Digital Text Analysis Working Group
Dr. J. Berenike Herrman, "Pinning Down Franz Kafka - Comparative Style Analysis Combining two Quantitative Measures"
Meeting of the GCDH Digital Text Analysis Working Group, Heyne Haus Seminarraum 2
In this presentation, I will report on the results of quantitative analyses of Franz Kafka's literary style using two measures of digital humanities research - keyword analysis and stylometric cluster analysis. I compared a substantial part Kafka's literary prose (~374,000 words / ~50 texts) to a big corpus of German literature (~134,000,000 words / ~93,000 texts), which allowed (a) examining prominent lexico-grammatical features of his writing, as well as (b) approaching the long-held assumption of stylistic singularity in stylometric terms. With regard to the linguistic features, keyword results point towards a prolific use of ambiguous and inconspicuous adverbs that allow undermining the stability of reality in the fictional world when applied across distinct local co-texts. As for the clustering of Kafka's texts among German literature, surprising stylometric results suggesting stylistic similarity with authors of 19th/early 20th Century children's literature such as Johanna Spyri ("Heidi") and Agnes Sapper ("Familie Pfäffling") will be addressed from both content and methodological angles.