The Etymology of Teaching

The educated person differs from the uneducated in almost everything he does. Three great metaphors have been devised to account for the behavior which distinguishes him.

Growth or development. The behavior is sometimes attributed to maturation, the embryo with its minimal contact with the environment providing a good analogy. The metaphor is most convincing in the early years. The behavior of a child is studied as a function of time; charts and graphs record the appearance of responses at various ages; and typical performances are established as norms. The results can be used to predict behavior but, since time cannot be controlled, not to change it. The emphasis is on the topography of behavior its form or structure. The metaphor assigns only a modest role to the teacher, who "cannot really teach but only help the student learn." To teach is to nourish or cultivate the growing child (as in a Kindergarten), or to give him intellectual exercise, or to train him in the horticultural sense of directing or guiding his growth.

Development does not easily account for many features of behavior which are obviously derived from the environment. A child may be born capable of learning to speak English, but he is certainly not born English-speaking. What grows or develops cannot be behavior as such. It is often said instead to be certain inner requirements or determinants of behavior, such as cognitive powers, faculties, or traits of character. Education is said to be the culture of the intellect or mind. A student grows in wisdom. He behaves more successfully when concepts emerge in his thinking.

Acquisition. The environmental variables neglected by growth or development find their place in a second metaphor in which the student gets his knowledge and skills from the world around him. He receives an education. The learning process is followed in curves of acquisition. The teacher plays the active role of transmitter. He shares his experiences. He gives and the student takes. The energetic student grasps the structure of facts or ideas. If he is less active, the teacher impresses facts upon him, or drills ideas into him, or inculcates good taste or a love of learning ("inculcate" originally meant to grind under the heel).

In an osmotic version of the metaphor of acquisition, the student absorbs knowledge from the world about him. He soaks up information. What the teacher says sinks in. Teaching is a form of alchemical wetting: the student is imbued with a love of learning, ideas are infused, wisdom is instilled. In a gastronomic version the student has an appetite or thirst for knowledge. He digests facts and principles (provided he is not given more than he can chew). In another version, teaching is impregnation. The teacher is seminal (a tout vent). He propagates knowledge. He engenders thoughts. He implants the germs of ideas, and the student conceives (provided he has a fertile mind). A medical version is based on infection or contagion.

As these expressions show, transmission is also a plausible metaphor only if we are speaking of inner states or entities. The teacher does not actually pass along some of his own behavior. He is said to impart knowledge, possibly only after subdividing it into meanings, concepts, facts, and propositions. (Theories of learning which emphasize acquisition make the same concession: behavior is only "performance"; what are acquired are associations, concepts, hypotheses, and so on - depending on the theory.) What is transmitted must also be stored (the teacher stocks the student's mind, and the student retains what he has acquired), but it is not behavior but only certain precursors or determiners which can be stored in memory.

These conceptual maneuvers are necessary because neither growth nor acquisition correctly represents the interchange between organism and environment. Growth is confined to one variable, the form or structure of behavior and acquisition adds a second -the stimulating environment; but two variables are still not enough, as the inadequacies of both stimulus-response and information theories show. Superficially, the exchange between organism and environment may be viewed as a matter of input and output, but difficulties arise. Some discrepancies may be attributed to overloading, blocking, and so on, but output still cannot be accounted for solely in terms of input. Certain inner activities - physiological in stimulus-response theories, cognitive in information theory, are therefore invented and given just those properties needed to complete the account.

Apart from theoretical difficulties, neither metaphor tells the teacher what to do or lets him see what he has done. No one literally cultivates the behavior of a child as one cultivates a garden or transmits information as one carries a letter from one place to another.

Construction. A student possesses a genetic endowment which develops or matures, and his behavior becomes more and more complex as he makes contact with the world around him, but something else happens as he learns. If we must have a metaphor to represent. teaching, instruction (or, better, the cognate construction) will serve. In this sense we say that the teacher informs the student, in the sense that his behavior is given form or shape. To teach is to edify in the sense of building. It is possible, of course, to say that the teacher builds precursors such as knowledge, habits, or interests, but the metaphor of construction does not demand this because the behavior of the student can in a very real sense be constructed.

