Ismail almazaidah, Areej Allawzi, Haneen Amireh, Mohannad Alkhalaileh
ABSTRACT
This study offers a comparative analysis of three Arabic translations of selected passages from T. S. Eliot’s The Waste Land, examining human translation (Abdul Wahid Lu’lu’a), rule-based machine translation (Google Translate), and AI-generated translation (ChatGPT). The research aims to evaluate each modality’s ability to preserve linguistic accuracy, poetic aesthetics, symbolic meaning, and cultural engagement. Methodologically, the study adopts a qualitative stylistic framework informed by translation theorists such as Nida and Taber, Halliday, Toury, and Venuti. Five sections of the poem were selected for close comparative analysis. Findings indicate that Lu’lu’a’s translation surpasses the others in maintaining Eliot’s intertextual richness and poetic rhythm. ChatGPT, while showing notable fluency and partial symbolic awareness, often lacks interpretive depth. Google Translate consistently underperforms due to literalism and syntactic inconsistency. The study concludes that human translation remains superior in capturing the cultural and philosophical density of modernist poetry. However, AI tools may serve as useful drafting aids in literary translation when used under expert supervision.This research contributes to evolving debates in AI and literary translation by highlighting the limits and possibilities of neural models when tasked with complex poetic texts.
