Bayesian Narrative: Probability, Plot and the Shape of the Fictional World

Abstract

‘Probability’ seems to be a term forgotten by literary theory. Central to neoclassical and Augustan criticism, probability describes the inferences of readers and their developing discernment of what is likely to happen in a narrative (Patey 1984). This article proposes to bring probability back into the current debates in narratology and literary theory by drawing on recent advances in probabilistic, Bayesian approaches to different aspects of human cognition. Considering the example of Frances Burney’s novel Evelina (1778), it presents a Bayesian model for the analysis of narrative through the ways in which the encounter with the text shapes readers’ probability judgements. A narrative’s ‘probability design’ cues readers to revise or maintain their expectations for its further development and leads readers to accept outcomes as inevitable that seemed distinctly unlikely at the beginning of the narrative (such as Evelina’s brilliant marriage to the aristocrat Lord Orville in Burney’s novel). Reconsidering narrative from a Bayesian, probabilistic point of view offers new perspectives on the emotional investments of readers in narrative, as well as plot and verisimilitude.