All three metaphors are embedded in our language and it is perhaps impossible to avoid them in informal discussion. Many examples will be found in the present text. Any serious analysis of the interchange between organism and environment must, however, avoid metaphor. Three variables compose the so-called contingencies of reinforcement under which learning takes place: (1) an occasion upon which behavior occurs, (2) the behavior itself, and (3) the consequences of the behavior. Contingencies so composed, together with their effects, have been thoroughly investigated in the experimental analysis of behavior upon which this book is based. Some familiarity with any science is, of course, helpful in considering its technological applications, and probably no part of a scientific analysis of human behavior is irrelevant to education, but a detailed familiarity is not assumed in what follows. Facts and principles will be presented as needed.

So far as we are concerned here, teaching is simply the arrangement of contingencies of reinforcement. Left to himself in a given environment a student will learn, but he will not necessarily have been taught. The school of experience is no school at all, not because no one learns in it, but because no one teaches. Teaching is the expediting of learning; a person who is taught learns more quickly than one who is not. Teaching is most important, of course, when the behavior would not otherwise arise. (Everything which is now taught must have been learned at least once by someone who was not being taught, but thanks to education we no longer need to wait for these rare events.)

Three Theories

Certain traditional ways of characterizing learning and teaching appear to be not so much wrong as incomplete in the sense that they do not fully describe the contingencies of reinforcement under which behavior changes.

“We learn by doing." It is important to emphasize that a student does not passively absorb knowledge from the world around him but must play an active role, and also that action is not simply talking. To know is to act effectively, both verbally and nonverbally. But a student does not learn simply by doing. Although he is likely to do things he has already done, we do not make it more likely that he will do something a second time by getting him to do it once. We do not teach a child to throw a ball simply by inducing him to throw it. It is not true, as Aristotle asserted, that we learn harp-playing by playing the harp and ethical behavior by behaving ethically. If learning occurs under such circumstances, it is because other conditions have been inadvertently arranged. Much more than going through the motions is involved when a child throws a ball or a student plays a harp or behaves ethically. Execution of the behavior may be essential, but it does not guarantee that learning will take place.

“Frequency theories" extend the notion of learning by doing. When one instance of a response makes no obvious difference, the teacher adds other instances. There are plausible analogies. If we spin the end of a stick against a stone, we may leave no mark, but if we spin it repeatedly, we drill a hole, and we drill our students in the same sense. A wheel passing over hard ground may leave no trace, but if it passes often enough, it leaves a rut or route, and this is the sense in which our students learn by rote. The teacher induces his student to exercise or practice so that his habits, like his muscles, will grow stronger with use. But it is what is happening frequently, not mere frequency," which is the important thing.

“Recency theories" also emphasize learning by doing. An organism is likely to do again what it has done because conditions responsible for the first response probably still prevail and may even have been improved. Having observed one occurrence, therefore, we often successfully predict a second, but only because we then have evidence that conditions are favorable.

“We learn from experience." The student is to learn about the world in which he lives and must be brought into contact with it. The teacher therefore provides the student with experiences, singling out features to be noted or sets of features to be associated, often by pairing a verbal response with the thing or event it describes: "This is a gazebo," or “Note that the fluid rises in the tube." From experience alone a student probably learns nothing. He will not even perceive the environment simply because he is in contact with it.

Combining experience with doing, we arrive at a two variable formulation in which "experience" represents stimulus or input and “doing" represents response or output. Possibly what is learned is a connection between the two. But why is a connection made? The usual answer (appropriate to a two-variable formulation) invokes hypothetical inner activities. The student does something. He “learns,” for example, as a kind of mental action; he processes the information he receives from the environment; he organizes his experiences; he forms connections in his mind. We are forced to assume that he does all this because we have neglected important variables in the environment to which the result could otherwise be traced.

“We learn by trial and error.” Certain stimuli standing in a different temporal relation to behavior remain to be taken into account. They compose another kind of experience, the significance of which is often expressed by saying that we learn by trial and error. The reference is to the consequences of behavior, which are often called, with some suggestion as to their effects, reward and punishment.

