Caring Kids: The Role of the Schools (*)

March 1991

Caring Kids

The Role of the Schools

By Alfie Kohn

“Education worthy of the name is essentially education of character,” the philosopher Martin Buber told a gathering of teachers in 1939.(1) In saying this, he presented a challenge more radical and unsettling than his audience may have realized. He did not mean that schools should develop a unit on values or moral reasoning and glue it onto the existing curriculum. He did not mean that problem children should be taught how to behave. He meant that the very profession of teaching calls on us to try to produce not merely good learners but good people.

Given that even the more modest task of producing good learners seems impossibly ambitious – perhaps because of a misplaced emphasis on producing good test-takers – the prospect of taking Buber seriously may seem positively utopian. But in the half-century since his speech, the need for schools to play an active role in shaping character has only grown more pressing. That need is reflected not only in the much-cited prevalence of teenage pregnancy and drug use but also in the evidence of rampant selfishness and competitiveness among young people. At a tender age, children learn not to be tender. A dozen years of schooling often do nothing to promote generosity or a commitment to the welfare of others. To the contrary, students are graduated who think that being smart means looking out for number one.

I want to argue, first, that something can be done to rectify this situation because nothing about “human nature” makes selfishness inevitable; second, that educators in particular should do something about the problem; and third, that psychological research, common sense, and the experience of an important pilot project in California offer specific guidance for helping children to grow into caring adults.


Much of what takes place in a classroom, including that which we have come to take for granted, emerges from a set of assumptions about the nature of human nature. Not only how children are disciplined, but the very fact that influencing their actions is viewed as “discipline” in the first place; not merely how we grade students, but the fact that we grade them at all; not simply how teachers and students interact, but the fact that interaction between students is rarely seen as integral to the process of learning – all of these facts ultimately rest on an implicit theory of what human beings are like.

Consider the fact that most conversations about changing the way children act in a classroom tend to focus on curbing negative behaviors rather than on promoting positive ones. In part, this emphasis simply reflects the urgency of preventing troublesome or even violent conduct. But this way of framing the issue may also tell us something about our view of what comes naturally to children, what they are capable of, and, by extension, what lies at the core of our species. Likewise, it is no coincidence, I think, that the phrase “it’s just human nature to be . . .” is invariably followed by such adjectives as selfish, competitive, lazy, aggressive, and so on, Very rarely do we hear someone protest, “Well, of course he was helpful. After all, it’s just human nature to be generous.”

The belief persists in this culture that our darker side is more pervasive, more persistent, and somehow more real than our capacity for what psychologists call “prosocial behavior.” We seem to assume that people are naturally and primarily selfish and will act otherwise only if they are coerced into doing so and carefully monitored. The logical conclusion of this world view is the assumption that generous and responsible behavior must be forced down the throats of children who would otherwise be inclined to care only about themselves.

A review of several hundred studies has convinced me that this cynicism is not realism. Human beings are not only selfish and self-centered, but also decent, able to feel – and prepared to try to relieve – the pain of others. I believe that it is as “natural” to help as it is to hurt, that concern for the well-being of others often cannot be reduced to self-interest, that social structures predicated on human selfishness have no claim to inevitability – or even prudence. This is not the place for rehearsing the arguments and data that support these conclusions – in part because I have recently done so at book length.(2) But I would like to mention a few recent findings from developmental psychology that speak to the question of whether educators can aim higher than producing a quiet classroom or a nondisruptive child.

To start at the beginning, newborns are more likely to cry – and to cry longer – when they are exposed to the sound of another infant’s crying than when they hear other noises that are equally loud and sudden. In three sets of studies with infants ranging in age from 18 to 72 hours, such crying seemed to be a spontaneous reaction rather than a mere vocal imitation.(3) In the view of Abraham Sagi and Martin Hoffman, who conducted one of the studies, this finding suggests the existence of “a rudimentary empathic distress reaction at birth.”(4) Our species may be primed, in other words, to be discomfited by someone else’s discomfort.

As an infant grows, this discomfort continues and takes more sophisticated forms. Marian Radke-Yarrow, Carolyn Zahn-Waxler, and their associates at the National Institute of Mental Health have been studying toddlers for nearly 20 years, having in effect deputized mothers as research assistants to collect data in the home instead of relying on brief (and possibly unrepresentative) observations in the laboratory. A 10- to 14-month-old child, they have found, can be expected to show signs of agitation and unhappiness in the presence of another person’s distress, perhaps by crying or burying her head in her mother’s lap. As a child develops the capacity to undertake more deliberate behavior, in the period between 18 and 24 months, his response to distress will become more active: patting the head, fetching a toy, offering verbal expressions of sympathy, finding an adult to help, and so forth.(5)

I should add that, like all parents, researchers have also observed hostile and selfish actions on the part of children. To say that sympathy or helping behavior is pervasive and precocious is not to claim that every child is an angel or to deny that toddlers – particularly in a society preoccupied with possessiveness – will sometimes snatch back a toy (“Mine!”) or throw it across the room. But it is to argue that the antisocial is no more basic or natural than the prosocial.

By the time children are of preschool age, comforting, sharing. and helping are regular occurrences. One study of preschoolers during free play discovered that 67 of the 77 children shared with, helped, or comforted another child at least once during only 40 minutes of observation. After counting such behaviors in similar experiments of her own, Arizona State University psychologist Nancy Eisenberg became curious about why children were acting this way. To find out, she came up with a technique that few research psychologists had thought to use: she asked the children.

