Abstract

The purpose of this thesis is to determine the precise relationship between perceptual processes and cognitive structures such as concepts and beliefs. In order to accomplish this task, an account of cognition is developed which provides a way to distinguish cognitive from non-cognitive modes of processing information. I argue that a vast number of perceptual processes which have often been taken to involve cognitive mediation can more properly be understood as computational processes which do not involve cognitive structures. The criteria for empirically determining this issue as well as its philosophical implications are carefully explored. ;I proceed by examining the recent debate in perceptual psychology concerning the kinds of mediating processes which best explain visual perception. The debate is between theorists who hold that perception in general is cognitively mediated and those who argue that it is not . I maintain that this debate involves a number of philosophically significant issues, including questions about the objects of perception, the units of sensory stimulation, the concept of information, and the nature of perceptual laws. However, a number of confusions and ambiguities have hampered progress and require clarification. In particular, I argue that neither party in the dispute develops a clear account of cognition nor of related processes such as inference and computation. ;I claim that when these issues are clarified the constructivist and direct theories may be viewed as complementary in that each emphasizes a different aspect of the perceptual process. While constructivists have tended to conflate computational processes with cognitive ones, holders of the direct theory have often ignored the role of computational processes and hence have oversimplified the perceptual process. ;In response to these problems, I argue that the role of cognition in perception can be determined only by examining specific perceptual processes and the kind of information which is presented to the perceptual systems. Finally, I develop an account of perceptual processing which is consistent with direct realism.