The Problem of Consciousness:

A Debate with B. F. Skinner

Brand Blanshard

Opening Remarks by

Professor Blanshard

Behaviorism is the view that mind can be adequately studied through the behavior of the body. Sometimes it is offered as a method only, implying nothing as to whether a distinct realm of consciousness exists. But it clearly cannot stop there. If there are in fact conscious events distinct from bodily events, a method that disregards them and confines itself to the body cannot be adequate to the study of mind. If behaviorism is to be adequate as a method, it must also be sound as a philosophy; it will give us an adequate science of mind only if mental behavior is bodily behavior. And this bold conclusion is one that both Watson and Professor Skinner have had the courage to draw.

What is the difference between the older and the newer behaviorism? It is not a difference in philosophy; it is a difference in policy over what it is in behavior that may be most profitably studied. Watson held that the events to be studied under such heads as sensation, perception, and imagination were events within the body, and particularly in the nervous system; he held what Lovejoy called a hypodermic theory of consciousness. Professor Skinner’s approach is different. Without denying that events in the nervous system do largely determine behavior, he is not much interested in them, for he holds that they are all themselves conditioned by stimuli from the outside. The best course for the psychologist, therefore, is to bypass these minute and often conjectural nervous changes and to correlate observable stimuli directly with observable bodily responses. This is the new program for psychology, which is offered with the promise that if it could be carried through, it would explain every kind of human behavior.

The layman will at once have a question. Is he to understand that conscious motives, feelings and ideas have no part in determining conduct or therefore in explaining it? To this question Professor Skinner’s answer is essentially the same as Watson’s: ideas, motives, feelings in the traditional sense are not, even in principle, observable events, and science can take no account of them. Does this mean that they may be real events, though beyond scientific range? No. There are no events that, if factual at all, are beyond the interest of science. Are we then to dismiss distinctively conscious events altogether? Watson said Yes, on the ground that he could find no such events in his testtubes. Professor Skinner would put it less crudely, but I think his conclusion is in substance the same. He says that to resort to “nonphysical events” is to offer a “fictional explanation” (Skinner, 1953, references at end); the belief in a mind or consciousness irreducible to any form of bodily change seems to him an anachronism, a last lingering survival of primitive animism (Skinner, 1953).

I believe this view to be radically mistaken. I shall urge three points against it: (1) that there is such a thing as consciousness, irreducible to physical change; (2) that its denial, instead of according with modern science, is in conflict with it; and (3) that to reject the efficacy of consciousness makes nonsense of practical life (Blanshard, 1939).

First, then, there is such a thing as consciousness, distinct from bodily change. For a philosopher to be called upon to prove this is a strange sort of challenge. Modem philosophy began with what I think is the valid insight that consciousness is the most certain thing in the world. Descartes showed that you can doubt with some plausibility the existence of rocks and rivers; you can doubt the existence, at least when unobserved, of your own hands and feet. But you cannot coherently doubt that you are conscious, for to doubt it is to be conscious; you establish the fact of consciousness in the very act of doubting it. The behaviorist may reply that by “doubting” Descartes must have meant some physical response, and in taking it for a form of consciousness he was merely deluded. But this will not do either, for a delusion is clearly a fact of consciousness; matter in motion cannot be deluded.

But the behaviorist has persuaded himself that matter in motion is all there is. And if you talk about sensing or perceiving or thinking, he is commonly prepared to tell you what movements of matter you ought to mean by these words. For Watson, a toothache was a change in a dental nerve. What is a toothache for Professor Skinner? He says that when a person reports “my tooth aches,” he is making a verbal response to an event within his body. What sort of event? Though Professor Skinner adheres to his policy of saying little about such events, this much seems to be clear: the event is a physical event, a movement of matter, since it can be investigated by science, and this is the only kind of event that science recognizes. On the point of greatest interest, Professor Skinner’s behaviorism is thus at one with Watson’s.

And its major difficulties are the same. The most obvious one is that the experience of pain, for example, is self-evidently not the same thing as a physical movement of any kind. That their identification is a confusion can be shown in various ways. First their properties are different. If a pain were any kind of physical motion, we could ask what its direction and velocity were, whereas it makes no sense to talk of the direction or velocity of a toothache. On the other hand, we speak of the pain as dull or excruciating, while a dull or excruciating motion is meaningless again.

Secondly, no behaviorist behaves as if his theory were true. For example, he has occasionally, like the rest of us, had to visit his dentist. The dentist says, “Mr. B., this may be pretty painful; shall I use a little cocaine?” “Yes, by all means,” says Mr. B. He clearly wants to avert something; what? On his theory, it is a particular physical event, about whose nature he is not very clear. But why should he select this event for his aversion? There is nothing in it as a physical change to make it more objectionable than a million others. What makes it objectionable is plainly just one thing: it happens to carry with it, no one knows why, an excruciating pain, which is as different from the physical change that conditions it as are any two things in the world. If the behaviorist admits that it is this he is trying to avert, he has broken out of his behaviorism; if he denies it, his own conduct refutes him.

