Econlib

The Library

Other Sites

 Front Page arrow Titles (by Subject) arrow BRUCE GOLDBERG, Skinner's Behaviorist Utopia - New Individualist Review

Return to Title Page for New Individualist Review

Search this Title:

Also in the Library:

Subject Area: Economics
Subject Area: Political Theory
Subject Area: History
Order this book from Liberty Fund

BRUCE GOLDBERG, Skinner’s Behaviorist Utopia - Ralph Raico, New Individualist Review [1961]

Edition used:

New Individualist Review, editor-in-chief Ralph Raico, introduction by Milton Friedman (Indianapolis: Liberty Fund, 1981).

About Liberty Fund:

Liberty Fund, Inc. is a private, educational foundation established to encourage the study of the ideal of a society of free and responsible individuals.


Skinner’s Behaviorist Utopia

BRUCE GOLDBERG*

THAT A STATEMENT or a theory is true or false is a matter which is, presumably, to be decided by the employment of the various canons of scientific (in the widest sense of that term) observation and inference. A theory, for example, is tested by its ability to explain the facts on the basis of which it was introduced, and also by its ability to explain new facts which were not explicitly taken into account in its construction. And there are considerations irrelevant in determining the acceptability of putative truths. One obviously irrelevant factor is the aesthetic satisfaction anyone might get from contemplating a given theory. Whatever the poet may have thought, beauty is not the same as truth. Less obvious, and perhaps worth mentioning for that reason, is that the moral consequences of a statement are not relevant determinants of its truth or falsity. It may be that the general acceptance of some statement would lead to universal misery. Conceivably this could be a reason for keeping the statement secret. It could not be a reason for concluding that it was false. All men are mortal, whatever Angst might be occasioned by recognition of the fact.

That these considerations are not truth determinants would seem to be hardly susceptible of dispute. And yet, without explicit avowal, antagonists of scientific theories have all too often allowed aesthetic or moral upset to count against those theories. The Einsteinian conception of time as a relativistic magnitude did indeed shatter a well-established Weltanschauung. The relatively simple picture of Newtonian mechanistic interactions embedded in an absolute temporal framework died hard. That death, however, should have been an easy one. The major test of a scientific theory is its ability to explain the phenomena, and the new theory was better at this than the old. And for most (although not absolutely all) physicists, this was enough. But for others—theologians, philosophers, aestheticians, even laymen (though certainly not for all the members of any of these groups)—this was not enough. The aprioristic defenses of the old order began. “It’s perfectly self-evident,” ran one defense, “that time is absolute throughout the universe. One simply can’t conceive of it being any other way.” “It’s in the nature of time,” ran another, “to be the same everywhere.” And in the background was the—unexpressed—objection that the adjustment was too difficult to make, that the new picture was not anything so much as repulsive.

Nowhere does this anti-scientific resistance to new theorizing assume more vigor than with respect to explanations in psychology. While physicists may displease, they are in possession of a mathematical apparatus which frightens. The critics, for all their resistance, remain more or less quiet and, at times, even a little ashamed. However, the language of, say, Freudian theory is ordinary English (or German). The technical terms are no more awe-inspiring than those of a competent aesthetician: “drive cathexis” is no more intimidating than “aesthetic distance.” This releases inhibitions, and the defenders of received opinion feel that they have a free hand.

But there is a feature of psychological theory more important in this connection than that its propositions are unmathematized.1 Psychological theory concerns people. It attempts, among other things, to explain why they act the way they do. And often the answers are such that the actor would not unreflectively acknowledge their correctness. Further, the critic may have a stake in the answers’ being incorrect. How often has Freudian theory been attacked because of its “repulsive” doctrines of infantile sexuality? Or on the ground that our traditional views of human responsibility would have to be discarded? But such objections are absolutely worthless. The repulsiveness of Freudian theory is no more relevant to its truth or falsity than its country of origin. But it is one thing to point up the utter irrationality of an objection, and quite another to dispel the fears which prompt it.

As objectionable as Freudian theory was, however, it was sufficiently anthropomorphic to pose less of a threat to “man’s dignity” than the mechanistic view associated with a certain other psychological theory. While Freud, the standard version goes, moved the springs of action from the conscious mind to the unconscious, later twentieth century psychology removd it from the psyche altogether. This new dehumanization of man is called behaviorism. Behaviorism not only eliminated the mental as a factor—it combined this with a thoroughgoing determinism which left no room for free action at all. The moral conscience revolted. Behaviorism was castigated as evil, and therefore, presumably, false. But the argument is no more legitimate here than it was in the preceding cases. Behaviorism simply cannot be shown to be incorrect by showing that its alleged consequences are undesirable, morally repugnant, or even evil. Those who adopt this approach deserve our ears no more than the critics of another theory, who insisted that man’s dignity required that his planet be at the center of the universe.

THESE OBSERVATIONS ON what is relevant and what is irrelevant in the evaluation of a scientific theory such as behaviorism are necessary, for in this essay I shall be concerned with the behaviorist-based recommendations for social organizations of Professor B. F. Skinner, the eminent Harvard University psychologist, as set forth in his utopian novel, Walden Two.2 If his proposal for a society rigidly controlled by behaviorist psychologists is to be rejected (as I think, in reason, it must be), it will have to be on the basis of logical analysis, and not of foolish sloganizing.

The blurb on the book informs us that Walden Two “provocatively pictures a society in which human problems are solved by a scientific technology of human conduct—and in which many of our contemporary values are obsolete.” That anyone has presented us with a solution for human problems should make us listen, especially when the donor is a psychologist of unquestionable achievement. If in the end we will be skeptical, no part of the cause should be lack of gratitude for the attempt.

Walden Two is an attempt, in fictional form, to outline a system of social organization based on behaviorist psychological theory. “The methods of science have been enormously successful wherever they have been tried,” Skinner says in another place;3 “let us then apply them to human affairs.” The scene of most of the book is a utopian community (Walden Two). Frazier, the creator of the experiment, defends his theories of social organization against Professor Burris, a slightly skeptical but generally sympathetic antagonist, and Professor Castle, a tender-minded philosopher and a bumptious, nasty, and unreasonable caviller, who interrupts the discussion from time to time with generally strawmannish objections.

A good deal is said in this novel, and I shall not try to examine every point that is made. What I shall do is consider certain of the central ideas of Walden Two, on the falsity of which the claims of the book would founder. Occasionally I shall bring to bear illustrative evidence from one of Skinner’s non-fiction works, Science and Human Behavior.

