Eine LLM-Analyse von Stilmitteln in Batty Webers Abreißkalender

This study explores the application of large language models (LLMs) for automated literary analysis in the field of Digital Humanities. The focus is on Batty Weber’s Abreißkalender, a collection of daily columns from the early 20th century that reflect key social and historical themes. We examine how a GPT model can assist a close reading, particularly in identifying and interpreting the theme of war. Methodologically, we follow Underwood’s (2023) approach to analyzing temporal structures in literary texts. Our process combines the extraction of stylistic devices such as metaphors, metonymy, and comparisons with semantic context analysis. The results indicate that LLMs could provide valuable support for computational text analysis and enhance the efficiency of literary research. Finally, we discuss the implications of LLMs for future humanities research.