Thirdly, he continually deserts his program in his experimental practice. He agrees that, as a psychologist, he is concerned with such events as feeling rage or sorrow, as seeing colors and hearing sounds. What he insists on is that these names be now given not to conscious experiences, which cannot be scientifically studied, but to physical responses that can. But what responses are these old names now to mean? It would be ridiculous to give the name “rage” to the limp and drooping behavior of sorrow; one must give it to the kind of behavior that we know as the normal expression of rage. But what that implies is that even in assigning the new terms, we must fall back on the old meanings. “Rage” is now to mean the kind of behavior that we can link uniquely to the emotion of rage. We must resort to the conscious state as our only reliable index to the behavior that is supposed to supplant it. Perhaps the oddest achievement of this kind is the Watsonian physiological psychology. Watson would identify seeing blue with the response of the optic nerve to a definite kind of stimulation, namely that of light waves about seven microns long, and seeing red with the response to waves about four microns long. But how did he know which of these minutely different responses to call “seeing blue” and which “seeing red”? It was by starting with the conscious sensation and then looking for its nervous correlate—which meant, of course, that at every step of his experimental inquiry he assumed the very difference between sensation and nervous event which it was his main purpose to deny.

Behaviorists, old and new, have been forced into inconsistency in another way. They are concerned to explain why some responses are acquired and others stamped out. A child, if it tastes sugar, will take it readily in the future, and if given quinine will in the future avoid it. Why? Common sense and traditional psychology answer, Because it finds one pleasant and the other disagreeable. The behaviorist must profess not to know the meaning of such words. “‘Pleasant’ or ‘satisfying’,” writes Professor Skinner, “apparently do not refer to physical properties of reinforcing events, since the physical sciences use neither these terms nor any equivalents” (Skinner, 1953). The business of science is simply to report that the sugar response is “reinforced” and the quinine response “extinguished.” But why should sugar reinforce a response and quinine inhibit it? There is nothing in the physical properties of either, taken alone, that throws the least light on this. Hence even behaviorists as skilful as Professor Skinner in excluding reference to conscious experience find themselves willy-nilly using terms that mean such experience—or else mean nothing at all. Professor Skinner writes, for example, “though we have been reinforced with an excellent meal in a new restaurant, a bad meal may reduce our patronage to zero”; and on the next page, that in extinguishing a response, “the currently preferred technique is punishment” (Skinner, 1953). Now do such terms as “excellent,” “bad,” and “punishment” refer to pleasure and pain or not? If not, they are wholly irrelevant in explaining why one response is maintained and another dropped. If they do help in explaining this, as Professor Skinner here implies that they do, it is because they are drawing their force from an interpretation that behaviorism denies them.

Why should behaviorists reject this interpretation? They answer, as we have seen, that it is because it is vetoed by science, and they want above all things to be scientific. “The methods of science,” writes Professor Skinner, “have been enormously successful wherever they have been tried. Let us then apply them to human affairs.” What Professor Skinner seems here to be saying is that he is moved by the desire to explain things scientifically, though he cannot mean by this what most people mean. The behaviorists of the sixties, like the positivists of the thirties, dismiss as meaningless all references that go beyond physical things and events, entities that at least in principle can be publicly observed. If there was anything purely psychical about us, like old-fashioned desires and sensations, no one else could ever hope to observe them, or therefore meaningfully speak of them.

This curious “physicalism” of the early positivists has now been abandoned by the positivists themselves; Ayer, Feigl, Hempel and Carnap have all renounced it. Behaviorists apparently still cling to it. They still think it demanded by science, though in fact it is deeply at odds with science. That is my second thesis. It will be enough to support it with a single instance.

Consider the behaviorist treatment of images. Watson took the heroic course of denying that there were such things, thereby, as someone remarked, exalting a personal defect into an ontological principle. I shall not pause to argue whether images exist; I shall assume that they do. No one among us who is a fair visualizer will have any trouble in summoning up, for example, the face of Einstein, with its cloud of white hair, its moustache and wrinkles, and its sad, dark eyes. What is the behaviorist to do with such images? He cannot say that when we talk of them, we mean only certain motions among the particles in the brain, for we know too well that we do not. Nor can he say, with common sense and traditional psychology, that they exist in the realm of consciousness, for now there is no such realm. Professor Skinner struggles manfully with the problem, admitting that “perhaps the most difficult problem in the analysis of behavior is raised by responses beginning ‘I see. ..’, ‘I hear. . .’, and so on, when customary stimuli are lacking” (Skinner, 1953). He recognizes that such reports may be true; he has read the famous chapter of Galton in which a portrait painter is described whose imagery was so vivid that he could paint from it after his sitter was gone. Professor Skinner’s explanation of such performances is that seeing is a response which may be conditioned to occur in the absence of its normal stimulus, like the salivating of Pavlov’s dog. This is no doubt true, but it leaves the critical question untouched. What is the status of the imaginary or hallucinatory object that the subject admittedly “sees”? It is in the physical world for on the behaviorist account, there is nowhere else for it to be. But it is not in the subject’s head, for no surgeon, nor the subject himself with the help of mirrors, will ever find it there. And of course it is not in consciousness; Professor Skinner avoids using the word “imagine” and insists on using “see”, even in cases of hallucination. Now if the object thus admittedly seen is not in consciousness and not in the head, where is it? One can only conclude that it is out there in nature. But if this conclusion is offered as consistent with modem physics, it is not hard to predict the comment of the physicist. It would certainly be eloquent, but it might not be quotable.