The initial picture of the inhabitants of Walden Two with which we are presented is an attractive one. “These were delightful people,” Professor Burris muses; “their conversation had a measure and a cadence more often found in well-wrought fiction than in fact. They were pleasant and well-mannered, yet perfectly candid; they were lively, but not boisterous; affectionate, but not effusive” (page 28). And in the course of the book, we come to learn that the inhabitants of Walden Two possess most of the desirable character traits one can think of and almost none of the bad ones. Indeed, with the exception of Frazier himself (who did not have the benefit of a Walden Two upbringing) everyone seems to be supremely happy and well-adjusted. This is not to be scoffed at. Only those with no social concern at all (and they don’t count) could be indifferent to the possibility of establishing a form of social organization which has such results.

How does one go about producing such desirable characters? Well, babies at Walden Two are reared in community nurseries. At the age of one year, they graduate to community playgrounds, being subjected during these formative years to an intensive and highly scientific program of conditioning. All this provokes much resistance from the antagonistic Castle. But the questions of the techniques of Walden Two seem to me less interesting than the question of whether Skinner has even provided a coherent account of what can be accomplished by those techniques. If that account is itself internally inconsistent, the details are unimportant.

What I mean is this. Skinner’s general program is the behavioral conditioning of certain kinds of emotions and behavioral responses. This conditioning is to have the result that only certain kinds of emotions appear in the members of Walden Two, while others, the undesirable ones, disappear through lack of positive reinforcement. The result is lots of people with good emotions, and very few or none with bad ones. It might immediately be objected that this could never come about, that human nature cannot be changed, or something of the sort. But this Castle-type move is a bad one. Talk about human nature is far too vague to permit a reasonable decision to be made as to whether or not it could be changed.

“As to emotions—we aren’t free of them all, nor should we like to be. But the meaner and more annoying—the emotions which breed unhappiness—are almost unknown here, like unhappiness itself. We don’t need them any longer in our struggle for existence, and it’s easier on our circulatory system, and certainly pleasanter, to dispense with them” (page 101). To the objection that emotions are fun, Frazier replies, “Some of them, yes. The productive and strengthening emotions—joy and love. But sorrow and hate—and the high-voltage excitements of anger, fear and rage—are out of proportion with the needs of modern life, and they’re wasteful and dangerous. Mr. Castle has mentioned jealousy—a minor form of anger, I think we may call it. Naturally we avoid it” (page 102). And so it is conditioned out, along with the other unpleasant emotions, presumably simply by never receiving any positive reinforcement on any occasion of its occurrence. This is the account one must examine, not, I should add, to determine whether or not it would be advisable to condition out the emotions referred to but to determine whether such a conditioning out would be even a theoretical possibility.

The behaviorist picture (at least as Skinner presents it) is in some ways an excessively simplistic one. There seems to be the idea that, with respect to any given person and the various emotions he is capable of experiencing, one could, given a proper technology, pluck out some while leaving the others intact. Occurrences of emotions are, on this account, something like pains. Just as one can eliminate a given pain (by administering a drug, say), so one can eliminate a given emotion (by a proper administration of behavioral engineering). In this sense, the account of the emotions is an atomistic one. Each emotion can be considered separately, and the relevant conditioning techniques can be applied to it. The good ones stay and the bad ones go.

THIS PICTURE IS A radically misconceived one, and it requires no experimentation at all to show this. Is it even possible to become clear about what the picture suggests? Skinner wants to retain joy (because it is productive) and eliminate sorrow (because it isn’t). Under what circumstances might one experience joy? Well, suppose one has not seen one’s mother for twenty years. One goes to meet the plane on which she is expected to arrive. The door opens, she descends the stairs and comes running to meet one, arms outstretched, tears in her eyes. Presumably, on this occasion one is happy, joyful, and the inspiration provided enables one that very evening to make some important new contribution to knowledge.4 But suppose that just as one is about to embrace one’s mother, an unknown assailant shoots her in the back. Doesn’t one then feel at least sorrow, or even grief? Is one neutral? Indifferent? Isn’t it obvious that joy and sorrow are not atomistic states with no stronger connection than that they are both emotions? If the death of one’s mother (in the situation described above) did leave one indifferent, how could the embrace produce joy? In order for one to experience joy in the situation described, it must (logically must) be true that whether or not one’s mother lives or dies makes some difference to one. But if that’s true, then one could not remain indifferent if she does die.

Perhaps another example will make the point clearer. Skinner is pro-happiness. The people in Walden Two are a happy group. They are often engaged in creative enterprises, and they are happy in their work. But unhappiness is unproductive. Presumably it’s also to be conditioned out. Again, the question arises about the theoretical possibility of such a state of affairs, i.e., a state of affairs in which people are only happy. Suppose I have been working on a project for ten years, have constructed an elaborate theory which has only to receive its final experimental confirmation. The laboratory technicians bring in the results of a series of experiments which they have been running for the past two weeks. Now what is Skinner’s claim? Am I indifferent to the results of my experiments? Suppose I am. In that case, it’s odd to say that I’m very happy, even overjoyed, when I learn that my theory has been confirmed. Suppose that I’m not. Then it’s equally odd to say that I’m not even a little bit unhappy when I learn that the experiments have falsified the theory. Given that people sometimes fail to achieve what they want very much (and not even Walden Two promises to fulfill all desires), it follows that they are sometimes unhappy. Isn’t it patently obvious that happiness and unhappiness are not independent atomic states? But if that’s true, then it doesn’t even make sense to suppose that one could condition people never to experience the one and always (or almost always) to experience the other.

The principle illustrated by these examples is that many emotions have what might be called quasi-logical counterparts. Pairs such as love-hate, joy-sorrow, happiness-unhappiness are familiar. To suppose that (whatever technique one might employ) one could condition out one member and leave the other intact is to believe a fiction. And this has nothing to do with the nature of man or the limitations of scientific technique. It is, rather, a matter of the logic of emotional predication. “Whenever a particular emotion is no longer a useful part of a behavioral repertoire, we proceed to eliminate it,” Frazier tells us (page 103). Presumably, on this account we could, if we wanted, “engineer” a person to feel only grief or only love or only nostalgia. We decide which emotions we want, and then instill them. But this is something we shall never be able to do (at least not in the way implied in Walden Two). And the reason is that to say that we can do it doesn’t even make sense.

