In comparing very different cultural, theoretical, or methodological standpoints, the nature of truth itself becomes a problem. If the standpoints have different conceptions of truth, a comparative approach that respects both involves the contradiction of conflicting legitimate claims to truth. But if we reject this contradiction, we eliminate the possibility that standpoints can have legitimately different conceptions of truth. And with that we reject the sense of a genuine comparison in this respect, rather than a reading of one framework in the light of the other. Davidson and Rorty have mounted especially powerful arguments against the very sense of this kind of contradiction between frameworks, and so against the sense of a comparative approach in this respect. Through a detailed discussion of their work, this paper argues that the contradictory conception of truth is the right one. It also argues that this contradiction is manageable. As a result, a properly comparative framework is both possible and necessary, even where the nature of truth itself is concerned. In particular, this conception makes room for ideas of truth as both absolute and relative, and also (contradictorily but with due respect for many of the cultural and theoretical frameworks available for comparison) for non-contradictory conceptions of truth.