- It is easy to agree with the negative point Searle makes about mind and AI in this stimulating paper. What is difficult to accept is Searle's own conception of mind.
- His negative point is that the mind can never be a computer program. Of course, that is what behaviorists have said all along ("residual behaviorism" in the minds of AI researchers notwithstanding). His positive point is that the mind is the same thing as the brain. But this is just as clearly false as the strong AI position that he criticizes.
- Perhaps the behaviorist viewpoint can best be understood through two examples, one considered by Searle and one (although fairly obvious) not considered. The combination robot example is essentially a behavioral one. A robot behaves exactly like a man. Does it think? Searle says "If the robot looks and behaves sufficiently like us, then we would suppose, until proven otherwise, that it must have mental states like ours"(italics mine). Of course we would. And let us be clear about what this robot would be required to do. It might answer questions about a story that it hears, but it should also laugh and cry in the right places; it should be able to tell when the story is over. If the story is a moral one the robot might change its subsequent behavior in situations similar to the ones the story describes. The robot might ask questions about the story itself, and the answers it receives might change its behavior later. The list of typically human behaviors in "response" to stories is endless. With a finite number of tests we can never be absolutely positive that the robot understood the story. But proof otherwise can only come from one place - the robot's subsequent behavior. That is, the robot may prove that it did not understand a story told to it at time-X by doing or saying something at a later time that a normal human would not do who heard a similar story under similar conditions. If it passes all our behavior tests we would say that, pending future behavioral evidence, the robot understood the story. And we would say this even if we were to open up the robot and find a man translating Chinese, a computer, a dog, a monkey, or a piece of stale cheese.
- The appropriate test is to see whether the robot, upon hearing the story, behaves like a normal human being. How does a normal human being behave when told a story? That is a valid question - one in which behaviorists have been interested and one to which Searle and his fellow mentalists might also profitably devote their attention when they finish fantasizing about what goes on inside the head. The neural mythology that Searle suggests is no better than the computer-program mythology of the AI researchers.
- Searle is willing to abandon the assumption of intentionality (in a robot) as soon as he discovers that a computer was running it after all. Here is a perfect example of how cognitive concepts can serve as a mask for ignorance. The robot is said to think until we find out how it works. Then it is said not to think. But suppose, contrary to anyone's expectations, all of the functional properties of the human brain were discovered. Then the "human robot" would be unmasked, and we might as well abandon the assumption of intentionality for humans too. It is only the behaviorist, it seems, who is willing to preserve terms such as thought, intentionality, and the like (as patterns of behavior). But there are no "mental states underlying. . . behavior" in the way that a skeleton underlies our bodily structure. The pattern of the behavior is the mental state. These patterns are results of internal and external factors in the present and in the past - not of a controlling mental state - even one identified with the brain.
- That the identification of mind with brain will not hold up is obvious from the consideration of another example which I daresay will be brought up by other commentators - even AI researchers - so obvious is it. Let's call it "The Donovan's brain reply (Hollywood)." A brain is removed from a normal adult human. That brain is placed inside a computer console with the familiar input-output machinery - tape recorders, teletypewriters, and so on. The brain is connected to the machinery by a series of interface mechanisms that can stimulate all nerves that the body can normally stimulate and measure the state of all nerves that normally affect muscular movement. The brain, designed to interact with a body, will surely do no better (and probably a lot worse) at operating the interface equipment than a standard computer mechanism designed for such equipment. This "robot" meets Searle's criterion for a thinking machine - indeed it is an ideal thinking machine from his point of view. But it would be ridiculous to say that it could think. A machine that cannot behave like a human being cannot, by definition, think.
Text Colour Conventions (see disclaimer)
- Blue: Text by me; © Theo Todman, 2019
- Mauve: Text by correspondent(s) or other author(s); © the author(s)