Explaining Why Things Look the Way They Do
How are we able to perceive the world veridically? If we ask this question as a part of the scientific investigation of perception, then we are not asking for a transcendental guarantee that our perceptions are by and large veridical; we presuppose that they are. Unless we assumed that we perceived the world for the most part veridically, we would not be in a position to investigate our perceptual abilities empirically. We are interested, then, not in how it is possible in general for us to perceive the world veridically, but instead in what the relation is between our environment and its properties, of which we have knowledge, on the one hand, and our perceptual mechanisms, on the other, that results in very many, even most of our perceptions being veridical in everyday life. In this paper, I am concerned with a certain kind of answer to our question that has been popular in psychological studies of our perceptual abilities at least since Helmholtz (1867). The answer is that we do it by taking account unconsciously of various perceptual cues about objects and events in our environment, and then reasoning to what the environment must be like on the basis of these cues, our general knowledge of the environment, and how it impinges on our perceptual organs. It is doubtful that anyone has ever held a pure inference theory. I call any theory an inference theory that appeals at least in part to unconscious inferences, from cues provided by stimulus to the nature of the perceiver's environment, in explaining how things look. While the power of these accounts is undeniable, they are, I think, deeply mistaken. When I say this, I do not mean merely that they are as a matter of fact false, or that the evidence in fact is overwhelmingly against them. I mean that the explanans employed could not explain the explanandum; the appearance of explanatory force is an illusion. It is not that all the particular explanations of this type have failed, but that no explanation of this type could be correct. In the paper, I first give some examples of the sort of explanation that is the target of this investigation. I then concentrate on two examples drawn from explanations of visual perception, the visual perception of size and motion, though many of the points I make generalize to other sensory modalities, and to explanations of non-perceptual cognitive capacities as well. My initial aim is to present these explanations in as strong a light as possible. I then develop an a priori argument to show that they cannot be correct, develop some detailed criticisms of them. The a priori argument defends a version of what John Searle has called ‘connection principle’. I criticize Searle’s argument for the principle he articulates, and reject the formulation of it he offers, but argue on independent grounds that we must accept a revised version of the connection principle if we are to make sense of the special relation each of us bears to his own mental states. I argue independently of this that detailed examination of inference explanations turn out to be incoherent when we try to understand how such inferences are related to the person who is supposed to be making them. However, I provide a diagnosis of their appeal, and argue also that, despite these explanations being necessarily false, they can be reinterpreted so that they have a legitimate use in psychological investigations of perception. The empirical evidence that psychologists have accumulated is not unimportant, but it is not evidence for the existence of unconscious thoughts and inferences. I conclude by replying to some possible objections to my argument.