Abstract

This study conducts a comparative analysis of human and AI translations of Arabic dialect poetry into English, focusing on Gulf, Levantine, and Egyptian Arabic dialects. The study reveals distinct strengths and limitations by evaluating both translation approaches across five critical criteria- form, word usage, poetic techniques, language conventions, and illustrations. Human translators excel in preserving the nuanced structure, rhythm, and cultural depth that are essential in poetry. For example, in the translation of Prince Khalid AlFaisal’s Gulf Arabic poem, ((ما هقيت ‘I Thought Not’, human translators maintained the emotional rhythm and poetic structure that AI translations rendered in a more prose-like format. Additionally, human translators demonstrated greater sensitivity to culturally embedded words and idioms, as seen in their handling of idiomatic expressions and symbolic metaphors in Talal Haidar’s Levantine Arabic poem ‘Alone’ (وحدن) and Ahmad Fouad Najem’s Egyptian Arabic poem Alexandria (إسكندرية). While often accurate in literal terms, the AI translations struggled with interpreting poetic devices like metaphor and alliteration and lacked the interpretive depth needed for conveying cultural and emotional nuances. Nonetheless, AI tools show potential as initial translators for generating draft translations that human experts can refine. The study concludes that, in the face of AI advances, human translators remain indispensable for capturing Arabic dialect poetry's aesthetic, cultural, and emotional layers. This underscores the need for a collaborative model where AI serves as an assistant rather than a replacement for human expertise in literary translation. This research sheds light on AI's role and limitations in the nuanced domain of poetry translation, thereby contributing to the field.

Statistics