In this article the author proposes a definitional framework for empirical research on a question that is problematic both within and between most of our institutions: namely, how translation classes (and institutions) should relate to foreign language classes (and institutions) at university level. He suggests that the relation between these two general teaching activities can be formalized in terms of a simple descriptive distinction between binary and non-binary errors, and that this distinction can generate working hypotheses which are potentially useful for the tempering of conflictual institutional relations.