INTRODUCTION

For many years Isaac Levi has been a staunch defender of a strictly normative and prescriptive conception of rationality. The origin and motivation for this crucial commitment, as it transpires particularly clearly in The Covenant of Reason (Levi 1997; henceforth CR), has been Levi's exploration and development of the Peirce-Dewey “belief-doubt” model of inquiry. On the latter, justifiable change in state of belief is a species of rational decision making. This is what motivates Levi's concern with “rationality” in the first place (CR, 20). In fact, no substantive commitment on what rationality substantively is – or on what it is to be rational – emerges from this theoretical interest. In particular, we are not told what beliefs or values we should have, which ones it is rational to have, or how we should base our beliefs on “evidence.” Rather, principles of rationality are primarily justified instrumentally through their regulative use as formal constraints on well-conducted inquiry and problem solving, no matter the domain, be it science, politics, economics, technology, or art, or even simply the personal decisions we face in daily life. Given their exceeding generality, we can only expect constraints on the coherence of choice to be both formal and weak. Principles of rationality are to be kept immune from revision if a general theory of how rational changes in point of view are to be justified is to be possible at all (CR, 24).