The above is not mere philosophic pedantry. One of the attractions of Walden Two is that it seems to offer us an escape from the unpleasant, and this through the application of an allegedly scientific theory. But that theory, whatever its other merits, cannot at least succeed in this. If the inhabitants of Walden Two can experience joy, then they can experience sorrow. And it is a logical mistake to suppose that things could be otherwise.

WE HAVE, THEN, eliminated one major piece of the theoretical groundwork of Walden Two, at least as Skinner views his system. There is another view advanced by Skinner, of even greater importance than the view that one can selectively condition emotions. This is the idea that free will is an illusion. Skinner regards a behavioral technology as being incompatible with free will Since he regards the former as possible, he denies the latter. Now, the problem of free will is an enormously complicated one. What I shall try to do here is merely to show that Skinner’s view of what it would mean to have free will is a confused one, and that nothing he says even tends to show that free will is an illusion.

Skinner regards adherence to the idea that man has free will as essentially a relic of pre-scientific ways of thinking. And this is a fairly widespread notion. Without discussing the merits of this claim—my own opinion is that it is much too vague for argumentation on either side—it might be worthwhile to see just how Skinner proposes to demonstrate the falsity of the free will doctrine. In a way, this is a difficult thing to do, for Skinner seems not altogether clear about what view he is attacking. Sometimes it is the view that human actions are spontaneous, at other times it is that their actions are the actions of responsible agents, at still other times it is the view that human actions are uncaused. This last is probably the most substantial, so let us deal with that one. Skinner argues that it is a mistake to regard actions as uncaused. The most we are justified in saying is that for various actions we do not know what the causes are. However, the advancing march of science gives us every reason to believe that in time the causes of human behavior will present no greater problem than the causes of heat transfer in gases. The issue of free will is often discussed in these terms. That is, those who assert that we have free will generally regard this as committing them to the view that human action is uncaused. Those who opt for a general determinism regard this as implying that human action is unfree.5 One of the unfortunate aspects of this controversy is that both sides seem to have thought that they had pretty clear ideas about the meaning of the key expression, i.e., determinism. The allegedly unproblematic explication of this notion has most often been: “the theory that every event has a cause.” But a number of contemporary philosophers have seen that this explication is by no means as clear as has been thought. Some have gone so far as to assert that they do not know what the thesis of determinism is.6 And there is good reason for this. For example, if one sees how Skinner fills out his account of what determinism with respect to behavior is, one cannot but find it surprising that he does regard the truth of determinism as incompatible with human freedom.

In Chapter Three, “Why Organisms Behave,” of Science and Human Behavior, Skinner says that “we are concerned . . . with the causes of human behavior. We want to know why men behave as they do. Any condition or event which can be shown to have an effect upon behavior must be taken into account. By discovering and analyzing these causes we can predict behavior; to the extent that we can manipulate them, we can control behavior” (page 23).

Initially such a picture of a general determinism of behavior seems to many people to be a frightening one. The suggestion is that we are mere ciphers in a causal stream over which we have no control. And this is a picture to which Skinner frequently alludes in a favorable way in the course of his writings. We can manipulate causes and control people. A Brave New World (if one is frightened) or a Walden Two (if one is pleased) looms up before us.

But is there reason, given Skinner’s view (which is a widely held one), for either fear or pleasure? Is there, indeed, anything excitingly new being said?

What are the causes of behavior on Skinner’s showing? “Any condition or event which can be shown to have an effect upon behavior. . . .” Suppose I am a devotee of Shakespeare and a friend tells me that a new production of King Lear has just come to the Lido. Excited, I rush down to reserve tickets for the next performance. Now it is unquestionably true that my friend’s informing me that the play was being performed is an event which had an effect on my behavior. If he hadn’t told me I wouldn’t have done what in fact I did do, viz., go to the theater to reserve the tickets. Of course I might later have learned on my own that the play was being performed and would have then bought the tickets, but I wouldn’t have done that when, as a matter of fact, I did. Suppose that I go to the play, and, after having thoroughly enjoyed it, applaud wildly when the performers take their curtain calls. Again, the appearance of the performers on the stage is an event which had an influence on my behavior. If they hadn’t come out I would have returned home without applauding, wondering what went wrong, and perhaps feeling a bit sad because I didn’t have the opportunity to show my appreciation. We have here, then, two examples of events which had an influence on my behavior. Were they causes of it?

It is not altogether easy to answer this question. For Skinner, of course, the answer is clear. They are causes. According to the quoted passage any condition or event which has an influence on behavior is a cause. But there are difficulties. Consider another kind of case. I tell someone, “I was just in the psychology laboratory with John and I caused his leg to rise.” When asked how, I reply that I hit his knee in the familiar spot with a small hammer. But suppose that, when asked how I caused John’s leg to rise I reply, not that I hit his knee with a small hammer, but that I asked him if he would raise his leg. Ordinarily, we would say that if this is how I got John’s leg to rise (by asking him to raise his leg), then I didn’t cause his leg to rise, although what I said was certainly relevant to bringing it about that John’s leg did rise. It would at least be misleading, under these circumstances, to say that I did cause his leg to rise. In a context like this one, it would in general be assumed that if I caused John’s leg to rise, I did something which in some way put it out of John’s control that his leg did rise. When this assumption turns out to be wrong, i.e., when it turns out that I simply asked John to raise his leg and he did so, then it sounds very odd indeed to say that I caused his leg to rise. This is not always true. If John were hypnotized and raised his leg at my command, then it would be perfectly in order for me to say that I caused what occurred. One can easily think of other cases in which it wouldn’t sound odd to say that I caused John’s leg to rise. But in the case where I simply ask John to raise his leg (and he’s fully conscious—not hypnotized, etc.), it does. But why should this be true if, as Skinner says, a cause is any condition or event which can be shown to have an effect upon behavior? Why does it sound at all odd to say that the performers’ appearing on the stage caused me to applaud?

Above I hinted at what seems to me to be at least a partial explanation. It seems to be a legitimate inference that if someone is caused to do something, then that thing is in some way out of his control that he is in some way not at liberty not to do it. And Skinner appears to be aware of the legitimacy of this inference. Notice that he says that to the extent to which we can manipulate behavior we can control it. If I am controlling someone’s behavior, then he certainly isn’t. This is surely the source of the doctrine that if someone’s behavior is determined, then he isn’t free. If the behavior is determined, then it’s caused; if it’s caused, then it is out of his control; and if it is out of his control, then he’s not free. Assuming this argument to be legitimate, then one cannot at the same time grant that a bit of behavior was determined and that it was free. From the fact that the behavior was determined, it would follow that it was not free. But does it follow from the fact that there were events which influenced the behavior that the behavior was unfree? Put in another way, from the fact that the agent was influenced by certain circumstances to act in a certain way, does it follow that the action was out of his control? As we ordinarily speak about behavior, this certainly does not follow. Let us take still another example.

