| OCR Text |
Show Rationality and the Structure of the Self, Volume II: A Kantian Conception 67 and necessary criteria of truth in the sense that whatever contradicts them must be false. But they do not establish that whatever fails to contradict them is true, because a representation that satisfies them still may be contradicted by its object (1C, A 59/B 84). The linguistic holism that inferentialism endorses might weave a tight and complex web of sentential inferences indeed, which nevertheless bore no truth-preserving relationship to the denotations of its singular terms and predicates. If the inferential relationships mapped in the Table of Judgments provide no sufficient condition for determining whether or not a representation that satisfies them succeeds in denoting its object, then as far as Kant is concerned, it provides no leverage for a materially robust inferentialist truth criterion. So whereas the denotational relationship offers agreement with the object but no general criterion of truth, the inferential relationship offers a general and necessary criterion of truth but no guarantee of agreement with, i.e. denotation of the object. The sufficient condition of truth that such agreement would provide cannot be stated in a general form. Hence neither is adequate, either singly or conjointly, to provide an answer to the question of what truth is. From this conclusion Kant can now argue that an answer to that question can be found only within the limited realm of empirical experience itself, in which we agree to leave investigation of the cognitive and metaphysical preconditions for having such experience out of account. While we can ascertain whether an empirical assertion is or is not true to the facts we observe, we cannot ascertain whether or not the facts as we empirically observe them are or are not true to the noumenal reality to which we futilely intend our assertions to refer. What we have seen from close analysis of passage (C') above is that Kant's view synthesizes key elements of both inferentialism and representationalism. It is inferentialist in its defense of a restricted set of judgment forms that bear logical interrelationships and circumscribe the scope and types of judgments it is humanly possible to make. It is representationalist in insisting on a causally direct and unmediated relationship between certain concepts that enter into such judgments, and the real world objects those concepts represent. Brandom is quite right to argue, as he does in passage (B), above, that Kant requires conceptual classification of an object as a necessary condition for experiencing it. Brandom is also right to insist that concepts must combine in the right ways in order for us to make judgments about those objects. Where he goes wrong is in thinking that we could make such judgments, and could understand their singular terms and predicates, without any independent representational relationship to the objects those subsentential expressions denote. Kant does not make that mistake because of the foundational role he accords the notion of a concept as a function for unifying representations. In the following sections I hope not to make that mistake either, and for much the same reasons. © Adrian Piper Research Archive Foundation Berlin |
