Positive Reinforcement at Home, in Institutions Positive Reinforcement at Home, in Institutions
By Murray Sidman
Excerpt from Coercion and its Fallout (1989), Chapter 16, "Is There Any Other Way?" pp. 214-223

Positive Reinforcement at Home.

We enter parenthood without anyone ever having taught us how to fulfill that responsibility. We learn quickly that children make special demands. Before they can talk they learn to express and enforce those demands in the only ways possible for them. At first they cry and scream. Later, if they have not been taught other ways of communicating. they turn from crying and screaming to aggression, destructiveness, and more subtle forms of misbehavior. These bring the immediate satisfaction of their needs to the top of the adult priority list.

Even infants can develop an arsenal of coercive practices. Reinforcing those practices perpetuates them, often transforming the children from lovable bundles of joy into objects to be escaped from, avoided, and aggressed against. This, often in combination with other stresses, lies behind those more and more frequently reported incidents of child abuse. Punishment may produce the peace a desperate parent needs -- at the expense of the inevitable side effects -- but it offers the chastened child no altemative course of action, no way to adapt constructively. Providing a diversion instead of a punishment would leave the child interacting productively and happily with the environment. Instead of scolding or isolating a crying child, we can often stop the tears by bringing out a plaything. Parents who react not by punishing but by giving their children opportunities to obtain positive reinforcers find themselves with happy, self-reliant, competent children. Households that practice positive reinforcement enjoy an additional benefit: occasions for punishment rarely arise. If positive reinforcers become available whenever a child acts badly, does that not convey a message? Will the child not learn to misbehave, to do more of the same? That does happen. Anyone who has wondered about the possibility is well on the way to a useful understanding of how conduct is controlled. The principle was caught nicely in a cartoon that had one child telling another, "My parents don't pay much attention to me, but they are really all right. All I have to do is say "bleep" a few times, and they give me anything I want." Used unskillfully, positive reinforcement can strengthen conduct that is just as unwanted as any of coercion's side effects. If we never provide affection, attention, and other reinforcers except when our children misbehave. the result will be continued misbehavior. It is not difficult to create little monsters. If positive reinforcement generally prevails in the household, however, occasional misbehavior will remain just that -- occasional; the children will learn that they do not have to act up in order to get us to attend to their wants.

Observant parents will learn to recognize signs of impending trouble. Children do not ordinarily just burst into misbehavior without having given signals that all is not right. They may complain, cling, reject favorite playthings and activities, imitate a younger sibling, or display any of a number of negativisms that characteristically precede an outbreak. The alert parent will not wait for the outbreak, but will immediately try to get the child to do something good, and then keep that behavior going with positive reinforcement.

Positive reinforcers given unskillfully can cause problems. Handing out reinforcers independently of anything a child does will teach the child that anything goes. The extreme outcome of completely unconditional giving is the spoiled brat, who may continue that way into adulthood. We all know adult brats who expect everything to be provided for them no matter what they have or have not done, who act as they please, paying no heed to the consequences their actions have for anyone else.

We do not, of course, want just a "you scratch my back and I'll scratch yours" relationship with our children. We want to use positive reinforcers to teach them how to live happy and productive lives, but without making them feel that they always have to do something special to earn our love and support. We want them to feel secure, to know that support, affection -- the whole supply of reinforcers -- is still available even though they might have done something wrong. We have to strike a balance. We have to maintain positive contingencies and at the same time generate trust and security. Our children must know that we will still be there even if they fail to meet a contingency successfully.

Nobody teaches us how to do this. But if we have learned that relationships thrive on positive reinforcement, we will know at least to look for noncoercive ways to instill security and self-confidence in our children even while maintaining contingencies. As long as we consider coercion necessary, we will never even ask the question. Behavior analysis provides no quantitative formulas, but it does state the problem in terms that permit solution.

The secret is to set up realistic contingencies that the child can meet. Do not ask for complexities too soon. Nothing breeds security and self-confidence like success. Also, provide real reinforcers, consequences that satisfy the child and not just the parent. Sometimes a pat on the head is not enough; a real hug may be needed. And sometimes a cookie will be more effective than a kiss. Finally, the inevitable failures should be used as occasions for teaching, not for punishment. Teaching, itself, should be a reinforcing interaction, both for child and parent.

Sometimes, things seem to have gone too far to be handled in any way except by punishment. The child is driving us up the wall with his whining; his temper tantrums are getting scary; his jabs at the baby are keeping us on edge. What do we do? First, take a quick look at what the child has been getting for acting that way; those reinforcers are keeping him going. His whining gets him anything he wants; his temper tantrums make him the focus of attention; being mean to baby brother keeps Mom closely attentive.

