Abstract：The present corpus-based study systematically examines the differences between human and machine translation products in terms of multidimensional linguistic features, including lexical, syntactic, and discourse-related aspects, with the aim of investigating human intelligence advantages in translation competence and, consequently, identifying new requirements and orientations of translation teaching in the age of artificial intelligence. By employing both quantitative and case analyses, the current study highlights the following three human intelligence advantages in translation competence: (1) accurate bilingual switching through contextual cues and world knowledge, (2) enhanced reader comprehension through the use of simplified language and im- proved text cohesion, and (3) effective pragmatic communication through targeted wordings in specific situations. The findings offer insights for translation teaching in the modern age, particularly in terms of developing translators’ targeted competences to enhance both human translation and AI-supported translation.
Key words: translation competence; human translation; machine translation; corpus