Images have always offered special difficulty for the behaviorists, and I do not want to exploit that difficulty unduly. So perhaps I should add that precisely the same difficulty is occasioned by ordinary perception. Professor Skinner uses the example of seeing a rainbow in the sky. What is it that we are here responding to? He does not, as Watson did, develop the fact of light waves impinging on nerve ends. He holds, and rightly, that these are not what we see; what we are seeing is the reds, greens and yellows of the bow in the sky. But then these reds, greens and yellows are not really there at all. Would any responsible physicist admit for a moment that they are? He would admit, of course, that the vibrations are there, but that is a quite different matter. The whole modem tradition of physics from Newton down has relegated the reds, greens, and yellows to sensations, sense-data in our minds, caused indeed by outer vibrations, but utterly different from them. Bertrand Russell did attempt fifty years ago to work out a philosophy of science that would place even colors, tastes, and smells out there in nature, but in his last important book in philosophy (Russell, 1948), he comes full circle. He holds that everything we directly perceive exists in consciousness, and that the entire world of physical science is a speculative construction built on that foundation. And about the secondary qualities such as colors, sounds and smells, physics would surely agree with him. Regarding these qualities, Professor Skinner seems to me caught in a dilemma. If he puts them in physical nature, he is at odds with science. If he puts them in consciousness, he abandons behaviorism. And I doubt if he can find an ontological purgatory between these two.

My third thesis is that behaviorism leaves a vacuum at the heart of our moral and practical life. It makes us out to be hollow men in a wasteland. It tells us that we are machines—enormously complicated machines, but in the end nothing more. Let us assume for the moment that this is true, and ask what would be the value of a world in which only such machines existed and that unscientific embarrassment, consciousness, did not exist. The answer I suggest is simple: it would have no value at all. Consciousness, however frail and evanescent, is the seat of all goods and evils, of all values of all kinds, and they would go out with it like a candle.

Run down the list of things which, in the opinion of major thinkers, make life worth living. What are they? They are such things as pleasure or happiness, wisdom or understanding, friendship, the sense of beauty, the sense of duty. All of them, you will note, are forms of consciousness. The wisdom prized by Spinoza was not that of a computer, even the giant one pictured by the New Yorker which, to the dismay of its mechanics, is issuing from its depths a slip reading “Cogito ergo sum.” What Spinoza meant was conscious insight, the experience of understanding. The pleasure stressed by Bentham and Mill, the friendship prized by Epicurus, the beauty prized by Schopenhauer, were the experience of these things; and when Kant put good will at the top of them, he meant the conscious recognition and choice of duty. These men would all have been bewildered by the suggestion that the goods they spoke of lay in the play of nerves or limbs; they would have said that if by some miracle an unconscious robot were to duplicate the story of St. Francis or Casanova, there would be no good or evil about it. With that judgment I agree. In an account of human conduct that confines itself to physical change, everything of intrinsic value has been left out.

Nor is that all. The behaviorist is committed to ignoring not only the goods that give life value, but also the motives that make men seek them.* [* The criticism has been developed in three recent books, Joseph Wood Krutch’s The Measure of Man, Floyd W. Matson’s The Broken Image, and Jacques Barzun’s Science the Glorious Entertainment. ] For him the only causes of human behavior lie in its physical conditions. Professor Skinner points out and deplores that it is a “common practice to explain behavior in terms of an inner agent which lacks physical dimensions and is called ‘mental’ or ‘psychic’’’; he gives as an instance of this mistake the remark about a lecturer that “what he says is often disorganized because his ideas are confused” (Skinner, 1953). There can be “no violation,” he says “of the fundamental principle of science which rules out ‘final causes’” (Skinner, 1953). Conscious purposes and intentions, even if they existed, could not affect bodily behavior; “since mental or psychic events are asserted to lack the dimensions of physical science, we have an additional reason for rejecting them” (Skinner, 1953).

Put abstractly in this way, the doctrine seems not implausible. But consider what it means in the concrete. Here are a few of the things that strictly and literally follow: Your conscious interest or desire to hear a discussion of behaviorism today had no part in making you come here. St. Francis’s love for his fellows, if that means a feeling as distinct from physical conditions, had no influence on how he treated them. Hitler’s hatred of the Jews contributed nothing toward his orders to have them exterminated. No gourmet has ever chosen one item on a menu rather than another because of his desire for a pleasanter food. Newton’s theoretical interest had no part in keeping him at his desk, nor did his ideas of gravitation ever affect in the slightest what he said or put down on paper. No one has ever done anything because he saw it to be his duty, or even because he mistakenly thought it to be his duty. The novelists, dramatists, and historians of the world have been governed by a unanimous illusion. They have represented Othello as moved to action by feelings of jealousy, and Romeo by his love of Juliet, and Silas Marner by his love of money, and Napoleon by his love of power, and Churchill by his love of England. A psychologist with full understanding would have freed himself from such delusions. He would see that all these “mental states” are equally irrelevant in explaining how people behave.

Of course this disparity between the judgment of the world and that of the behaviorist does not prove the behaviorist wrong. But it does show who must assume the burden of proof. And if one asks which is the more probable: that every major moralist, historian, and man of letters has been talking nonsense about human life, or that a set of brilliant young psychologists should have been carried away by excess of zeal, one cannot hesitate long.

References

1. Skinner, B. F., Science and Human Behavior (MacMillan, 1953), pp. 278, 29, 81, 70, 71, 265-6, 30 31.

2. Russell, Bertrand, Human Knowledge, Its Scope, and Limits (London: Allen and Unwin), 1948.

3. Blanshard, Brand, The Nature of Thought (London: Allen and Unwin; N.Y.: Humanities Press, 1939); Chapter IX is an examination of the older behaviorism.

