Georges Dicker

Phenomenalism and the Causal Theory of Perception: A Combined Theory

We have seen that neither the classical causal theory of perception nor ontological phenomenalism provides an adequate epistemology of perception. The classical causal theory implies a causal analysis of perceiving and a realist metaphysic that cannot be denied without doing violence to both the concept of perceiving and common-sense, but it fails on epistemological grounds. Ontological phenomenalism is very attractive from a purely epistemological point of view, but fails because it is incompatible with a causal analysis of perceiving and implies other metaphysical absurdities as well. Analytical phenomenalism is an attempt to secure the epistemological advantages of ontological phenomenalism without incurring its metaphysical paradoxes. Whether analytical phenomenalism succeeds in providing an adequate epistemology of perception, then, depends (as I said in the previous chapter) on two questions. First, is analytical phenomenalism compatible with the causal analysis of perceiving and the realist metaphysic implied by the causal theory; i. e. can one consistently be both a phenomenalist and a (causal) realist? Second, does analytical phenomenalism provide an adequate account of the justification of perceptual beliefs concerning physical things? In this chapter, I shall argue that affirmative answers may be given to both of these questions. Sections 1–3 will be concerned primarily with the first question, and Sections 4–6 primarily with the second.

The Argument from Perceptual Relativity

Perhaps the best-known method of introducing sense-data is by means of the Argument from Perceptual Relativity. This argument can be found in the writings of philosophers from antiquity to present times, and there are many different ways of formulating it.1 I shall begin by expounding (only) what seems to me to be the strongest version of the argument; then I shall present the most important contemporary objection to the argument; and finally I shall try to specify the considerations on the basis of which the issue raised by this objection should ultimately be resolved.

A Philosophical Problem Concerning Perception and Knowledge

It is obvious that there is a fundamental relationship between perception and knowledge. This is not to say that the two are identical. At any given time, we each know many things that we do not perceive. I know at this moment that there is ice at the North Pole, but I am not perceiving this to be the case. Further, at any given time we each know many things that we are not even thinking about. At this moment I know certain facts of history and geography, for example, which are in no sense present to my mind; and I shall continue to know these facts even while sound asleep. For these and other reasons, perception cannot be simply identified with knowledge.

The Argument from Hallucination

In this chapter I want to consider a final argument for the introduction of sense-data, the Argument from Hallucination. I shall begin by discussing a formulation of the argument given by D. M. Armstrong in his book, Perception and the Physical World. Armstrong’s formulation is an instructive one for two reasons. First, it does not confuse hallucinations with other phenomena such as illusions and perceptual relativity, as do some other, looser presentations of the argument.1 Second (and more significantly, from the point of view being developed here), Armstrong is careful to point out that the purpose of the argument is to show that the immediate object of perception is always a sense-datum. Unlike many other writers, Armstrong recognizes that the sense-datum theory need not deny that physical things are perceived, but only that they are immediately perceived. He correctly insists that unless this point is kept clearly in mind, philosophers who accept the sense-datum theory will seem to be defending the absurd view that we never perceive physical objects, while those who reject the theory will seem to be defending the platitude that we do perceive physical objects.

The Argument from Causation

In this chapter I shall consider the main ‘scientific’ argument for the introduction of sense-data, the argument from Causation. I shall try to show that the argument has force only insofar as it is really an epistemological argument, and that its force derives from its calling attention to a dimension of the epistemological problem of perception that we have so far not considered.

The Causal Theory of Perception

A causal theory of perception has been held by many philosophers, e. g. by Descartes, Locke, Kant, and Russell, to mention only a few. Yet these philosophers’ theories of perception differ greatly — so greatly that it would be misleading to say that they held ‘the same theory of perception’. But since my aim is to analyze the epistemological problem of perception and to appraise possible solutions to it (rather than to expound the views of any particular philosopher), I shall not go into special features of individual philosophers’ versions of the causal theory except insofar as this serves my purpose. Rather, I shall seek to formulate the fundamental features of the causal theory — fundamental in the sense that ignoring them would lead to overlooking either some aspect of the problem or some possible solution to it. In this task, I shall be guided by H. P. Grice’s valuable article, ‘The Causal Theory of Perception’.1 In this article Grice provides an excellent definition of the theory’s fundamental features.