Modéliser une tradition de fables ésopiques : architecture TEI, taxonomies et structuration des connaissances dans l’édition numérique de l’Isopet 1-Avionnet

This article examines knowledge modeling within the framework of digital scholarly editing, drawing on the corpus of medieval fables from the Isopet 1-Avionnet as part of the FabuLab project. It proposes a stratified TEI architecture that distinguishes between a conceptual level - comprising a master TEI file that consolidates taxonomies and reference frameworks - and a documentary level corresponding to the encoding of manuscript witnesses. This structure makes it possible to separate the definition of categories from their use in texts, while ensuring their articulation through mechanisms of reference and annotation. The study shows that this organization promotes consistency in annotations, reduces redundancy, and enables corpus exploration based on analytical categories. The article thus advocates a conception of digital scholarly editing as a space for the formalization of knowledge, in which analytical categories are integrated into the data model.