Eisenberg and a colleague simply followed 4- and 5-year-olds around a preschool and watched for unprompted prosocial behavior. Each time such an act was observed, the child was asked why he or she did it. (“How come you gave that to John?”) None of the children intended to conform to adult expectations or expressed any fear of punishment. Very few said that they expected to benefit in some way by helping – such as by impressing their peers. Among the most frequent explanations heard was the simple observation that the other child had needed help.(7) This, when you come right down to it, is the heart of altruism.(8) And it is enough to suggest that parents and educators hoping to raise a child who is responsive to the needs of others already have “an ally within the child,” in Martin Hoffman’s lovely phrase.


If we had to pick a logical setting in which to guide children toward caring about, empathizing with, and helping other people, it would be a place where they would regularly come into contact with their peers and where some sort of learning is already taking place. The school is such an obvious choice that one wonders how it could be that the active encouragement of prosocial values and behavior – apart from occasional exhortations to be polite – plays no part in the vast majority of American classrooms. This would seem to stem either from a lack of interest in the idea or from some objection to using the schools in particular for this purpose. Both factors probably play a role, but I will concentrate here on the latter and consider three specific reservations that parents, teachers, policy makers, and others may have – or at least may hear and thus need to answer – about classroom-based programs to help children develop a prosocial orientation.

The first objection is that an agenda concerned with social and moral issues amounts to teaching values – a dangerous business for a public institution. In response, we must concede that a prosocial agenda is indeed value-laden, but we should immediately add that the very same is true of the status quo. The teacher’s presence and behavior, her choice of text, the order in which she presents ideas, and her tone of voice are as much part of the lesson as the curriculum itself. So, too, is a teacher’s method of discipline or classroom management saturated in values, regardless of whether those values are transparent to the teacher. In short, to arrange our schools so that caring, sharing, helping, and empathizing are actively encouraged is not to introduce values into a neutral environment; it is to examine the values already in place and to consider trading them in for a new set.

It is sometimes said that moral concerns and social skills ought to be taught at home. I know of no one in the field of education or child development who disagrees. The problem is that such instruction – along with nurturance and warmth, someone to model altruism, opportunities to practice caring for others, and so forth – is not to be found in all homes. The school may need to provide what some children will not otherwise get. In any case, there is no conceivable danger in providing these values in both environments. Encouragement from more than one source to develop empathic relationships is a highly desirable form of redundancy.

The second concern one hears – and this one dovetails with the broader absence of interest in the prosocial realm – is the fear that children taught to care about others will be unable to look out for themselves when they are released into a heartless society. The idea that someone exposed to such a program will grow up gullible and spineless, destined to be victimized by mean-spirited individuals, can be traced back to the prejudice that selfishness and competitiveness are efficacious social strategies – a sterling example of what sociologist C. Wright Mills used to call “crackpot realism.” In fact, those whose mantra is “look out for number one” are actually at a greater disadvantage in any sort of society than those who are skilled at working with others and inclined to do so. Competition and the single-minded pursuit of narrowly conceived self-interest typically turn out to be counterproductive.

By contrast, a well-designed program of prosocial instruction will include training in cooperative conflict resolution and in methods of achieving one’s goals that do not require the use of force or manipulation. But even without such a component, there is nothing about caring for others that implies not caring for or looking after oneself. A raft of research has established that assertiveness, healthy self-esteem, and popularity are all compatible with – and often even correlates of – a prosocial orientation.(9)

The final objection to teaching children to be caring individuals is that the time required to do so comes at the expense of attention to academics – a shift in priorities apt to be particularly unpopular at a time when we entertain ourselves by describing how much students don’t know. In fact, though, there is absolutely no evidence to suggest that prosocial children — or the sort of learning experiences that help to create them — are mutually exclusive with academic achievement. To the contrary, the development of perspective-taking – the capacity to imagine how someone else thinks, feels, or sees the world – tends to promote cognitive problem solving generally. In one study, the extent to which girls had these skills at age 8 or 9 was a powerful predictor of performance on reading and spelling tests taken two years later – an even better predictor, in fact, than their original test scores. (10)

Not only are the ingredients of a prosocial orientation conducive to academic excellence, but the educational process itself does not require us to choose between teaching children to think and teaching them to care. It is possible to integrate prosocial lessons into the regular curriculum; as long as children are learning to read and spell and think critically, they may as well learn with texts that encourage perspective-taking. Indeed, to study literature or history by grappling with social and moral dilemmas is to invite a deeper engagement with those subjects. Meanwhile, literally hundreds of studies have shown that cooperative learning, which has an important place in a prosocial classroom, enhances achievement regardless of subject matter or age level.(11) So consistent and remarkable have these results been that schools and individual teachers often adopt models of cooperative learning primarily to strengthen academic performance. The development of prosocial values is realized as an unintended bonus.


Education of character in Buber’s sense asks of teachers something more than the mere elimination of behavior problems in the classroom. The absence of such problems is often seen as an invitation to move past behavioral and social issues and get on with the business at hand, which is academic learning. I am arguing, by contrast, that behavioral and social issues, values and character, are very much part of the business at hand. But whether we are talking about addressing misconduct or about taking the initiative to help students become more responsive to one another, a teacher can take any of several basic orientations. Here are four approaches to changing behaviors and attitudes, presented in ascending order of desirability.

