The Steep and Thorny Way to a Science of Behavior

B. F. Skinner

A critic contends that a recent book of mine (Skinner, 1971) does not contain anything new, that much the same was said more than four centuries ago in theological terms by John Calvin. You will not be surprised, then, to find me commending to you the steep and thorny way to that heaven promised by a science of behavior. But I am not one of those ungracious pastors, of whom Ophelia complained, who "recking not their own rede themselves tread the primrose path of dalliance." 'No, I shall rail at dalliance, and in a manner worthy, I hope, of my distinguished predecessor. If I do not thunder or fulminate, it is only because we moderns can more easily portray a truly frightening hell. I shall merely allude to the carcinogenic fallout of a nuclear holocaust. And no Calvin ever had better reason to fear his hell, for I am proceeding on the assumption that nothing less than a vast improvement in our understanding of human behavior will prevent the destruction of our way of life or of mankind.

Why has it been so difficult to be scientific about human behavior? Why have methods that have been so prodigiously successful almost everywhere else failed so ignominiously in this one field? Is it because human behavior presents unusual obstacles to a science? No doubt it does, but I think we are beginning to see how these obstacles may be overcome. The problem, I submit, is digression. We have been drawn off the straight and narrow path, and the word diversion serves me well by suggesting not only digression but dalliance. In this article I analyze some of the diversions peculiar to the field of human behavior which seem to have delayed our advance toward the better understanding we desperately need.

A Science 0f Behavior

I must begin by saying what I take a science ot behavior to be. It is, I assume, part of biology. The organism that behaves is the organism that breathes, digests, conceives, gestates, and so on. As such, the behaving organism will eventually be described and explained by the anatomist and physiologist, As far as behavior is concerned, they will give us an account of the genetic endowment of the species and tell how that endowment changes during the lifetime of the individual and why, as a result, the individual then responds in a given way on a given occasion. Despite remarkable progress, we are still a long way from a satisfactory account in such terms. We know something about the chemical and electrical effects of the nervous system and the location of many of its functions, but the events that actually underlie a single instance of behavior – as a pigeon picks up a stick to build a nest, or a child a block to complete a tower, or a scientist a pen to write a paper – are still far out of reach.

Fortunately, we need not wait for further progress of that sort. We can analyze a given instance of behavior in its relation to the current setting and to antecedent events in the history of the species and of the individual. Thus, we do not need an explicit account of the anatomy and physiology of genetic endowment in order to describe the behavior, or the behavioral processes, characteristic of a species, or to speculate about the contingencies of survival under which they might have evolved, as the ethologists have convincingly demonstrated. Nor do we need to consider anatomy and physiology in order to see how the behavior of the individual is changed by his exposure to contingencies of reinforcement during his lifetime and how as a result he behaves in a given way on a given occasion. I must confess to a predilection here for my own specialty, the experimental analysis of behavior, which is a quite explicit investigation of the effects upon individual organisms of extremely complex and subtle contingencies of reinforcement.

There will be certain temporal gaps in such an analysis. The behavior and the conditions of which it is a function do not occur in close temporal or spatial proximity, and we must wait for physiology to make the connection. When it does so, it will not invalidate the behavioral account (indeed, its assignment could be said to be specified by that account), nor will it make its terms and principles any the less useful. A science of behavior will be needed for both theoretical and practical purposes even when the behaving organism is fully understood at another level, just as much of chemistry remains useful even though a detailed account of a single instance may be given at the level of molecular or atomic forces. Such, then, is the science of behavior from which I suggest we have been diverted – by several kinds of dalliance to which I now turn.

Feelings and Their Relation to Behavior

Very little biology is handicapped by the fact that the biologist is himself a specimen of the thing he is studying, but that part of the science with which we are here concerned has not been so fortunate. We seem to have a kind of inside information about our behavior. It may be true that the environment shapes and controls our behavior as it shapes and controls the behavior of other species – but we have feelings about it, And what a diversion they have proved to be. Our loves, our fears, our feelings about war, crime, poverty, and God – these are all basic, if not ultimate, concerns. And we are as much concerned about the feelings of others. Many of the great themes of mythology have been about feelings – of the victim on his way to sacrifice or of the warrior going forth to battle. We read what poets tell us about their feelings, and we share the feelings of characters in plays and novels. We follow regimens and take drugs to alter our feelings. We become sophisticated about them in, say, the manner of La Rochefoucauld, noting that jealousy thrives on doubt, or that the clemency of a ruler is a mixture of vanity, laziness, and fear. And along with some psychiatrists we may even try to establish an independent science of feelings in the intra-psychic life of the mind or personality.

