S. L. Mastros
March 26, 2001
Metaphysics; Dr. Barbara Montero

In his Argument For the Identify Theory, David Lewis claims that a causal definition of experience (as a universal) and a belief in the causal efficacy of physics together create a necessary argument for a materialist model of experience. I will examine some consequences of his argument. I feel the logical form of his argument, if accepted as valid, allows for other, less pleasing, arguments to be made.

Lewis begins with the hypothesis that experiences are defined by their causal roles. A causal role can be explained as the conjunction of a series of statements about the typical causes and effects of an experience. This is a very appealing hypothesis, and gives us the explanatory power to deal with things often thought well outside the domain of materialism.

Consider a poltergeist. Traditionally, poltergeists (or, literally "noisy ghosts") are associated with a number of phenomena. So, we could construct a causal role definition of poltergeist in the following way; "A poltergeist is that thing which typically occurs near adolescent girls, and typically causes dishes to break and typically causes magnetic media to degrade and typically causes milk to sour and is typically abated by the act of leaving honeyed bread and milk out overnight and so forth". Now, consider the home in which all (or at least, most) of these phenomena occur. By Lewis's theory, we are completely justified in saying there is a poltergeist in the home. This is a powerful tool for materialists, who might otherwise not be able to deal with the concept 'poltergeist' in any informative way.

Lewis then moves on to his second hypothesis, that of the explanatory adequacy of science. By this, he means to say that all physical phenomena may be explained by way of physical antecedents and physical laws, and that these laws and antecedents are knowable by men. I have previously written on such matters, and will accept this hypothesis as given.

He also makes an implicit assumption that science has shown experiences to be correlated with neural states. That is, the causal role of pain has been shown to coexist with a given (or, more likely, multiple given) brain states. This too touches on topics well ouitside my focus, and I will accept it.

Here comes the thrust of Lewis's argument. Since any experience is nothing more than the object(s) which fill a causal role, and since we have empirically established that neurological states fill the causal role, he concludes that any experience is nothing more than a neurological state. There are many reasons to consider this a compelling argument; I will discuss them but briefly.

Defining experiences casually gives an outstanding new explanatory power to materialism which is difficult to acquire without some sort of similar functional/behaviorist model. It is this seemingly magical ability to describe the things we are so accustomed to thinking of as mental in a completely physical way which has made materialism the golden child of modern ontological theories. Moreover, it gives the materialist an enviable ability to apply Occam's razor; it is hard to defend the position that while a materialist position is all that is required, there is reason to believe in immaterial objects.

put some more praise here

However, there are fundamental flaws in Lewis's reasoning which, if accepted allows for some dire consequences. Consider first the logical structure of the argument, which breaks down as follows:

P = { x | C(x) }
C(y)
therefore, y = P

When viewed in this form, it becomes obvious that the claim Lewis is actually entitled to make is y is an element of P, that is to say, that neurological states are part of what we call experience. This may tempt the question -- "What else could fill the causal role but a physical thing?" And here we find the central flaw in Lewis's reasoning. Implicit in his definition of causal role is the assumption that the causes and effects by which we define an experience are solely physical. (This is necessary to be able to apply the explanatory adequacy of science as he describes) To say that we define experiences in a wholly physical way is to make a materialist claim in a very strong fashion. I think few dualists, or even epi-phenomenalists, would be ready to say such a thing. If we include immaterial objects in our causal role definition, the adequacy of science to account for physical events becomes useless in an attempt to reconcile experience and the material world.

Consider for a moment a dualist's (or an idealist's) definition of pain in casual-role form: "Pain is that experience which typically causes the experiencer to have an unpleasant feeling and which typically makes the experiencer believe that she wishes to avoid the experience and which typically causes the desire to make "ouch" type noises and so forth..." Clearly science can tell us little about what sorts of things fill this causal role.