Reply by Professor Skinner

Professor Blanshard correctly paraphrases the behavioristic principle that ideas, motives, and feelings have no part in determining conduct and therefore no part in explaining it, but he is wrong in saying that it is because these things are not observable. No major behaviorist has ever argued that science must limit itself to public events. The physicalism of the logical positivist has never been good behaviorism, as I pointed out twenty years ago (Skinner, 1945).¹ In an adequate science of behavior nothing that determines conduct can be overlooked no matter how difficult of access it may be. To make inferences about private events is not to abandon or destroy an “objective” position. As a behaviorist, however, I question the nature of such events and their role in the prediction and control of behavior.

I do not expect to find answers in any “hypodermic” theory of consciousness. It is the introspective psychologist who would escape from dualism through physiology. The organism is not empty, and it is important to study what goes on inside it, but most physiologists are looking for the wrong things. No matter how much they may improve their techniques, they will never find sensations, thoughts, or acts of will. On that point Professor Blanshard and I agree, but I do not agree that mental events are thereby shown to be irreducible to physical change.

Behaviorism begins with the assumption that the world is made of only one kind of stuff—dealt with most successfully by physics but well enough for most purposes by common sense. Organisms are part of that world, and their processes are therefore physical processes. In studying behavior, especially the behavior of men, we exhibit further instances of the thing we are studying. In thus behaving about behaving we may raise some tricky problems, but they have their counterparts in “thinking about thinking.” A special problem arises from the inescapable fact that a small part of the universe is enclosed within the skin of each of us. It is not different in kind from the rest of the universe, but because our contact with it is intimate and in some ways exclusive, it receives special consideration. It is said to be known in a special way, to contain the immediately given, to be the first thing a man knows and according to some the only thing he can really know. Philosophers, following Descartes, begin with it in their analysis of mind. Almost everyone seems to begin with it in explaining his own behavior. There is, however, another possible starting point—the behavior of what Max Meyer used to call the Other-One. As a scientific analysis grows more effective, we no longer explain that behavior in terms of inner events. The world within the skin of the Other-One loses its preferred status. But what about the so-called introspective evidence?

Let us consider a few examples. The so-called conative aspects of consciousness have to do with the initiation and direction of action. A man is said to act when he wills to do so and to act in a given way to fulfill a purpose. A competing, external explanation is suggested by the concept of the stimulus. In determining which response will occur and when, a stimulus usurps some conative functions. But a great deal of behavior is not simply elicited by stimuli. It is true that stimulus-response theorists still try to leave the initiation and selection of behavior to a “total stimulating situation,” but we often know that such a situation has acted as a stimulus only after the fact. In the concept of operant behavior, not to be confused with stimulus-response theories, the stimulus is merely the occasion for action. The initiation and selection of a response are matters concerning its probability, and probability of response is determined by other variables—in particular, by the contingencies of reinforcement. I have suggested that operant reinforcement is simply a more effective formulation of purpose (Skinner, 1963).² To the extent that an analysis of contingencies of reinforcement permits us to predict and control when and how a man will act, we have nothing to gain by speaking of a purposive act of will.

So-called cognitive functions of consciousness are also displaced by an adequate analysis of the Other-One. A hungry pigeon will repeatedly peck a colored disk if reinforced with food when it does so, but it adds nothing to say that it pecks because it knows that it will get food. Similarly, if it has been reinforced for pecking a red disk, it will also peck a yellow one, though not so rapidly, but it adds nothing to say that it has seen some similarity in the two disks and has mentally generalized from one to the other. Observable contingencies of reinforcement also account for discrimination, abstraction, and concept formation, as well as for other kinds of changes in behavior said to show cognitive processes. To the extent that they do so adequately, no explanatory role survives to be played by mental processes as such.

But the question is, do we not see conative and cognitive activities? Before answering it, we should ask what it means to see anything. Here the mentalist has almost succeeded in sealing off a behavioristic analysis. It has long been supposed that sensing or perceiving is something man does to the environment; we are only beginning to discover what the environment does to man. The modern concept of stimulus control reverses the very direction of action and, for that and other reasons, is not readily seen to be equivalent to “direct experience.” A convenient account of stimulus control has been written by Herbert Terrace (1965).³ Instead of asking our subject to describe “what he sees,” using his everyday vocabulary supplemented by a few technical terms, we can bring his behavior under the control of explicit stimuli. The subject can be an animal as well as a man. A nonverbal psychophysics, with either animals or men, is proving to be much more feasible than early critics of behaviorism foresaw. The mentalist will say that we are still getting our subject to tell us what he sees—even when he is an animal. But nothing comparable to experience enters into the formulation, although the same contingencies of reinforcement account for the same behavior, including behavior said to report the content of consciousness.

If the mentalist insists that we have still left out “seeing” itself, we must ask him what he means. What has been left out? What is seeing? What does a perceiver do—either to, or about, or in response to the world he perceives? “To experience” often seems to mean simply “to be in contact with.” We know the world in the social sense of having been introduced to it. Knowledge is a form of “acquaintance” (the word comes from the same root as cognition). But the action is often more positive. The perceiver ‘apprehends the world almost as one apprehends a criminal. He makes it his own almost as if he were ingesting it, as one ingests the body of a god in the rites of Mithra. He knows the world almost in the biblical sense of possessing it sexually.

