Science Fiction versus Fantasy: A Semantic Categorization and its Contribution to Distinguishing Two Literary Genres

Abstract

In this study, we examine the methodological issues that arise in the semantic and stylistic classification of recurrent narrative patterns, with particular attention paid to “big data” analysis based on large corpora of text. We discuss the semantic approach adopted in the PhraseoRom project to investigate recurrent lexico-syntactic trees and show how effective this semantic annotation is from a literary perspective when applied to distinguishing two fictional genres, science fiction and fantasy (in French and in English), from each other. We further demonstrate the efficacy of this approach in letting researchers define and delimit particular fictional subgenres. Since the genres in question bear certain similarities, they offer a good test for the categories used in the semantic classification. Lastly, we show that our approach allows identifying generic profiles characterized by particularly striking phraseology and that it provides interesting elements for a contrastive analysis.