The concept of trial and error has a long history in the study of problem solving and other forms of learning in both animals and men. Learning curves are commonly plotted to show changes in the number of errors made in performing a task. A sampling of behavior is generally called a trial. The formula is easily applied to daily affairs, but it is quite inadequate in describing the role played by the consequences of behavior in contingencies of reinforcement. No doubt we often learn from our errors (at least we may learn not to make them again), but correct behavior is not simply what remains when erroneous behavior has been chipped away. When we characterize behavior as “trying," we inject a reference to consequences into what should be a description of the topography of response. The term “error" does not refer to the physical dimensions of the consequences, even those called punishment. The implication that learning occurs only when errors are made is false.

These classical theories represent the three essential parts of any set of contingencies of reinforcement: learning by doing emphasizes the response; learning from experience, the occasion upon which the response occurs; and learning by trial-and-error, the consequences. But no one part can be studied entirely by itself, and all three parts must be recognized in formulating any given instance of learning. It would be difficult to bring the three theories together to compose a useful formulation. Fortunately, we do not need to do so. Such theories are now of historical interest only, and unfortunately much of the work which was done to support them is also of little current value. We may turn instead to a more adequate analysis of the changes which take place as a student learns.

The Science of Learning and the Art of Teaching

Some promising advances have recently been made in the field of learning. Special techniques have been designed to arrange what are called contingencies of reinforcement - the relations which prevail between behavior on the one hand and the consequences of that behavior on the other - with the result that a much more effective control of behavior has been achieved. It has long been argued that an organism learns mainly by producing changes in its environment, but it is only recently that these changes have been carefully manipulated. In traditional devices for the study of learning - in the serial maze, for example, or in the T-maze, the problem box, or the familiar discrimination apparatus - the effects produced by the organism's behavior are left to many fluctuating circumstances. There is many a slip between the turn-to-the right and the food-cup at the end of the alley. It is not surprising that techniques of this sort have yielded only very rough data from which the uniformities demanded by an experimental science can be extracted only by averaging many cases. In none of this work has the behavior of the individual organism been predicted in more than a statistical sense. The learning processes which are the presumed object of such research are reached only through a series of inferences.

Recent improvements in the conditions which control behavior in the field of learning are of two principal sorts. The Law of Effect has been taken seriously; we have made sure that effects do occur and that they occur under conditions which are optimal for producing the changes called learning. Once we have arranged the particular type of consequence called a reinforcement, our techniques permit us to shape the behavior of an organism almost at will. It has become a routine exercise to demonstrate this in classes in elementary psychology by conditioning such an organism as a pigeon. Simply by presenting food to a hungry pigeon at the right time, it is possible to shape three or four well-defined responses in a single demonstration period - such responses as turning around, pacing the floor in the pattern of a figure eight, standing still in a corner of the demonstration apparatus, stretching the neck, or stamping the foot. Extremely complex performances may be reached through successive stages in the shaping process, the contingencies of reinforcement being changed progressively in the direction of the required behavior. The results are often quite dramatic. In such a demonstration one can see learning take place. A significant change in behavior is often obvious as the result of a single reinforcement.

A second important advance in technique permits us to maintain behavior in given states of strength for long periods of time. Reinforcements continue to be important, of course, long after an organism has learned how to do something, long after it has acquired behavior. They are necessary to maintain the behavior in strength. Of special interest is the effect of various schedules of intermittent reinforcement. Most of the basic schedules have been investigated and in general have been reduced to a few principles. On the theoretical side we now have a fairly good idea of why a given schedule produces its appropriate performance. On the practical side we have learned how to maintain any given level of activity for daily periods limited only by the physical endurance of the organism and from day to day without substantial change throughout its life. Many of these effects would be traditionally assigned to the field of motivation, although the principal operation is simply the arrangement of contingencies of reinforcement.

These new methods of shaping behavior and of maintaining it in strength are a great improvement over the traditional practices of professional animal trainers, and it is not surprising that our laboratory results are already being applied to the production of performing animals for commercial purposes. In a more academic environment they have been used for demonstration purposes which extend far beyond an interest in learning as such. For example, it is not too difficult to arrange the complex contingencies which produce many types of social behavior. Competition is exemplified by two pigeons playing a modified game of ping-pong. The pigeons drive the ball back and forth across a small table by pecking at it. When the ball gets by one pigeon, the other is reinforced. The task of constructing such a "social relation" is probably completely out of reach of the traditional animal trainer. It requires a carefully designed program of gradually changing contingencies and the skillful use of schedules to maintain the behavior in strength. Each pigeon is separately prepared for its part in the total performance, and the social relation is then arbitrarily constructed. The events leading up to this stable state are excellent material for the study of the factors important in nonsynthetic social behavior. It is instructive to consider how a similar series of contingencies could arise in the case of the human organism through the evolution of cultural patterns. Cooperation can also be set up, perhaps more easily than competition. Two pigeons have been trained to coordinate their behavior in a cooperative endeavor with a precision which equals that of the most skillful human dancers.

