The fundamental insight that lies at the foundation of functionalism is that different types of material objects can perform the same function. The original functionalists combined this insight with a second, more contentious claim--namely, that mental terms refer to certain functions performed by the brain--to conclude that the brain is irrelevant to the existence of consciousness. After all, if the designated functions can be performed by other forms of matter--animal neurons, silicon chips, or alien whatever--then the brain is not required in order for consciousness to exist. More recently, a more tepid set of functionalists, most notably David Chalmers, have emerged. In face of the qualia problem, they accept that the mental terms associated with consciousness need not refer to the functions themselves. Still they believe that such functions, for reasons that we do not fully understand, give rise to consciousness. Thus, the ultimate conclusion of the functionalists remains the same. As long as the appropriate functions can be duplicated in other forms of matter, then these other types of matter possess consciousness.

Critics of functionalism, most notably John Searle and Ned Block, simply run the argument the other way around. Since a collection of beer cans or the population of China can have the same functional organization as the human brain but would obviously lack consciousness, mental terms cannot refer to a set of functions nor necessarily arise from them.

I want to suggest that the functionalist argument in all its forms depends upon a powerful, yet unexamined and perhaps mistaken assumption: namely, that other types of material objects can perform the same functions as the human brain without being a human brain. It is easy enough to overlook the critical role that this assumption plays because, if one describes a set of functions broadly enough, it is true: other types of material objects can perform the same functions as the human brain. However, as one adds specificity to the functions that must be performed, it delimits the kinds of material objects that can perform them. In the end, it might turn out that only wet matter can perform all of the functions associated with consciousness. To put a sharper edge on this last comment, it could well be that silicon based computers will forever be left out from the experience game.

Suppose we begin by considering the very kind of example that gave rise to functionalism. I am in my garage working on the mechanics of my lawn mower when I come across a nut that needs to be loosened. I could, of course, seek out the proper tool, one that was designed to perform just that function: namely, a wrench. But my wrenches are in the basement and, being the lazy sort, I cast about for alternatives. I consider using a pair of pliers from the kitchen drawer, but, given my inherent laziness, even that seems too far to walk. Ah, what the heck, why not just use my fingers?

What we have here is an example of three material objects, all of which can be used to perform the same function: loosening a nut. Yet no one would want to suggest that just because my fingers can be used to loosen a nut, they are a wrench; or that, a pair of pliers and a wrench are one in the same kind of object. That is the point of functionalism. Three different kinds of material objects--a wrench, a pair of pliers, and a set of fingers--can all perform the same function. Very much in the same way, we are told, a set of silicon chips, correctly organized, could duplicate the functions of the human brain. Since those functions either constitute thought or give rise to it (depending upon the version of functionalism at issue), we would then have a conscious, thinking machine.

Unfortunately, as much as my lawn mower experience helps illustrate the basic insight that lies at the foundation of functionalism, it also raises further questions. When I try to loosen the nut with my fingers, I find that it is secured too tightly. Seeking recourse, I go to the kitchen to grab the pair of pliers. Unfortunately, while the pliers can do something that a single wrench cannot, grasp a number of nuts of very different size, when I try the pliers on this particular nut, the nut is secured too tightly and the angle of grip is wrong so that the pliers begin to strip the corners off. Alas, I regretfully do what I wanted to avoid--walk down to the basement to get a wrench. It is the only tool available that functions in the way that I need; it alone can loosen this particular nut.

This example illustrates a basic principle. The greater the particularly with which one specifies a function, the more one delimits the number of material objects that can perform that function. We now know, for example, that both Gary Kasparov and Deep Blue can play the game of chess at a higher level of proficiency than any other thing presently on Earth. That makes it true, in a very broad sense, that both are capable of carrying out the same set of logical operations. But no one has ever suggested that Gary Kasparov could carry out these functions in the same manner that Deep Blue does--by examining umpteen million board positions within a three minutes time span. Why then do the functionalists assume the reverse--that a properly programmed computer could not only play chess as well as Gary Kasparov, but do so by duplicating the manner in which his brain functions?