One of the factors influencing a general’s decision to attack may be the intelligence information that the enemy is shortly going to concentrate its forces against his weak flank. Is his subsequent behavior (giving instructions to his subordinates to attack, preparing certain false reports which he is going to allow to fall into enemy hands, etc.) something which is out of his control? Surely not. Such a situation is just the sort of situation in which we say someone’s behavior is under his control. Under what circumstances would we say that his behavior was out of his control? Well, suppose that he received the intelligence information and instead of calmly preparing plans for the coming attack, went berserk. Suppose he ran out of the Command Headquarters raving about the imminent destruction of his forces, and ordered each individual soldier (with or without a weapon) to attack the enemy at once. Suppose, too, that he himself starts running in the direction of the enemy camp, knife in hand, screaming that he is going to kill the mad beasts. Here, most probably, we would say that his behavior was not under his control. Surely our estimate of the situation must be different when he calmly sits down to prepare the attack, and gives his subordinates intelligent instructions, involving what, in the opinion of all, is a master military move. And yet there are events which influenced this behavior, just as there are events which influenced the berserk behavior. But if we allow that, in the case of the intelligent action, simply from the fact that there were indeed events which influenced his behavior it does not follow that the behavior was out of his control, then we must either reject the argument that causality implies non-freedom or reject Skinner’s assumption that any event which influences behavior is a cause of that behavior.

And indeed how would one go about trying to show that any event which influences behavior is a cause of that behavior? Skinner, in the passage I quoted, seems to want to do this by definition. But this can show absolutely nothing. Presumably Skinner must regard the assertion that all behavior is, because determined, unfree, as an interesting assertion about behavior, i.e., about actions which people perform. That is, presumably he doesn’t regard the contention that behavior is unfree as a simple definition. For if it were a mere definition, then while we might agree (accepting his definition for a moment) that “behavior” will imply unfreedom, we could not say about any of the events which we normally call cases of human action that they were in fact cases of “behavior.”

Let us assume, then, that the statement at issue is not a simple definition. Now, to say that behavior is determined, is to say that it is caused. Therefore, by the original argument, caused behavior is unfree. And if any event or condition which influences behavior is a cause, then behavior which is to any extent influenced by events or conditions is unfree. What is it to say that a bit of behavior was influenced by an event or condition? Presumably, that the event or condition would be relevant in explaining why the behavior occurred, why the man acted as he did. From all of the above, it follows logically that the only behavior which could be free is that behavior for which there is no explanation at all. The only behavior which could be free is that behavior which is completely irrelevant to the circumstances in which it occurred, in the sense that nothing could be brought to bear to explain why the agent acted as he did. And from this it follows that every case of goal-directed behavior (that is, behavior in which the agent acts to achieve some end) is, to the extent to which the end was relevant to his behavior, unfree. And finally, it follows that all behavior, in so far as it is rational, is unfree. This, I submit, is absurd, for if behavior is unfree, then it is impossible that it be rational. Any argument which has as its conclusion the assertion that all rational behavior is unfree (in the last analysis, because it is rational) must be rejected. Skinner’s argument about free will thus results in a palpable contradiction. His reasons for thinking that free will is an illusion are no reasons at all.

I have spent a good deal of time on what might be called the philosophical underpinning of Walden Two, the two theses, namely, that it is possible selectively to condition emotions and behavior, and that all behavior is unfree. If the truth of these theses were granted, much (though not all) of what Skinner recommends would be extremely difficult to deny. Mere charges that it is all diabolical would, in this case (and actually, in any case), be pointless. Now that we have seen that they are false, we can go on to see how they lead to the advocacy of Walden Two.

IN WALDEN TWO, things are controlled from the top. There are various managers (of play, of social activity, of work, etc.) who decide what sorts of enterprises are conducive to the psychological welfare of the members of the community, and are therefore to be permitted. The society is thus a rigidly controlled one, and the reason is that “when a science of behavior has once been achieved, there’s no alternative to a planned society. We can’t leave mankind to an accidental or biased control” (page 264). Why, one might ask, is the only alternative to planning (controlling) the lives of people by Skinnerian managers that of leaving them to the (accidental and/or biased) control of others? The historical tradition of liberalism would seem clearly to have shown us another alternative, that of letting people shape their own lives as they see fit. But, it should by now be clear, this is not an alternative which Skinner, in terms of the framework he has constructed, could even suggest. Since all behavior, whether we like it or not, is, in fact, controlled, we might as well let it be controlled by the good guys. Why is all behavior controlled? In the last analysis, because there are events and conditions which influence that behavior. Because of his mistaken idea that no behavior can be free, Skinner is rendered incapable of seeing the distinction between planning someone’s life and letting him plan it himself.

This is a most important point. Prima facie there would seem to be two alternatives (with lots of gradations in between)—either people are controlled, or they are not. But for Skinner, this last is not even a possibility. The argument about free will makes it impossible that anyone is uncontrolled. The only real alternatives are good control and bad control (page 263). But this is simply the product of a bit of sham reasoning. What, one wants to ask, do you say about the possibility of establishing a society in which people are not subject to a rigid conditioning process in the hands of “behavioral engineers,” but in which they are able to encounter many and diverse influences and make up their own minds about which they regard as the more important and which the less? Skinner’s answer is clear. This, he would presumably say, is just the situation I am trying to avoid; it is the one I was talking about when I described a situation in which there was accidental and biased control.

It seems to me that therein lies, for the superficial reader, the main attraction of Walden Two. Only the unreasonable would prefer a system of biased and accidental control to one of intelligent control. Walden Two is, after all, better than Russia. It is to expose the view that these are our only choices that I spent as much time as I did examining Skinner’s argument about free will.

In actuality, there are a number of alternatives open to us who would recommend a form of social order. There are at least the three just alluded to: a Russian system controlled from the top by foolish and arbitrary men, a Walden Two in which this control is exercised by intelligent men of good will, and an uncontrolled society.7 These are, of course, not the only alternatives, but if we recognize the fallaciousness of Skinner’s free will argument, we see that there are at least these three.

