No Contest: The Case Against Competition, rev. ed.
(Houghton Mifflin, 1992)
Chapter 10


Learning Together

A Defense and Analysis of Cooperative Learning

By Alfie Kohn

I consider…the tendency toward a selfish individualism one of the strongest counts against our customary…sit-alone-at-your-own-desk procedure….Whole-hearted purposeful activity in a social situation as the typical unit of school procedure is the best guarantee of the utilization of the child’s native capacities now too frequently wasted.

— William H. Kilpatrick, 1918

INTERVIEWER:  What is it like to work in a group?
JUSTIN (age 10):  You have four brains.


Back to School

Every so often, parents and teachers and policy makers get swept up in yet another earnest discussion about whether values should be taught in the schools.  Kids need them, someone will exclaim, brandishing fresh evidence that today’s young people routinely violate standards of decent conduct.  No, values ought to be taught at home, comes the predictable reply.  But the point is that they’re not being taught at home, the first speaker will rejoin.  Well, then, whose values would we teach?, another challenger wants to know.

Despite their differences, the speakers in this imaginary colloquy share a profound naivete.  To argue about whether it is time for values to be introduced into the classroom is like asking (to take Moliere’s charming example) whether you ought to start speaking in prose:  The point is you have already been doing so, even if you never thought of it that way.  Teachers do not need to lecture children about stealing in order to exude values.  Their choice of stories and the order in which they are taught and the tone of voice in which a character is mentioned; the fact that children must raise their hands to speak or obtain permission to go to the bathroom or address one person in the room by his or her last name; the objects that decorate the walls and who decided they should be there and how students’ work is evaluated (and for what purpose) — all these and many other aspects of life at school already vibrate with values whether we realize it or not.  There is no question, then, of introducing values into a neutral environment, but only of critically examining existing values in light of others that could be there instead.

I have already argued that few values are more persistently promoted in American classrooms than the desirability of trying to beat other people.  Sometimes this lesson is presented with all the subtlety of a fist in the face, as with the use of spelling bees,[1] grades on a curve (a version of artificial scarcity in which my chance of receiving an A is reduced by your getting one), awards assemblies, and other practices that redefine the majority of children as losers.

At other times, competition is promoted tacitly, perhaps even unwittingly, by pitting students against one another for the teacher’s attention and approval.  This may occur through the use of manipulative behavior management strategies — for example, a public announcement such as:  “I like the way Joanne is sitting so nice and quiet.”  (A contest has been created for Nicest, Quietest Pupil, and everyone except Joanne has just lost.)  Or it may follow from the conventional arrangement of asking a question of the whole class:

The teacher asks the question, the students who think they know the answer raise their hands, and the teacher calls on one of them.  We’ve all seen it many times: when one student is called on, the other students who have their hands up register their disappointment with a little ‘Oh.’  It’s a structure that sets the kids against each other.[2]

Anyone who doubts that competition is the subtext of most whole-class question-and-answer sessions need only continue watching the faces of the children who were not recognized.  Are they rooting for Jeremy, who now has the floor, to succeed?  Hardly.  They are hoping he says something stupid because this will present them with another opportunity to triumph.  The teacher’s face is scanned for signs of dissatisfaction with Jeremy’s answer; once found, their hands shoot up again, fingers reaching anxiously for the fluorescent lights.  Some students participate energetically in this scramble to be the first one with the right response, while others stare dully and look beaten (which, at some point, they have been).  Our attention, however, is properly focused not on the temperament of the individual participants but on the structure that has turned learning into a contest.

(I should note, however, that having a conversation with a large group of people does not in itself establish a competitive interaction.  One way to avoid transforming a discussion into a contest is to ask deeper questions for which there are no clear-cut answers (“What would happen if there were no countries?” instead of “What’s the capital of Italy?”) or to encourage students to do some of the asking themselves.  Where there are unambiguously right answers, formats other than whole-class discussions might be preferable.  In any case, these conversations always can be conducted so that each participant’s contribution is valued, an atmosphere of safety prevails such that students feel able to take risks, and verbal fluency or simple recall are not the only skills seen as worth having.)

Alongside competition stands another value, stressed with equal vigor by educators:  individualism.  Between the American flag in the corner and that preprinted strip of impossibly proper cursive letters that stretches above the blackboard, there might as well be a sign that proclaims:  I WANT TO SEE WHAT YOU CAN DO, NOT WHAT YOUR NEIGHBOR CAN DO.  Children sit at separate desks, as if on their own private islands, instructed to keep their eyes on their own work.  Helping is construed as cheating, since it goes without saying that one is evaluated only on the basis of one’s solitary efforts.  Quite a bit goes without saying, in fact, since even talking with one’s classmates is usually regarded as inappropriate.  The fact that each child is supposed to be responsible for his or her own assignments and behavior means that even when students are not led to see each other as obstacles to their own success, each is, at best, irrelevant to the other’s learning.

