زهرا فاخر عجب شیر

Recently, the use of machine translation (MT) to support the second language (L2) writing has increased. Since translation quality via MT matures every year, updated studies are required. The present study explored the quality of MT (Google Translate) outputs from Persian to English by comparing them with the texts translated to English by 83 intermediate-level English as a foreign language (EFL) learners. After the instruction on the narrative genre was delivered to the participants, they watched a pictorial narrative prompt and wrote a narrative text directly in English. In the next session, they received training on MT use, watched another narrative prompt, wrote drafts in Persian, and submitted them to MT for translation. A comparison of the texts written directly in English by the participants and MT products showed that the use of MT was beneficial in measures related to mechanical aspects, lexical sophistication, and some grammatical aspects. Direct L2 writing, on the other hand, rendered better performance in lexical accuracy and some grammatical structures as well as the general understanding of the text. Based on the results, pedagogical implications for the use of MT in L2 educational contexts were presented. Keywords