Should we just stop giving him those reinforcers? Do we just ignore him? That is a frequent suggestion, but it will not work. Children pick up objectionable ways to obtain reinforcers because they cannot get them any other way. They need those reinforcers. Instead of ignoring the child, give him those same reinforcers for doing something else. Instead of waiting for him to whine before reading him a story, make a practice of reading to him when he has been playing quietly for a while; instead of making him beat his head against the wall to get attention, give him applause and approval when he says his rhymes, dances, and plays constructively; even when busy doing other things, keep frequent contact with the child, looking in on his activities and interacting with him; do not make him threaten the baby in order to attract interest.

Things can, of course, go too far. Sometimes preoccupied, sometimes not understanding, we make mistakes and end up having to deal with an emotional outburst that no rational treatment can possible overcome. And then, too, emergencies arise that have to be dealt with immediately or else somebody will be hurt. Punishment may indeed be necessary to put a quick halt to a dangerous situation. These are not occasions to be concerned about. In a relationship based on strong and frequent positive reinforcement, an infrequent punishment is not going to cause any long-term damage. If these mistakes or emergencies start to happen frequently, however, they are danger signals, indicating a deteriorating relationship.

Positive reinforcement is not just for children. Loving support, helpfulness, kindness, and all the reciprocal amenities and responsibilities of conjugal living will keep a marriage going. But love given under coercion will only keep the coercion going. Waiting to be coerced into giving positive reinforcers is the same as asking to be coerced. Like parents who guarantee their children's misbehavior by reinforcing it, spouses who submit to a demanding, unappreciative, and hurtful mate will bring more of the same on themselves. It should come as no surprise when a coerced spouse, seizing any opportunity for negative reinforcement, escapes to more appreciative arms. How many alcoholics and workaholics, drug addicts and TV addicts are actually escapees from spousal coercion?

Noncontingent love, too, can spoil an adult just as effectively as it can a child. At any age, love that is always given unconditionally will teach the recipient that giving is a one-way street. Parents or grandparents who enjoy all the advantages of the family without having to pull their own weight can become egocentric, unappreciative, inconsiderate, and generally coercive, demanding more and more attention to their own wants.

The elderly need positive reinforcement just as much as children do to build and maintain their own sense of security and worth. Young parents, whose children need them, can receive almost automatic positive reinforcement from their children's progress, but their own parents may occupy an anomalous position. Grownup children, no longer dependent, may still give their parents all respect and consideration but ask nothing in return. Unless the older ones have a successful and reinforcing life of their own, a huge void may open up in their existence. No longer needed - with no community to show its appreciation for what they can do -- they will have little reason to behave. They may end up depressed and uncaring.

It does the elderly no favor to treat them only with respect and kindness but to ask nothing of them. They need the positive reinforcers that have always come from the use of their skills and from the interactions that these made possible. Request their help with baby-sitting, fmances, advice, influence, cooking, kitchen chores, housekeeping, transportation, home repairs, gardening, telephone calls, letter writing, and other family obligations; encourage and record reminiscences and family-tree data; get them to come along on family trips. Even physiologically-caused senility can be ameliorated to some extent by making requests and even demands that they can answer successfully. Deprivation of opportunities to achieve positive reinforcers is equivalent to unavoidable shock, a form of noncontingent punishment to which we unknowingly subject our elderly.

Positive Reinforcement in Institutions

Those who pose threats to themselves or to society at large, we frequently commit to institutions. There, we permit them only limited social relationships, deprive them of freedom of movement and of opportunities for decision making, and forbid most of the amenities that they could enjoy outside. We often justify these institutions as instruments for beneficial change: "Schools" for the handicapped are supposed to teach their pupils new skills to help them overcome their limitations; "hospitals" for the mentally ill are supposed to cure them; "correctional institutions" are supposed to rehabilitate lawbreakers.

Locating these facilities in areas that are relatively unpopulated and difficult to get to (at least initially, before cities or suburbs have grown up around them) indicates, however, what we really intend them for. They are supposed to keep the retarded, the insane, and the criminal out of sight. We hand these "humane" facilities over to members of the helping professions -- physicians, nurses, psychologists, behavior analysts, physical therapists, speech therapists, teachers, social workers, and correctional officers -- and wash our hands of the problems.

Their geographic isolation, their walls, fences, gates, and security towers, and the public tendency to ignore the very fact of their existence leave these institutions almost completely without control from the outside. Whatever humanitarian impulses might have led to their initial establishment, their freedom from public accountability turns most of them into little more than warehouses for society's misfits. The immediate priorities of staff and administrative convenience, inmate docility, and obedience of rules and regulations replace longer-term educational, therapeutic, or correctional goals. Coercion then becomes the technique of choice forgetting the inmates to "behave."