And do feelings not have some bearing on our formulation of a science of behavior? Do we not strike because we are angry and play music because we feel like listening? And if so, are our feelings not to be added to those antecedent events of which behavior is a function? This is not the place to answer such questions in detail, but I must at least suggest the kind of answer that may be given. William James questioned the causal order: Perhaps we do not strike because we are I angry but feel angry because we strike. That does not bring us back to the environment, however, although James and others were on the right track. What we feel are conditions of our bodies, most of them closely associated with behavior and with the circumstances in which we behave, We both strike and feel angry for a common reason, and that reason lies in the environment. In short, the bodily conditions we feel are collateral products of our genetic and environmental histories. They have no explanatory force; they are simply additional facts to be taken into account.

Feelings enjoy an enormous advantage over genetic and environmental histories. They are warm, salient, and demanding, where facts about the environment are easily overlooked. Moreover, they are immediately related to behavior, being collateral products of the same causes, and have therefore commanded more attention than the causes themselves, which are often rather remote. In doing so, they have proved to be one of the most fascinating attractions along the path of dalliance.

Environment and Its Relation to Behavior

A much more important diversion has, for more than 2,000 years, made any move toward a science of behavior particularly difficult. The environment acts upon an organism at the surface of its body, but when the body is our own, we seem to observe its progress beyond that point; for example, we seem to see the real world become experience, a physical presentation become a sensation or a percept. Indeed, this second stage may be all we see. Reality may be merely an inference and, according to some authorities, a bad one. What is important may not be the physical world on the far side of the skin but what that world means to us on this side.

Not only do we seem to see the environment on its way in, we seem to see behavior on its way out. We observe certain early stages – wishes, intentions, ideas, and acts of will – before they have, as we say, found expression in behavior. And as for our environmental history, that can also be viewed and reviewed inside the skin, for we have tucked it all away in the storehouse of our memory. Again this is not the place to present an alternative account, but several points need to be made. The behavioristic objection is not primarily to the metaphysical nature of mind stuff. I welcome the view, clearly gaining in favor among psychologists and physiologists and by no means a stranger to philosophy, that what we introspectively observe, as well as feel, are states of our bodies. But I am not willing to give introspection much of a toehold even so, for there are two important reasons why we do not discriminate precisely among our feelings and states of mind and hence why there are many different philosophies and psychologies.

In the first place, the world within the skin is private. Only the person whose skin it is can make certain kinds of contact with it. We might expect that the resulting intimacy should make for greater clarity, but there is a difficulty. The privacy interferes with the very process of coming to know. The verbal community which teaches us to make distinctions among things in the world around us lacks the information it needs to teach us to distinguish events in our private world. For example, it cannot teach us the difference between diffidence and embarrassment as readily or as accurately as that between red and blue or sweet and sour.

Second, the self·observation that leads to introspective knowledge is limited by anatomy. It arose very late in the evolution of the species because it is only when a person begins to be asked about his behavior and about why he behaves as he does that he becomes conscious of himself in this sense. Self knowledge depends on language and in fact on language of a rather advanced kind, but when questions of this sort first began to be asked, the only nervous systems available in answering them were those that had evolved for entirely different reasons. They had proved useful in the internal economy of the organism, in the coordination of movement, and in operating upon the environment, but there was no reason why they should be suitable in supplying information about those very extensive systems that mediate behavior. To put it crudely, introspection cannot be very relevant or comprehensive because the human organism does not have nerves going to the right places,

One other problem concerns the nature and location of the knower. The organism itself lies, so to speak, between the environment that acts upon it and the environment it acts upon, but what lies between those inner stages – between, for example, experience and will? From what vantage point do we watch stimuli on their way into the storehouse of memory or behavior on its way out to physical expression? The observing agent, the knower, seems to contract to something very small in the middle of things.