Contact and possession are important elements in the concept of experience because they are related to the physical privacy of the world within one’s skin. They have led to the view that the world beyond one’s skin can be known only from copies. When the copies are discovered to be bad, it is concluded that the world can never be known as it really is. Richard Held has described the search for copies of reality and for an explanation of their shortcomings in his paper Object and Effigy (1965).⁴ Physiologists and neurologists have tried to support the notion of inner copies by tracing the stimulus into the nervous system, where it can be more intimately possessed, but their facts seem to show that the organism begins at once to analyze stimuli and to respond to them in ways which cannot be regarded as the construction of effigies. The behaviorist has no interest in experience as contact or possession, and he quite sensibly leaves the environment where it is. And even if private copies of it existed in the external world, they would not answer the question we are asking. Put the thing you see wherever you like—at the surface of the organism, in the heart of the nervous system, or in the mind—what does it mean to say that you see it?

The behaviorist must give his own answer, and like everyone else he finds it difficult. The study of operant behavior suggests that we look to the contingencies of reinforcement. How are responses brought under stimulus control? What responses are involved? A pigeon may learn to discriminate between colors in pecking at disks for food, but surely there is nothing about seeing color which involves pecking disks. Usually, however, an enormous number of responses come under the same kind of stimulus control, and it is possible that seeing color is something common to them. Precurrent responses of observing or attending to stimuli may be important. In verbal behavior certain generalized contingencies maximize control by stimuli and minimize control by other variables, producing the kind of response I have called a tact—and may be especially relevant. Naming a color is no closer to seeing a color than pecking a colored key, but the psychophysicist usually accepts it as an equivalent activity. A similar analysis of contingencies may eventually explain why a subject responds to one feature of a stimulus rather than to another or sees one thing as if it were something else.

All these issues, together with their uncertainties, arise when we are said to see our acts of will, our thoughts, our sensations, our images. Things of this sort should be seen with special clarity. They are inside our bodies; we are in contact with them; we possess them; there is no need for copies. But the fact is, they are not seen clearly at all. Two people rarely agree about them, and for good reason. A private event may be close to the perceiver, but it is remote from the environment which teaches him to perceive. We learn to tell the difference between colors when people reinforce our behavior appropriately while manipulating colored objects. We cannot learn the difference between a resolution and a wish, or between guessing and believing, in the same way. Certain techniques permit the verbal environment to circumvent the fact of privacy, but only to a limited extent (Skinner, 1945; 1957).⁵ In particular, we describe our past and present behavior, even when it is not visible to others, and we appear to describe our future behavior when we announce an intention. We often do so simply because we are in a favorable position to observe the variables of which our behavior is a function, and many of these are public, but whether public or private, all the events described are physical.

Professor Blanshard must deny this. “The experience of pain. . . ,” he says, “is self-evidently not the same thing as a physical movement of any kind.” His example is as nearly self-evident as possible, but I am afraid it does not quite make the grade. Painful stimuli, being inside the body, are both particularly strong and hard to observe in other ways, and they do not need to be copied. The experience of an external object would serve Professor Blanshard less well as an example. The expression “physical movement,” taken from a contemporary analysis of matter, also contributes to the plausibility, for a motion is not likely to be dull or excruciating. But things are. Indeed, these two adjectives are applied to pain just because they apply to the things which cause pain. A dull pain is the kind of pain caused by a dull object, as a sharp pain is caused by a sharp object. The term “excruciating” is taken from the practice of crucifixion.

Ideas, motives, and feelings are more important to Professor Blanshard’s argument than pain, and they are by no means so self-evidently perceived. The processes or states which such terms describe acquire control very slowly. Only a long and complicated history of reinforcement leads one to speak of sensations, images, and thoughts. Such a history is characteristic only of certain cultures. Our own culture shows wide variations—it produces the thoroughgoing extravert on the one hand and the introspective psychologist and philosopher on the other. Some sort of history of reinforcement is essential. Descartes could not begin, as he thought he could, by saying “Cogito ergo sum.” He had to begin as a baby—a baby whose subsequent verbal environment eventually generated in him (though not, I am sure, in millions of his contemporaries) certain subtle responses, one of which was “cogito.” It is a loose term, not because the events it is said to describe are necessarily vague, but because they are almost inaccessible to the verbal community which builds all descriptive verbal repertoires.

Professor Blanshard seizes upon my admission that the problem of images is difficult. It is difficult for the mentalist too. What do you see when what you see is not really there—as in a memory, a fantasy, or a dream? But first, what do you see when what you see is there? Someone shows you a picture of a group of scientists, among them Einstein. He asks, “Is Einstein there?” and you say, “Yes,” as you have been taught to do in thousands of comparable cases. But suppose he asks, “Do you see Einstein?” and you say, “Yes.” What have you reported? Did you, in response to his question, simply look at Einstein a second time? If so, how do you distinguish between “seeing Einstein” and “seeing that you are seeing Einstein”?

