In this paper I shall outline Roderick Chisholm's objection to C.I. Lewis's phenomenalist response to skepticism and then attempt to prove that regardless of Lewis's replies, Chisholm's objection succeeds. Before explaining the objection, though, I must first give a brief explanation of Lewis's view.

Lewis's analysis of the meaning of statements of objective facts involves the claim that all such statements can be translated to equivalent statements about one's own experience. In other words, if we subtract from our beliefs everything about which we could possibly mistaken, we would be left with certain affirmations about our own sensory experience. These affirmations can be stated in the form of "terminating judgments" which, in turn, can be said to be equivalent in meaning to what we think of as statements about the world. These terminating judgments each take the form of a conditional. Thus, a statement like, "There is a doorknob in front of me." is equivalent to a group of conditionals based solely on my own experience, such as, "If I were to feel like I was reaching out my hand, I would feel like I was having the experience of touching a doorknob." This is only one of the countless such conditionals I am actually considering when I make such an objective statement about the world.

Chisholm's response is actually quite simple and involves an objection to the logical structure of Lewis's argument. If it is true, Chisholm argues, that statements of fact are equivalent in meaning to the set of terminating judgments in question, then it must be true that the former implies the latter. Thus, let us call the statement about the presence of the doorknob p and the set of terminating judgments, of which I cited one possible example, q, and Lewis would have to admit that p implies q. However, if this is true, then p, in conjunction with any other statement, s, must also imply q.

Chisholm's argument rests of the realization that there are countless statements that we could plug in for s that would make the implication false. For example, s could be a condition on p that would make q false. Sticking to our doorknob example, suppose we add the condition, s, such that s says, "there is an invisible force that blocks my hand." In this case, it would not be true that I would, upon feeling like I was reaching out my hand, have the feeling of touching a doorknob.

That, in a nutshell, is Chisholm's objection. But let us now consider some possible replies Lewis could and does offer, which will both help clarify the objection and show why it is immune to any such replies. First, one could say that in the case of the invisible force, the antecedent of q is no longer fulfilled. In other words, if there were such a force, it would not be possible for me to feel as if I were reaching out my hand, since it would be blocked half way. This is the easiest objection to respond to, since we can just find another meaning to assign to S in order to make sure that the antecedent of q is fulfilled. Chisholm asks us to suppose that we have a "malady" which causes us to believe we are moving when in fact we are not. In such a situation, I again would feel like I was reaching my hand out, but again I would not have the feeling of touching a doorknob (since in fact my hand never would come into contact with the doorknob.)

This example shows what Chisholm means by the "relativity of perception." He claims that Lewis has not acknowledged the basic fact that perceptual experiences necessarily depend on certain conditions on the subject having such experiences. So now let us suppose Lewis grants that such conditions must be met but says that we can just expand p to include such conditions. In the doorknob example, Lewis could say that we just have to add to the statement that "there is a doorknob in front of me," a group of other statements like, "There is no invisible force between me and the doorknob" and "I don't have any strange maladies which might cause me to believe I am moving my hand when I'm not."

This seemingly simple solution would actually cause a bigger problem: the number of such conditions would be infinite! One need only think for a moment to see that we could think of any number of such maladies and forces, since they don't actually have to exist, they only have to be conceivable. Moreover, and more gravely, such conditions would give us more statements about the world for which we would then have to find new terminating judgments. In other words, a condition like "there is no invisible force between me and the doorknob." is just another objective statement about the world for which there would have to be a whole new set of conditionals, which could only be met if certain conditions were met, etc. We can see that this just multiplies the problem

These problems reveal Lewis's basic misconception, which once understood will show why he is in a hopeless situation. Lewis is looking for strict logical implication in empirical facts about the world, and even if that world is restricted to our own sensory experience, such implication is impossible. Statements of objective fact have no necessary implications, as David Hume so eloquently proved, since we can always imagine a state of affairs under which such implications would not hold.

Lewis still has one possible response, which we shall consider and then show why it too fails. He could give up altogether such a claim to strict implication. He could do this by adding an element of probability to each of the terminating judgments. Thus q would become, "If I feel like I'm reaching out my hand, then, in all probability, I would feel like I was touching a doorknob." This seems to get Lewis out of trouble, but again only for a minute, until we realize that, under certain hypothetical conditions, it is not even true that such a probability exists. Thus, suppose again that there is an invisible force in between me and the doorknob, then there is actually not even a small probability that my reaching out would result in my feeling the doorknob. In other words, probability changes depending on what conditions you add, which completely undermines Lewis's terminating judgments, making them meaningless.

Now the only thing left for Lewis would be to abandon the idea of implication altogether and say that it isn't true that p implies q. However, in this case Lewis's whole conception of meaning would no longer be of any use. The whole point was to eliminate the gap between our own experience, about which we cannot be mistaken, and statements about the world. If Lewis were to abandon the implication from statements about our sensory experience to statements about the world, then the gap between the two would be reintroduced and he would be "back to the old drawing board."

November 4, 1997