The two standard classroom arrangements in the U.S. require children to workagainst each other or — as a result of the ostensibly progressive reform of a few decades ago known as individualized learning — apart from each other.  Earlier in this book, I indicated that some teachers are making use of an alternative to these two models, known as cooperative learning.  My scant three paragraphs on this subject do not begin to do justice to the topic, however — with respect to the research supporting its use, the various models for implementing it, or the extent of its potential impact on education.  Since writing that brief account in chapter 3, I have learned a great deal more about cooperative learning and have become an enthusiastic and outspoken proponent of its use.  Meanwhile, the movement itself has taken root in many schools around the country — indeed, around the world.  Cooperative learning (CL) is, in my judgment, one of the most promising alternatives to structural competition not just in the classroom but in any arena.  If there is a single concrete image that represents the transcendence of mutually exclusive goal attainment, it is a picture of three or four children sitting around a table, animatedly exchanging information and ideas.

Between Student and Student

From Morton Deutsch comes the notion of promotive (or positive) interdependence:  My success is facilitated by, or even dependent on, your success.  The explicit application of this model to classroom learning in the U.S. began quite recently.  David Johnson, a former student of Deutsch’s, and his brother Roger published the first major book on CL in 1975.[3]  The 1970s also saw early work on a reward-driven model of team learning at Johns Hopkins University by David DeVries and his doctoral student, Robert Slavin, as well as the development of the Jigsaw approach by Elliot Aronson and the Group Investigation method by Shlomo and Yael Sharan in Israel.  (From another perspective, though, students have been working together — although not necessarily cooperatively — for as long as there have been too few Bunsen burners to go around in science classrooms.  In fact, the practice of having students learn from each other can be traced to the 17th century and probably all the way back to Aristotle.)[4]

CL means working toward a common goal in the classroom — learning in pairs or small groups in the context of positive interdependence.  This context implies that merely dividing a class into teams and announcing that students should work with their groupmates (“You four do this worksheet together”) is not sufficient for, much less equivalent to, cooperative learning.  Because of this, teachers who have merely put children in groups and are unimpressed with the results have not yet given CL a chance to prove itself.  Virtually everyone who has thought seriously about CL has, at one time or another, warned educators that it takes time and skill to foster positive interdependence successfully — particularly in light of the competitive and individualistic norms that students have internalized from other settings.

If CL is something more than casual groupwork, it is something less than altruism — or at least less than the narrow version of altruism that requires an individual to sacrifice his or her own interests.[5]  In asking children to work together, we are not demanding that they ignore their own academic well-being in order to make sure that someone else understands the material.  Positive interdependence means that when you succeed, I succeed, too; my interest in your learning is matched by your interest in mine.

No catalogue of misconceptions about CL would be complete without reference to the idea that working in groups is just a gimmick to perk up a classroom now and then, providing a break from serious instruction.  (“OK, kids, it’s the third Friday of the month.  Remember, that means today we work in teams!”)  While teachers doubtless will want to continue making some use of whole-class discussion[6] and individualized work, CL can — and, I would argue, ought to — become the standard or “default” classroom arrangement starting in kindergarten.

Finally, to echo my earlier remarks on constructive conflict, CL does not entail harmony or concurrence about controversial issues — even after a discussion has been concluded and an assignment completed.  Teachers who properly want to avoid constructing adversarial encounters need not demand a premature consensus:  “Your assignment is to come to an agreement about whether abortion should be legal.”  One likely result of such an instruction is that the student with the loudest voice, or the one holding the pencil, will write down whatever she wants, everyone else in the group will give in so as to be done with the whole business, and no one will learn anything.  Conversely, teachers who properly want to preserve controversy need not set up a win/lose structure to do so.  In a classroom whose members have come to feel part of a community and who have been helped to develop a range of social skills, disagreement can occur without debate, and conflict without competition.  (It makes sense for teachers and students to create a cooperative context before tackling tasks that will provoke strong disagreement.  Conflict should be introduced gradually, with care taken to ensure that students’ social skills (and the bonds between students) can accommodate mildly opposing views before asking them to thrash out more incendiary issues.)

Perhaps it is easier, then, to say what CL is not than to say what it is — at least beyond a one-sentence formula that is about as helpful as saying a poem is a careful arrangement of words.  Yes, we can add that in classrooms where CL is used, teachers communicate the message, “I want to see what you and your neighbor can do together.”  They are likely to ask students to turn first to a peer to check out an idea or request an explanation.  They may evaluate some projects as group efforts, which is what they are.  These teachers probably have arranged the room so children are clustered around tables most of the day instead of seated at separate desks.  They recognize that “socializing” is not something relegated to recess and lunch, something that distracts from learning; rather, they know that learning emerges not only from what transpires between student and teacher, or between student and text, but also from what happens between student and student.

