Anthology of Computers and the Humanities · Volume 3

Patterns of Canon: A Multilingual Network Study

Judith Brottrager1 ORCID , Jean Barré2 ORCID , Yuri Bizzoni3 ORCID and Pascale Feldkamp Moreira3 ORCID

  • 1 Technical University of Darmstadt, Darmstadt, Germany
  • 2 LaTTiCe-CNRS, École Normale Supérieure - PSL University, Paris, France
  • 3 Center for Humanities Computing, Aarhus University, Aarhus, Denmark
Download PDF Download Citation

Permanent Link: https://doi.org/10.63744/Y0dv4ooACREY

Published: 21 November 2025

Keywords: canonicity, network analysis, computational literary studies, multilingual text analysis, multilingual stylistics, multilingual stylometry

Abstract

This study examines whether canonical literature exhibits consistent structural signatures in networks of textual similarity across languages. Using four diachronic corpora of prose fiction (5,000 texts, 17th–21st century), we construct time sensitive similarity networks, with edges weighted by textual proximity. Canonical status—derived from multiple markers of canonization—is analyzed in relation to various centrality measures. We find that global metrics such as betweenness and degree centrality show limited association with canonization, while sub-cluster centrality offers a more reliable signal. Our analysis also reveals temporal shifts: in some periods, canonical status aligns with structural centrality, while in others the correlation weakens or even reverses, with canonization favoring more distinctive or peripheral works.