Apparently, aiming to comply with the norm ‘Believe that P if and only if the proposition that P is true’ can hardly differ from aiming to comply with the norm ‘Believe that P if and only if the proposition that P is epistemically justified’. So one may be tempted to agree with Richard Rorty that the distinction between truth and justification is pragmatically useless because it cannot make any difference ‘when the question is about what I should believe now’. I (...) resist this conclusion by arguing that the distinction between truth and justification is pragmatically useful even if the two properties are indeed normatively coincident. The argument I offer turns on the claim that truth plays an explanatory role that justification is inherently incapable of playing. However, my contention is not just that the distinction between truth and justification is pragmatically useful because truth is a bona fide explanatory notion. It is that the distinction between truth and justification is pragmatically useful because the realization that the former plays an explanatory role that the latter is inherently incapable of playing gives access to reasons which would otherwise escape our attention. If truth is a bona fide explanatory notion, the distinction between truth and justification is pragmatically useful because it is precisely when the question is about what I should believe now that attending to such a distinction will often make a difference – and it will make a difference even if the two properties are in fact normatively coincident. (shrink)