In the formulation of a science with which I began, it is the organism as a whole that behaves. It acts in and upon a physical world, and it can be induced by a verbal environment to respond to some of its own activities. The events observed as the life of the mind, like feelings, are collateral products, which have been made the basis of many elaborate metaphors. The philosopher at his desk asking himself what he really knows, about himself or the world, will quite naturally begin with his experiences, his acts of will, and his memory, but the effort to understand the mind from that vantage point, beginning with Plato’s supposed discovery, has been one of the great diversions which have delayed an analysis of the role of the environment.

Observation of a Behaving Organism

It did not, of course, take inside information to induce people to direct their attention to what is going on inside the behaving organism. We almost instinctively look inside a system to see how it works. We do this with clocks, as with living systems. It is standard practice in much of biology. Some early efforts to understand and explain behavior in this way have been described by Onians (1951) in his classic Origins of European Thought. It must have been the slaughterhouse and the battlefield that gave man his first knowledge of anatomy and physiology. The various functions assigned to parts of the organism were not usually those that had been observed introspectively. If Onians is right, the phrénes were the lungs, intimately associated with breathing and hence, so the Greeks said, with thought and, of course, with life and death. The phrénes were the seat of thumós, a vital principle whose nature is not now clearly understood, and possibly of ideas, in the active sense of Homeric Greek. (By the time an idea had become an object of quiet contemplation, interest seems to have been lost in its location.) Later, the various fluids of the body, the humors, were associated with dispositions, and the eye and the ear with sense data. I like to imagine the consternation of that pioneer who first analyzed the optics of the eyeball and realized that the image on the retina was upside down!

Observation of a behaving system from without began in earnest with the discovery of reflexes, but the reflex arc was not only not the seat of mental action, it was taken to be a usurper, the spinal reflexes replacing the Rückenmarkseele or soul of the spinal cord, for example. The reflex arc was essentially an anatomical concept, and the physiology remained largely imaginary for a long time. Many years ago I suggested that the letters CNS could be said to stand, not for the central nervous system, but for the conceptual nervous system. I had in mind the great physiologists Sir Charles Sherrington and Ivan Petrovich Pavlov. In his epoch·making Integrative Action af the Nervous System, Sherrington (1906) had analyzed the role of the synapse, listing perhaps a dozen characteristic properties. I pointed out that he had never seen a synapse in action and that all the properties assigned to it were inferred from the behavior of his preparations. Pavlov had offered his researches as evidence of the activities of the cerebral cortex though he had never observed the cortex in action but had merely inferred its processes from the behavior of his experimental animals. But Sherrington, Pavlov, and many others were moving in the direction of an instrumental approach, and the physiologist is now, of course, studying the nervous system directly.

The conceptual nervous system has been taken over by other disciplines – by information theory, cybernetics, systems analyses, mathematical models, and cognitive psychology, The hypothetical structures they describe do not depend on confirmation by direct observation of the nervous system, for that lies too far in the future to be of interest. They are to be justified by their internal consistency and the successful prediction of selected facts, presumably not the facts from which the constructions were inferred.

These disciplines are concerned with how the brain or the mind must work if the human organism is to behave as it does. They offer a sort of thermodynamics of behavior without reference to molecular action. The computer with its apparent simulation of Man Thinking supplies the dominant analogy. It is not a question of the physiology of the computer – how it is wired or what type of storage it uses – but of its behavioral characteristics. A computer takes in information as an organism receives stimuli and processes it according to an inbuilt program as an organism is said to do according to its genetic endowment. It encodes the information, converting it to a form it can handle, as the organism converts visual, auditory, and other stimuli into nerve impulses. Like its human analogue it stores the encoded information in a memory, tagged to facilitate retrieval. It uses what it has stored to process information as received, as a person is said to use prior experience to interpret incoming stimuli, and later to perform various operations – in short, to compute. Finally, it makes decisions and behaves: It prints out.

