<![CDATA[The translation of Al-Mutanabbi’s poetry presents significant challenges due to its linguistic complexity and cultural depth, often exceeding the capabilities of neural machine translation. This study analyzes the comparative application of Peter Newmark’s translation procedures between a human translator (Hadijah Rima) and DeepL Pro in translating selected classical poems. Employing a qualitative descriptive-comparative method, the research identifies and quantifies translation strategies across 385 data units. The findings reveal a sharp contrast in approach: the human translator predominantly relies on Adaptation (36.1%) and Semantic Translation to preserve aesthetic value and contextual meaning, whereas DeepL Pro is dominated by Literal Translation (90.9%). This over-reliance on literalism leads to significant semantic distortions and "hallucinations" in AI, exemplified by the fatal error of translating the metaphor 'Al-Bidh' (Swords) into "Eggs," thereby stripping the text of its intended nuance. The study concludes that while AI offers speed, it ultimately fails to render expressive texts accurately due to a lack of cognitive flexibility. Consequently, human post-editing remains essential to ensure cultural and semantic integrity in literary translation.]]>
Copyrights © 2026