1. Punishing. A reliance on the threat of punishment is a reasonably good indication that something is wrong in a classroom, since children have to be bullied into acting the way the teacher demands. Apart from the disagreeable nature of this style of interaction – which cannot be disguised, incidentally, by referring to punishment as “consequences” – it is an approach distinguished mostly by its ineffectiveness. Decades of research have established that children subjected to punitive discipline at home are more likely than their peers to break rules when they are away from home.

Isolating a child from his peers, humiliating her, giving him an F, loading her with extra homework, or even threatening to do any of these things can produce compliance in the short run. Over the long haul, however, this strategy is unproductive.

Why? First, at best, punishment teaches nothing about what one is supposed to do – only about what one is not supposed to do. There is an enormous difference between not beating up one’s peers, on the one hand, and being helpful, on the other.

Second, the child’s attention is not really focused on the intended lesson at all (“pushing people is bad”), much less on the rationale for this principle, but primarily on the punishment itself. Figuring out how to get away with the misbehavior, how to avoid detection by an authority, is a perfectly logical response. (Notice that the one who punishes becomes transformed in the child’s eyes into a rule-enforcer who is best avoided.) Social learning theory tells us that this attention to the punishment is also likely to teach the child to be punitive and thus exacerbate the behavior problems; a teacher’s actions do indeed speak louder than words.

Finally, punishment breeds resistance and resentment. “The more you use power to try to control people, the less real influence you’ll have on their lives,” Thomas Gordon has written.(12) Since such influence is associated with helping children to develop good values, the use of power would seem ill-advised.

2. Bribing. There is no question that rewards are better than punishment. On the other hand, what these two methods share is probably more important than the respects in which they differ, and herein lies a tale that will be highly disconcerting to educators enamored of positive reinforcement.

Psychological – and particularly developmental – theory and research have come a long way since the simplistic behaviorism of the last generation, but many well-meaning teachers continue to assume that what works for training the family pet must be appropriate for shaping children’s actions and values as well.

Gold stars, smiley faces, trophies, certificates, high grades, extra recess time, candy, money, and even praise all share the feature of being “extrinsic” to whatever behavior is being rewarded. Like sticks, carrots are artificial attempts to manipulate behavior that offer children no reason to continue acting in the desired way when there is no longer any goody to be gained. Do rewards motivate students? Absolutely. They motivate students to get rewarded. What they fail to do is help children develop a commitment to being generous or respectful.

In fact, the news is even worse than this. Not only is bribing someone to act in a particular way ultimately ineffective, but, like the use of threats, it can actually make things worse. Consider the effects of rewards on achievement. Yale psychologist Robert Sternberg recently summed up what a growing number of motivation researchers now concede: “Nothing tends to undermine creativity quite like extrinsic motivators do. They also undermine intrinsic motivation: when you give extrinsic rewards for certain kinds of behavior, you tend to reduce children’s interest in performing those behaviors for their own sake.”(13) Once we see ourselves as doing something in order to get a reward, we are less likely to want to continue doing it in the absence of a reward – even if we used to find it enjoyable.

Readers of the Kappan were first exposed to research demonstrating this phenomenon more than 15 years ago,(14) and the data have continued to accumulate since then, with some studies concentrating on how extrinsic motivators reduce intrinsic interest and others showing how they undermine performance, particularly on creative tasks.(15) A number of explanations have been proposed to account for these remarkably consistent findings. First, people who think of themselves as working for a reward feel controlled, and this lack of self-determination interferes with creativity. Second, rewards encourage “ego involvement” to the exclusion of “task involvement,” and the latter is more predictive of achievement. Third, the promise of a reward is “tantamount to declaring that the activity is not worth doing for its own sake,” as A. S. Neill put it;(16) indeed, anything construed as a prerequisite to some other goal will probably be devalued as a result.

What is true for academic learning also applies to behavior. A little-known series of studies has pointed up the folly of trying to encourage prosocial behavior through the use of extrinsic incentives. Children who received rewards for donating to another child – and, in another experiment, adults who were paid for helping a researcher – turned out to be less likely to describe themselves in words suggesting intrinsic motivation to help than were people who received nothing in return.(17) In another study, women offered money for answering a questionnaire were less likely to agree to a similar request two or three days later, when no money was involved, than were women who had not been paid for helping with the first survey.(18)

The implication is that, when someone is rewarded for prosocial behavior, that person will tend to assume that the reward accounts for his or her actions and thus will be less likely to help once no one is around to hand out privileges or praise. Indeed, elementary school students whose mothers believed in using rewards to motivate them were less cooperative and generous than other children in a recent study.(19) Such findings are of more than theoretical interest given the popularity of Skinnerian techniques for promoting generosity in schools. A recent New York Times article described elementary schools where helpful children have their pictures posted in hallways, get to eat at a special table in the cafeteria, or even receive money.(20) Such contrivances may actually have the effect of undermining the very prosocial orientation that their designers hope to promote.

3. Encouraging commitment to values. To describe the limitations of the use of punishments and rewards is already to suggest a better way: the teacher’s goal should not be simply to produce a given behavior – for example, to get a child to share a cookie or stop yelling – but to help that child see himself or herself as the kind of person who is responsible and caring. From this shift in self-concept will come lasting behaviors and values that are not contingent on the presence of someone to dispense threats or bribes. The child has made these behaviors and values his or her own.

