Translating Cultural Reference Items In Children’s Literature: A Comparative Study Of Human And Ai Strategies In The Magical World Of Harry Potter
Auteurs : Samir Aynaz . Niyosha Shojaee .
Résumé
This study investigates and compares the translation strategies employed by Artificial Intelligence (AI) and a human translator in translating Cultural Reference (CR) items from English into Persian, with a specific focus on Harry Potter and the Half-Blood Prince novel, a prominent work of children’s fantasy literature. Drawing on Klingberg’s (1986) model for the classification of CRs and Cintas and Remael’s (2014) taxonomy of translation strategies, the research adopts a descriptive-comparative and qualitative methodology. A total of 179 CRs were extracted from the source text and their translations by both a human translator (i.e., Vida Eslamiyeh) and an AI translator (i.e., ChatGPT-4) were analyzed. The findings revealed that the most frequent CR type was Names (37.4%), followed by references related to Food, Furnishings, Mythology, and Flora and Fauna. In terms of translation strategies, the human translator most commonly used Calque (27.9%), Cultural Substitution (20.1%), and Explicitation (17.3%), whereas the AI translator predominantly applied Loan (45.8%), Calque (21.2%), and a hybrid Loan + Explicitation strategy. The statistical analysis using the chi-square test demonstrated a significant difference between the strategies employed by AI and the human translator. These results highlight the different tendencies of the two translators: while the AI translator showed strengths in consistent literal translation, it often lacked the cultural sensitivity and contextual appropriateness necessary for literary translation. Conversely, the human translator demonstrated greater awareness of cultural nuances, opting for strategies that preserved both the ideological and aesthetic dimensions of the original text. This study underscores the continued importance of human translators in culturally rich literary contexts and provides implications for the integration of AI tools in literary translation practices, translator training, and the development of more culturally aware AI systems.
Mots clés
Artificial Intelligence ; ChatGPT-4 ; Cultural References ; Translation Strategies
