MACHINE TRANSLATION AND METAPHOR: THE LIMITS OF ARTIFICIAL INTELLIGENCE
Keywords:
Machine translation; metaphor; artificial intelligence; semantic equivalence; cognitive linguistics; Uzbek-English comparison; translation theory; cultural context; neural networks; figurative languageAbstract
This article explores the challenges and limitations of artificial intelligence in translating metaphorical expressions across languages, with a focus on machine translation systems. Metaphors, as culturally and cognitively rich linguistic units, often resist literal interpretation, making their accurate rendering a complex task for AI. Through comparative analysis of Uzbek and English metaphors, the study highlights semantic distortions caused by literal translation and underscores the need for context-aware and culturally sensitive models. The research contributes to the field of computational linguistics and translation studies by identifying gaps in current technologies and proposing directions for improvement.
References
1. Newmark, P. A Textbook of Translation. London: Prentice Hall, 1988. – 292 p.
2. Schäffner, C. Metaphor and Translation: Some Implications of a Cognitive Approach. Journal of Pragmatics, 2004, vol. 36, no. 7, pp. 1253–1269.
3. Guerberof-Arenas, A., Toral, A. The Impact of Machine Translation on Literary Texts: A Study of Metaphor Translation. Translation Spaces, 2022, vol. 11, no. 1, pp. 45–68.
4. Li, Y., Chen, M. Comparing Human and AI Translation of Metaphors: A Neural Perspective. Computational Linguistics, 2025, vol. 51, no. 2, pp. 210–230.
5. Vinall, K., Hellmich, E. Metaphors We Teach By: Language Educators’ Perceptions of Machine Translation. Language Learning & Technology, 2021, vol. 25, no. 3, pp. 1–20.