In a more serious vein these techniques have made it possible to explore the complexities of the individual organism and to analyze some of the serial or coordinate behaviors involved in attention, problem solving, various types of self-control, and the subsidiary systems of responses within a single organism called personalities. Some of these are exemplified in what are called multiple schedules of reinforcement. In general, a given schedule has an effect upon the rate at which a response is emitted. Changes in the rate from moment to moment show a pattern typical of the schedule. The pattern may be as simple as a constant rate of responding at a given value; it may be a gradually accelerating rate between certain extremes; it may be an abrupt change from not responding at all to a given stable high rate. It has been shown that the performance characteristic of a given schedule can be brought under the control of a particular stimulus and that different performances can be brought under the control of different stimuli in the same organism. In one experiment performances appropriate to nine different schedules were brought under the control of appropriate stimuli presented at random. When Stimulus 1 was present, the pigeon executed the performance appropriate to Schedule 1. When Stimulus 2 was present, the pigeon executed the performance appropriate to Schedule 2. And so on. This result is important because it makes the extrapolation of our laboratory results to daily life much more plausible. We are all constantly shifting from schedule to schedule as our immediate environment changes.

It is also possible to construct very complex sequences of schedules. It is not easy to describe these in a few words, but two or three examples may be mentioned. In one experiment the pigeon generates a performance appropriate to Schedule A where the reinforcement is simply the production of the stimulus characteristic of Schedule B, to which the pigeon then responds appropriately. Under a third stimulus, the bird yields a performance appropriate to Schedule C where the reinforcement in this case is simply the production of the stimulus characteristic of Schedule D, to which the bird then responds appropriately. In a special case, first investigated by L. B. Wyckoff, Jr., the organism responds to one stimulus where the reinforcement consists of the clarification of the stimulus controlling another response. The first response becomes, so to speak, an objective form of "paying attention" to the second stimulus. In an important version of this experiment, we could say that the pigeon is telling us whether it is paying attention to the shape of a spot of light or to its color.

One of the most dramatic applications of these techniques has been made by Floyd Ratliff and Donald S. Blough, who have skillfully used multiple and serial schedules of reinforcement to study complex perceptual processes in the infrahuman organism. They have achieved a sort of psychophysics without verbal instruction. In an experiment by Blough, for example, a pigeon draws a detailed dark-adaptation curve showing the characteristic breaks of rod and cone vision. The curve is recorded continuously in a single experimental period and is quite comparable with the curves of human subjects. The pigeon behaves in a way which, in the human case, we should not hesitate to describe by saying that it adjusts a very faint patch of light until it can just be seen.

In all this work, the species of the organism has made surprisingly little difference. It is true that the organisms studied have all been vertebrates, but they still cover a wide range. Comparable results have been obtained with pigeons, rats, dogs, monkeys, human children, and psychotic subjects. In spite of great phylogenic differences, all these organisms show amazingly similar properties of the learning process. It should be emphasized that this has been achieved by analyzing the effects of reinforcement and by designing techniques which manipulate reinforcement with considerable precision. Only in this way can the behavior of the individual organism be brought under such precise control. It is also important to note that through a gradual advance to complex interrelations among responses, the same degree of rigor is being extended to behavior which would usually be assigned to such fields as perception, thinking, and personality dynamics.

