The question of translatability as introduced by Benjamin, Quine and Derrida leads from quite opposite starting-points to extreme positions with respect to the understanding and conception of language. The question of plurality of languages appears in a different light depending on whether one looks at the aspect of meaning (in its sense of reference), the aspect of demarcation (in its sense of difference) or the aspect of sense (in its sense of correspondence). Accordingly, Quine puts the focus on indeterminacy, Derrida on distinctiveness, Benjamin on supplementation. It is obvious that knowledge about language can be gained from translation, and vice versa. In this paper, the author aims at examining the ways in which the various underlying language conceptions are useful for a theory of translation. Accordingly, he also examines to what extent each conception of translation supports its underlying conception of language.