ChatGPT in Translating Cultural Nuances in O Night by Gibran Khalil Gibran

Abstract

This study investigates the challenges of using ChatGPT in translating the cultural nuances in the O Night poem by Gibran Khalil Gibran from Arabic into English. Then, the study compares ChatGPT’s translation with human translation. The sample consisted of 40 culture-bound expressions that were purposefully selected from the poem. They are analyzed using a mixed-methods approach based on Wang's (2023) framework of the challenges embedded in AI translation tools. The findings reveal that the majority of the culture-bound expressions translated by ChatGPT are inadequate with the percentage (75%). This inadequacy is attributed to the fact that ChatGPT relies on denotative meaning due to its inability to translate the connotative meaning. Another challenge in ChatGPT is manifested in using an inappropriate linguistic style. Besides, its poor logical structure and its inability to capture the cultural nuances require human intervention to post-edit and refine the translation provided by ChatGPT. To overcome these challenges, the study recommends using contextual translation to convey the connotative and expressive meaning of poetic verses.