Schoolroom Teaching

From this exciting prospect of an advancing science of learning, it is a great shock to turn to that branch of technology which is most directly concerned with the learning process -education. Let us consider, for example, the teaching of arithmetic in the lower grades. The school is concerned with imparting to the child a large number of responses of a special sort. The responses are all verbal. They consist of speaking and writing certain words, figures, and signs which, to put it roughly, refer to numbers and to arithmetic operations. The first task is to shape these responses to get the child to pronounce and to write responses correctly, but the principal task is to bring this behavior under many sorts of stimulus control. This is what happens when the child learns to count, to recite tables, to count while ticking off the items in an assemblage of objects, to respond to spoken or written numbers by saying "odd," "even," or "prime." Over and above this elaborate repertoire of numerical behavior, most of which is often dismissed as the product of rote learning, the teaching of arithmetic looks forward to those complex serial arrangements of responses involved in original mathematical thinking. The child must acquire responses of transposing, clearing fractions, and so on, which modify the order or pattern of the original material so that the response called a solution is eventually made possible.

Now, how is this extremely complicated verbal repertoire set up? In the first place, what reinforcements are used? Fifty years ago the answer would have been clear. At that time educational control was still frankly aversive. The child read numbers, copied numbers, memorized tables, and performed operations upon numbers to escape the threat of the birch rod or cane. Some positive reinforcements were perhaps eventually derived from the increased efficiency of the child in the field of arithmetic and in rare cases some automatic reinforcement may have resulted from the sheer manipulation of the medium -from the solution of problems or the discovery of the intricacies of the number system. But for the immediate purposes of education the child acted to avoid or escape punishment. It was part of the reform movement known as progressive education to make the positive consequences more immediately effective, but anyone who visits the lower grades of the average school today will observe that a change has been made, not from aversive to positive control, but from one form of aversive stimulation to another. The child at his desk, filling in his workbook, is behaving primarily to escape from the threat of a series of minor aversive events - the teacher's displeasure, the criticism or ridicule of his classmates, an ignominious showing in a competition, low marks, a trip to the office "to be talked to" by the principal, or a word to the parent who may still resort to the birch rod. In this welter of aversive consequences, getting the right answer is in itself an insignificant event, any effect of which is lost amid the anxieties, the boredom, and the aggressions which are the inevitable by-products of aversive control.

Secondly, we have to ask how the contingencies of reinforcement are arranged. When is a numerical operation reinforced as "right"? Eventually, of course, the pupil may be able to check his own answers and achieve some sort of automatic reinforcement, but in the early stages the reinforcement of being right is usually accorded by the teacher. The contingencies she provides are far from optimal. It can easily be demonstrated that, unless explicit mediating behavior has been set up, the lapse of only a few seconds between response and reinforcement destroys most of the effect. In a typical classroom, nevertheless, long periods of time customarily elapse. The teacher may walk up and down the aisle, for example, while the class is working on a sheet of problems, pausing here and there to call an answer right or wrong. Many minutes intervene between the child's response and the teacher's reinforcement. In many cases, for example, when papers are taken home to be corrected, as much as 24 hours may intervene. It is surprising that this system has any effect whatsoever.

A third notable shortcoming is the lack of a skillful program which moves forward through a series of progressive approximations to the final complex behavior desired. A long series of contingencies is necessary to bring the pupil into the possession of mathematical behavior most efficiently. But the teacher is seldom able to reinforce at each step in such a series because she cannot deal with the pupil's responses one at a time. It is usually necessary to reinforce the behavior in blocks of responses - as in correcting a worksheet or page from a workbook. The responses within such a block must not be interrelated. The answer to one problem must not depend upon the answer to another. The number of stages through which one may progressively approach a complex pattern of behavior is therefore small, and the task so much the more difficult. Even the most modern workbook in beginning arithmetic is far from exemplifying an efficient program for shaping mathematical behavior.

Perhaps the most serious criticism of the current classroom is the relative infrequency of reinforcement. Since the pupil is usually dependent upon the teacher for being told that he is right, and since many pupils are usually dependent upon the same teacher, the total number of contingencies which may be arranged during, say, the first four years, is of the order of only a few thousand. But a very rough estimate suggests that efficient mathematical behavior at this level requires something of the order of 25,000 contingencies. We may suppose that even in the brighter student a given contingency must be arranged several times to place the behavior well in hand. The responses to be set up are not simply the various items in tables of addition, subtraction, multiplication, and division; we have also to consider the alternative forms in which each item may be stated. To the learning of such material we should add hundreds of responses such as those concerned with factoring, identifying primes, memorizing series, using short-cut techniques of calculation, and constructing and using geometric representations or number forms. Over and above all this, the whole mathematical repertoire must be brought under the control of concrete problems of considerable variety. Perhaps 50,000 contingencies is a more conservative estimate. In this frame of reference, the daily assignment in arithmetic seems pitifully meagre.

