Accepting Selinker's view about the study of interlanguage (IL) in regard to ESP (1984), I shali examine partial performance of two. top M. Sc, students at Mosul University, Iraq. This study does not scem to lend support to his ciaim that control over topic, rather than linguistic struo- tures, must come first in the teaching of cS? (cf. Selinker 1986). As will be denonstrated below, I tend to be much more for doing away with prio- rities than taking a counter stand. This position is confirmed by two inter- views with the subjects. The two Subjects who have been experiencing a considerable number of "approximative systems" (Nemser 1974: 56) at Successive stages of learning both general English and ESP are themsolves aware of 'topic' and 'structure'. The claim is nothing new; it has so often been expressed (and put into practice for over quarter of a century at the University) that in ISP courses, teaching comprehension comes first, gra- mmar sccond and, finally, speech. No wonder, this strategy is also favou- red by the learners who would like to find themselves understand what they read, the res ult of which is poorly written and spoken English