How are we to understand the visual situation described by 'x looks F to S '1, where F is restricted to the so-called 'purely visual properties'namely, the various colors? In particular, how does the visual situation described by 'x looks F to S' enable S to acquire knowledge about the world? By examining (in Section I) the deficiencies in some recent attempts to answer these questions. I think we can take some significant steps toward formulating an adequate theory of visual perception, and hence, of perception in general. 2 Some aspects of such a theory are developed in Section II.
According to what might be called the 'belief-acquiring theory' of beinglooked-to (BAT), x's looking F to S is identical, at least in part, with S's non-inferentially acquiring, by visual means, the belief or would-be belief that something in his field of vision is F (mere would-be beliefs occurring only when the percipient has some reason to believe that his eyes are deceiving him). A theory of visual knowledge follows upon specifying certain further conditions that must be fulfilled in a visual situation for the percipient to know, of x, that it is F. Variations on this general theme have recently been developed by David Armstrong, 3 Moreland Perkins, 4 and George Pitcher. 5
According to both Armstrong and Pitcher, this analysis of beinglooked-to, though requiring the filling out of details, is substantially complete. But there is something very implausible about this claim, and neither Armstrong nor Pitcher say anything that significantly lessens this implausibility. It is hard to believe that whatever it is about perceptual experiences that is missing from non-perceptual beliefs can be supplied by the specification either that the beliefs are being acquired by means of the appropriate stimulus-sensitive organs, 6 or that they are richer in content