"The beginning of the war" in the reports of Richard Sorge: background meanings and the voice of a scout (experience in digital analysis)

Abstract

This article presents the results of a big data analysis on the prediction of the start date of the Great Patriotic War, presented in Richard Zorge's intelligence reports from 1931 to 1941. The article utilizes a multidisciplinary research approach: military-historical texts are analyzed using computational linguistic methods for big data text processing. Methodologically, this study employs topic modelling, which reveals the general distribution of topics within Zorge's correspondence, as well as distributional semantic analysis, specifically tracking the chronological frequency of bigrams within the corpus of the texts. This quantitative approach contextualizes the thematic evolution of the “German theme” in Zorge’s reports over time, mapping diachronic shifts in conceptual associations and semantic density. The study also examines Zorge's speech features as his “voice”, as revealed in the texts of his reports, to explicate the linguistic degree of his subjectivity, hidden at the background behind the standardized style of intelligence reports.