Overrepresentation and underrepresentation of cohesive devices in EFL learners’ translated and free narrative writings

This study aimed at investigating the L2 rhetorical organization of translation and free writing tasks in terms of cohesive devices used by Iraqi intermediate EFL learners in the narrative genre. To do so, 30 Iraqi intermediate EFL learners at Kufa university took part in the study. The participants were asked to translate three narrative texts from Arabic to English and write three narratives related to the general topics given to them as prompts. The narratives were coded and rated by two experts based on Halliday and Hasan’s (1976) framework. The data obtained from translation narratives were compared with a standard translation for each text. Results of the one-sample t-test performed on the scores of translation narratives indicated that Iraqi leaners produce shorter passages in the target language than in the source language; however, they used significantly more times than expected for certain types of cohesive devices. Moreover, comparisons between translated narratives and free narratives indicated no significant difference between the translated and composed narratives. It is argued that patterns of cohesive devices used in English output of the Iraqi EFL learners are compatible with properties of their first language. The findings also show that, unlike English grammatical properties, cohesive devices are not a problematic area and would not lead to fossilized errors in the performance of Iraqi EFL learners.