A student manipulated by currently fashionable behavioral techniques, however, is unlikely to internalize the values underlying the desired behaviors. At the heart of Assertive Discipline, for example, is control: “I want teachers to learn that they have to take charge,” Lee Canter explained recently.(21) I don’t. I want children to become responsible for what they do and for what kind of people they are. The teacher has a critical role to play in making sure that this happens; in criticizing manipulative approaches I am not suggesting that children be left alone to teach themselves responsibility. But the teacher ought to be guided less by the need to maintain control over the classroom than by the long-term objective of helping students to act responsibly because they understand that it is right to do so.

I will have more to say below about strategies for facilitating this internalization, but first I want to mention a version of this process that I believe is even more desirable – the ideal approach to helping children become good people.

4. Encouraging the group’s commitment to values. What the first two approaches have in common is that they provide nothing more than extrinsic motivation. What the first two share with the third is that they address only the individual child. I propose that helpfulness and responsibility ought not to be taught in a vacuum but in the context of a community of people who learn and play and make decisions together. More precisely, the idea is not just to internalize good values in a community but to internalize, among other things, the value of community.

Perhaps the best way to crystallize what distinguishes each of these four approaches is to imagine the question that a child is encouraged to ask by each. An education based on punishment prompts the query, “What am I supposed to do, and what will happen to me if I don’t do it?” An education based on rewards leads the child to ask, “What am I supposed to do, and what will I get for doing it?” When values have been internalized by the child, the question becomes “What kind of person do I want to be?” And, in the last instance, the child wonders: “How do we want our classroom (or school) to be?”


Educators eager to have children think about how they want their classrooms to be – which is to say, educators who do not feel threatened at the prospect of inviting children to share some of the responsibility for creating norms and determining goals – need to think in terms of five broad categories: what they believe, what they say, what they do, how they relate to students, and how they encourage students to relate to one another. Let us consider each in turn.

What educators believe. The famous Pygmalion effect refers to the fact that a teacher’s assumptions about a child’s intellectual potential can affect that child’s performance. Such self-fulfilling prophecies, however, are by no means limited to academics; they also operate powerfully on a child’s actions and values. Write off a student as destructive or disruptive, and he or she is likely to “live down to” these expectations. Conversely – and here is the decisive point for anyone concerned with promoting generosity – attributing to a child the best possible motive that is consistent with the facts may set in motion an “auspicious” (rather than a vicious) circle. We help students develop good values by assuming whenever possible that they are already motivated by these values – rather than by explaining an ambiguous action in terms of a sinister desire to make trouble.

However, what we assume about a given student is also colored by our assumptions regarding human nature itself. While I am not aware of any research on this question, it seems reasonable to suppose that an educator who thinks that self-interest motivates everything we do will be suspicious of individual instances of generosity. Someone who takes for granted that a Hobbesian state of nature would exist in a classroom in the absence of a controlling adult to keep children in line, who believes that children need to be leaned on or “taught a lesson” or bribed to act responsibly, is likely to transfer these expectations to the individual child and to produce an environment that fulfills them. The belief that children are actually quite anxious to please adults, that they may simply lack the skills to get what they need, that they will generally respond to a caring environment can create a very different reality. What you believe matters.

What educators say. An immense body of research has shown that children are more likely to follow a rule if its rationale has been explained to them and that, in general, discipline based on reason is more effective than the totalitarian approach captured by the T-shirt slogan “Because I’m the mommy, that’s why.” This finding applies not only to discouraging aggression but to promoting altruism. From preschool to high school, children should learn why – not merely be told that – helping others is good. Pointing out how their actions affect others sensitizes students to the needs and feelings of others and tacitly communicates a message of trust and responsibility. It implies that, once children understand how their behavior makes other people feel, they can and will choose to do something about it.

How such explanations are framed also counts. First, the level of the discourse should be fitted to the child’s ability to understand. Second, the concept of using reason does not preclude passion. A prohibition on hurting people, for example, should not be offered dispassionately but with an emotional charge to show that it matters. Third, prosocial activity should not be promoted on the basis of self-interest. “Zachary, if you don’t share your dump truck with Linda, she won’t let you play with her dinosaur” has an undeniable appeal for a parent, but it is a strategy more likely to inculcate self-regarding shrewdness than genuine concern for others. The same goes for classroom exhortations and instruction.

A series of studies by Joan Grusec of the University of Toronto and her colleagues is also relevant. Her research provides a concrete alternative to the use of rewards or praise to elicit generosity. “Children who view their prosocial conduct as compliance with external authority will act prosocially only when they believe external pressures are present,” she has written. Far preferable is for children to “come to believe that their prosocial behavior reflects values or dispositions in themselves.”(22)

This result is best achieved by verbally attributing such values or dispositions to the child. In one experiment, in which children gave away some of their game winnings after watching a model do so, those who were told that they had made the donation “because you’re the kind of person who likes to help other people” were subsequently more generous than those who were told that they had donated because they were expected to do so.(23) In another study, the likelihood of children’s donating increased both when they were praised and when they were led to think of themselves as helpful people. But in a follow-up experiment, it was the latter group who turned out to be more generous than those who had received verbal reinforcement. In other words, praise increased generosity in a given setting but ceased to be effective outside of that setting, whereas children with an intrinsic impulse to be generous continued to act on that motivation in other circumstances.(24)

A study of adults drives home the point. Subjects who were told that a personality test showed that they were kind and thoughtful people were more likely to help a confederate who “accidentally” dropped a pile of cards than were those who were told that they were unusually intelligent or those given no feedback at all. This finding is important because it implies that being led to think of oneself as generous does not affect behavior merely because it is a kind of reinforcement or a mood-enhancer; this label apparently encourages prosocial action because it helps to build a view of the self as altruistic.(25)

This is not to suggest that a teacher’s every utterance must be – or can be – geared toward internalization. Simply making sure that a classroom is a safe environment conducive to learning can require the sort of behavioral interventions on a day-to-day basis that don’t do much to strengthen a child’s prosocial self-concept. But the more teachers attend to the latter, the fewer problems they are likely to have over the long run.

