Where does this leave 'truth' in the 'game of asking and giving reasons' for stated historiographical theses? Provided that redefining the notion of truth by reference to some novel semantic or epistemic entities or processes is discounted and that the notion of epistemic authority is taken as a more fundamental concept than that of truth, I do not see why the 'truth' would be necessary in the game of asking and giving reasons. That is, if we manage by some other means to attribute to our assertions the kind of epistemic authority that compels a rational being to accept the assertions, reference to 'truth' is superfluous. And epistemic authority can be seen to derive from the inferential practice of providing reasons itself. Brandom suggested that Wittgenstein's substitution of his earlier question 'What are the facts?' with the question 'What are we entitled to say?' leads to the 'de-emphasis of the notion of truth' (1976, 138).
Indeed, if putting forward a claim as true is to endorse it, to ask others to accept it, endorsement can be carried out by providing reasons for its defense in the discursively regimented 'space of reasons' (cf. Brandom 1994, 170). Because of the problems with the correspondence theory in the case of historical theses, no deeper substance can be provided to 'truth'-clauses than the endorsement itself. Notice, however, that the truth of some subordinated statements may be required, as discussed in the final chapter of this book.