Beyond this, though, it is difficult to say with certainty what CL is for the simple reason that it is strikingly different depending on which theorist, trainer, or teacher you ask.  From a distance, the distinction between CL and traditional classroom arrangements is so pronounced that those who promote learning in pairs or groups seem distinguished principally by this commitment.  But seen from up close, those who huddle together under this conceptual umbrella often differ from one another in the way they conceive of cooperation and, for that matter, learning itself.

The Effects of Cooperative Learning

Before exploring some of the different varieties of CL, I want to say why that discussion — and, for that matter, the whole movement — is so terribly important.  Several versions of CL have been tested in literally hundreds of studies, and the results of that research should be much better known than they are at present.  The gains realized from CL are so impressive and consistent that it is difficult to understand how any talk of school reform could omit the critical question of whether students should be learning alone or together.

SELF-ESTEEM    Consider first the issue of how a student feels about himself and his abilities.  The evidence in chapter 5 might be summarized as follows:  Competition is to self-esteem as sugar is to teeth.  Far from providing a lasting sense of confidence in oneself, a competitive environment leads one to ask:  “What have I done for me lately?”  But CL does not merely avoid the abrasive effect of competition:  At its best, it has a fortifying effect of its own.

Recent reports of CL’s impact on self-esteem are remarkable.  Positive outcomes were found in “eleven of the fifteen studies in which the effects of cooperative learning on self-esteem were studied,” Robert Slavin wrote in 1990.[7]  In a much larger review of the research comparing competitive and cooperative arrangements on self-esteem — a review that, for better or worse, lumped together studies of many different types and of varying quality –the Johnsons reported that a fair number failed to find any statistically significant difference between the two conditions.  But of those studies that did turn up a significant difference, exactly one showed an advantage for competition while 81 showed an advantage for cooperation.  This sort of consistency is, to put it mildly, uncommon in the social sciences.[8]

On the other hand, it should not be concluded that CL is a magical cure for low self-esteem.  First, children sometimes feel lousy about themselves because they are abused or humiliated at home or do not get enough to eat; these problems require changes far deeper and wider than bringing CL to the classroom.[9]  Second, the research on self-esteem (from which conclusions about the effects of educational arrangements are drawn) leaves a great deal to be desired in terms of rigor.  While many of us have gotten in the habit of casually tossing the term around, a global version of self-esteem (how you feel about yourself in general) is far less useful, scientifically speaking, than specific aspects of self-esteem (how positive you feel about your capacity to do math problems, for instance).  Moreover, the tests used to measure self-esteem are often idiosyncratic and badly designed, which means that the result of one self-esteem experiment cannot meaningfully be compared to the result of another that used a different test.[10]  Finally, even some of the better pencil-and-paper measures are intended to tap how one feels about oneself at the time the test is taken, and this tells us little about one’s persistent, stable, underlying beliefs or feelings about oneself — assuming such things can be identified or measured at all.

Notwithstanding all of these (and other) limitations on the theory and practice of self-esteem, there is reason to believe that to whatever extent the time a child spends in school can make a difference in how she sees herself, CL is far more likely to promote a positive self-image than are conventional methods of instruction.[11]  Learning in well-functioning groups provides an environment of social support and enhances the chances of academic success; these effects, in turn, help children to develop more confidence in themselves, more resilience in the face of failure, and more of a sense that their fate is (at least sometimes) in their own hands rather than at the mercy of external forces.

SOCIAL INTERACTION     No one is likely to be astonished by the fact that a structure of positive interdependence (your success equals my success) inclines us to look more favorably on the people with whom we are interacting than a structure of negative interdependence (your success equals my failure).  The only question for those with an interest in education is whether reducing hostility, improving social skills, promoting an acceptance of people from different backgrounds and with different abilities, and coming to view others as potential collaborators (rather than as obstacles to one’s own success) are judged to be sufficiently important goals to warrant the use of CL.  To ask this question is, I think, to answer it.  In fact, David Johnson has said that his work is motivated more by a desire to help children accept people who are different from themselves than by a desire to boost achievement.[12]

Once again, recent reviews of evidence confirm that CL leads children to view each other more favorably, increases the likelihood of cross-ethnic friendships, enhances acceptance of handicapped students, and promotes perspective taking (that is, the capacity to imagine someone else’s point of view).[13]  Moreover, “cooperation seems to promote better relationships when intergroup competition is absent.”[14]  At a time when managers frequently complain that employees seem unable to get along with each other or work together effectively, the fact that children learn social skills from CL seems to be a persuasive argument in itself for making American education less competitive and more cooperative.