An institution that is operated mostly for the benefit of the staff attaches little significance to the deleterious side effects of coercion. And so we find coercion prevailing in the institutional management of the retarded, of the mentally ill, and of criminals of all types. When public or judicial pressure for reform does arise, it is short-lived and usually ineffective because it concentrates on physical facilities and administrative procedures. Rarely does an investigation evaluate the rationale and application of behavior management techniques. Through misunderstanding and incompetence, some institutional managers and members of the helping professions twist and alter the concept of reinforcement beyond recognition, attempting to transform even positive reinforcement into an instrument of coercion.

The Misuse of Deprivation. Those whom we have placed in positions of control over ourselves and others -- teachers, military officers, prison guards, police, government officials -- are so accustomed to coercion that they often can comprehend no other way. If they do try positive reinforcement, their first impulse is to take something away from their controllees so they can then give it back in return for "good behavior." That is exactly what happened in some infamous prison projects that claimed to be using positive reinforcement. They imposed solitary confmement on inmates and then let them out for short periods if they showed the proper contrition; deprived them of food and then handed them morsels if they acted subserviently; denied them privacy and then gave them a few moments by themselves if they had not been seen engaging in suspicious social interchanges with other prisoners; gave them menial jobs and switched them to more desirable work if they performed uncomplainingly and without resistance. And then, with any lapse, real or perceived, they reimposed the deprivations.

Such techniques are, of course, completely coercive. They are based on socially imposed deprivation, and on the escape and avoidance that such deprivation generates. Punishment by shocks or by deprivation makes escape reinforcing. If we deprive prisoners, students, children, or others of their basic needs, rights, and privileges in order to create reinforcers, those reinforcers are negative, not positive. They may serve temporarily to keep orderliness in cell blocks, barracks. and classrooms, but they will also generate the long-term side effects of coercive control.

Deprivation, however, does contribute to the effectiveness of positive reinforcers: We have little interest in food right after a good meal, but food influences our actions powerfully as mealtime approaches; the sexual appetite of sailors after a long sea voyage is legendary; although individuals vary widely, what we do to get money and what we do with money after we get it depend strongly on how much we already have. Nevertheless, even though deprivation makes positive reinforcers stronger, it is still not necessary to impose deprivations deliberately in order to make use of positive reinforcers. No one has enough of everything; it does not usually take much extra effort to discover reinforcers that are already effective without additional deprivation.

My concern here is with the use of deprivation as an instrument of coercion. In certain extreme cases, deprivation for a brief time can produce desirable consequences that are unavailable any other way. After everyone else has given up, you can still set a retarded child on the road to effective learning: First, make her hungry; then, use food as a reinforcer for some basic behavior like self-feeding and following simple instructions. Once the child has learned those, you can develop other reinforcers and discontinue food deprivation. In cases of extreme retardation, or when previous incompetent treatment has made a child unresponsive to standard methods of instruction, both the child and the community will find the temporary hardship beneficial.

Even then, one uses deprivation only to enhance the attractiveness of a positive reinforcer, not to punish unsatisfactory behavior. Once the child learns some adaptive behavior, one quickly discontinues the deprivation, with no threat to impose it again. Taking away food, possessions, privileges, or rights just so that these can be given back in return for good behavior, and then taken away again to punish bad behavior, subverts the principle of positive reinforcement. Anyone who uses deprivation this way can expect the controllees to escape, fight back, and exert countercontrol, just as they would react to any coercive regimen.

It is far more effective simply to take advantage of naturally occurring deprivations. Many exist even without social intervention; that is the way the world works. Food, sex, and other biologically determined deprivations are built in. Without producing them ourselves or making them any more severe than they would be in the normal course of things, we can often put these deprivations to good use in teaching basic skills to beginners and to those with learning deficiencies.

As mealtimes approach, for example, food becomes a stronger and stronger positive reinforcer. Retarded people and some of the mentally ill seem sensitive only to a small number of reinforcers, but food is one of the most reliable. The use of food as a reinforcer at mealtimes is a proven and powerful way to teach basic skills to the mentally retarded. It is just as useful in teaching normal children.

Such teaching does not require us to deprive our pupils of meals if they fail to learn. Teaching methods are now available that guarantee learning, so meals need not be missed because of unsuccessful teaching. Even if we have not yet worked out a completely effective instructional program, pupils who have trouble learning do not have to go hungry. While we are perfecting our instructional plan, we can always let them earn a full meal by reviewing what they had learned before.