The result of all this is, of course, well known. Even our best schools are under criticism for their inefficiency in the teaching of drill subjects such as arithmetic. The condition in the average school is a matter of widespread national concern. Modern children simply do not learn arithmetic quickly or well. Nor is the result simply incompetence. The very subjects in which modem techniques are weakest are, those in which failure is most conspicuous, and in the wake of an ever-growing incompetence come the anxieties, uncertainties, and aggressions which in their turn present other problems to the school. Most pupils soon claim the asylum of not being "ready" for arithmetic at a given level or, eventually, of not having a mathematical mind. Such explanations are readily seized upon by defensive teachers and parents. Few pupils ever reach the stage at which automatic reinforcements follow as the natural consequences of mathematical behavior. On the contrary, the figures and symbols of mathematics have become standard emotional stimuli. The glimpse of a column of figures, not to say an algebraic symbol or an integral sign, is likely to set off, not mathematical behavior, but a reaction of anxiety, guilt, or fear.

The teacher is usually no happier about this than the pupil. Denied the opportunity to control via the birch rod, quite at sea as to the mode of operation of the few techniques at her disposal, she spends as little time as possible on drill subjects and eagerly subscribes to philosophies of education which emphasize material of greater inherent interest. A confession of weakness is her extraordinary concern lest the child be taught something unnecessary. The repertoire to be imparted is carefully reduced to an essential minimum. In the field of spelling, for example, a great deal of time and energy has gone into discovering just those words which the young child is going to use, as if it were a crime to waste one's educational power in teaching an unnecessary word. Eventually, weakness of technique emerges in the disguise of a reformulation of the aims of education. Skills are minimized in favor of vague achievements - educating for democracy, educating the whole child, educating for life, and so on. And there the matter ends; for, unfortunately, these philosophies do not in turn suggest improvements in techniques. They offer little or no help in the design of better classroom practices.

The Improvement of Teaching

There would be no point in urging these objections if improvement were impossible. But the advances which have recently been made in our control of the learning process suggest a thorough revision of classroom practices and, fortunately, they tell us how the revision can be brought about. This is not, of course, the first time that the results of an experimental science have been brought to bear upon the practical problems of education. The modern classroom does not, however, offer much evidence that research in the field of learning has been respected or used. This condition is no doubt partly due to the limitations of earlier research. But it has been encouraged by a too hasty conclusion that the laboratory study of learning is inherently limited because it cannot take into account the realities of the classroom. In the light of our increasing knowledge of the learning process we should, instead, insist upon dealing with those realities and forcing a substantial change in them. Education is perhaps the most important branch of scientific technology. It deeply affects the lives of all of us. We can no longer allow the exigencies of a practical situation to suppress the tremendous improvements which are within reach. The practical situation must be changed.

There are certain questions which have to be answered in turning to the study of any new organism. What behavior is to be set up? What reinforcers are at hand? What responses are available in embarking upon a program of progressive approximation which will lead to the final form of the behavior? How can reinforcements be most efficiently scheduled to maintain the behavior in strength? These questions are all relevant in considering the problem of the child in the lower grades.

In the first place, what reinforcements are available? What does the school have in its possession which will reinforce a child? We may look first to the material to be learned, for it is possible that this will provide considerable automatic, reinforcement. Children play for hours with mechanical toys, paints, scissors and paper, noise-makers, puzzles - in short, with almost anything which feeds back significant changes in the environment and is reasonably free of aversive properties. The sheer control of nature is itself reinforcing. This effect is not evident in the modern school because it is masked by the emotional responses generated by aversive control. It is true that automatic reinforcement from the manipulation of the environment is probably only a mild reinforcer and may need to be carefully husbanded, but one of the most striking principles to emerge from recent research is that the net amount of reinforcement is of little significance. A very slight reinforcement may be tremendously effective in controlling behavior if it is wisely used.

