Beyond Literal Meaning: Neural Machine Translation Constraints in Translating the Poetic Depth of Al-Mutanabbi’s "Tell My Beloved"

Abstract

This study examines the limitations of neural machine translation (NMT) in the context of literary translation, using a comparative analysis of a Google Translate version and a human translation of the renowned Arab poet Al-Mutanabbi's poem "Tell My Beloved." By juxtaposing these two renditions, the research aims to elucidate the challenges that current NMT technology faces in effectively capturing the nuanced expressiveness, cultural subtleties, and aesthetic qualities inherent in literary works. The analysis reveals that while NMT models excel at conveying the literal, denotative meaning of the text, they often struggle to replicate the poetic sensibility, contextual understanding, and interpretive depth required for the translation of seminal works of poetry. The findings underscore the continued importance of human expertise and creative intervention in the translation of literary texts, even as NMT technologies continue to advance. This study contributes to the ongoing academic discourse surrounding the role of artificial intelligence in the field of literary translation and the limitations of machine translation in preserving the richness and integrity of the original work.