Network science and narratives: Basic model and application to victor Hugo’s les Misérables

Abstract

Propelled by the recent advances in digitization of books and computational methods for automated text analysis, we are witnessing a promising opportunity for a serious scientific study of narratives. The importance of such an endeavor stems from the fact that a good story, albeit often fictional and artificial, is composed of highly believable characters who interact and experience a sequence of events together in a realistic world setting, and thus a better understanding of narratives may yield new insights for comprehending various real social phenomena as well as literary fiction. Here we present the basic scientific framework for modeling narrative as complex networks, which allows us to study how the narrative structure is reflected in the network of characters and how they allow us to understand the dynamics of narrative progression. This paper contains the fundamental network model of narratives and its properties that serves as the starting point for a more comprehensive future work.