ACHIEVEMENT     The most gratifying (and perhaps surprising) finding from CL research is the news that there is no trade-off between helping students to feel better about themselves and each other, on the one hand, and helping them to learn more effectively, on the other.  Working together offers “bottom-line” benefits in terms of achievement just as surely as it enhances self-esteem and the quality of relationships.  These benefits are enjoyed by students of all ages, in all subjects, and at all kinds of schools.

Anyone who finds personal testimonies compelling ought to seek out a teacher who has been using CL over a period of time and ask what the effect has been on students’ comprehension, problem-solving strategies, creativity, or even simple recall of facts.  The change, in fact, can be literally difficult to believe.  One teacher in New York told me that her students’ academic performance improved so drastically after they began working in teams that her principal noticed the difference in her grade reports.  Rather than asking what accounted for the children’s higher achievement, however, he demanded to know why she had suddenly become an easier grader.

Research confirms the improvement, even on relatively superficial measures like standardized tests.  Slavin picked 68 achievement comparisons, making a point of excluding studies he thought were inferior and including a disproportionate number of his own.  Of these, 49 (nearly three quarters of the total) found superior results from CL as opposed to standard classroom arrangements.[15]

At the other end of the continuum, methodologically speaking, the Johnsons reviewed every available study from 1898 to 1989 — some 369 in all — that examined task achievement in cooperative, competitive, or independent conditions.  (Only some of these were conducted in classrooms.)  In comparing the effects of cooperation and competition, no significant differences were found about one third of the time.  When there was a difference that probably was not due to chance, cooperation produced better results than competition did in 87 percent of the comparisons.  Cooperation produced better results than individualized task achievement in almost exactly the same proportion where there were significant differences.[16]  Elsewhere, the Johnsons have observed:

That working together to achieve a common goal produces higher achievement and greater productivity than does working alone is so well confirmed by so much research that it stands as one of the strongest principles of social and organizational psychology.[17]

This principle has been noticed even at the farthest reaches of higher education.  A 1992 assessment of teaching and learning at Harvard University, based on interviews with 570 undergraduates, led to the central conclusion that intellectual growth was associated primarily with “interpersonal activities with faculty members, or with fellow students, built around substantive, academic work.”  The report urged the creation of study groups where they do not already exist and noted that many students avoided taking science classes not because of the heavy workload but because of the competition for grades.[18]

Chapter 3 offers a few reasons to explain why competition in general proves less successful than cooperation (although it’s not necessary to arrive at the position I have taken in this book, that competition is inherently counterproductive, in order to recognize the merit of cooperative learning).  Here I would like to summarize and supplement those accounts, widening the focus of our inquiry to ask how it is that CL is so effective at helping children to learn.  Some of these reasons are negative:  CL works because it avoids having children compete, thus avoiding a number of counterproductive effects:

1.  ANXIETY:  Competition promotes anxiety of a kind and level that interferes with performance.

2.  EXTRINSIC STATUS:  Like other extrinsic motivators (that is, artificial incentives outside the task itself) such as money, grades, certificates, extra dessert, and gold stars, competition leads children to complete a task as quickly as possible, which means they avoid taking risks and exploring ideas in the way that is necessary for creative problem-solving and deep learning.  Competition and other extrinsic motivators tend to reduce interest in, and ultimately impede performance on, many kinds of tasks.[19]

3.  ATTRIBUTIONS:  Whether they win or lose, children typically attribute the results of a competitive encounter to luck or fixed ability.  The result is a diminished sense of empowerment and responsibility for their learning.[20]

4.  PREDICTABILITY:  If the chance to win a contest is the chief reason students have been given for learning something, most of them have figured out in September who is likely to come out on top.  The presumptive winner therefore has been given no reason to do anything more than is necessary to defeat everyone else.  More important, all of the likely losers have been given no rationale for bothering with the subject matter at all.[21]

The remainder of the reasons are positive:  CL works because of benefits that cooperation itself offers — namely:

5.  EMOTIONAL BENEFITS:  CL’s positive effect on self-esteem and relationships with peers will tend to translate into a positive effect on achievement.  Our experience in the workplace bears this out:  People who feel good about both themselves and their colleagues are more likely, all things being equal, to do a better job.  Self-esteem, relationships, and performance are integrally related such that enhancing any one variable can positively affect either of the others.

6.  NERD STIGMA:  A student in the conventional classroom who gives evidence of being too interested in the lesson is liable to be on the receiving end of any number of unlovely labels that have replaced “teacher’s pet.”  But CL helps to create a new set of norms.  “When students are working together toward a common goal, academic work becomes an activity valued by peers.”[22]  CL, in other words, makes it cool to learn, and this increases the probability that effective learning will indeed take place.

