Fail Better? The Computational Study of Characters Types in Narratives
Characters are a fundamental element of narrative texts and have been a central focus of computational literary studies (CLS) since the field’s inception. However, the very qualities that make characters so compelling to readers — their vast diversity and complexity — are exactly what make them uniquely challenging for both traditional literary studies and CLS. To date, there remains no historically variable, let alone data-driven, descriptive model for characters, particularly regarding character abstractions.
This talk presents an experimental approach to modeling these abstractions, drawing on research in semantic change. While achieving a degree of generalization appears possible — at least within specific corpora — the resulting representations pose new methodological challenges: how do we interpret them, and how can they be formally validated?
This lecture will also be live streamed via Zoom. You are welcome to join us either in South Hall or online.
For online participants
Online participants must have a Zoom account and be logged in. Sign up for your free account here. If this is your first time using Zoom, please allow a few extra minutes to download and install the browser plugin or mobile app.
Speaker
Fotis Jannidis
Fotis Jannidis is professor of computer philology and modern German literary history at the Julius Maximilian University of Würzburg. His research focuses on the computational analysis of narrative texts, particularly dime novels and serial fiction, as well as poetry. He served as the coordinator of the German Research Foundation (DFG) Priority Programme for Computational Literary Studies from 2020 to 2026, and is currently co-spokesperson for the DFG research unit “Semantic Shifts in Low-Resource Domains.” He received his Ph.D. in German literary studies from LMU Munich in 1996. Further information and a complete list of publications are available at jannidis.de.