A possibility which needs to be considered is that in reporting that you see Einstein, you are reporting a response rather than a stimulus. No matter how obscure its dimensions, the behavior called seeing must be involved, and you must be reporting it, rather than the presence of the thing seen, when you report that you see something. You may be reporting the same thing when you report that you see something which “isn’t really there”—when you are merely “imagining how Einstein looked.” Seeing something in memory is not necessarily seeing a copy. The whole concept of “stored data structures” —supported without warrant by the analogy of the computer—may be quite wrong. What is stored may be responses—in this case the responses involved in seeing. When I recall how something looked, I may simply be recalling how I once looked at something. There was no copy inside me when I first looked at it, and there is none now. I am simply doing again what I once did when I looked at something, and I can tell you that I am doing so.

In another criticism of behaviorism, Professor Blanshard⁶ has written, “The bats’ heads taken by the toper to be emerging from the wallpaper are clearly not physical objects—what in the world would physics do with them? They are not processes in the nervous system; they are not muscular responses; and it is absurd to turn one’s back on them and insist that they are nothing at all.” But why is it absurd? We have no more reason to assume the existence of imaginary or hallucinatory objects than of memories as sense data. The toper is quite literally “seeing things.” His behavior would be most readily evoked by actual bats’ heads emerging from wallpaper, but it can occur, especially after toping, in the absence of any such stimuli. Physiologists may someday find the precursors or mediators of that behavior, but they will never find anything that looks like a bat, for there is nothing of that sort inside the skin at any time—whether external stimuli are present or not.

The same explanation may hold for seeing something which does not closely correspond with an external stimulus. Under certain conditions, analyzed in the study of perception, the response of seeing a straight line may be evoked by a curved line. We do not need to say that the straight line is in “the world as it appears to be” and the curved line in “the world as it really is.” There is no world as it “appears to be”; there are simply various ways of seeing the world as it is—in this case, a way of seeing a straight line in response to a curved line as a stimulus.

Professor Blanshard feels that I am forced into another absurd position in having to assert that “Hitler’s hatred of the Jews contributed nothing toward his orders to have them exterminated” and that Newton’s ideas of gravitation never “affected in the slightest degree what he said or put down on paper.” But again, why are these statements absurd? If we are to speak of Hitler’s hatred, it is necessarily as an inference from a long series of verbal and nonverbal actions. Hitler himself undoubtedly had other information, for he must have observed small-scale actions of the same sort not seen by anyone else, as well as responses of his autonomic nervous system. But no one part of this complex was the cause of any other part—unless, indeed, following William James, we could say that the action caused the feelings. A much more reasonable view is that the whole pattern was caused by environmental events in Hitler’s personal history. It is too late to discover enough of these to make a convincing case (only historians and psychoanalysts explain individual behavior that way), but it is important to emphasize that the real causes lay in the environment, because if we want to do anything about genocide, it is to the environment we must turn. We cannot make men stop killing each other by changing their feelings. Whatever UNESCO may say to the contrary, wars do not begin in the minds of men. The situation is much more hopeful. To prevent war we must change the environment. In doing so we may well reduce the so-called mental tensions which accompany, and are erroneously said to foster; war-like acts.

And so with Newton. We infer his ideas from the things he said and wrote. Newton himself knew about things he almost said and wrote, as well as things he said or wrote, and revoked, but the ideas he did not quite express were not the causes of the ideas he expressed (in behavioristic terms, his covert responses were not the causes of his overt). The point is important if we are to induce young people to have ideas. For more than two thousand years teachers have been trying to stimulate minds, exercise rational powers, and implant or tease out ideas, and they have very little to show for it. A much more promising program is to construct an educational environment, verbal and nonverbal, in which certain kinds of verbal responses, some of them original, will be emitted. We are closer to such an environment than most educators know.

Did Shakespeare actually represent Othello as moved to action by feelings of jealousy? We should quite justly complain that he had not motivated his character if he had done so. He paints a detailed picture of jealous behavior ending in the smothering of the innocent Desdemona. Most of that behavior, as one should expect in a play, is verbal. Othello tells us about his actions past, present, and future, and his emotional states. Some of this is public and some private, but no one part is the cause of any other part. If he had had time, he might have described the wound he inflicts upon himself, but the felt pain was no more responsible for his death than his feelings of jealousy were responsible for his jealous acts. A common cause was involved in each case. We must turn to the machinations of “honest Iago” to understand Othello’s behavior, and it is a standard criticism of the play that Iago’s motives are not clear. Certainly it will not do for Professor Blanshard to tell us that he was moved to action by his “villainy.”

I am surprised that Professor Blanshard cites the contrary judgment of the world. Philosophers seldom enjoy the support of that judgment, and it seems dangerous to value it too highly. When physicists began to assert that fantastic amounts of energy could be extracted from a lump of mud, I doubt that they were much disturbed by the fact that “every major moralist, historian, and man of letters” would have called it nonsense. Against the judgment of the world, the behaviorist may set the opportunity to discover comparable sources of energy making it possible at long last to deal effectively with human behavior. I continue to argue the behavioristic position because I believe it has vast implications. Current education supplies a useful example. Powerful techniques of teaching, derived from an experimental analysis of behavior, are being widely opposed in the name of traditional philosophies of education philosophies which are avowedly mentalistic and which receive substantial support from mentalistic, particularly cognitive, psychologists. Those who subscribe to them have no really effective new techniques to offer, but their discussions of education are warmly received because they frequently allude to the mysteries of mind. Millions of school children are being sacrificed on the altar of cognitive theory. Men have suffered long enough from that strange quirk in their behavior which keeps them from applying the methods of natural science to their own lives.

