The Human Versus Artificial Intelligence Translations: A Contrastive Analysis of Ghassan Kanafani's Men in the Sun

Abstract

• 
  

  This study is an attempt to address a gap in research on evaluating human versus AI translations of modern Arabic literary works. To achieve this objective, the study conducts a qualitative, quantitative descriptive-analytical comparison between Hilary Kilpatrick’s published English translation of Ghassan Kanafani's novella Men in the Sun (1963) and those generated by three AI tools: ChatGPT, Gemini, and DeepSeek. Using a corpus of randomly chosen 17 extracts from the novella and guided by Nord's functionalist error typology, the study focuses specifically on evaluating the data translations in terms of fluency, cultural adequacy, and literary devices. The findings reveal that the human translation has proved superior in that it, by means of adaptation, attends to the nuanced pragmatic and cultural aspects in the ST to preserve the original author's intent, a task that all AI models appear to frequently fail to undertake. The AI translations, though often linguistically fluent, exhibit pragmatically and culturally serious errors, tending to misinterpret contextual variables, idioms, and culture-specific references. Among the AI tools, it is DeepSeek that produced a translation with more lexical precision and stylistic alignment. The study concludes, then, that while AI is a powerful translation tool, it currently lacks the components that can efficiently handle deep contextual knowledge essential for literary translation, which reinforces the indispensability of the human translator’s role as a cultural mediator.