There is nothing new about any of this. The same things were done thousands of years ago with clay tiles. The overseer or tax collector kept a record of bags of grain, the number, quality, and kind being marked appropriately. The tiles were stored in lots as marked, additional tiles were grouped appropriately, the records were eventually retrieved and computations made, and a summary account was issued. The machine is much swifter, and it is so constructed that human participation is needed only before and after the operation. The speed is a clear advantage, but the apparent autonomy has caused trouble. It has seemed to mean that the mode of operation of a computer resembles that of a person. People do make physical! records which they store and retrieve and use in solving problems, but it does not follow that they do anything of the sort in the mind. If there were some exclusively subjective achievement, the argument for the so called higher mental processes would be stronger, but as far as I know, none has been demonstrated. True, we say that the mathematician sometimes intuitively solves a problem and only later, if at all, reduces it to the steps of a proof, and in doing so he seems to differ greatly from those who proceed step by step, but the differences could well be in the evidence of what has happened, and it would not be very satisfactory to define thought simply as unexplained behavior.

Again, it would be foolish of me to try to develop an alternative account in the space available. What I have said about the introspectively observed mind applies as well to the mind that is constructed from observations of the behavior of others. The accessibility of stored memories, for example, can be interpreted as the probability of acquired behaviors, with no loss in the adequacy of the treatment of the facts, and with a very considerable gain in the assimilation of this difficult field with other parts of human behavior.

An Inner Life

I have said that much of biology looks inside a living system for an explanation of how it works. But that is not true of all of biology. Sir Charles Bell could write a book on the hand as evidence of design. The hand was evidence; the design lay elsewhere. Darwin found the design, too, but in a different place. He could catalog the creatures he discovered on the voyage of the Beagle in terms of their form or structure, and he could classify barnacles for years in the same way, but he looked beyond structure for the principle of natural selection. It was the relation of the organism to the environment that mattered in evolution. And it is the relation to environment that is of primary concern in the analysis of behavior. Hence, it is not enough to confine oneself to organization or structure, even of the most penetrating kind. That is the mistake of most of phenomenology, existentialism, and the structuralism of anthropology and linguistics. When the important thing is a relation to the environment, as in the phylogeny and ontogeny of behavior, the fascination with an inner system becomes a simple digression.

We have not advanced more rapidly to the methods and instruments needed in the study of behavior precisely because of the diverting preoccupation with a. supposed or real inner life. It is true that the introspective psychologist and the model builder have investigated environments, but they have done so only to throw some light on the internal events in which they are interested. They are no doubt well intentioned helpmates, but they have often simply misled those who undertake the study of the organism as a behaving system in its own right. Even when helpful, an observed or hypothetical inner determiner is no explanation of behavior until it has itself been explained, and the fascination with an inner life has allayed curiosity about the further steps to be taken.

I can hear my critics: “Do you really mean to say that all those who have inquired into the human mind, from Plato and Aristotle through the Romans and scholastics, to Bacon and Hobbes, to Locke and the other British empiricists, to John Stuart Mill, and to all those who began to call themselves psychologists – that they have all been wasting their time?" Well, not all of their time, fortunately. Forget their purely psychological speculations, and they were still remarkable people. They would have been even more remarkable, in my opinion, if they could have forgotten that speculation themselves. They were careful observers of human behavior, but the intuitive wisdom they acquired from their contact with real people was flawed by their theories.

It is easier to make the point in the field of medicine. Until the present century very little was known about bodily processes in health and disease from which useful therapeutic practices could be derived. Yet it should have been worthwhile to call in a physician. Physicians saw many ill people and should have acquired a kind of wisdom, unanalyzed perhaps but still of value in prescribing simple treatments. The history of medicine, however, is largely the history of barbaric practices – bloodlettings, cuppings, poultices, purgations, violent emetics – which much of the time must have been harmful. My point is that these measures were not suggested by the intuitive wisdom acquired from familiarity with illness; they were suggested by theories, theories about what was going on inside an ill person. Theories of the mind have had a similar effect, less dramatic, perhaps, but quite possibly far more damaging. The men I have mentioned made important contributions in government, religion, ethics, economics, and many other fields. They could do so with an intuitive wisdom acquired from experience. But philosophy and psychology have had their bleedings, cuppings, and purgations too, and they have obscured simple wisdom. They have diverted wise people from a path that would have led more directly to an eventual science of behavior. Plato would have made far more progress toward the good life if he could have forgotten those shadows on the wall of his cave.

