Perception, causation and information

In this work I answer the question, what is it for a perceiver P to see a physical object x by having a visual experience E. First of all, I consider whether a causal connection between x and E is necessary, and I conclude that it is. I then consider various causal theories of seeing. The best of these is Michael Tye's 'systematic variation' theory. Tye holds that P sees x if there is a causal connection between x and E such that varying certain of x's properties would produce systematic variations in certain of E's properties. I show that a deeper explanation of seeing can be given in terms of what I call 'significant information'. I propose that P sees x when x causes P to have a visual experience E which provides significant information about x for P. E provides significant information about x for P when a perceiver of P's kind can have a sufficient number of determinate true beliefs about x's properties on the basis of E. I then argue that there is no one kind of perceiver that P is, and this entails that whether E provides significant information about x for P depends upon what kind of perceiver P is being considered as. Thus, whether P sees x is relative to what kind of perceiver P is being considered as. I then show that the 'systematic variation' intuition underlying Tye's theory is wrong. Rejecting this intuition allows me to overcome David Lewis's 'censor' problem, and what I call the 'idle mechanism' problem, which other causal theories cannot do. Rejecting this intuition also raises some problems for my analysis. I show that they can be overcome, and in doing so, I show that E provides significant information about x to P only when it is most likely that there is only a small range of properties that an object could have such that it could cause E in the circumstances P is in, and x has such properties. A problem for all causal theories is that in certain situations they entail that P sees x when the ordinary person would not hold that P sees x. Rather than adding conditions to rule out P seeing x in such situations, as Tye attempts, I argue that we should accept that P sees x in these cases. I go on to show that we should accept that P sees x by seeing television pictures of x and even photos of x. I also consider various other matters related to seeing, and I provide an account of illusion and hallucination and strong and weak seeing. Finally, I reevaluate the importance of seeing to epistemology.