And, granting that Walden Two is preferable to Soviet society, we must raise the question concerning whether it is better than an uncontrolled one. Frazier, in effect, raises this very question, “. . . what do you say to the design of personalities? Would that interest you? The control of temperament? Give me the specifications, and I’ll give you the man! . . . Let us control the lives of our children and see what we can make of them” (page 292). Notice that Skinner relies, in characterizing his own alternative, on the view which we discovered earlier was not even coherent. It is not true, irrespective of the techniques one employs, that one can generate people according to any arbitrary list of specifications: one could not (logically) make people who were only happy. But leaving this aside, the question is whether we should allow a central committee in control of society consciously to design personalities. Prima facie the answer to the question would seem to depend on which values the central committee is concerned to instill. It might seem that if these values are good ones, then we can accept Skinner’s recommendation, and that if they are bad ones, we should reject it. What are the character traits and social values which Skinner accepts? In general, they are the very best. Happiness, productivity, lovingness, etc. This is another reason, I suspect, for the attraction which Walden Two exerts. If the answer to the question did depend on whether or not Skinner’s choices were good ones, then there would be little more to say.8 But it does not. What must be decided is the question of the desirability of a controlled society.

An initial question, preparatory to deciding the matter, might be the following: given that we can produce artists, scientists, musicians, at will, how could anybody reasonably decide about the desirable proportions? This seems to be the sort of value question which the behavioral scientist is in no better position to answer than any of the rest of us. Suppose that the behaviorist were to reply that this isn’t a value question at all. It’s purely an economic matter, i.e., the question is one of the allocation of resources: we want to provide the people in our behaviorist utopia with the maximum possible satisfaction of their various desires, and our decision about what sort of people to produce (artists, musicians, scientists, etc.) will depend not on our values, but on the values of the community.

Now there would be difficulties in proceeding in this way. I mean that the behaviorist in his controlled socialist society would have the same problems in the effective allocation of resources as any other socialist planner. But this sort of objection is, at this point, not the relevant one. Can we really allow that, as far as we have presented it, the behaviorist’s problem is one of resource-allocation? I think not. Remember that the important idea about Walden Two is that we are conditioning values, i.e., likes and dislikes. We are not simply taking an arbitrary society and discussing the question about how best to satisfy the needs of that society given its (and not our) set of values. What this means in the present case is that the problem about how many musicians to produce is not an economic question. We are deciding not merely how to satisfy desires—we are deciding what those desires shall be, i.e., we must decide how many people are going to want to hear music, how many people enjoy art, etc.

Thus the behavioral scientist is faced with a question of value, a question of the order in which various goods are to be placed on a preference scale, and the behavioral scientist is in no better position to resolve this problem than we are. Indeed what criteria could be employed to decide such a question? This is surely not a scientific question, in the sense that there are generally accepted ways of finding out the answer. finding out the answer. The very best one could do is suggest an answer, and that answer will be an expression of the particular value scheme embraced by the suggester. There is no reason at all to suppose that the behaviorist is possessed of any expertise in this matter which is denied to the rest of us.9

IT BEGINS TO LOOK as though distinctly unscientific elements of arbitrariness (which arbitrariness, remember, Skinner was concerned to avoid) are being introduced into the design of the behaviorist utopia. But there are further difficulties. Skinner argues that a free society (in the ordinary sense) is an extremely undesirable thing. The reason is that all sorts of elements of arbitrariness and haphazardness are introduced by people pursuing incompatible (and often unpraiseworthy) goals. The various efforts which these people make in the attempt to persuade their fellows that the goals are desirable ones are causal factors which can change, in the last analysis, the value orientation of that society. This arbitrariness, the possibly of bad values being substituted for good ones as a result of the free (again in the ordinary sense) interplay of competing goals, is perhaps the chief evil against which Walden Two is directed. This is the reason Skinner advocates a controlled society.

We must then investigate whether or not it is possible, even on Skinner’s terms, to eliminate arbitrariness in this sense. Let us suppose that we have a scientific procedure for determining not only how to condition various abilities, character traits, and values, but also for determining which ones we should condition. Now, it is not so easy to know what we are supposing here. With respect to musical ability, for example, are we supposing ourselves to be conditioning general ability in this area, ability at baroque music, ability at Bach, Vivaldi, or what? The same question arises for musical tastes. Do we condition people to like baroque music or do we condition a fondness for music in general? This is by no means an academic question for the behavioral engineer. We shall see that his failure to solve this problem will result in the emergence of new values in society, which the behaviorist psychologist cannot have taken into consideration in his scheme of conditioning, and for the control of which he will have to resort to methods which have nothing of the respectable air of science about them.

We can distinguish various levels of what we might call value generality, and the behavioral engineer must decide, assuming him to have the suitable techniques, on which level to operate. For purposes of the discussion, we might call music in general the first, or highest, level of generality; the various periods—baroque, classical, romantic, etc.—the second level of generality; particular composers, the third level; and so on. Let us call an ability or a liking on the first level an Ability1 or a Value1, and so on.

Now the question is: at what value or ability level are our conditioning procedures supposed to operate? Presumably the behaviorist is not going to condition each miniscule preference or ability in each member of society. I mean he is not going to decide that so and so many people should be devotees of the late Beethoven quartets, and that so and so many people are going to be admirers of Bach piano sonatas played without the pedal. Let us assume, then, that conditioning procedures are going to be applied at the first level: we are going to produce so and so many musicians, so and so many artists, etc. We are not going to produce as a matter of deliberate policy a given number of baroque pianists, a given number of impressionists, and so on. Thus, Value2 choices are not, as the Value1 choices are, the direct result of policies and practices instituted by the planners.

This is the situation we are envisaging. What could we reasonably expect from such a situation; what would be the natural result of this absence of Value2 conditioning? Well, one thing we could reasonably expect would be that, in an important sense, Value2 would not be anything like as stable as Value1. While the number of people who have musical interests and abilities is something that we can expect to be constant (because we are fixing it) the number of people who like any kind of music and the number who are skilled in the production of any kind of music is something that is likely to vary.

So, to restrict ourselves to this specific case, we can expect there to be periods in which one sort of music is in the ascendancy and periods when this sort of music is replaced by another in the favor of the inhabitants of the society. More interesting is the fact that we can expect new values to be created on this level. Since, to the extent to which values on this level are unconditioned there is an area for what we might call a free interplay of competing values, there is reason to expect that there will be some sort of building on what is already in existence. Gifted musicians will see the possibility of innovating, of building upon the stock of values and knowledge which they already possess, in the same way that this was done in the old society, in the way, for example, in which Mahler is said to have built on Wagner, and Schoenberg on Mahler. And we can also expect that some of these innovations will receive acceptance on the part of some of the music lovers of the society. Remember that we are supposing their Value2 choices to be unconditioned.

