Drawing on the preference hypothesis claimed by Blum-Kulka (1986/2004), Baker (1992) holds that every language has its own stylistic conventions and preferences in using certain textual patterns, mainly involving aspects of cohesion and coherence. Closely and critically examining authentic extracts belonging to different text types in English/Arabic translation, this study has questioned several general textual assumptions about Arabic. In particular, English discourse has been shown to be as lexically repetitive as Arabic discourse and Arabic discourse has been argued to be as hypotactic as English discourse, thus questioning the oft- cited claims that Arabic tends for formal lexical repetition rather than lexical variation and that English discourse is much more hypotactic than Arabic discourse. Also, both languages prove to be pronouns-dense, but they display different profiles in the use of subject vs. other pronominal elements. Arabic discourse, however, proves to be more conjunctions-dense due to the frequent employment of wa and fa as default markers which carry little semantic content and are mainly meant to smooth and naturalize the flow of discourse. Finally, some key textual decisions, e.g. the use of one conjunction rather another, may result in presenting a different mental image of the state of affairs in question and, subsequently, it seriously affects text coherence in translation.
cohesion, coherence, textuality, repetition, parataxis, hypotaxis