Assessing Quality in Machine-Translated English-Turkish Climate Communication Subtitles: A Longitudinal Study

Authors : Rumeysa Nur Çakmak, Muhammed Baydere

Pages : 22-46

Doi:10.37599/ceviri.1793812

View : 137 | Download : 129

Publication Date : 2025-12-31

Article Type : Research Paper

Abstract :As climate change communication increasingly relies on digital media, the quality of machine-translated subtitles becomes a critical factor in preventing misinformation. This study investigates the quality of English–Turkish machine-translated subtitles in climate change communication, a domain characterized by technical terminology and the need for high semantic precision. Focusing on two Intergovernmental Panel on Climate Change (IPCC) videos released in 2013 and 2022, we conduct a longitudinal analysis comparing their automatic subtitles from 2022 and 2025. Using Jan Pedersen’s (2017) FAR model—which assesses functional equivalence (semantic and stylistic accuracy), acceptability (grammatical and idiomatic correctness), and readability (technical presentation)—we offer a systematic assessment of translation quality and analyze how advances in machine translation (MT) have affected subtitle accuracy, fluency, and usability. Results demonstrate substantial improvement, with total error scores reduced by 63.04% for the first video and 53.40% for the second. These gains were uneven: the largest reduction in errors was in readability for the first video (73.77%) and in acceptability for the second (76.25%), while functional equivalence showed more limited progress (27.91%-53.18%) and remained the leading source of errors in 2025, with persistent issues in conveying nuanced meaning. The findings have implications for automated subtitle practices, translator training, and sustainability-focused media strategies. We argue that for producers of climate communication, a shift towards “a human-centered augmented translation” (O’Brien, 2023) model is necessary to reduce the risk of misinformation and improve the effectiveness of climate communication. Furthermore, for consumers, we contend that translation literacy, specifically MT literacy, is not merely a technical skill but a fundamental public competency, crucial for ensuring accessible and trustworthy communication. We recommend fostering this skill through educational initiatives beyond translation programs such as targeted workshops, short modules in relevant courses, and cross-departmental certificate options. This, in turn, will facilitate the capacity of non-specialist audiences to critically assess and comprehend MT-mediated digital content broadly, and information pertaining to climate conveyed via this modality specifically. The study highlights the role of high-quality translation in strengthening global climate literacy and supporting informed policy engagement by linking the evaluation of MT with environmental communication.
Keywords : makine çevirisi, görsel-işitsel çeviri, kalite değerlendirmesi, makine çevirisi okuryazarlığı, sürdürülebilirlik

ORIGINAL ARTICLE URL