as in other language activities (e.g. language teaching) which involve two language or more, it is very difficult to avoid interference or more specifically negative transfer which affects the performance, in our case the translated text, in a way that reproduces inaccurate linguistic structures and poor collocations on the semantic level, sometimes. This study gains support from Selinker's view of interlanguage [Selinker (1986) vs.(Gorgis, 1994)) that control over meaning of a target language text rather than grammatical stuctures comes first in nguage activities. Relevant to the notion of negative transfer is that of "fossilization" where it could be more abvious at advanced stages of learning; where the properties of the native language, in our case the kurdish language, combine together with the target language, Arabic, to form an intermediate system which is to be used persistently and consistently by the learner, in our case the translator (Gorgis, 1994). Still, the identification of the likely fossilized patterns (grammatical structures) in a given performance (the translated text) remains difficult (ibid).