Knowledge-based Accounts of Rationality

Philosophy Department of the University of Geneva
Swiss Centre for Affective Sciences
Institüt für Philosophie, Berne



Project members: Julien Dutant, Pascal Engel, Fabrice Teroni, Jacques Vollet

What we ought to do and to believe depends on what we know. The idea may appear obvious. It reflects the way in which many decision makers, scientific experts and ordinary citizens think. For instance, they will say that we should not market a drug which we do not know to be safe enough. Yet dominant theories of rationality reject it. Standard decision theory, bayesian accounts of scientific rationality and many ethical theories are internalist: they hold that what is rational for us to do and to believe depends on our internal mental states alone, such as our beliefs and experiences, rather than on what we know (Wedgwood, 2002). However, internalist views of rationality are increasingly challenged within epistemology (Williamson, 2000; Hawthorne and Stanley, 2008; Sutton, 2007; Pritchard, 2012). The present project builds on these developments in order to elaborate a full-blown account of rationality in terms of knowledge. It aims at reconciling our fundamental theories of rationality with the way in which scientific experts, decision makers and ordinary citizens think about rational decision and belief.

The first stage of the project investigates knowledge-based accounts of rationality applied to particular domains: belief, action and emotion (modules 1.b, 1.c and 1.d, respectively). It will also review the related debate on the knowledge norm of assertion (module 1.a). This stage will allow us to bring together the scattered discussions of knowledge-based norms in recent literature. In the case of emotions, it will also open up new lines of research.

The second stage states and discusses fully general accounts of rationality in terms of knowledge. For their general structure, we draw inspiration from analyses of rationality in terms of reason and belief in ethics (module 2.a). We set out a range of motivations for such accounts (module 2.b). Some are general, such as the fact that there appear to be normative differences between knowledge and true reasonable belief that internalists leave out, or the fact that knowledge-based accounts provide a better explanation of various norms of inquiry. Others are specific, such as the fact that internalists have a hard time explaining the rationality of memory-based belief. We consider the main internalist challenge to knowledge-based accounts (module 2.c.). We lay out various options of response and their implications.

The third stage deals with advanced developments of the view and further issues. First, we aim at integrating knowledge-based accounts with dominant formal models of theoretical and practical rationality. Most models are probabilistic; we correspondingly explore the notion of probabilistic knowledge. Second, we discuss the normative sceptical paradox, a potential source of difficulty for knowledge-based accounts. Third, we investigate whether knowledge-based accounts of rationality require certain assumptions on the nature of knowledge and whether they shed light on the value of knowledge and the concept of knowledge.