Thus new and unplanned values have come into existence in the society. What interest, if any, should the planners take in this fact? Presumably they are going to be very interested indeed. After all, this new sort of music, when it is performed, is going to be one of the causal factors influencing the behavior of the inhabitants of the society. And there will be all sorts of factors, on this level, which will be determinants of the behavior of the inhabitants of the society—new art forms, new ways of painting, new ways of writing and performing plays; innovations in technology; new ways of building homes, new alloys for use in such building. The possibility of a literally endless number of changes in the established order opens up. These new elements are going to be competing with each other for the attention and approval of society at large. Were the planners to step back disinterestedly from this development the very element of arbitrariness and randomness which Skinner was concerned to eliminate would reappear.

So the planners must act. If it could be shown that some new Value2—some new style of painting, for instance—was likely to have what in the judgment of the planners would be an adverse influence on the populace, it must be stopped. The whole point of the society is that such adverse influences are to be eliminated. Thus, control is introduced at this lower level of value generality.

OR IS IT? When we speak of “control” at this level are we speaking about the same sort of thing with which we began? The answer is no. In our initial attempt at describing the behaviorist utopia, how did we suppose control to be introduced? What counted as control? Clearly, it was the conditioning of certain likes and dislikes, certain abilities and character traits. This is what Skinner seems to mean throughout Walden Two. This, for Skinner, is an innocuous sort of control. The planners of the society are not acting against the wishes of the members—they are shaping those wishes. People do what they want to do, although what they want to do is determined by what some others want them to want. It is in this sense, for Skinner, that Walden Two is a controlled society in which all the inhabitants are free. But the control of which we are now speaking is nothing like this at all. It consists simply in the proscription, by the authorities, of certain kinds of activities. If the planners decide that a certain type of music is likely to have an injurious effect, they will forbid it. This is political control pure and simple.

Interestingly enough, Skinner admits that control of the kind takes place in Walden Two (page 164). If someone in the society comes up with something new, he does not try to bring it about that the society at large accept his innovation. He goes to the relevant planning board. It is there that the decision about the acceptability of what he has done is made. If the planners decide against his idea they will prohibit its circulation. Thus is met the threat of “arbitrariness.”

It is of absolute importance to recognize the difference between control in this latter sense and behavioral conditioning. Walden Two seems to present us with a picture of a society in which the ordinary, “nasty” elements of political control are absent. I suppose that many readers of the book have found this to be one of its most attractive features. But, as we have shown above, this is an illusion. As long as the planners of the behaviorist utopia do not condition values and abilities down to the last thinkable atom, it will be absolutely necessary to control the society (in the ordinary sense) in order to insure that the chosen values maintain their supremacy. Whatever merits Walden Two may have, the absence of rigid political control is not one of them.

This last point obviously raises the question of freedom, now in the political sense, and not in the sense of freedom of the will. Skinner argues that the former question is one that finds an easy solution in Walden Two. If people do what they want to do (whether or not their wants have been selected by someone else) then they are free. Suppose we accepted this account. Would it then follow that the inhabitants of the behaviorist utopia are free? Surely not. Consider a citizen of the society going to the planning board convinced that his new idea will result in enormous progress in sundry fields. He wants (if the word is not too weak) to have the idea introduced. But after many hours of pleading with the planners (whom he comes to look on, at least in part, as thick-headed), he is turned down. He asks, as a last desperate move, to be permitted to give a public address on the merits of his new scheme. This too is turned down. The planners, we may suppose, feel that any propagation of the ideas they have just heard will have a markedly deleterious effect on the rest of society. This sort of situation is clearly possible in the behaviorist utopia. Skinner admits as much when he says that new ideas will have to be cleared through the relevant planning boards. But the situation we have described is a situation in which someone is not allowed to do what he wants to do. Thus we can grant Skinner all he asks for (in the way of a definition of freedom) and still the society he envisages is an unfree one. Not only are wants and desires (on some level) conditioned by the behavioral engineers, but any deviation from the established pattern is suppressed by those behavioral engineers when it does not accord with their own views. Walden Two may now begin to look, not like an excitingly new departure in the theory of social organization, but like an old-fashioned totalitarian society of a kind with which we are already familiar.

Thus we have seen that at every important stage in the argument for the behaviorist utopia, the position adopted by Skinner is one that cannot be maintained. But let us suppose now that none of this was shown. Let us suppose, that is, that Skinner’s argument, up to this point, has been a good one. Would even that show that his recommendations are worthy of acceptance? Is a controlled society (assuming that it can do most of what Skinner says) a desirable thing? Again, I submit, the answer is no. But, given our assumption, it is not possible to show this by pointing up internal contradictions in the theory behind Walden Two. We are assuming, in some vague sense, that Walden Two can “work.”

Is a working Walden Two a more desirable social organization than a free society? The answer to this question obviously depends, in part, on the sort of criteria one employs in assessing the merits of a particular form of social organization. Given certain ends, Walden Two is more desirable. Frazier indicates, in the course of his advocacy of Walden Two, that an important feature of the society is that it provides the opportunity to make controlled social experiments. Obviously, Walden Two would provide a more effective medium in which to conduct controlled social experiments (of a certain kind) than would a free society. But this is, after all, a rather restricted goal. And the fact that it is so restricted renders it untenable as a justifying goal for a form of social organization. One doesn’t construct a society for the purpose of experimenting with its members. Skinner recognizes this, and what he regards as justifying the society is something else. Walden Two is supposed to represent an extremely progressive society, and one remarkably well adapted to achieve the goals set for it. It is in these terms that the issue between Walden Two and a free society must be discussed.10

ONE OF THE MOST important criteria by which we judge societies, of course, is a great degree of satisfaction of material wants—that is, the society must be an economically successful one. And the economic question is especially important in regard to Walden Two: much of the glamor of the life there proceeds from the fact that the inhabitants, having to work no more than four hours a day, have a great deal of leisure for the pursuit of culture and general self-improvement. Unfortunately, I am afraid that there is not much use in discussing this aspect of Skinner’s ideas. He states, in his book, that the economics of a society are “child’s play,” and, indeed, he proceeds to deal with economic problems much as a child would. His ideas here are simply naive socialism, of the silliest sort. One example: once in a while, the children in Walden Two are sent out into the world and given a

sort of detective assignment. The game is to establish a connection in the shortest possible time between any given bit of luxury and some piece of poverty or depravity. The children may start with a fine residence, for example. By going in the service drive they may be able to speak to a colored laundress hanging out clothes. They induce her to let them drive her home. That’s enough [page 206].