If the natural reinforcement inherent in the subject matter is not enough, other reinforcers must be employed. Even in school the child is occasionally permitted to do "what he wants to do," and access to reinforcements of many sorts may be made contingent upon the more immediate consequences of the behavior to be established. Those who advocate competition as a useful social motive may wish to use the reinforcements which follow from excelling others, although there is the difficulty that in this case the reinforcement of one child is necessarily aversive to another. Next in order we might place the good will and affection of the teacher, and only when that has failed need we turn to the use of aversive stimulation.

In the second place, how are these reinforcements to be made contingent upon the desired behavior? There are two considerations here - the gradual elaboration of extremely complex patterns of behavior and the maintenance of the behavior in strength at each stage. The whole process of becoming competent in any field must be divided into a very large number of very small steps, and reinforcement must be contingent upon the accomplishment of each step. This solution to the problem of creating a complex repertoire of behavior also solves the problem of maintaining the behavior in strength. We could, of course, resort to the techniques of scheduling already developed in the study of other organisms, but in the present state of our knowledge of educational practices scheduling appears to be most effectively arranged through the design of the material to be learned. By making each successive step as small as possible, the frequency of reinforcement can be raised to a maximum, while the possibly aversive consequences of being wrong are reduced to a minimum. Other ways of designing material would yield other programs of reinforcement. Any supplementary reinforcement would probably have to be scheduled in the more traditional way.

These requirements are not excessive, but they are probably incompatible with the current realities of the classroom. In the experimental study of learning it has been found that the contingencies of reinforcement which are most efficient in controlling the organism cannot be arranged through the personal mediation of the experimenter. An organism is affected by subtle details of contingencies which are beyond the capacity of the human organism to arrange. Mechanical and electrical devices must be used. Mechanical help is also demanded by the sheer number of contingencies which may be used efficiently in a single experimental session. We have recorded many millions of responses from a single organism during thousands of experimental hours. Personal arrangement of the contingencies and personal observation of the results are quite unthinkable. Now, the human organism is, if anything, more sensitive to precise contingencies than the other organisms we have studied. We have every reason to expect, therefore, that the most effective control of human learning will require instrumental aid. The simple fact is that, as a mere reinforcing mechanism, the teacher is out of date. This would be true even if a single teacher devoted all her time to a single child, but her inadequacy is multiplied many fold when she must serve as a reinforcing device to many children at once. If the teacher is to take advantage of recent advances in the study of learning, she must have the help of mechanical devices.

A Teaching Machine

The technical problem of providing the necessary instrumental aid is not particularly difficult. There are many ways in which the necessary contingencies may be arranged, either mechanically or electrically. An inexpensive device which solves most of the principal problems has already been constructed. It is still in the experimental stage, but it suggests the kind of instrument which seems to be required. The device is a box about the size of a small record player. On the top surface is a window through which a question or problem printed on a paper tape may be seen. The child answers the question by moving one or more sliders upon which the digits 0 through 9 are printed. The answer appears in square holes punched in the paper upon which the question is printed. When the answer has been set, the child turns a knob. The operation is as simple as adjusting a television set. If the answer is right, the knob turns freely and can be made to ring a bell or provide some other conditioned reinforcement. If the answer is wrong, the knob will not turn. A counter may be added to tally wrong answers. The knob must then be reversed slightly and a second attempt at a right answer made. (Unlike the flash card, the device reports a wrong answer without giving the right answer.) When the answer is right, a further turn of the knob engages a clutch which moves the next problem into place in the window. This movement cannot be completed, however, until the sliders have been returned to zero.

The important features of the device are these: reinforcement for the right answer is immediate. The mere manipulation of the device will probably be reinforcing enough to keep the average pupil at work for a suitable period each day, provided traces of earlier aversive control can be wiped out. A teacher may supervise an entire class at work on such devices at the same time, yet each child may progress at his own rate, completing as many problems as possible within the class period. If forced to be away from school, he may return to pick up where he left off. The gifted, child will advance rapidly, but can be kept from getting too far ahead either by being excused from arithmetic for a time or by being given special sets of problems which take him into some of the interesting bypaths of mathematics.