The Individual as a Creator

Still another kind of concern for the self distracts us from the program I have outlined. It has to do with the individual, not as an object of self-knowledge, but as an agent, an initiator, a creator. I have developed this theme in Beyond Freedom and Dignity. We are more likely to give a person credit for what he does if it is not obvious that it can be attributed to his physical or social environment, and we are likely to feel that truly great achievements must be inexplicable. The more derivative a work of art, the less creative; the more conspicuous the personal gain, the less heroic an act of sacrifice. To obey a well enforced law is not to show civic virtue. We see a concern for the aggrandizement of the individual, for the maximizing of credit due him, in the self-actualization of so called humanistic psychology, in some versions of existentialism, in Eastern mysticism and certain forms of Christian mysticism in which a person is taught to reject the world in order to free himself for union with a divine principle or with God, as well as in the simple structuralism that looks to the organization of behavior rather than to the antecedent events responsible for that organization. The difficulty is that if the credit due a person is infringed by exigences of the conditions of which his behavior is a function, then a scientific analysis appears to be an attack on human worth or dignity. Its task is to explain the hitherto inexplicable and hence to reduce any supposed inner contribution which has served in lieu of explanation. Freud moved in this direction in explaining creative art, and it is no longer just the cynic who traces heroism and martyrdom to powerful indoctrination. The culminating achievement of the human species has been said to be the evolution of man as a moral animal, but a simpler view is that it has been the evolution of cultures in which people behave morally although they have undergone no inner change of character.

Even more traumatic has been the supposed attack on freedom. Historically, the struggle for freedom has been an escape from physical restraint and from behavioral restraints exerted through punishment and exploitative measures of other kinds. The individual has been freed from features of his environment arranged by governmental and religious agencies and by those who possess great wealth. The success of that struggle, though it is not yet complete, is one of man’s great achievements, and no sensible person would challenge it. Unfortunately, one of its byproducts has been the `slogan that "all control of human behavior is wrong and must be resisted." Nothing in the circumstances under which man has struggled for freedom justifies this extension of the attack on controlling measures, and we should have to abandon all of the advantages of a well developed culture if we were to relinquish all practices involving the control of human behavior. Yet new techniques in education, psychotherapy, incentive systems, penology, and the design of daily life are currently subject to attack because they are said to threaten personal freedom, and I can testify that the attack can be fairly violent.

The extent to which a person is free or responsible for his achievements is not an issue to be decided by rigorous proof, but I submit that what we call the behavior of the human organism is no more free than its digestion, gestation, immunization, or any other physiological process. Because it involves the environment in many subtle ways it is much more complex, and its lawfulness is, therefore, much harder to demonstrate, But a scientific analysis moves in that direction, and we can already throw some light on traditional topics, such as free will or creativity, which is more helpful than traditional accounts, and I believe that further progress is imminent.

The issue is, of course, determinism. Slightly more than 100 years ago, in a famous paper, Claude Bernard raised with respect to physiology the issue which now stands before us in the behavioral sciences. The almost insurmountable obstacle to the application of scientific method in biology was, he said, the belief in “vital spontaneity," His contemporary, Louis Pasteur, was responsible for a dramatic test of the theory of spontaneous generation, and I suggest that the spontaneous generation of behavior in the guise of ideas and acts of will is now at the stage of the spontaneous generation of life in the form of maggots and microorganisms 100 years ago.

The practical problem in continuing the struggle? for freedom and dignity is not to destroy controlling forces but to change them, to create a world in which people will achieve far more than they have ever achieved before in art, music, literature, science, technology, and above all the enjoyment of life. It could be a world in which people feel freer than they have ever felt before, because they will not be under aversive control. In building such a world, we shall need all the help a science of behavior can give us. To misread the theme of the struggle for freedom and dignity, and to relinquish all efforts to control would be a tragic mistake.

But it is a mistake that may very well be made. Our concern for the individual as a creative agent is not dalliance; it is clearly an obstacle rather than a diversion, for ancient fears are not easily allayed. A shift in emphasis from the individual to the environment, particularly to the social environment, is reminiscent of various forms of totalitarian statism. It is easy to turn from what may seem like an inevitable movement in that direction and to take one’s chances with libertarianism. But much remains to be analyzed in that position. For example, we may distinguish between liberty and license by holding to the right to do as we please provided we do not infringe upon similar rights of others, but in doing so we conceal or disguise the public sanctions represented by private rights. Rights and duties, like a moral or ethical sense, are examples of hypothetical internalized environmental sanctions.

