Sometimes our intentions and beliefs exhibit a structure that proves us to be irrational. Is there anything wrong with that? Should we be rational rather than irrational? In other words, is rationality normative? Intuitively, the answer to this question is ‘yes’. Calling someone irrational amounts to a form of criticism. By doing so, we seem to imply that the person in question has made some kind of mistake, that her mental attitudes are in need of revision. In recent discussion, however, the normativity of rationality has come under attack; many philosophers now accept the skeptical position that we may have no normative reason to conform to rational standards.
The problem occurs with so-called structural requirements of rationality, which prohibit certain combinations of attitudes independently of the reasons that count for or against the particular attitudes involved. I agree with the skeptic that there are compelling arguments for the conclusion that structural requirements of rationality, if there are any, cannot be normative. In contrast to the skeptic, however, I argue for an anti-structuralist account of rationality, which maintains the normativity of rationality by denying that there are any structural requirements of rationality to begin with. I show that the assumption of structural requirements leads to serious problems independently of the question of normativity. I then sketch an alternative account of rationality according to which rationality requires us to respond to available reasons. Finally, I provide the outline of an explanation of the appearance of structural irrationality in terms of a response conception of rationality.