The device makes it possible to present carefully designed material in which one problem can depend upon the answer to the preceding problem and where, therefore, the most efficient progress to an eventually complex repertoire can be made. Provision has been made for recording the commonest mistakes so that the tapes can be modified as experience dictates. Additional steps can be inserted where pupils tend to have trouble, and ultimately the material will reach a point at which the answers of the average child will almost always be right.

If the material itself proves not to be sufficiently reinforcing, other reinforcers in the possession of the teacher or school may be made contingent upon the operation of the device or upon progress through a series of problems. Supplemental reinforcement would not sacrifice the advantages gained from immediate reinforcement and from the possibility of constructing an optimal series of steps which approach the complex repertoire of mathematical behavior most efficiently.

A similar device in which the sliders carry the letters of the alphabet has been designed to teach spelling. In addition to the advantages which can be gained from precise reinforcement and careful programming, the device will teach reading at the same time. It can also be used to establish the large and important repertoire of verbal relationships encountered in logic and science. In short, it can teach verbal thinking. The device can also be operated as a multiple choice self-rater.

Some objections to the use of such devices in the classroom can easily be foreseen. The cry will be raised that the child is being treated as a mere animal and that an essentially human intellectual achievement is being analyzed in unduly mechanistic terms. Mathematical behavior is usually regarded, not as a repertoire of responses involving numbers and numerical operations, but as evidences of mathematical ability or the exercise of the power of reason. It is true that the techniques which are emerging from the experimental study of learning are not designed to "develop the mind" or to further some vague "understanding" of mathematical relationships. They are designed, on the contrary, to establish the very behaviors which are taken to be the evidences of such mental states or processes. This is only a special case of the general change which is under way in the interpretation of human affairs. An advancing science continues to offer more and more convincing alternatives to traditional formulations. The behavior in terms of which human thinking must eventually be defined is worth treating in its own right as the substantial goal of education.

Of course the teacher has a more important function than to say right or wrong. The changes proposed should free her for the effective exercise of that function. Marking a set of papers in arithmetic, "Yes, nine and six are fifteen; no, nine and seven are not eighteen" is beneath the dignity of any intelligent person. There is more important work to be done in which the teacher's relations to the pupil cannot be duplicated by a mechanical device. Instrumental help would merely improve these relations. One might say that the main trouble with education in the lower grades today is that the child is obviously not competent and knows it and that the teacher is unable to do anything about it and knows that too. If the advances which have recently been made in our control of behavior can give the child a genuine competence in reading, writing, spelling, and arithmetic, then the teacher may begin to function, not in lieu of a cheap machine, but through intellectual, cultural, and emotional contacts of that distinctive sort which testify to her status as a human being.

Another possible objection is that mechanized instruction will mean technological unemployment. We need not worry about this until there are enough teachers to go around and until the hours and energy demanded of the teacher are comparable to those in other fields of employment. Mechanical devices will eliminate the more tiresome labors of the teacher but they will not necessarily shorten the time during which she remains in contact with the pupil.

A more practical objection: Can we afford to mechanize our schools? The answer is clearly Yes. The device I have just described could be produced as cheaply as a small radio or phonograph. There would need to be far fewer devices than pupils, for they could be used in rotation. But even if we suppose that the instrument eventually found to be most effective would cost several hundred dollars and that large numbers of them would be required, our economy should be able to stand the strain. Once we have accepted the possibility and the necessity of mechanical help in the classroom, the economic problem can easily be surmounted. There is no reason why the schoolroom should be any less mechanized than, for example, the kitchen. A country which annually produces millions of refrigerators, dishwashers, automatic washing machines, automatic clothes driers, and automatic garbage disposers can certainly afford the equipment necessary to educate its citizens to high standards of competence in the most effective way.

There is a simple job to be done. The task can be stated in concrete terms. The necessary techniques are known. The equipment needed can easily be provided. Nothing stands in the way but cultural inertia. But what is more characteristic of the modern temper than an unwillingness to accept the traditional as inevitable? We are on the threshold of an exciting and revolutionary period, in which the scientific study of man will be put to work in man’s best interests. Education must play its part. It must accept the fact that a sweeping revision of educational practices is possible and inevitable. When it has done this, we may look forward with confidence to a school system which is aware of the nature of its tasks, secure in its methods, and generously supported by the informed and effective citizens whom education itself will create.

From B.F. Skinner, The Technology of Teaching (1968).