The Evolution of Translation Practices: Comparing Traditional and AI-Assisted Techniques in Student Texts

Abstract

Artificial intelligence has revolutionized translation, making it more efficient and consistent. Yet, human oversight is still necessary to ensure accuracy, clarity, and cultural appropriateness. The present study examines the development of translation practices by comparing the use of Molina & Albir’s translation techniques in traditional and AI-assisted student translations. The goal is to discover application patterns of these techniques, and to evaluate the effects of AI on translation strategies and results. The study applied a qualitative participatory research design involving ten students of English Literature from Universitas Bumigora in the sixth semester. Data collection was carried out by means of document analysis of translation assignments produced under two conditions – traditional translations and AI-assisted translations with ChatGPT, Google Translate and DeepL. For each translation, techniques applied were classified, their frequency and functional impact were analyzed. The findings show that Literal Translation (36.16%) and Modulation (35.03%) were the most frequently used among the conventional translations, which together accounted for more than 71% of the observed strategies. Other techniques like Reduction, Established Equivalent, Amplification, Transposition, Borrowing and Adaptation were also applied selectively. Literal Translation (23.12%) and Modulation (15.05%) remained the most frequent techniques in AI-assisted translations but the use of other techniques such as Established Equivalent, Transposition, Amplification and Description were more evenly distributed, which suggests the students’ attempts to enhance the clarity, stylistic accuracy and contextual appropriateness of the AI output . The results suggest that traditional translation attaches more importance to literal fidelity and dynamic cognitive adaptation, while AI-assisted translation offers more structural consistency and requires students to focus on revising and improving the AI drafts.