David Berlinski - The Devil's Delusion

Although imaginary at its inception, a Turing machine brilliantly anticipated its own realization in matter, with Turing’s ideas giving rise to the modern digital computer.

The promotion of the computer from an imaginary to a material object serves the purpose of restoring it to the world that can be understood in terms of the physical sciences. As a physical device, nothing more than a collection of electronic circuits, the digital computer may be represented entirely by Clerk Maxwell’s theory of the electromagnetic field. The distinction between a computer and its program is duplicated in the distinction between a physical system governed by certain specific laws and its initial condition—the state from which it starts. We are returned to the continuous and infinite world in which mathematical physics tracks the evolution of material objects moving through time in response to the eternal forces of nature itself.

If something is gained by the assimilation of the brain to a computing device, something is lost as well, and that is the recognition that deep down every one of these metaphors is profoundly limited. A certain “power to alter things,” Albertus Magnus remarked, “indwells in the human soul.” The existence of this power is hardly in doubt. It is evident in every human act in which the mind imposes itself on nature by taking material objects from their accustomed place and rearranging them, and it is evident again whenever a human being interacts with a machine. Writing with characteristic concision in the Principia, Isaac Newton observed that “the power and use of machines consist only in this, that by diminishing the velocity we may augment the force, and the contrary.” Although Newton’s analysis was restricted to mechanical forces, his point is nonetheless general. A machine is a material object, a thing, and as such its capacity to do work is determined by the forces governing its nature and by its initial conditions. Before an inclined plane can do work, it must be inclined.

Those initial conditions must themselves be explained, and in the nature of things, they cannot be explained by the very device they serve to explain. An inclined plane does not incline itself. This is precisely the problem that Newton faced in the Principia. The magnificent system of the world that he devised explained why the orbit of the planets should be conic sections, but Newton was unable to account for the initial conditions that he had himself imposed on his system.

This pattern, along with its problem, recurs whenever machines are at issue, and it returns with a vengeance whenever computers are invoked as models for the human mind.

If the brain is a computer, then the very same thesis about the human mind should be in force whether we describe the human mind as a digital computer or whether we describe the human mind in terms of a device that is logically identical to a digital computer—an abacus, say. The thing is a trifle. Made of wood, it consists of a number of wires suspended in a frame and a finite number of beads strung along the wires. Nevertheless, an idealized abacus has precisely the power of a Turing machine, and so both the abacus and the Turing machine serve as models for a working digital computer. By parity of reasoning, they also both serve as models for the human mind.

Yet the thesis that the human mind is an abacus seems distinctly less plausible than the thesis that the human mind is a computer. And for an obvious reason: It is absurd. It is precisely when things have been reduced to their essentials that the interaction between a human being and a simple machine emerges clearly. That interaction is naked, a human agent handling an abacus with the same directness of touch that he might employ in handling a lever, a pulley, or an inclined plane.

With the nakedness of interaction revealed, a characteristic problem is revealed as well. While an abacus may represent certain human intellectual operations such as addition or subtraction, it cannot represent its own initial conditions. Regarding her big-nosed customers with indifference, it is the Chinese cashier at the Imperial Gardens who does that. The force that she brings to bear on an abacus is muscular, and so derived from the chemistry of the human body, these causes ultimately emptying out into the great ocean of physical interactions whose energy loosens and binds the world’s large molecules.

No known chain of causes accommodates the inconvenient fact that by setting the initial conditions of a simple machine, a Chinese cashier has brought about a novel, an unexpected, an entirely idiosyncratic distribution of matter. The initial state of any mechanical artifact represents what the anthropologist Mary Douglas called “matter out of place.” Explaining even the simplest of human acts, the trivial tap or touch that sets a polished wooden bead spinning down a wire, requires tracing the causal chain backward. But that leads only to a wilderness of causes, each of them displacing material objects from their proper settings, so that in the end the mystery is simply shoveled back until the point is reached when it can be safely ignored.

A chain of physical causes is thus not obviously useful in explaining how the human mind imposes itself on matter. But neither does it help to invoke the hypothesis that another abacus is needed to fix the initial conditions of the first. If each abacus requires yet another abacus in turn, the road lies open to the madness of an infinite regress. Daniel Dennett has argued in Brainstorms: Philosophical Essays on Mind and Psychology that if receding computers are, like himself, diminished in their capacity, the regress may end in some trivial mechanical device, one that he describes as “stupid.” But if those receding computers are too thick to function as models of the mind, how do they do what they are said to do?

And if they are not, how have we been advanced?

If we are able to explain how the human mind works neither in terms of a series of physical causes nor in terms of a series of infinitely receding mechanical devices, what then is left? There is the ordinary, very rich, infinitely moving account of mental life that without hesitation we apply to ourselves. It is an account frankly magical in its nature. The human mind registers, reacts, and responds; it forms intentions, conceives problems, and then, as Aristotle dryly noted, it acts.

And in none of this do we seem to be doing anything that can be explained or expressed in terms of what the brain does, or what any machine can do.

“Mind is like no other property of physical systems,” the physicist Erich Harth has reasonably remarked. “It is not just that we don’t know the mechanisms that give rise to it. We have difficulty in seeing how any mechanism can give rise to it.”

One of the curiosities about the current enthusiasm for various pseudo-scientific accounts of the human mind is that deep down those most willing to promote its premises are least willing to accept its conclusions.

Whatever scientists may say on those all too frequent occasions when they are advising the rest of us what to think, one thing that they do not say is that they believe what they are telling us to think. The result at times is moving. Writing to the widow of his old friend Michele Besso, Einstein remarked that “now he has departed from this strange world a little ahead of me. That means nothing. People like us, who believe in physics, know that the distinction between past, present, and future is only a stubbornly persistent illusion.” Whatever the illusion, he acknowledged ruefully, it is one “stubbornly held.”