AI vs. Human: Translating Cultural Nuances in Naguib Mahfouz's The Cairo Trilogy

This study investigates the performance of the AI translation system ChatGPT in comparison with human translators in rendering culturally embedded expressions from Naguib Mahfouz’s Cairo Trilogy into English. Cultural expressions – particularly idioms, metaphors, and religious references – present major challenges to translators, who must interpret meaning not only at the linguistic level but also at the pragmatic, emotional, and socio-cultural levels. Adopting a qualitative content analysis approach, the study conducts a comparison of AI-generated and professionally published human translations. The evaluation employs Baker’s equivalence-based framework, assessing semantic accuracy, pragmatic equivalence, cultural adequacy, naturalness, and functional alignment. Through this analytical lens, the study explores how effectively each translator – human or AI – handles the interpretive demands of culturally and pragmatically rich expressions. The findings reveal that while ChatGPT performs well in producing fluent, structurally coherent translations, it frequently struggles with figurative language and culturally specific nuances. These limitations stem from the absence of embodied cultural knowledge and self-evaluative judgement, which are essential for interpreting implicit meanings and cultural resonance. Human translators demonstrate stronger interpretive and cultural sensitivity, particularly in challenging areas such as metaphorical imagery and religious references. The study concludes that although AI systems increasingly contribute to efficient and accessible translation workflows, culturally intricate texts like the Cairo Trilogy still require human expertise. A hybrid human – AI translation model, therefore, emerges as the most effective approach for balancing efficiency with cultural and pragmatic depth.