Les formules figées dans la traduction des fables d’Ésope du français au turc :

Exemples de Nurullah Ataç et de l’Intelligence Artificielle

Abstract

This study shows that literary translation, particularly the translation of fables, extends well
beyond linguistic concerns and constitutes a deeply cultural endeavour. Through a
comparative analysis of 100 Aesop’s fables translated by Nurullah Ataç and by the
DeepSeek-V3 AI model, we found that the human translator employs significant stylistic
adaptations to align the text with the target culture. These include, notably, the
transformation of indirect speech into direct speech (91 instances) and the incorporation of
traditional Turkish narrative formulas. Such interventions, entirely absent from the AI generated translations, underscore the current limitations of artificial intelligence in
capturing the cultural and stylistic subtleties that are integral to literary translation. While
AI tools offer advantages in terms of speed and efficiency, our findings highlight the
indispensable role of human translators especially for texts demanding intercultural
sensitivity and stylistic nuance that automated systems have yet to emulate.