Enough for what? To prove that all wealth in a free enterprise system is gotten unfairly? To prove that wealth in a society is like a pie, and that if some have more others must necessarily have less? Does Skinner know what an economic cost is, and that there are fortunes to be made by entrepreneurs who can discover ways of cutting costs which their competitors have not yet realized? Does Skinner know that there exists a question of the practicability of rational allocation of resources in the absence of a market for capital goods—in such a situation as obtains in Walden Two, for instance? Offhand, one would have considered an economics primer required reading for a society-builder. But in any case, the economic problems of a utopian socialist community are much beyond the scope of our discussion here. Let us, therefore, pass on to the more general question of social control vs. the decentralization characteristic of a free society.

I have mentioned that Skinner admits the necessity for political control over new ideas. This concession occurs in his discussion of the “Walden Code,” a system of maxims regulating the conduct of the members of the community in an incredibly detailed way: the Code even governs how introductions between people are to be made, establishes what is to count as rudeness in a conversation and prohibits (for the sake of psychological health) the deliberate expression of gratitude between members of the society. Now, in regard to this Code, Skinner states:

As to disagreement, anyone may examine the evidence upon which a rule was introduced into the Code. He may argue against its inclusion and may present his own evidence. If the Managers refuse to change the rule, he may appeal to the Planners. But in no case must he argue about the Code with the members at large. There’s a rule against that.11

What this involves, then, is that every innovation in social life will have to gain the approval of the behavioral engineers. Thus, the distinction between Walden Two and the open society becomes as clear-cut as possible, and we are in a position to introduce what is probably the chief argument for freedom.

Let us attempt to state this argument in the form of an analogy—the analogy between society as a whole considered as an organization for the acquisition and communication of knowledge, and any given semi-orgnized field within society serving the same functon. If advances in social knowledge (in the best ways of meeting the problems that arise in human life) are best promoted by a system of rigid controls such as Skinner’s why not apply this system to science and the arts? If we established a commission of the best physicists, say, who would decide what lines of thought subordinate physicists would be allowed to follow up, is it likely that breakthroughs in this science would be as frequent as in these past centuries of “anarchy,” when each physicist has done as he wished? In other fields, to cite illustrations is virtually to close the case as far as these disciplines are concerned: who would wish that Bradley and Bosanquet—the outstanding British philosophers of their day—had been commissars of philosophy, empowered to decide whether Russell and Moore would be allowed to present their ideas to the public? Who would wish to have set Haydn in a similar way as arbiter over the young Beethoven?

These examples illustrate the principle that the progress of ideas is not served by casting over any given field—and a fortiori, over society as a whole—the mental limitations of one mind or one group of minds, no matter how superior a position they occupy in regard to other individual minds.

Now, it may be objected that the analogy is a false one; that the behaviorist psychologists would occupy, in relation to their wards, an extremely superior position, not at all comparable to that of Haydn to Beethoven. That is, the behaviorist might argue: “Of course, we would not wish a young innovating genius to have his hands tied by the established and conservative leaders in his field. But this would not be the case with human behavior. It is not the ideas of a young Beethoven that will be ‘suppressed’ by our behavioral engineers; it is simply the ‘ideas’ of the masses of uninspired, generally mediocre men and women, and they will assuredly gain in our system.”

But such a rebuttal fails to see the full implications of the problem of innovation. Very often, innovation depends not only on qualities internal to the innovator (which conceivably could be predicted by a system of testing), but on a unique combination of circumstances in which the innovator finds himself. As Hayek describes the situation:

. . . we have no way of predicting who will at each step first make the appropriate move or what particular combinations of knowledge and skills will suggest to some man the suitable answer, or by what channels his example will be transmitted to others who will follow the lead. It is difficult to conceive all the combinations of knowledge and skills which thus come into action and from which arises the discovery of appropriate practices or devices that, once found, can be accepted generally. But from the countless number of humble steps taken by anonymous persons in the course of doing familiar things in changed circumstances spring the examples that prevail. They are as important as the major intellectual innovations which are explicitly recognized and communicated as such.12

TO PUT ANYONE in charge of a field of human activity—or of human activity in general—is necessarily to place him in a position of arbiter over every potential innovator. The original question remains: “Why doesn’t Skinner propose establishing commissions in each science and art, to pass judgment on the efforts of all its practitioners?”

It seems clear, then, that innovation will not be promoted by a controlled society. But perhaps a controlled society is in the best position to make use of the knowledge already available (we are speaking now of “social knowledge”—the sort of thing with which the “Walden Code” deals). Actually, it is difficult to know what to say here, because Skinner appears to be maintaining a position little short of incredible. When he says, for instance, that the expression of gratitude is not to be allowed between members of society—does he mean this literally? Will it, under all possible circumstances, represent an infraction of the Code to express gratitude? What if a man saves my life at the risk of his own? Or, to put it more strongly, what if he submits to torture, in order to conceal my place of hiding from my would-be murderers? Does Skinner mean to say that I will have to clear it with a board of behavioral engineers before I express my gratitude? Perhaps the rule concerning gratitude, when it emerges from the laboratories, will contain a clause permitting its expression in certain cases. But then, how many clauses will it contain? In general, how will it be possible for a group of men—endowed, as we may assume them to be, with a great deal of scientific psychological knowledge—to foresee all the possible combinations of circumstances which occur in social life, circumstances which often include the pitting of one generally-accepted rule against another?