What educators do. Children of all ages, from before the time they can read until after the time they start seeking distance from adults, learn from what they see. Studies show that children who watched, even briefly, as someone donated to charity were themselves likely to donate more than other children – even if months had elapsed since the exposure to the model.(26) The extent to which a teacher expresses concern about people in distress and takes the initiative to help – which applies both to how the teacher treats the students themselves and how he or she refers to people outside the classroom – can set a powerful example and be even more effective than didactic instruction in promoting a sense of caring in students.

There is no shortage of suggestions about how to devise lessons that address social and ethical issues, ranging from explicit training in perspective-taking or moral reasoning to discussions about values that can, in turn, include either “clarification” of the beliefs that students already hold or old-fashioned lectures on character or morality. Most of the debate on the subject occurs between proponents of just such programs, each accusing the other of being relativistic or of seeking to indoctrinate. Far less consideration is given to the possibility of integrating such issues into the regular curriculum.

A distinction, though not a sharp one, can be made between teaching morality (or about morality) as such and helping children to be positively connected to others. The latter is my focus here, and some writers have argued that, particularly for younger children, it ought to be the primary focus in the schools, too. “Unless the young child has acquired a positive propensity towards other persons,” says one educator, “subsequent moral education will become virtually impotent.”(27)

As an alternative to special units devoted to one of these approaches, children can use texts in conventional subject areas that encourage perspective-taking. This option should allay the concern that moral instruction will distract us from academics.

How educators relate to students. Preceding and underlying specific techniques for encouraging particular behaviors is the practice of nesting all kinds of discipline and instruction in the context of a warm, nurturant, and empathic relationship with students. Children whose parents are interested in and supportive of them usually distinguish themselves as socially competent and psychologically healthy on a range of measures, and there is no reason to think that the teacher/student relationship is any different.

Warm, caring, empathic adults do several things at once. They provide the child with a benevolent, safe place in which to act. (If a child’s experience with others leaves him or her feeling threatened rather than safe, this is likely to foster psychological damage control at the expense of any inclination to help others.) I hope that few educators take seriously the absurd dictum that teachers should display no warmth until well into the school year – after firm control of the classroom has been won. Instead, teachers should establish themselves from the beginning as the students’ allies, adults with whom they can work to solve the problems that emerge during the normal course of development. In meeting a child’s emotional needs we give him or her the emotional freedom to meet the needs of others.

How educators encourage students to relate to one another. Anyone interested in children as social beings must recognize the need to attend to the interactions among them in the classroom. In most American schools, children are forced to work either against one another (by competing) or apart from one another (by learning individually). The chance to work with one another, to learn social skills and caring, is left to happen by itself during recess, at lunch, or after school. This single fact goes a long way toward explaining why people in our society tend to regard others as potential obstacles to their own success. David Johnson and Roger Johnson of the University of Minnesota have emphasized that the relationship between student and student is at least as important as that between student and teacher or between student and curriculum. Their reference to student/student relationships is offered in the context of creating good learners, and it is all the more true in terms of creating good human beings.

How, specifically, should teachers encourage student interaction? First, students can be put in pairs or small groups so as to help one another learn. This concept, known as cooperative learning, embraces many disparate models for implementation: some depend on grades and other extrinsic incentives to insure that students work together, some involve cooperation among as well as within groups, some provide for a strict division of labor in completing assignments. A substantial number of studies have found that cooperative learning of various types has the potential to help students feel good about themselves, feel good about each other, feel good about what they are learning, and learn more effectively.

Cooperation, by virtue of being an interaction in which two or more people work together for mutual benefit, is not itself an example of prosocial behavior as the term is usually used. Neither does its successful use presuppose the existence of prosocial motives in all children. Rather, by creating interdependence and a built-in incentive to help, cooperative learning promotes prosocial behavior. Having children learn from one another creates powerful bonds between them and sends a message very different from that sent by a classroom in which each child is on his or her own – or, worse still, one in which the success of each is inversely related to the success of the others.(28)

In one study, fifth-graders who studied grammar in cooperative learning groups were more likely to give away prize tokens to a stranger than were those who studied on their own; in another, kindergartners who participated in cooperative activities acted more prosocially than their peers in a traditional classroom.(29) But the consequences are not limited to generosity per se. Carefully structured cooperative learning also promotes a subjective sense of group identity, a greater acceptance of people who are different from oneself (in terms of ethnicity or ability level), and a more sophisticated ability to imagine other people’s points of view.(30) Cooperation is an essentially humanizing experience that predisposes participants to take a benevolent view of others. It allows them to transcend egocentric and objectifying postures and encourages trust, sensitivity, open communication, and prosocial activity.

