LANGUAGE PLAY IN UKRAINIAN POETRY: TRANSLATIONAL CHALLENGES AND AI POTENTIAL

Abstract

The article presents a comprehensive analysis of language play in contemporary Ukrainian poetry and examines the specifics of its rendering into English from a translation studies perspective. It focuses on lexical, phonetic, and morphological types of language play that are typical for the individual poetic style of Halyna Kruk, represented in her collection A Crash Course in Molotov Cocktails. Language play is considered as a multidimensional phenomenon that performs not only an aesthetic but also cognitive-emotional and discursive functions, enhancing irony, expressiveness, and the traumatic sensitivity of poetic expression shaped by the experience of war. The research material comprises 87 instances of language play and their English translations. The methodology combines linguo-stylistic, semantic, and translation analysis with quantitative assessment. The study determines the percentage of preserved instances of language play in translations made by a human and demonstrates that lexical wordplay is rendered most successfully, while phonetic and morphological forms are significant losses. It is argued that the primary factors leading to the reduction of language play include the absence of polysemous equivalents in the target language, cultural discrepancies, and the inherent difficulty of reproducing homophonic effects Special attention is paid to evaluating the potential of artificial intelligence (GPT-5 mini) in translating complex cases of language play. The analysis shows that the model is capable of compensating for stylistic losses by activating English idioms, imitating parallel structures, and exploiting the internal motivation of words, thus preserving the pragmatic and stylistic density of the poetic utterance. The findings suggest that artificial intelligence can serve as an effective auxiliary tool in the translation of stylistically complex contemporary poetry, particularly in cases requiring creative compensation strategies. The results of the study contribute to the ongoing discussion on human–machine collaboration in literary translation and highlight the role of artificial intelligence in expanding the range of available translation solutions. Furthermore, the proposed approach may be applied to the analysis and translation of other genres characterized by a high degree of linguistic creativity.