re-late – re-use – re-vise. Large Language Models and Their Potential for Variant and Intertextuality Studies in the Digital Humanities

The development of AI-based methods in recent years has also caused a veritable revolution in the digital humanities; the consequences and course of which are unclear and unpredictable. This applies not least to the field of digital editing and computational literary studies, where large language models (LLMs) in particular are the subject of intense debate, without any common denominator of best practices emerging as yet. The aim of our workshop is to explore the potential applications of large language models (LLMs) in the field of digital editing and computational literary analysis, using a very specific classic use case from these disciplines, namely in the field of recognition, visualization, and analysis of text variants and intertextuality, in order to link methodological innovations to questions relevant to the humanities. LLMs open up new ways of identifying semantically and contextually similar text passages in extensive corpora—even across language and edition boundaries. Such models can be used effectively, particularly in philological work with variants of tradition, adaptations, and intertextual references, and can also be used complementary to existing qualitative and quantitative methods.

The workshop is aimed at researchers and students from the fields of literary studies, editing, digital humanities, and related disciplines. In addition to methodological keynote talks, the focus will be on practical application examples in which the use of LLMs will be systematically discussed, methodologically reflected upon, and tested in an application-oriented manner.