Second, teachers can move the idea of discipline not only away from punishments and rewards but also away from the premise of these strategies – namely, that teachers should simply be figuring out by themselves how to elicit compliance with a set of rules or goals that they alone devise. The realistic alternative is not for the teacher to abdicate responsibility for what happens in the classroom but rather to bring in (and guide) children so that they can play a role in making decisions about how their classroom is to be run and why. (Must hands always be raised or only during certain kinds of discussions? What is the best way for the class as a community to balance principles of fairness and the spontaneity that encourages participation?)

Discipline would thus be reconfigured as collaborative planning and mutual problem solving. Such an approach will be preferred by anyone who favors the idea of autonomy and democratic decision making – but it can also be argued that purely practical considerations recommend it, since children are more likely to follow rules that they have helped to create than rules dictated to them. This, of course, assumes that following rules is in itself a desirable goal. More broadly, educators need to ask themselves and each other about the ultimate objective of discipline. Even if one of the conventional programs of behavior control succeeded in keeping children quiet, do quiet children learn more effectively or merely make fewer demands on the teacher? (The Johnsons like to say that a principal walking through the school corridors should be concerned if he or she hears no sound coming from a classroom; this means that real learning probably is not taking place.) And which approach is most likely to help children come to care about one another?

To invite children to participate in making decisions not only about classroom procedures but also about pedagogical matters (what is to be learned, how, and why) and housekeeping matters (how to celebrate birthdays or decorate the walls) is to bring them into a process of discussion, an opportunity to cooperate and build consensus. To this extent, it is a chance for them to practice perspective-taking skills, to share and listen and help. In short, involving children in planning and decision making is a way of providing a framework for prosocial interactions that supports other such opportunities; it turns a routine issue into another chance to learn about and practice caring – and, not so incidentally, thinking as well.(31)

Finally, educators can provide students with opportunities to be responsible for one another so that they will learn (prosocial values and skills) by doing. Ideally, this can include interaction with those of different ages. For an older child to guide someone younger is to experience firsthand what it is to be a helper and to be responsible for someone who is dependent on him or her. For the younger child, this cross-age interaction presents an opportunity to see a prosocial model who is not an adult.


One of the most exciting and innovative educational programs now in operation, the Child Development Project (CDP), is devoted specifically to helping children become more caring and responsible.(32) The experience of the CDP offers lessons in the systematic application of the ideas discussed here; indeed, I owe my formulation of some of these ideas to the work done by Eric Schaps, Marilyn Watson, and others involved with the project.

The CDP is the first long-term, comprehensive, school-based project in prosocial education. After being invited a decade ago to work in the San Ramon Valley (California) Unified School District, about 30 miles east of San Francisco, the staff carefully matched two sets of three elementary schools in the district for size and socioeconomic status. A coin flip then determined which of these sets would receive the program and which would serve as the comparison group. The first teachers were trained before the start of the 1982-83 school year. Staff researchers focused on a group of children in the experimental schools (then in kindergarten and now in junior high school) to assess whether their attitudes, behavior, and achievement differed significantly from those of their counterparts in the comparison schools. In the fall of 1988, the program was introduced into two elementary schools in nearby Hayward, a district more ethnically diverse than the white, affluent suburbs in San Ramon Valley, and Schaps is now seeking funding to take the program to eight more sites around the country.

“How do we want our classroom to be?” is exactly the question that the CDP would have children ask. Rejecting punishment and rewards in favor of strategies geared toward internalization of prosocial norms and values, the CDP invites teachers and students to work together to turn their classrooms into caring communities. The primary components of the program intended to bring this about are these:

* a version of cooperative learning that does not rely on grades or other extrinsic motivators;

* the use of a literature-based reading program that stimulates discussion about values and offers examples of empathy and caring even as it develops verbal skills;

* an approach to classroom management in which the emphasis is on developing intrinsic motives to participate productively and prosocially, in which teachers are encouraged to develop warm relationships with the children, and in which periodic class meetings are held so that children can play an active role in planning, assessing progress, and solving problems; and

* a variety of other features, including pairing children of different ages to work together, setting up community service projects to develop responsibility, giving periodic homework assignments specifically designed to be done (and to foster communication) with parents, and holding schoolwide activities that may involve whole families.

In their writings, members of the CDP staff have distinguished their way of teaching values from the approaches of better-known models. Unlike certain kinds of character education, the CDP approach emphasizes helping students understand the reason for a given value rather than simply insisting that they accept it or behave in a certain way because they have been told to do so. Unlike purely child-centered approaches, however, the CDP is committed to the importance of adult socialization: the teacher’s job is to teach, to guide, to enforce, to facilitate cooperation, to model behaviors — in short, to be much more than a passive bystander. Prosocial values come from a synthesis of adult inculcation and peer interaction, and these values – in contrast to the programs developed by some theorists in the area of moral reasoning – emphasize caring for others as well as applying principles of fairness.

Prior to the implementation of the CDP, students randomly selected from the three experimental and the three comparison schools proved to be similar not only demographically but also on a range of social attitudes, values, and skills. Once the program was implemented, however, structured interviews and observations turned up significant differences between students participating in the program and those in the comparison schools on some, though not all, measures.