7.  INTEREST IN THE SUBJECT:  Not only is it more fun to be part of an effective and supportive group than to work against, or apart from, others, but CL also leads students to become more enthusiastic about what they are learning.  (This is surely an independent benefit:  Beyond how children score on today’s test, we want them to become hooked on playing with words and numbers and ideas.)  The Johnsons cite 22 studies to support the claim that CL promotes “more positive attitudes toward the subject area being studied and the instructional experience.”[23]  One of their favorite stories concerns a teacher who advised them to warn other participants in their training sessions against using CL just before lunch.  Puzzled, they asked why.  The teacher explained that each class in her school was supposed to arrive in the cafeteria at a different, precisely assigned time in order to prevent long lines from developing.  When her students were working in groups, though, it was difficult to drag them away from what they were doing:  She would tell them it was time to stop for lunch and they would protest:  “Wait!  Wait!  We’re almost finished!” [24]

By contrast, students in most American classrooms are eager to be free of them.  They count the minutes until the end of the period, the days left before the weekend, the weeks they must endure until the next vacation.  Perhaps this is because they hear their parents articulating similar sentiments about their work by thanking God for Friday.  But maybe students genuinely find life at school to be a collection of tedious tasks and humiliating evaluations from which any reasonable person would want to escape.  John Goodlad’s mammoth study of more than a thousand representative classrooms across the country confirmed that “the kinds of classroom practices found most often were well liked by relatively small percentages of students”; the older the child, the less satisfaction was expressed.[25]

Of course, everyone is aware that most kids do not care for most aspects of school.  Everyone also is aware that too many students are graduated without the intellectual skills or knowledge that we expect the schools to have provided; indeed, it has become a popular entertainment of late to describe how much students do not know.  But rarely do we connect these two pieces of data.  If children seem unhappy about going to school, we typically attribute this to the fact that kids just are wont to complain, that they don’t like anything — or at least anything good for them.  Then we insist that they had better get used to things that aren’t any fun.  (The premise here seems to be that the chief purpose of school is not to get children excited about learning but to get them acclimated to doing mind-numbing chores.)

It is also possible, however, to conclude that the problem may just lie with what happens in school rather than with some character flaw in the individual child.  The extent of “on-task” behavior in a classroom, by the same token, may tell us something about the teacher as well as about the students.  When a teacher complains that children are off task, our first response should be to ask, “What is the task?”  (In the long run, though, individual teachers probably are not to blame given that decisions about what children must learn and how they must learn it are frequently made by administrators, school board members, parents, and politicians.)  Moreover, the sharp line drawn between enjoying the process, on the one hand, and buckling down to learn how to spell, on the other, may reflect a philosophy of self-denial — or, more accurately, other-denial — that yields the worst of both worlds since it often produces neither enjoyment nor effective results.

Goodlad ticks off what students are asked to do:  passively listen to teachers’ ceaseless talking (in his survey, the average teacher “out-talked the entire class of students by a ratio of about three to one”), submit to close and constant monitoring, work separately and silently on textbooks and worksheets, and so on.

How would I react as an adult to these ways of the classroom?  I would become restless.  I would groan audibly over still another seatwork assignment.  My mind would wander off soon after the beginning of a lecture.  It would be necessary for me to put my mind in some kind of ‘hold’ position.  This is what students do.[26]

More to the point, students come to find boring[27] and detestable not only the place called school but the subjects taught there and activities such as reading and solving problems.  This is why CL, by virtue of making learning (and what is being learned) more enjoyable, can also make learning more successful.

(A couple of important qualifications are needed here.  First, CL is not absolutelynecessary for helping students to become more excited about school; enthusiasm can and does describe non-cooperative classrooms, too.  Second, CL is not sufficient for changing attitudes about school.  It does not in itself respond adequately to the fact that so many children view school as a series of unpleasant tasks, sweetened only by the chance to see their friends.  As I will argue later, using team learning to more effectively cram the same pointless and disconnected bits of information into children’s heads, or relying on the same carrot-and-stick approach to instruction (except this time with groups), will not transform the experience of learning in the way that CL has the potential to do.  In fact, a student’s view of CL may itself be contaminated by its continued association with an otherwise stagnant system of pedagogy.)

8.  INTELLECTUAL INTERACTION:  Most importantly, CL succeeds because “none of us is as smart as all of us.”[28]  A well-functioning group — and, of course, not all groups are well-functioning — can be more successful, particularly on open-ended, challenging tasks, than any member of the group could be on his own.  Ten-year-old Jason, quoted at the beginning of this chapter, captured the phenomenon as succinctly as any researcher has.  The whole of the group truly becomes more productive than the sum of its parts.

Cognitive and social psychologists have tried to figure out just what it is about groups that accounts for the heightened quality of performance.[29]  At the most basic level, everyone benefits from sharing their talents and skills and resources, a process that is either discouraged or prohibited outright in non-cooperative classrooms.  When the different members of a group work separately to collect information and think through a problem, then reconvene to exchange the results, each individual has access to everyone else’s labors.

