The comparative study of Idioms Translation by Artificial Intelligence and Human Translation in the Everything is Well

This comparative study of idioms in translation is critically important because idioms represent one of the most challenging aspects of translation, serving as linguistic and cultural markers that require sophisticated processing beyond literal meaning. The research systematically compares how Artificial Intelligence (AI) machine translation tools and human translators handle English idioms within the literary text Everything is well. Using a descriptive research design with chi-square statistical analysis via SPSS, the study analyzes translation outputs from both AI systems and human translators. The analysis employs Fernando's (۱۹۹۶) taxonomy of idioms, Baker's (۲۰۱۱) translation strategies, and Nida's (۱۹۶۴) concepts of equivalence as analytical frameworks. Descriptive statistics reveal that semi-idioms constituted the most frequent type (۴۳.۵%) in the source text, followed by literal idioms and pure idioms. Chi-square analysis confirms significant differences in translation strategy preferences between AI and human translators, particularly regarding the use of omission, which human translators employed more frequently. Both groups showed a strong preference for paraphrasing as the dominant strategy, indicating a shared priority on meaning transfer over formal equivalence. The analysis of how translation strategies shaped the use of formal and dynamic equivalence in idiom translation by human and Ai translator, results showed that both human and AI translators shifted between literal and natural renderings. The human translator more consistently achieved dynamic equivalence, especially for culturally nuanced expressions. The AI system produced natural equivalents when familiar patterns existed in its training data but relied on literal forms for less common idioms. These findings have important implications for understanding ۱ www.eltconf.ir the evolving capabilities of AI in literary translation and highlight the continued necessity of human judgment for nuanced cultural and stylistic decisions that statistical approaches alone cannot fully capture.

کلیدواژه ها:

Artificial Intelligence (AI) ، human translation ، Idiom Translation ، Translation Strategies