Children taking part in the CDP engaged in a greater number of spontaneous prosocial behaviors in class, seemed better able to understand hypothetical conflict situations, and were more likely to take everyone’s needs into account in dealing with such situations. They were more likely to believe that one has an obligation to speak up in a discussion even if one’s position seems unlikely to prevail (which should answer those concerned about the assertiveness of caring children). While the CDP`s emphasis has not required any sacrifice of conventional achievement (as measured by standardized test scores), neither has it given participants a consistent academic advantage over students in comparison schools. (In part, this finding may be due to a ceiling effect: students in the district already score in the top 10% of California schoolchildren, so there is not much room for improvement.) By the time the CDP group reached sixth grade, though, they were outscoring their counterparts in the comparison schools on a measure of higher-order reading comprehension (essays written about stories and poems).

It remains to be seen whether and in what ways the values and behaviors of children from schools using the CDP will continue to distinguish them from those who attended comparison schools now that they are all in junior high school. But this pilot project provides real evidence for the larger point I am making here: it is both realistic and valuable to attend to what students learn in the classroom about getting along with their peers. Children can indeed be raised to work with, care for, and help one another. And schools must begin to play an integral role in that process.


1. Martin Buber, Between Man and Man, trans. Ronald Gregor Smith (New York: Macmillan, 1965), p. 104.

2. Alfie Kohn, The Brighter Side of Human Nature: Altruism and Empathy in Everyday Life (New York: Basic Books, 1990).

3. Marvin L. Simner, “Newborn’s Response to the Cry of Another Infant,” Developmental Psychology, vol5, 1971, pp. 136-50; Abraham Sagi and Martin L. Hoffman, “Empathic Distress in the Newborn,” Developmental Psychology, vol12, 1976, pp. 175-76; and Grace B. Martin and Russell D. Clark III, “Distress Crying in Neonates: Species and Peer Specificity,” Developmental Psychology, vol18, 1982, pp. 3-9.

4. Sagi and Hoffman, p. 176.

5. See, for example, Carolyn Zahn-Waxler and Marian RadkeYarrow, “The Development of Altruism: Alternative Research Strategies,” in Nancy Eisenberg-Berg, ed., The Development of Prosocial Behavior (New York: Academic Press, 1982).

6. Marian Radke-Yarrow and Carolyn Zahn-Waxler, “Dimensions and Correlates of Prosocial Behavior in Young Children,” Child Development, vol. 47, 1976, pp. 118-25.

7. Nancy Eisenberg-Berg and Cynthia Neal, “Children’s Moral Reasoning About Their Own Spontaneous Prosocial Behavior,” Developmental Psychology, vol15, 1979, pp. 228-29. Eisenberg and another colleague have observed that appeals to authority or punishment (which were completely absent here) are what one would expect if the children were at Lawrence Kohlberg’s first stage of moral reasoning and that the apparently altruistic needs-oriented explanations have often – and presumably unfairly – been coded as stage 2, that is, as an immature, “preconventional” way of thinking about moral problems (see Nancy Eisenberg-Berg and Michael Hand, “The Relationship of Preschoolers’ Reasoning About Prosocial Moral Conflicts to Prosocial Behavior,” Child Development, vol50, 1979, pp. 356-63).

8. The tendency to define altruism so narrowly that only Mother Teresa would qualify for the label both reflects and perpetuates a cynical view of human nature. It would never occur to us to define aggression so as to exclude everything short of mass murder.

9. Kohn, ch. 3.

10. Norma Deitch Feshbach and Seymour Feshbach, “Affective Processes and Academic Achievement,” Child Development, vol58, 1987, pp. 1335-47. For more research on cognitive skills and perspective-taking, see David W. Johnson and Frank P. Johnson, Joining Together: Group Theory and Group Skills, 3rd ed. (Englewood Cliffs, N.J.: Prentice-Hall, 1987), p. 244.

11. For example, see David Johnson et al., “Effects of Cooperative, Competitive, and Individualistic Goal Structures on Achievement: A Meta-Analysis,” Psychological Bulletin, vol89, 1981, pp. 47-62; David W. Johnson and Roger T. Johnson, Cooperation and Competition (Edina, Minn.: Interaction Book Co., 1989), especially ch. 3; and Robert E. Slavin, Cooperative Learning: Theory, Research, and Practice (Englewood Cliffs, N.J.: Prentice-Hall, 1990), especially ch. 2.

12. Thomas Gordon, Teaching Children Self-Discipline (New York: Times Books, 1989), p. 7.

13. Robert J. Sternberg, “Prototypes of Competence and Incompetence,” in Robert J. Sternberg and John Kolligian, Jr., eds., Competence Considered (New Haven: Yale University Press, 1990), p. 144.

14. Mark R. Lepper and David Greene, “When Two Rewards Are Worse Than One: Effects of Extrinsic Rewards on Intrinsic Motivation,” Phi Delta Kappan, April 1975, pp. 565-66.

15. See, for example, Edward Deci and Richard Ryan, Intrinsic Motivation and Self-Determination in Human Behavior (New York: Plenum Press, 1985); Mark R. Lepper and David Greene, eds., The Hidden Costs of Reward (Hillsdale, N.J.: Erlbaum, 1978); and the work of John Nicholls, Teresa Amabile, Judith M. Harackiewicz, Mark Morgan, and Ruth Butler. I have reviewed some of this research in “Group Grade Grubbing Versus Cooperative Learning,” Educational Leadership, February 1991, pp. 83-87.

16. Quoted in Mark Morgan, “Reward-Induced Decrements and Increments in Intrinsic Motivation,” Review of Educational Research, vol54, 1984, p. 5.