In the long run, the aggrandizement of the individual jeopardizes the future of the species and the culture. In effect, it infringes the so called rights of billions of people still to be born, in whose interests only the weakest of sanctions are now maintained. We are beginning to realize the magnitude of the problem of bringing human behavior under the control of a projected future, and we are already suffering from the fact that we have come very late to recognize that mankind will have a future only if it designs a viable way of life. l wish I could share the optimism of both Darwin and Herbert Spencer that the course of evolution is necessarily toward perfection. It appears, on the contrary, that that course must be corrected from time to time. But if the intelligent behavior that corrects it is also a product of evolution, then perhaps they were right after all. But it could be a near thing.

Diversions and Obstacles

Perhaps it is now clear what I mean by diversions and obstacles. The science I am discussing is the investigation of the relation between behavior and the environment – on the one hand, the environment in which the species evolved and which is responsible for the facts investigated by the ethologists and, on the other hand, the environment in which the individual lives and in response to which at any moment he behaves. We have been diverted from, and blocked in, our inquiries into the relations between behavior and those environments by an absorbing interest in the organism itself. We have been misled by the almost instinctive tendency to look inside any system to see how it works, a tendency doubly powerful in the case of behavior because of the apparent inside information supplied by feelings and introspectively observed states. Our only recourse is to leave that subject to the physiologist, who has, or will have, the only appropriate instruments and methods. We have also been encouraged to move in a centripetal direction because the discovery of controlling forces in the environment has seemed to reduce the credit due us for our achievements and to suggest that the struggle for freedom has not been as fully successful as we had imagined. We are not yet ready to accept the fact that the task is to change, not people, but rather the world in which they live.

A Different Approach

We shall be less reluctant to abandon these diversions and to attack these obstacles, as we come to understand the possibility of a different approach. The role of the environment in human affairs has not, of course, gone unnoticed. Historians and biographers have acknowledged influences on human conduct, and literature has made the same point again and again. The Enlightenment advanced the cause of the individual by improving the world in which he lived – the Encyclopedia of Diderot and D’Alembert was designed to further changes of that sort, and by the nineteenth century the controlling force of the environment was clearly recognized. Bentham and Marx have been called behaviorists, although for them the environment determined behavior only after first determining consciousness, and this was an unfortunate qualification because the assumption of a mediating state clouded the relation between the terminal events.

The role of the environment has become clearer in the present century. Its selective action in evolution has been examined by the ethologists, and a similar selective action during the life of the individual is the subject of the experimental analysis of behavior. In the current laboratory, very complex environments are constructed and their effects on behavior studied. I believe this work offers consoling reassurance to those who are reluctant to abandon traditional formulations. Unfortunately, it is not well known outside the field. Its practical uses are, however, beginning to attract attention. Techniques derived from the ‘ analysis have proved useful in other parts of biology – for example, physiology and psycho-pharmacology – and have already led to the improved design of cultural practices, in programmed instructional materials, contingency management in the classroom, behavioral modification in psychotherapy and penology, and many other fields.

Much remains to be done, and it will be done more rapidly when the role of the environment takes its proper place in competition with the apparent evidences of an inner life. As Diderot put it, nearly 200 years ago, “Unfortunately it is easier and shorter to consult oneself than it is to consult nature. Thus the reason is inclined to dwell within itself." But the problems we face are not to be found in men and women but in the world in which they live, especially in those social environments we call cultures. It is an important and promising shift in emphasis because, unlike the remote fastness of the so called human spirit, the environment is within reach and we are learning how to change it.

And so I return to the role that has been assigned to me as a kind of twentieth century Calvin, calling on you to forsake the primrose path of total individualism, of self actualization, self adoration, and self love, and to turn instead to the construction 0f that heaven on earth which is, I believe, within reach of the methods of science. I wish to testify that, once you are used to it, the way is not so steep or thorny after all.


Onians, R. D. The origins of European thought. Cambridge, England: University Press, 1951.

Sherrington, C. S. Integrative action of the nervous system. New Haven, Conn.: Yale University Press, 1906.

Skinner, B, F, Beyond freedom and dignity. New York: Knopf, 1971.