Abstract

This paper examines whether artificial intelligence can replace human literary translators by comparing AI-generated translations with James Riordan's canonical English translation of Chingiz Aitmatov's novella "Jamila" (1957). The study focuses on the translation of Kyrgyz realia -culturally bound words that carry deep national connotations and pose significant challenges for cross-cultural transmission.

The research analyzes ten key realia from the text, including chiy (steppe grass used for traditional mats), tandyr (clay oven for baking bread), tulpar (legendary winged horse from the Epic of Manas), and tökmö akyn (poet-improvisator who composes spontaneously). Using V.N. Komissarov's classification of lexical transformations (transcription, transliteration, calquing, modulation, and explication), the study compares Riordan's translation strategies with outputs from contemporary AI models. Primary sources include Aitmatov's original Kyrgyz text, the Russian translation by A. Dmitrieva, and Riordan's English version, alongside AI-generated translations of selected passages.

The analysis reveals that while AI demonstrates impressive competence in conveying denotative meanings and producing fluent target texts, it consistently fails to capture the connotative and cultural layers embedded in Kyrgyz realia. For instance, AI translates tоkmo akyn simply as "poet," losing the crucial aspect of improvisation that Aitmatov deliberately parallels with the spontaneous nature of love in the story. Riordan, despite some inevitable losses, demonstrates deeper cultural understanding by contextualizing these terms within the narrative, occasionally adding explanatory phrases like "the legendary steed" for tulpar.

I argue that literary translation requires not only linguistic competence but also deep cultural immersion and interpretive skills that current AI models lack. The case of "Jamila" a novella that bridges Kyrgyz culture with global readership -provides compelling evidence that while AI can serve as a valuable tool for translators, it cannot replace the human element in conveying the "soul" of a literary work.

This research contributes to two growing scholarly conversations: the ongoing debate about AI's role in translation studies and the broader field of Central Asian literary translation. It challenges the assumption that AI can fully automate creative translation tasks and underscores the enduring value of human cultural expertise in cross-cultural literary transmission.

Keywords: literary translation, artificial intelligence, Chingiz Aitmatov, Kyrgyz realia, Jamila, Central Asian literature