SEMANTIC TRANSFORMATION AND PRAGMATIC ADAPTATION OF CULTURE-SPECIFIC UNITS IN GENERATIVE AI-BASED ENGLISH–UZBEK LITERARY TRANSLATION

Abstract

The rapid development of generative artificial intelligence has significantly transformed contemporary translation practices, particularly in the processing of literary texts. Nevertheless, the translation of culture-specific units remains one of the most challenging areas for AI-based translation systems because such units encode not only lexical meanings but also historical, social, symbolic, and pragmatic information. This study examines semantic transformation and pragmatic adaptation of culture-specific units in generative AI-based English–Uzbek literary translation. Particular attention is paid to cases in which culturally marked expressions undergo semantic narrowing, generalization, explicitation, substitution, or partial loss during AI-mediated translation. The study argues that generative AI demonstrates considerable contextual sensitivity but may still neutralize culturally significant meanings when direct cross-cultural equivalents are unavailable. Therefore, an effective AI-assisted literary translation model should integrate semantic equivalence with pragmatic and linguocultural adaptation.