This article provides an assessment of AI-assisted translation of appliance manuals from English to Arabic, with a particular focus on instruction manuals of Moulinex Company products. It aims to identify different levels of linguistic inaccuracies that appear in ChatGPT-4o’s translation of these texts and demonstrates how these inaccuracies cause a loss in the communicative function of these texts. In light of this assessment, the article highlights the importance of adopting supportive human translations to maintain the pragmatic functions of technical texts by paying enough attention to their distinctive linguistic and textual features. Method: To achieve its objectives, the article adopts a descriptive-analytical approach through which the data are classified into two main categories: lexical and textual. The assessment of these levels is conducted with reference to Nida’s theoretical model of functional equivalence. Results: The research reveals that AI generates problematic versions of the translated data from English to Arabic, which stems from undermining the technical context at the lexical level and disregarding the Arabic preferences of cohesion and coherence at the textual level. Conclusion: Based on these findings, the researchers recommend adopting a hybrid human-AI approach to translate appliance instruction manuals.

Title

An-Najah University Journal for Research - B (Humanities)

Publisher

An-Najah National University

Publisher Country

Palestine

Publication Type

Prtinted only

Volume

--

Year

--

Pages

--