EUPHEMISTIC STRATEGIES IN THE LEXICO-DISCURSIVE CONSTRUCTION OF VIOLENCE: A CRITICAL DISCOURSE ANALYSIS OF AI-MEDIATED WAR NARRATIVES

The paper focuses on the role of euphemistic language in the lexico-discursive construction of violence in AI-mediated war discourse. The research examines how lexical choice, syntactic structure, and discursive framing transform violence into sanitized, abstract, and institutionally acceptable forms. The proposed study is situated at the crossroads of Critical Discourse Analysis (CDA), media linguistics, and the emerging research on algorithmically generated discourse. Precision strike, neutralize, engagement, and collateral damage are not simply expressions used in place of another in the war discourse; they regulate affect, background agency, and transform bodily harm into technical procedure. Although the use of euphemism in political and media speech has been well-researched, its functionality in AI-mediated narratives is relatively under-researched. The study is based on the qualitative CDA framework guided by the three-dimensional model of Fairclough and the socio-cognitive approach of van Dijk. Moreover, the current research examines AI-generated and AI-mediated war-related discourse to detect the repetitive patterns of lexical mitigation, nominalization, passivization, abstraction, and neutralizing frame construction. The analysis reveals that AI systems often recreate institutional repertoires of war language instead of producing ideologically empty descriptions. Their outputs often appear neutral, balanced, and informative, yet this neutrality is itself discursively produced through linguistic strategies. Such linguistic strategies suppress asymmetry, diffuse responsibility, and normalize violence. The paper thus argues that AI is not only a technical medium of transmission but also a discursive agent that exists within inherited systems of representation. The article extends the field of euphemism study to the area of AI discourse, thus, making a contribution to the current field of linguistics and emphasizing the necessity of critical examination of the algorithmically mediated representation of conflict.

Перевод пока недоступен