Eventually, the conduct learned at mealtimes enables retarded pupils to function adaptively at other times, too. Their newfound abilities -- carrying a tray from serving counter to table, using a fork and spoon, picking up spilled food, saying "please" and "thank you," -- make it possible to take them to cafeterias and restaurants. There, new choices become available to them, and they experience new environments. While on route to their treat, they can be taught travel skills. Their world begins to open up.

And then, new reinforcers become effective as they learn how to interact with different environments and with people who are important to them. They learn to recognize signs of approval as precursors of other reinforcers, so people's reactions take on significance, becoming reinforcers in their own right. When that happens, positive reinforcers like food need not always be forthcoming immediately; delay of gratification becomes possible. Food, one of the few effective reinforcers at first, gets these seemingly behaviorless inmates of the local institution for the retarded started. Before long, we fmd ourselves able to abandon food and use newly learned reinforcers to teach them more advanced behavior.

Time-out and Its Abuses. A controversial form of punishment in institutions for the retarded and mentally ill is the "time-out" procedure. What is time-out? What does it accomplish? Does it differ in any important way from other kinds of punishment?

The basic feature of a time-out is the withdrawal of positive reinforcement. This usually means removing someone physically from an environment that makes positive reinforcers available to another place where no reinforcment is possible. In practice, time-out may range from standing an obstreperous child in the comer to putting a violent patient into solitary confmement -- the classical padded cell. The withdrawal of positive reinforcement is just as coercive as the application of a shock, but because time-out inflicts no pain, it is often justified as a benign kind of punishment.

This reasoning is similar to justifying the use of drugs instead of straitjackets, ropes, or chains to immobilize an uncooperative patient. The cruelty lies less in the method than in the outcome. Isolation, physical restraint, and chemical restraint remove the victims from contact with all of the reinforcers that make life meaningful and worthwhile; drugs tum them into zombies, and padded cells tum them into raving maniacs. Both kinds of punishment put an end to all learning except for various forms of escape and avoidance that serve as mechanisms of countercontrol. When the power of the authorities is too great for reprisal or deception, depression takes over.

It is often forgotten that even a relatively mild time-out will be an effective punisher only if the punishee is removed from a positively reinforcing environment. That is what the name, "time-out," refers to; it means time away from reinforcement. Removing a disruptive child into a seeming time-out is not going to prevent future disturbances unless the original situation was reinforcing in the first place. If it was not, taking the child out of it may actually reinforce the disruptive behavior.

And then, our interaction while removing the child may provide stronger positive reinforcement than anything the child was getting in the original situation. When that happens, time-out itself becomes a positive reinforcer, making future disruptive behavior even more likely. We will strengthen the very conduct we intended to punish.

A child whom we have to place repeatedly in time-out is sending us a message: "I do not like it here; I would prefer you to carry me, kicking and screaming, into the bare room next door where you will have to sit with me and hold me in order to keep me from banging my head against the wall." Our response to that message has to be an examination of our own conduct.

If we have been trying to teach, we will probably find that we have not been successful. Because our pupil has not been learning, we have been unable to reinforce, and the pupil has found other ways to "succeed." The remedy is not to place the child in time-out, taking away further opportunities to learn, but to revise our teaching. Go back to the last thing the child learned successfully, so that positive reinforcement again becomes possible, and start all over. Proceed more slowly this time, and take advantage of newly available methods for reducing and even eliminating errors from the learning process.*

More often than not, even children medically diagnosed as hyperactive will participate constructively in classes for long periods of time, causing no disturbance or distraction as long as they are being reinforced for successful learning. Effective teaching will make it unnecessary to punish a child for misbehavior.

*A large technical literature shows that errors are not a necessary part of the learning process, but behavior analysts have not yet presented that material in easily available form for nonprofessionals. Behavior shaping -- teaching new behavior by reinforcing gradually closer approximations to what is desired -- can transform trial-and-error to trial-and-success in teaching motor skills like the production of tones on musical instruments or the pronunciation of words. Teaching long sequences of actions like shoe tying, spelling, or "top-down" computer programming can proceed errorlessly if the teacher starts from the end of the sequence and works backwards. With skillful environmental shaping -- teaching new relations between behavior and environment by changing the environment gradually from familiar to unfamiliar forms -- children can learn errorlessly to copy, write, and name letters of the alphabet; medical students can learn the basic structure of the nervous system so errorlessly that they find it difficult at first to believe they are actually learning anything. Procedures that establish equivalence relations among spoken words, written words, and pictures give children simple reading and speaking vocabularies that they were never explicitly taught and that they use even the very first time without error. Errorless teaching is an active field of research, with new methods and applications coming along rapidly.

Return to Index by Author
Return to Project NoSpank Table of Contents at www.nospank.net