The answer to this question, rather strangely, lies in a highly abstruse mathematical theorem: namely, Church's thesis. That thesis tells us that a strikingly simple device, a Turing machine, properly programmed, can carry out any mathematical operation that can be performed. Since digital computers can mimic the operations of a Turing machine (presuming that one does not put any restrictions on the size of the computer nor the time in which it has to complete its task), they too have the property of being universal operators. Finally, given that digital computers can perform any mathematically describable function, that means that they can duplicate the functional organization of the human brain.

The problem with this argument lies in the final step. It presumes that only the functional organization of the brain, something which can be duplicated within mathematical formulae, is relevant to the existence of consciousness. That trivializes the role played by the physical. The fact that the brain receives sensory input in the form of nerve impulses, that the electrochemistry of its axons is constantly in flux, that chemicals are manufactured and dispersed between the synapses, that dendrites are grown and atrophied, and that new nerve impulses are transmitted outward, all of this is deemed irrelevant to the existence of consciousness; according to the functionalists, the only thing matters is the organizational structure within which these transactions take place.

This is a remarkable assertion. It is somewhat akin to presuming that the success of a corporation depends entirely upon its organizational structure; that it makes no difference to the company whether the actual positions within the organizational chart are filled by bright, energetic individuals or a confederacy of dunces. What could ever have led the functionalists (who, after all, are trying to defend materialism) to ignore the actual physical components within the brain? In all likelihood, what we have here is a vestige from the days in which dualism dominated the philosophical scene. In historical context, the functionalists had to explain how the ‘mental’ (a presumably distinct category of events) differed from the ‘physical’. Since at least some ‘mental’ output could be described as the outcome of logical operations (two plus three equals five, knight to king bishop three), they declared that every ‘mental’ operation is equally the embodiment of a mathematically describable logical operation. When physical operations are clearly part of an event (as in the sensing of an object or the manipulation of it), the functionalists, very much like the dualists, presume that some exterior devise (a television camera or robot arm) could take care of this without interfering with the truly ‘mental’.

Relegating the ‘physical’ to the exterior, however, is not that easy. Perhaps this can best be illustrated by considering an argument put forth by David Chalmers. He envisions a typical human being, Robert, having the neurons of his brain replaced one by one by silicon chips which perform the same functions. The point of the argument is to show that as long as the silicon chips carry out the same functions as the neurons which they replace, we can transform a human being, Robert, into a machine, Robot, without this individual losing consciousness. It is important to note, however, that the neurons do more than channel electrons; they are also chemical factories. Chalmers himself notes this in passing by saying, "We can imagine that [the replacement chip] is equipped with tiny transducers that take in electrical signals and chemical ions and transforms these into a digital signal upon which the chip computes, with the result converted into the appropriate electrical and chemical outputs." (Emphasis added).

Obviously, something more than mere computing (that to which Church’s thesis applies) is taking place in this thought experiment. Apparently, the ‘silicon chip’ which is to replace a neuron has a pharmaceutical laboratory available through which it can create and dispense appropriate chemicals. Thus, it is disingenuous to suggest that a mere silicon chip can serve as a neuron replacement. In order to have the replacement ‘chip’ carry out all the appropriate functions (including the physical as well as the logical), one would have to have something that looks very much like and behaves like a neuron. What we do not know is to what extent the physical functions are responsible for the existence of consciousness.

Who knows? Perhaps when we solve the greatest remaining mystery of the universe--what is consciousness and how does it interact with (the rest of?) matter--we will discover that building a consciousness machine is no more difficult than building a radio transmitter. Perhaps the functionalists are right in asserting that it can be done with a wide variety of materials, including silicon. After all, radio waves can be produced, not only by vacuum tubes and transistors, but intrastellar gasses. We cannot, however, just assume that silicon can carry out all the appropriate functions, or, do so in the appropriate way. Once we know exactly what functions must be performed, it might well be that getting a silicon based computer to exhibit consciousness is as impossible as getting Gary Kasparov’s neuron based brain to examine a million board positions per minute. Silicon may simply be an inappropriate material to work with.