When one student catches on to something quickly and helps her teammate to understand, both tutor and tutee benefit.  This, as Noreen Webb discovered in a series of studies, happens consistently — provided that the first student does not just announce the solution but explains how she got it and justifies her conviction that it is correct.[30]  As a group of British researchers put it, “Tutors may gain a deeper understanding of the material learned by virtue of having to teach it, and ‘learning how to learn’ strategies may spill over into learning contexts other than the immediate learning task.”[31]  Researchers also have found that simply telling a student that he will be asked to teach someone else what he is about to read (as opposed to reading it in order to perform well on a test) leads to higher interest in the material and greater conceptual understanding of the subject matter.[32]

Beyond explicit tutoring, when people work in groups there is a tendency for one person’s idea to evoke another idea from someone else.  The second idea emerges as a reaction to the first and it might not have occurred at all if its creator had been working on her own.  In one study, fifth- and sixth-graders were given batteries, bulbs, and wire and asked to make a variety of different circuits so that the bulbs would light up.  The students working in groups found the task more engaging, seemed to get less frustrated, and, by building on each other’s ideas, created both more circuits and more unusual kinds of circuits than their classmates who worked individually.[33]

This phenomenon applies not only to ideas about the topic being studied but also to ideas about how to examine the topic (meta-level or second-order ideas).  Group work leads to “more frequent discovery and development of…higher quality cognitive reasoning strategies.”  First graders, for example, were more likely to figure out the abstract categories to which words belonged (in order to memorize these words more effectively) when they worked together.[34]

Finally, CL not only permits conflict, as I mentioned above, but actually relies on it to some extent for its success.  The process of finding that someone else thought a story’s character had a motivation very different from the one you had inferred — or that another person assumed that dinosaurs became extinct for a reason that never dawned on you — nudges you to think through the problem in a new way, to take account of this brand-new perspective and try to resolve it with your own.[35]

Notice that none of this analysis supports the idea that children should — or, with CL,do — become interchangeable members of a collective, relinquishing their selves to some amorphous blob of a group.  It is competition that creates conformity; cooperation thrives on the diversity of its participants and the distinct contributions made by each.  I do think teachers ought to make time for students to do independent work, too, but the more compelling point is that cooperative interaction may simply be the most powerful way to help each child find his own voice, make his own discoveries, devise his own connections to ideas and texts.  “Talking is not merely a way of conveying existing ideas to others; it is also a way by which we explore ideas, clarify them and make them our own.”[36]


All of this impressive empirical evidence supporting cooperative learning — and, for that matter, the reliance on empirical evidence in general — needs to be placed firmly in perspective.  Some researchers construct an imaginary brick wall, with those who respect the findings of science on one side and those who are in thrall to ideology on the other.  But this dichotomy, which has its roots in a philosophy called positivism, confuses science with truth itself.  It errs by overlooking the fact that all research, like all educational practice, is saturated in values, like it or not.  But more than this, it suggests that there is something disgraceful about having and defending values; this is why many scientists label them “biases” or “ideologies” and treat them like last week’s seafood dinner.

Having just finished reviewing the data that bear on the use of cooperative learning, I am obviously not arguing that it doesn’t matter what the research turns up.  It does matter, and it is with considerable relief that I can report that CL makes good sense when judged according to pragmatic and widely shared criteria.  But the fact that I am relieved suggests that I am already drawn to CL because of other, deeper values that I hold.  Frankly, it took me a few years to acknowledge this, to confess that I am not indifferent to how the evidence turns out (even though I hope I am able to assess it fairly).

Cooperative learning is worth defending even apart from its quantifiable consequences.  Ideally, as Canadian educator Judy Clarke reminds us, the idea of interdependence is grounded in

the belief in every person’s worth.  This belief frames people in relationships of positive inter-connectedness and therefore reflects an ethical orientation to life…..Without this moral foundation, co-operative learning may simply appear to teachers as a set of techniques…to ‘master’ and file in one’s repertoire.[37]

Without this moral foundation, CL likewise may appear to researchers as simply one arrangement among many to be evaluated dispassionately, much as one might test different thicknesses of glass for the classroom windows.

I believe there is something intrinsically preferable about having children work with each other as opposed to against or apart from each other.  I am glad that I can reassure parents and teachers that it is a sensible arrangement from any number of perspectives, but I stand with Robert Bellah and his colleagues when they remark that

learning is never the result of the efforts of isolated, competitive individuals alone….The evident weakness in American schools has much to do with the weakening of their community context….Education can never merely be for the sake of individual self-enhancement.  It pulls us into the common world or it fails altogether.[38]

Arguably, it has failed altogether.  But while there is no shortage of critics willing to charge American education with failure, often they miss the point about what has gone wrong and why.  The problem is not so much that students cannot find Turkey on a map but that they do not find themselves part of a community of learners; they do not find themselves, moreover, in a place where each person’s worth is affirmed.  The fact that cooperative learning can alter these sad realities is worth at least as much as anything the studies might tell us about achievement gains and the like.