17. Cathleen L. Smith et al., “Children’s Causal Attributions Regarding Help Giving,” Child Development, vol50, 1979, pp. 203-10; and C. Daniel Batson et al., “Buying Kindness: Effect of an Extrinsic Incentive for Helping on Perceived Altruism,” Personality and Social Psychology Bulletin, vol4, 1978, pp. 86-91.

18. Miron Zuckerman, Michelle M. Lazzaro, and Diane Waldgeir, “Undermining Effects of the Foot-in-the-Door Technique with Extrinsic Rewards,” Journal of Applied Social Psychology, vol9, 1979, pp. 292-96.

19. Richard A. Fabes et al., “Effects of Rewards on Children’s Prosocial Motivation,” Developmental Psychology, vol25, 1989, pp. 509-15.

20. Suzanne Daley, “Pendulum Is Swinging Back to the Teaching of Values in U.S. Schools,” New York Times, 12 December 1990, p. B-14.

21. Quoted in David Hill, “Order in the Classroom,” Teacher Magazine, April 1990, p. 77.

22. Joan E. Grusec and Theodore Dix, “The Socialization of Prosocial Behavior: Theory and Reality,” in Carolyn Zahn-Waxler, E. Mark Cummings, and Ronald Iannotti, eds., Altruism and Aggression: Biological and Social Origins (Cambridge: Cambridge University Press, 1986), p. 220.

23. Joan E. Grusec et al., “Modeling, Direct Instruction, and Attributions: Effects on Altruism,” Developmental Psychology, vol14, 1978, pp. 51-57.

24. Joan E. Grusec and Erica Redler, “Attribution, Reinforcement, and Altruism: A Developmental Analysis,” Developmental Psychology, vol. 16, 1980, pp. 525-34.

25. Angelo Strenta and William DeJong, “The Effect of a Prosocial Label on Helping Behavior,” Social Psychology Quarterly, vol44, 1981, pp. 142-47.

26. See James H. Bryan and Nancy H. Walbek, “Preaching and Practicing Generosity,” Child Development, vol41, 1970, pp. 329-53; James H. Bryan and Perry London, “Altruistic Behavior by Children,” Psychological Bulletin, vol72, 1970, pp. 200-211; Martin L. Hoffman, “Altruistic Behavior and the Parent-Child Relationship,” Journal of Personality and Social Psychology, vol31, 1975, pp. 937-43; and Marian Radke-Yarrow, Phyllis M. Scott, and Carolyn Zahn-Waxler, “Learning Concern for Others,” Developmental Psychology, vol8, 1973, pp. 240-60.

27. Ben Spiecker, “Psychopathy: The Incapacity to Have Moral Emotions,” Journal of Moral Education, vol17, 1988, p. 103.

28. For an analysis of the harms of competition in the classroom and elsewhere, see Alfie Kohn, No Contest: The Case Against Competition (Boston: Houghton Mifflin, 1986).

29. David W. Johnson et al., “Effects of Cooperative Versus Individualized Instruction on Student Prosocial Behavior, Attitudes Toward Learning, and Achievement,” Journal of Educational Psychology, vol68, 1976, pp. 446-52; and Bette Chambers, “Cooperative Learning in Kindergarten: Can It Enhance Students’ Perspective-Taking Ability and Prosocial Behaviour?,” unpublished manuscript , Concordia University, Montreal, 1990.

30. See, for example, the research cited in David W. Johnson and Roger T. Johnson, “The Socialization and Achievement Crisis: Are Cooperative Learning Experiences the Solution?,” in Leonard Bickman, ed., Applied Social Psychology Annual 4 (Beverly Hills, Calif.: Sage, 1983), p. 137; and Elliot Aronson and Diane Bridgeman, “Jigsaw Groups and the Desegregated Classroom: In Pursuit of Common Goals,” Personality and Social Psychology Bulletin, vol5, 1979, p. 443.

31. Another classroom management issue is raised by Carolyn Zahn-Waxler. She warns that a teacher who routinely and efficiently takes care of a child’s distress in order to preserve order in the classroom may unwittingly be teaching two lessons: 1) that “people do not react emotionally to upset in others” and 2) that, “if someone is hurt, someone else who is in charge will handle it” (“Conclusions: Lessons from the Past and a Look to the Future,” in Zahn-Waxler, Cummings, and Iannotti, p. 3 10).

32. For more about the Child Development Project, see Alfie Kohn, “The ABCs of Caring,” Teacher Magazine, January 1990, pp. 52-58; and idem, The Brighter Side of Human Nature, ch. 6. For accounts written by members of the staff, see Victor Battistich et al., “The Child Development Project: A Comprehensive Program for the Development of Prosocial Character,” in William M. Kurtines and Jacob L. Gewirtz, eds., Moral Behavior and Development: Advances in Theory, Research, and Applications (Hillsdale, N.J.: Erlbaum, 1989); and Daniel Solomon et al., “Cooperative Learning as Part of a Comprehensive Classroom Program Designed to Promote Prosocial Development,” in Shlomo Sharan, ed., Cooperative Learning: Theory and Research (New York: Praeger, 1990).

Copyright © 1991 by Alfie Kohn. This article may be downloaded, reproduced, and distributed without permission as long as each copy includes this notice along with citation information (i.e., name of the periodical in which it originally appeared, date of publication, and author’s name). Permission must be obtained in order to reprint this article in a published work or in order to offer it for sale in any form. Please write to the address indicated on the Contact Us page. — © Alfie Kohn