An attempt is made to pinpoint the way in which perception is related to belief. Although, for familiar reasons, it is not true to say that we necessarily believe in the existence of the objects we perceive, nor that they actually have their ostensible characteristics, it is argued that the relation between perception and belief is more than merely contingentThere are two main issues to address. the first is that ‘collateral’ beliefs may impede perceptual belief. It is argued that this (...) still assigns an essential role to belief in perception, though the belief may be of an attenuated form. the second is Fred Dretske's claim that even attenuated belief may be entirely absent from perception. It is argued that (a) ‘non‐epistemic’perception can be understood only by employing the concept of ‘epistemic’perception; (b) that the former can occur only partially—i.e., within perceptions that are otherwise epistemic; and (c) that by switching attention from the perception of objects to the Phenomenological tradition's concern with the perception of world, we can see that perception must be entirely permeated with ‘doxastic’force. (shrink)

As he practices this method, what emerges through the volume is the unity of McDowell’s own views.

J. Dokic s'interroge sur le concept de perception : en quoi consiste-t-elle? comment fonctionne-t-elle?, etc. Cette analyse est suivie de deux textes commentés, l'un de George Berkeley "Les idées du haut et du bas", et "Le contenu non conceptuel" de John McDowell.