Operationalizing Canonicity: A Quantitative Study of French 19th and 20th Century Literature

Abstract

This article delves into the literary canon, a concept shaped by social biases and influenced by successive receptions. The canonization process is a multifaceted phenomenon, emerging from the intricate interplay of sociological, economic, and political factors. Our objective is to detect the underlying textual dynamics that grant certain works exceptional longevity while jeopardizing the transmission of the majority. Drawing on various criteria, we present an operational framework for defining the French literary canon, centered on its contemporary reception and emphasizing the role of institutions, particularly schools, in its formation. Leveraging natural language processing and machine learning techniques, we unveil an intrinsic norm inherent to the literary canon. Through statistical modeling, we achieve predictive outcomes with accuracy ranging from 70% to 74%, contingent on the chosen scale of canonicity. We believe that these findings detect what Charles Altieri calls a “cultural grammar”, referring to the idea that canonical works in literature serve as foundational texts that shape the norms, values, and conventions of a particular cultural tradition. We posit that this linguistic norm arises from biased latent selection mechanisms linked to the role of the educational system in the canon-formation process.