The Practice of Cooperation

Cooperative learning takes many forms.  It is present in a classroom where students edit each other’s essays or help each other to develop and practice vocabulary or multiplication skills.  It may mean finding an ad hoc partner to quickly exchange ideas about evolution before resuming a whole-class discussion.  Or it may mean that each student belongs to a permanent team of four that meets daily to make sense of current events.

A cooperative lesson plan may come from the teacher’s notebook or it may be devised by the class.  Perhaps the students find that their parents’ talk about politics has made them curious to learn more about political systems in other countries, which they proceed to do with their teammates.  Or maybe a student’s announcement that her garage is now home to a litter of tiny, mewling kittens leads to an impromptu lesson on the subject, with children meeting in groups to read stories about cats, describe how a house looks from a cat’s point of view, research specific scientific questions that they want answered (How are cats both similar to and different from lions?  When do cats fight with each other?  Do cats dream?) and then present their findings to the rest of the class.

All of these examples share the premise that learning is an active and an interactive process.  To that extent, it is decidedly not a silent process.  This is one very practical reason that it is difficult for individual teachers to decide on their own to switch to CL; if the whole school has not made the change, the teacher in the next room may not understand the noise level.  Furthermore, the Johnsons have proposed, only half in jest, that principals should wander through the halls of their schools, listening at each classroom door.  Whenever they hear nothing, they ought to make a point of asking the teacher, “Why isn’t any learning going on in here?”  As delicious as this reversal of customary practice may be, some trainers go even further, urging teachers to make sure their voices do not rise above and overpower the voices of students.

The latter idea, which suggests a pedagogical approach that would startle even some theorists and practitioners of CL, offers a useful point of departure for discussing different ways children might work together.  Some of these issues are quite specific while others get right to the heart of a philosophy of learning; some are widely accepted among those who promote or use CL while others are controversial.  The point here is to stimulate thought about the dimensions of cooperative learning, not to offer a comprehensive guide for implementing it.

ASSIGNING STUDENTS TO GROUPS:  The size of CL groups will vary depending on the age of the students, their skill and experience working in teams, the type of assignment, and the time available.  As a general rule, it is more challenging to integrate more people.  As many as six students may be able to exchange information, but fewer — certainly no more than four — will be better able to produce a common product.  Many tasks, particularly when attempted by younger children, can best be done in pairs.[39]

More complicated is the question of how to assign students to groups.  Each of the most common methods has benefits and problems.

(1) Letting students sort themselves offers a strong vote of confidence in their capacity to make decisions, but the temptation will be powerful for them to work with people they already know and like rather than learning to cooperate with those who are new to them.  Moreover, some students will feel left out by a self-selection process, and the last thing teachers want is to introduce a competition for the most popular students, reminiscent of what happens when children choose up sides for team sports.  One way around this, especially appropriate for open-ended cooperative assignments, is to let children form groups on the basis of the questions they are interested in exploring.

(2) Teachers who construct groups may do so with an eye to making sure that each group includes students of different ethnic backgrounds and both genders.  The advantages of CL that depend on working with those who are different are maximized when the groups are structured to be heterogeneous in this way.

(3) Students with different levels of skill can be placed together deliberately, at least for some tasks, so that effective helping takes place.  (One researcher finds that mixing students from two distinct ability levels works better than combining those of high-, medium-, and low-ability in a single group.)[40]

(4) Finally, students can be placed randomly in groups[41] to avoid the contrived quality of assignments for heterogeneity.  To make sure that each student has the chance to work with a variety of others, groups should probably be shuffled more frequently if assignments are random.  Since students can belong to more than one group at a time, some assignments can last longer than others:  You may belong to one team only for the duration of a short assignment, for example, and to another, more permanent, multi-purpose “base” team for as long as a year.[42]  In any event, it is probably not a good idea for students to be able to switch groups just because they are having trouble resolving problems or getting along with their teammates.  Such difficulties should be seen as learning opportunities — a chance to figure out (with the teacher’s guidance) how to untangle conflicts; students should not be encouraged to assume that it is possible to bail out at the first sign of trouble.

