The notion of perceptual content is commonly introduced in the analysis of perception. It stems from an analogy between perception
and propositional attitudes. Both kinds of mental states, it is thought, have conditions of satisfaction. I try to show that
on the most plausible account of perceptual content, it does not determine the conditions under which perceptual experience
is veridical. Moreover, perceptual content must be bipolar (capable of being correct and capable of being incorrect), whereas
perception as a mental state is not (if it is veridical, it is essentially so). This has profound consequences for the epistemological
view that perception is a source of knowledge. I sktech a two-level epistemology which is consistent with this view. I conclude
that the analogy between perception and propositional attitudes, from which the notion of perceptual content is born, may
be more misleading than it is usually thought.
This revised version was published online in August 2006 with corrections to the Cover Date.