Machine Translation: Deficiency in Translating English Sentences with Different Temporal References into Arabic

Abstract

The study investigates the accuracy of machine translation from English sentences with different temporal references into Arabic. Google translator (GT) was used in this experiment. 12 English sentences with different temporal references divided into 3 sectors, present, past and future sector, were given to GT to and translated into Arabic. The output was again given to Google were to retranslate them back into English (round-trip translation). Furthermore, Arabic equivalent sentences for the English 12 sentences were given to GT to check their English translation. Results suggest that instead of facilitating English language in general and temporal references in particular, GT is a source of confusion for nonnative English speakers. It turns out also that GT is inaccurate in its translation of English sentences with different temporal references into Arabic.

Keywords: translation, Machine Translation, Arabic, temporal reference, tense