Analyzing Lithuanian parliamentary discourse on Ukrainian refugees by applying text analysis tools

Abstract

The article examines how members of the Lithuanian Seimas refer to Ukrainian refugees by investigating the linguistic patterns used by Lithuanian politicians in Lithuanian Seimas’ parliamentary sessions. The research applies the Natural Language Toolkit (NLTK), a Python library commonly employed in computational linguistics, and LitLatBERT, complemented by manual analysis. The study explores how Lithuanian politicians frame Ukrainian refugees linguistically by focusing on lexical aspects, mainly definitions of and references to refugees and modality. The article’s broader aim is to showcase how NLTK libraries and the transformer-based deep learning model LitLatBERT can be combined with methods of Critical Discourse Analysis (CDA) to analyze Lithuanian parliamentary discourse.