In order to understand eliminative materialism, we must first examine two other materialistic theories with which it contends—reductive materialism (also known as the identity theory) and materialistic functionalism. Once one has rejected dualism, at least the extreme form of it that there are two different kinds of substance, as incapable of explaining anything which could not just as easily be explained in terms of a single substance, it remains to be seen how the things which under dualist theory were assigned to the res cogitens will be explained in material terms. Reductive materialism claims that this problem will be resolved by saying that each thing hitherto called a mental state can be mapped directly to a brain state; functionalism states that because of the problem of multiple instantiability (that is, the same mental state may not have the same physical representation in all cases) that we should try to map mental states into a network with other mental states, stimuli, and behaviors.
Eliminative materialism is a stronger form of materialism than these two. Whereas reductive materialism and materialistic functionalism explain our commonsense psychological terms and ideas (collectively referred to as “folk psychology”) in terms of physical objects somehow, eliminative materialism claims that this folk psychology is fundamentally wrongheaded and should be eliminated entirely. For example, in Patricia Churchland's book Neurophilosophy, on page 384 she states that “folk psychology may be irreducible with respect to neuroscience—irreducible because dead wrong.”
At this point it is worth looking at what exactly it is that folk psychology attempts to do. An eliminative materialist, who typically has great respect for hard-science theories, will want his theory of psychology to operate in the same manner. That is, he will want it to give him theories which enable him to predict the events in the domain in question, and if they fail to predict, he will think they need to be replaced. However, folk psychology does not do this. As Elster's entire first chapter in Alchemies of the Mind is devoted to explaining, folk psychology very often explains by what he calls mechanisms rather than by laws. Mechanisms are a sort of middle ground between a law-like theory and total lack of understanding; they enable us to explain things, but often not to predict them.
The idea of a mechanism can be applied to many domains, but Elster claims that it is crucial to psychology in particular. This is problematic, of course, for those seeking a full law-like theory of psychology, but this is no indictment of theories which explain using mechanisms; while it may be claimed that these theories are less advanced than a law-like theory would be, they may also be much easier to come up with and work with, to the point of making it possible to actually make some headway rather than being stuck with the massive problem of analyzing the entire brain to understand psychology at all.
Eliminative materialists' reasons for favoring elimination over reduction or dualism generally run towards analogies to the “hard” sciences such as physics or chemistry. They see these sciences having great explanatory successes and believe that the same methods applied to psychology will also yield good results. These sciences sometimes seem to eliminate rather than reduce the theoretical terms of the theories that preceded them. Two examples of this, given by Paul Churchland in Matter and Consciousness on pages 43 and 48 respectively, are caloric and vital spirit.
Caloric theory held that warmth was an actual (though invisible) physical substance and that transfers of heat were due to its flowing from one object to another. When the microphysical properties of matter were examined, however, no material corresponding to caloric was found; what was rather observed was that materials having higher temperature were undergoing more rapid vibrations.
However, posing this as an example of total elimination seems rather uncharitable to caloric theory. Maybe there was no literal substance that had mass and took up volume that corresponded to caloric (though perhaps under relativity kinetic energy could affect mass and the laws of thermodynamic expansion do seem to imply at least that objects with higher temperature take up more volume); but despite this detail, caloric theory had (and would still have, were we to continue to apply it) explanatory power, and many statements made in terms of caloric could be relatively easily rephrased in terms of current thermodynamic theories (thus indicating a reduction rather than an elimination, at least to some extent.)
The case of vital spirit is also interesting. Vital spirit was purportedly much like caloric but for life rather than for heat; objects possessing it were alive, those lacking it were not. The definition of the term “alive” itself is, I would say, at the moment a matter of some contention. Two cases in which there is disagreement over whether or not to apply the term “alive” are viruses and sufficiently complex computer programs constructed using certain methods called (by their inventors and proponents) “artificial life.” This appears to be a point over which there will be more, not less, contention as advances are made in computer science, bioengineering and nanotechnology, unless the proponents of these various definitions are willing to acknowledge that their differences may be merely semantic rather than factual.
If we acknowledged such things as living computer programs, it would be patently obvious that vital spirit cannot exist as a physical substance, since the things that are classified as alive would not even share a physical substrate. Even if we do not acknowledge such arguments, though we may be able to find some substance which is shared by all organisms we can find, it seems unlikely that this substance would itself somehow be the “essence” of life and be able to instill the property of life on hitherto nonliving things. However, in the limited domain of earthly carbon-based biology, some ideas related to vital spirit persist in the form of the rejection of the theory that living things can (in commonplace circumstances) spring out of nonliving things, known as “spontaneous generation.” Furthermore, I would hardly say that the belief in vital spirit as a property has been refuted; indeed, I would say that life must be a property to function as it does in our theoretical framework.
Yet for all the objections that can be raised, the cases of caloric and vital spirit are at least somewhat plausible cases of elimination. Elimination in general, while it may be applied too broadly and with too much vigor, can be a reasonable move in theorizing. In most domains, if a theorist states that his new theory does not reduce but rather eliminates theoretical terms of an old theory in the same domain, it is worth giving him a fair hearing. However, I believe there to be special considerations in the case of mental states which rule out this move and make it unnecessary to consider seriously any theory which claims to eliminate mental states.
Specifically, I wish to point out a difference between the relation that the act of theorizing itself has to mental states and that it has to other entities considered for elimination, such as caloric and vital spirit. There is no special relation between the act of theorizing and caloric (or theories of heat and temperature in general), nor between theorizing and vital spirit (or biological theories in general). On the other hand, theorizing is permeated by mental terms. The theory itself is a proposition (a mental object), it contains theoretical terms which are also mental objects, and so on.
Despite the fact that acts of theorizing are permeated by mental acts and objects, theorizing can still tolerate various accounts of these mental acts and objects, as long as they have the connections to each other necessitated by the structure of theorizing. But suppose a theory to claim the notion of a proposition or a theoretical term as incoherent, to be replaced with some set of facts concerning neurophysiology. In that case, is the theory not denying its own validity as a theory (along with that of anything else)? Yet this is exactly what eliminativism claims to do. Eliminativism, in claiming to eliminate folk psychological theories and terms, draws on a framework for theorizing which depends in an essential way on the very folk psychology it seeks to eliminate.
At this point the eliminativist may object that I have gone too far. Perhaps total elimination will not work, but does this mean that all cases of elimination are invalid? Indeed it does not. However, the eliminativist cannot go back and forth snipping out arbitrary pieces of folk psychology and hope that the result will still allow him to express his own theory. He must be very careful not to remove any pieces on which the ability to theorize rests—and the roots of that ability may stretch further than he might anticipate.
Churchland, Patricia (1986). Neurophilosophy: Towards a Unified Science of the Mind/Brain. Cambridge, Massachusetts: The MIT Press.
Churchland, Paul (1988). Matter and Consciousness: A Contemporary Introduction to the Philosophy of Mind. Cambridge, Massachusetts: The MIT Press.
Elster, John (1999). Alchemies of the Mind: Rationality and the Emotions. New York, New York: Cambridge University Press.
Kenn Hamm
For copyright and other information, click here.
Last modified: 2 November 2000