Odoje, C.O.2026-02-1720181115-1609ui_art_odoje_human_2018https://repository.ui.edu.ng/handle/123456789/12223In: RALL:Research in African Languages & Linguistics, 17, pp. 1-13Machine translation (MT) as a multidisciplinary field has made significant attempts with resource-rich languages towards translation. There are concerted efforts to mobilize resources for resource-low languages to improve their translation outputs too. One of the recent arguments for improving the translation of MT is the consideration of very close languages. This paper, therefore, evaluated human translation and Google translate translation of German/English in line with the existing Yoruba/English translation. The Akungba Sentence Paradigm was used as an instrument, and a Nigerian who was rated A2 proficient in German was engaged as a human translator. Eleven (11) students of the Institute of Asian and African Studies, University of Hamburg, Hamburg, Germany who are proficient both in English and German volunteered to be human evaluators. It was found that English/German has high mean score than Yoruba/English since both English and German are Germanic languages, unlike Yoruba and English. The paper suggests that attention should be paid to related African languages to set a new benchmark for African languages’ MT translations.enMachine translation (MT)Human EvaluationHuman translationGoogle translate translationGerman/EnglishYoruba/EnglishOther