References

¹Skinner, B. F., “The Operational Analysis of Psychological Terms,” Psychology Review, 1945.

² Skinner, B. F., “Operant Behavior,” American Psychologist, 1963.

³ Terrace, Herbert, “Stimulus Control” to appear in W. K. Honig, ed., Behavior: Areas of Research and Application.

⁴ Held, Richard, “Object and Effigy” in Gyorgy Kepes, ed., Structure in Art and in Science, Braziller, N. Y., 1965.

⁵ Skinner, B. F., Verbal Behavior, Appleton-Century-Crofts, 1957.

⁶ Blanshard, Brand, “Critical Reflections on Behaviorism,” Proceedings of the American Philosophical Society, 1965.

Concluding Remarks by

Professor Blanshard

In my opening paper, I made three criticisms of Professor Skinner’s behaviorism. Has he answered them?

The first point was that a conscious event such as a toothache is not the same as a physical event—that is, a motion of particles—because its properties are self-evidently different. A motion has velocity, but not a pain; a pain may be dull or excruciating, while it is meaningless to say this of a motion. Professor Skinner replies that though a motion cannot be dull or excruciating, a physical object can be, and “these two adjectives are applied to pain just because they apply to the things which cause pain. A dull pain is the kind of pain caused by a dull object, as a sharp pain is caused by a sharp object.” On this there are three comments to be made.

First, it seems to be untrue. Dull toothaches or headaches are not normally caused by dull objects, or sharp ones by sharp objects. Secondly, granting that a pain may be caused by a physical thing, that does not show that the pain itself is a physical thing; cause and effect are not necessarily alike. Thirdly, even if the term “dull” is ascribable to a physical thing as well as to a pain, it is ascribed in a wholly different sense. The dullness or sharpness of a thing is a matter of its shape; that of a pain is not, for a pain has no shape. This reply can be extended to other cases. The fact that we describe mental events in terms of metaphors drawn from the physical is no evidence that the mental is physical and that we are not using metaphors at all.

In maintaining that there are distinct conscious events, I pointed out, further, that philosophers from Descartes to Russell have found their own consciousness the most certain thing in the world. I am still not clear how Professor Skinner would answer them. He seems to take several lines. One is that they would not have had this certainty if their psychology had been of the right kind and had based itself on the behavior of the Other-One. This of course is true) for we never observe the consciousness of other persons, but only their bodily behavior, and hence if we were confined to this sort of psychology, all talk of consciousness would be ruled out from the beginning. But the question is whether this sort of psychology does not itself leave something out. We all talk with confidence about our own desires and purposes, and the burden of proof is clearly on anyone who would ignore them.

So Professor Skinner offers a second line of reply; we do not need them in explaining behavior. The purpose by which we explain a given action has, if it exists at all, certain conditions of a physical kind. Why not, then, bypass the purpose and connect our behavior directly with these prior conditions? If a causes b, and b causes c, then a causes c, and if we can link these two directly, we shall not need to introduce anything inner or mental whatever. The chief novelty of Professor Skinner’s behaviorism lies, I suppose, in this attempt to bypass inner events, whether mental or physical, by correlating their causes directly with their effects and ignoring these events themselves. But surely you do not abolish the intermediate link in the chain by ignoring it. Every chug of the engine in your car has prior causes which are therefore the remoter causes of the car’s running, but that does not mean that the engine can be eliminated. Even if you did succeed in linking event 3 with event 1 in your explanation without mentioning 2, that does not prove that there is no 2.

Furthermore, we have the most impressive positive evidence that 2 does exist, the evidence of direct awareness. I am aware that I am now thinking about behaviorism with the purpose of criticizing it. Here we meet with Professor Skinner’s third line of reply: he questions this awareness. He says that we never see these inner processes. I agree, for I hold that they are not, like physical processes, the sort of events that can be seen. I can say this, but I am puzzled as to how Professor Skinner can say it. On his theory these processes are as truly physical events in space as the rolling of billiard balls, and such events can normally be seen. It is one of the paradoxes about behaviorists that they first tell us that pains and thoughts are just physical movements in space, and then when we draw the conclusion that we should be able to see them like other such movements, they draw back in dismay. Surely they should either give up the theory that these processes are physical movements—that is, give up behaviorism—or else go courageously forward to the view that we may hope, with a new electronic microscope, to see someone’s toothache or concept of liberty lurking or scurrying about in his head. This the behaviorist feels, with the rest of us, to be absurd. On his premises, I do not see why.

Still, consistently or not, Professor Skinner does deny that pains and thoughts can be seen, and, on that ground, apparently, questions whether there really are such things. But surely seeing is not the only kind of awareness. We may be vividly aware not only of shapes and sizes, but also of our activity of thinking, of our being hot, of being in doubt what should be done about Viet Nam, of the last time we saw Paris, of the funniness of a joke, or of 2-and-2’s making 4. I should not for a moment deny that there are difficulties in analyzing these kinds of awareness. I should only maintain that no argument by which people have sought to analyze them away is half so certain as their own disclosures to us. If any behavioristic or Christian Scientist tells me that my consciousness of an intense pain is an illusion, I can only say, No, thank you; if that is science, I prefer to stay with common sense or (heaven help me) philosophy.

