifications in the works of Wilkie Collins (1824-89) have been extensively investigated using different literary approaches; these are usually based on textual content and biographical considerations. Different critics place Collins’ works under the two main headings of detective fiction and sensation fiction. Such analyses have been generated by what is referred to as the ‘philological method’; that is, by an individual critic’s reading of the relevant material and their intuitive abstraction of generalizations from that reading. A problem with such an approach is that it is not objective, and it is therefore unreliable. The research question is thus asked in response to the subjectivity of previous genre classifications of the novels of Wilkie Collins and the lack of agreement among literary critics and researchers about such classifications. As such, I ask whether an objective and conceptually useful reading of the themes and subjects of Wilkie Collins’ prose fiction texts can be developed. As thus, computational lexical-semantics is suggested to understand the issues of thematic classification. For this purpose, vector space clustering (VSC) was used for capturing the lexical-semantic features of his novels and linking them explicitly to the relevant themes and genres. It is suggested that through this method, an objective, replicable, and reliable genre classification of Collins’ novels is possible. The results of this study can serve as a basis for future studies and criticisms of Wilkie Collins’ fiction.