Machine vs. human translation of political idioms into Arabic: a case study of Hilary Clinton’s book Hard Choices

This study compares translating English political idioms into Arabic by human translation (HT), Google Translate (GT), and ChatGPT (GPT). The corpus consists of one 100 idioms extracted from Hilary Clinton’s book Hard Choices, alongside their counterparts from its Arabic translation. The data is categorized and analyzed in terms of four approaches: ideational, functional, formal, and formal/functional, in addition to their combinations. The findings reveal distinct trends across HT, GT, and GPT. The ideational approach is the most utilized, particularly in HT (42%) and GPT (47%). However, the functional approach is more prominent in HT (24%) than in GPT (12%) and GT (9%). By contrast, the formal approach is most prevalent in GT (45%), followed by GPT (29%) and least in HT (17%). The formal/functional approach, where form and function align, is relatively more frequent in HT (12%) than in GPT (8%) and GT (9%). To conclude, while recent tools like GPT show improvements over traditional ones like GT, human translation remains functionally superior although it still suffers from some unacceptable literal translations.