Lexical and Corpus-Based Approaches to the Analysis of the Concept of 'Mind': A Comparison of Semantic and Frequency Models Based on the Image of Chatsky in A.S. Griboyedov’s Comedy Woe from Wit

Abstract

The core aim of this research is threefold: (1) to compare the findings derived from
traditional lexical/semantic analysis and a modern corpus-statistical analysis of the concept
in Chatsky’s speech; (2) to identify key points of convergence and divergence between these
two distinct methodological approaches; and (3) to determine empirically how
computational linguistics and Natural Language Processing (NLP) techniques can
systematically augment and refine traditional methods of conceptual research in literary
studies. The research methodology is structured in two sequential stages of analysis. Stage
One involves a lexical analysis of the concept of ум (mind), focusing on dictionary
definitions and establishing its semantic structure within Chatsky’s speech. Stage Two
comprises a rigorous corpus-statistical analysis employing NLP technologies. Using the
NLTK (Natural Language Toolkit) library, the analysis includes automatic extraction of all
relevant word forms, followed by tokenization and lemmatization. Key metrics utilized are
frequency analysis (with grouping by semantic categories), collocational analysis using
bigram statistics to determine stable word combinations, and concordance analysis to
examine the contextual environments of the concept. The primary finding demonstrates that
the corpus-statistical model reveals previously latent semantic dominants and associative
links, which are often obscured by qualitative reading alone. The study concludes that the
objective, frequency-based analysis offers a robust mechanism for uncovering hidden
patterns of conceptual functioning within dramatic discourse, ultimately providing a more
nuanced and empirical basis for the analysis of Chatsky's much-debated intellectual status.