Since it is clearly impossible, perhaps Skinner could allow a certain leeway in the interpretation of his gratitude rule. But it requires little insight to see that this would be a fatal breach in his system. Once people are permitted a degree of personal discretion in the application of a behaviorist “truth,” it becomes possible for them to modify the rule itself, in the way that judges have made dead-letters of various laws by their interpretations of them. Control over behavior would once more pass from the hands of the psychologists to those of society at large, the very situation Walden Two was created to eliminate. Thus, we must assume that no leeway will be allowed in the interpretation of the various behaviorist maxims. Now, what does this imply? It implies that the only place where society’s accumulated knowledge can be refined and fitted for application to an infinitely great variety of circumstances is in the laboratories of the behavioral psychologists. If someone thinks that a received rule ought to be modified in its application to the particular situation in which he finds himself, he is not allowed to do so. He must obey the rule as it has been handed down to him, and await the judgment of the behavioral engineers as to the permissibility of his suggested modification. Isn’t it obvious that this would bring about virtually the least flexible and least adaptable society imaginable? To arbitrarily exclude by far the larger segment of society from the work of adapting received rules and maxims to various situations is to eliminate all these minds as centers for acquiring, refining and passing on knowledge. Such a procedure would make sense if there were good reason to believe that the minds put in control were omniscient—but our previous discussion of free will and selective conditioning does not support that claim, at least as far as one important behaviorist is concerned, and it is unlikely that the case would be different with his fellows.

The argument for freedom is—quite unfortunately, from the viewpoint of the prospects for a free society—an enormously complicated and abstract one, and we have only touched the surface here.13 But, I hope that enough has been said to suggest that Walden Two is hardly a rival to a free society with respect to its ability both to adapt to changing conditions and to promote progress.

A word is perhaps in order for those who have read Walden Two and are made enthusiastic by its promise of a better existence for all, with less frustration and fewer misdirected lives than we see around us today. The important thing to realize is that the choice before us is not between whatever improvements behavioral psychology offers us in the problems of arranging our lives, and the benefits a free society offers us in the way of innovation and progress. Skinner’s utopia indeed precludes the advantages accruing from the free play of ideas and the clash of values of the open society, but the reverse is not the case. In an uncontrolled social setting, people are free to adapt their lives in the light of behavioral psychology, and, if a substantial part of what Skinner claims for his ideas is true, then the same society which gave birth to them will be free to apply them to advantage.

The real value of the book, I think, is that it performs one of the functions credited by John Stuart Mill to even the most erroneous doctrines: by energetically presenting the case for a centrally-directed social order, it leads us to reconsider fundamentals, and forces us to re-examine and refine the arguments for the open society. Thus, while we may thank Skinner for promoting this end—which was indeed no part of his intention—those of us who are reasonable must decline the offering.

[* ] Bruce Goldberg did graduate work in philosophy at Princeton and Oxford Universities, and is presently an instructor in philosophy at the University of Illinois. He is a contributor to Analysis and The Journal of Philosophy.

[1 ] I am, of course, ignoring those areas of psychology where mathematics is becoming increasingly important, e.g., learning theory.

[2 ] Originally published in 1948. I have used the paperback edition published by Macmillan in 1962.

[3 ]Science and Human Behavior (New York: Macmillan, 1953), p. 5.

[4 ] For joy, Skinner tells us, is a productive emotion (page 102). This seems hardly to be a “scientific” remark. What is it for an emotion to be “productive”? Is the view something like that joyful people do more productive work than sad ones? Is there any empirical evidence for this? Skinner provides no experimental evidence; and even if one were to make a historical survey which showed that joyful people are, in fact, more productive than sad people, that wouldn’t establish the claim. Their greater productivity might be attributable to all sorts of other factors. But Skinner probably doesn’t mean anything as precise as this. More likely, all he means by calling joy a productive emotion is that he has a favorable attitude towards joy and not towards sorrow. In any case, nowhere are any criteria presented for what is to count as a productive emotion.

[5 ] There have been exceptions in the history of philosophy. Hume and John Stuart Mill accepted both determinism and free will. But this is no place for a historical account of the question.

[6 ] Cf. P. F. Strawson, Freedom and Resentment (London: Oxford University Press, 1962), p. 187.

[7 ] It should be clear that by uncontrolled society I don’t mean one in which there is no control over behavior. Obviously, a society which prohibits certain acts (e.g., murder) is to that extent controlling behavior. By an uncontrolled society. I don’t mean an anarchic one. This is not the place to enter upon an extended discussion of the semantics of “control.” For my purposes it is enough to say that by an uncontrolled society I mean one in which the sorts of control that are characteristic of both Walden Two and Soviet society are absent—thought control, thoroughly propagandized education, etc.

[8 ] Of course, in this discussion I am accepting something which I spent much time earlier in denying, viz., the possibility of creating people to order.

[9 ] While Skinner seems to be vague enough on how questions of value are decided in Walden Two, he is at least candid in admitting that science plays no part in the decisions. “The philosopher in search of a rational basis for deciding what is good has always reminded me of the centipede trying to decide how to walk. Simply go ahead and walk! We all know what’s good, until we stop to think about it” (page 159). If Skinner means by this that universal agreement can be obtained on what is and what is not good, his statement is obviously false—suffering, promiscuous sexual behavior, etc., are things concerning the goodness of which a Christian and a hedonist might well disagree. But even assuming that we could cleanly separate good from bad things, this would in no way solve the problem of the behavioral engineer. Since he is charged with arranging the value systems of the members of society, and since very often a choice must be made among good things (at the very least because they can’t all be realized at the same time), he must have a good idea of the order in which goods are to be ranked among themselves—which are the more important ones, and which the less. This problem scarcely permits of a ready answer.

[10 ] Of course, a free society does not have goals set for it in the same sense that Skinner’s controlled society would, for there is no one in a position to set goals for everyone else. All we mean here by the goals of a free society are the ends which, in considering it, we would like it to achieve.

[11 ] It ought to be pointed out, incidentally, that the “scientific” nature of Walden Two is always stated rather than shown. That is, Frazier’s defense of a particular practice in his community takes the form not of the presentation of any pointer readings, but simply of suggestive inferences about human beings, much as might occur in a conversation between two educated persons with an interest in “what makes people tick.” (Sometimes, indeed, it descends somewhat beneath this level—see for instance his argumentation concerning the irrelevancy of history, with his vague but impossioned championship of “the Now” [page 239].) But, presumably, Skinner is not calling for the acceptance of any of the particular practices he describes (for which he gives no “laboratory evidence” at all); it is rather the method which he wants us to accept.

[12 ] F. A. Hayek, The Constitution of Liberty (Chicago: University of Chicago Press, 1960), p. 28.

[13 ] Much the most profound discussion of these problems which I have come across is contained in Hayek’s Constitution of Liberty. An interesting discussion of Skinner’s social ideas from a psychological point of view is provided by Carl R. Rogers, “The Place of the Individual in the New World of the Behaviorial Sciences,” in On Becoming a Person (Boston: Houghton Mifflin, 1961), pp. 384-402.