Navigating the Future of Translation: A Descriptive Qualitative Comparison of DeepL and Google Translate in Translating High School Social Texts

Abstract

This descriptive qualitative study investigates the translation performance of two prominent machine translation (MT) tools—DeepL and Google Translate—in the context of high school-level social texts in Indonesia. Five Indonesian texts covering topics such as environment, culture, education, and digital behavior were translated into English using both tools. The translations were assessed by human evaluators based on three key criteria: accuracy, fluency, and contextual appropriateness. The study also incorporates theoretical perspectives on neural machine translation and its role in education to frame the analysis. Findings indicate that DeepL consistently outperforms Google Translate, especially in fluency and contextual sensitivity, although both tools have strengths and limitations. This research contributes practical implications for language educators and students in using MT critically and effectively in learning environments.