The second thesis of my paper was that behaviorism, for all its desire to be scientific, is really in conflict with science. Only physical things and events exist, says Professor Skinner. Images, hallucinations and sensations of blue, if anything at all, must therefore be physical things or events. But no physicist would accept this. No natural scientist would introduce the bats’ heads seen by the toper, or our images of Einstein, or the hues we see in the rainbow, into physical space. The behaviorist is here ranging himself not against philosophy only, which he would do with equanimity, but against physics also, which he respects. What is Professor Skinner’s reply?

I have had some difficulty in following it, but apparently it is this. The image of Einstein is not an object seen or imagined at all, and hence the problem of finding a place for it does not arise. At some time in the past we have seen Einstein in person or in a picture, and when we now have an image of him, we are repeating the response or activity of seeing. “No matter how obscure its dimensions, the behavior called seeing must be involved, and you may be reporting it rather than the presence of the thing seen when you report that you see something.” That is, the seeing occurs, but without an object. To have an image is to engage in an activity of seeing into which the alleged object is telescoped or dissolved. And this activity of seeing is of course bodily activity.

What is most striking about this theory, to my mind, is its appeal from experienced fact to a materialist metaphysics; it could hardly have been suggested by the facts themselves. I report that I have an image of Einstein, so clear that I can see the fluffy hair and some of the wrinkles; I am told that in fact I am seeing nothing at all. I report that in my own experience seeing is always a seeing of something or other, that to see and yet see nothing is impossible. I am told that the impossible is perfectly possible, and this must be regarded as a case of it. I report that in my own experience there is a sharp difference between the object of a response and the response itself; I am told that here the object has been absorbed in the response in a way to which nothing in the facts seems to correspond. I report that in my own experience there is a difference between contemplating and acting, that I can envisage a face or a scene without doing anything about it. I am told that there is no such distinction, that contemplating is behaving. I report that seeing Einstein in imagination, even if it is in a sense behaving, seems at the farthest possible remove from the making of a bodily adjustment. I am assured that my having an image is a bodily adjustment and nothing more.

Now I believe in philosophy, but in a philosophy that starts from facts as given, not as moulded to suit the interest of a materialist or any other metaphysics. The behaviorist is committed to an antecedent philosophical theory which requires him to regard images, like everything else, as physical. But he meets with so stout a resistance from the physicist that he feels compelled to change his tactics and to say that the image is not a physical thing; it is really the physical activity of seeing. But this does not remove the difficulty; it doubles it; for now in addition to calling an image physical, he is trying to dissolve an object into the process of responding to it. It cannot be done. If my image of Einstein is not a physical thing, neither is it a physical process.

My third thesis was that in denying consciousness, behaviorism makes us robots or hollow men, since consciousness is the seat of all values and all motives. Regarding values, Professor Skinner says nothing, and I remain in the dark as to how values of any kind can exist in a world of matter in motion. Regarding motives, however, he is gratefully clear. He accepts my reading of behaviorism as implying that Hitler’s hatred had nothing causally to do with his acts of genocide, and adds that mental tensions are “erroneously said to foster war-like acts.” Newton’s “ideas” made no difference to what he wrote, and Othello’s feelings of jealousy were no more responsible for his jealous acts than his feelings of pain were responsible for his death. To the comment that here he would have against him “every major moralist, historian, and man of letters” Professor Skinner replies that these men would also have opposed modern science when it said you could extract vast energy from a lump of mud.

Now I do not think they would; humanists are more likely to be helplessly credulous when scientists speak in their own field. But even if they did, the testimony of humanists when in turn they speak in their own field can hardly thus be brushed aside. They feel threatened by a rising wave of computerized philistinism, which seems bent on liquidating the world they live in. They are coming to learn with bewilderment that the new science of mind rules out as an antiquated delusion the entire realm of mind once occupied by the humanities. It says that in the traditional humanistic sense the poetry of Keats, the thought of Kant, the music of Mozart, the morality of Schweitzer, the religion of Tillich, ever the scientific reflection of Newton and Darwin, literally never existed at all, and that there was really nothing about these men except their bodily complexity to distinguish them from an IBM computer. And the theory not only destroys the humanities in principle; it makes education in them or any other field pointless, for there is nothing about a complex robot to make it more worth being than a simple robot. “We cannot make men stop killing each other,” says Professor Skinner, “by changing their feelings.” Why educate their feelings at all, one wonders, if not even the hatred of a Hitler or the jealousy of an Othello can make the slightest difference in what they do? It seems equally pointless to educate men to reflect, for the foresight of the consequences of their conduct can never affect that conduct. Indeed it is hard to see why, in a behaviorist world, any consequences should be better than any others. Why should I not impose suffering on others if it is only a mentalistic unreality? Fortunately Professor Skinner is so unreasonable a behaviorist as to be a very kindly and considerate man.

He has quoted a name we both revere, and may I do so too in closing? William James, our greatest psychologist, was a man who believed in science thoroughly, but also believed in the existence, the importance, and the efficacy of consciousness. He concluded his Lowell Lectures with this sentence: “I, for one, as a scientist and practical man alike, deny utterly that science compels me to believe that my conscience is an ignis fatuus or outcast, and I trust that you too . . . will go away strengthened in the natural faith that your delights and sorrows, your loves and hates, your aspirations and efforts are real combatants in life’s arena, and not impotent, paralytic spectators of the game.”

Posted March 5, 2007

Back to Blanshard page