TEACHING SOCIAL SKILLS:  When students do not seem to be working together effectively, some educators are inclined to blame them for not being cooperative — or to give up on CL as unworkable.  Others, though, reason that children may have been presented with a task that requires more subtle interpersonal skills than are currently in their repertoires.  Many champions of CL emphasize that teachers must pay explicit attention to the phenomenon of working together — what it means and how it can be improved.  The Johnsons, for example, emphasize that “collaborative skills are directly taught in classrooms where teachers are serious about using cooperative learning” — not only because these skills are a prerequisite for realizing academic gains but also because they are valuable in their own right.[43]

One teacher recalled her frustration when she first tried to introduce CL to a class of sixth graders and concluded:

We must not…make the mistake of expecting that cooperative group behavior and thinking for oneself will occur in the absence of classroom instruction and practice aimed at these specific goals….We can’t order it to work; we have to make it work.  We have to teach children the skills of working thoughtfully and responsibly together…..I have heard teachers give it up after a single attempt….But these very same teachers would never say, ‘These children cannot read by themselves,’ and thereafter remove any opportunity for them to learn to read.[44]

Social skills such as learning to listen carefully, to make eye contact, or to criticize someone’s ideas without being insulting can be taught just as reading can be taught.  Each lesson, in fact, can be introduced by laying out an academic goal (mastering the future tense in Spanish) and a social goal (helping everyone in the group to feel included).  But I would offer two caveats to this approach.  First, students of every age should have a role in deciding which social skills they could benefit from working on.  It is important to let the specific needs of a given class determine how these issues are addressed rather than running through a packaged set of topics (If it’s Tuesday, it must be time to work on listening skills).  Second, as I will argue below, in the long run we need to move beyond a focus on teaching discrete skills to individuals in favor of creating a caring classroom community as the context for all lessons.

PROCESSING:  A concern with how we learn, not only what we learn, requires attention after each lesson as well as before.  A certain block of time — which, again, will vary with the age of the students and their experience with CL — should be set aside following the completion of a unit for each group to talk about how successful they have been at working together.  They can be asked to consider whether everyone contributed to the final project or one person did most of the work, whether someone dominated the discussion, whether everyone felt free to present ideas, whether tasks were divided up effectively.  They can evaluate the process and, just as important, reflect on how it might be improved the next time they work together.[45]  If they decide they have done well, the team members should have the opportunity to celebrate their success, taking pleasure in what they have done and how they have done it.

INDIVIDUAL ACCOUNTABILITY:  Many adults react skeptically to the idea of CL because their memory of doing group projects in school is that they ended up being stuck with most of the work while their teammates loafed.  (Two social scientists have observed with evident amusement that everyone seems to remember having been the one who had to carry the load; it is always “the others” who did nothing.)[46]  To be sure, if one person in the group writes the whole report or works most of the problems, the benefits of CL will not be realized.  The question, then, is how to ensure individual accountability.

Let us set aside the larger issue of how teachers can best find out, under any learning structure, what it is children have understood and what they need help with.  Instead, let us consider how teachers can make sure that every group member is participating in the CL process.  The simplest method is to give everyone a test on all of the material or to choose someone from each team randomly to explain what his group came up with and how they came up with it.

The problem with this solution is that it assumes “accountable for learning” means accountable to an external authority.[47]  It is taken for granted that the best — indeed, in many educators’ minds, the only — way to make sure that children apply themselves to a task is, in effect, to threaten them with a poor grade or public humiliation if they do not pull their weight.  These assumptions, in turn, reflect a set of beliefs about the nature of learning and human motivation that show up even more starkly in considering the basic issue of how to get students to work together.

POSITIVE INTERDEPENDENCE:  We want members of a team to turn to one another for help, to exchange ideas, to feel a measure of group unity.  One way to make this happen is to provide tasks that require students to work together toward a common goal.  (By contrast, giving children individualized assignments and sitting them at a table with an admonition to help each other does not provide a group objective.)  Another way to draw students together is to give a team only one copy of a story or assignment sheet and only one pencil.

As far as they go, these methods are not particularly objectionable.  Two other approaches, however, seem to me more problematic even though they are widely used by CL teachers.  The first is to assign a different role to each member of a group, so one student writes down all the ideas generated, another checks to make sure that everyone is participating and keeping up, someone else offers words of encouragement, and so on.  While there are times this might be useful — such as when CL is still new to students — roles may quickly come to be experienced as confining.  Spontaneous, cooperative interaction in the process of engaging with ideas will likely be frustrated by a child’s coming to perceive herself principally as the Recorder or Checker — a perception facilitated by teachers who have children wear badges to remind them of their responsibility.  In truth, a student’s primary responsibility is simply to learn in concert with others; anything that could eclipse that process ought to be used with extreme caution.  In a study of 33 classrooms using CL, researchers at the University of Missouri observed that,

in many cases, the designation of students as leaders, recorders, or materials managers seemed artificial.  Students tended to switch roles when necessary and, unless the teacher reminded them, often abandoned roles altogether.  Further, teachers often spent as much time reminding students of their roles as they did teaching a concept.[48]

The last method of creating positive interdependence is to use grades, certificates, gold stars, or other extrinsic rewards to induce students to work together.  Thus, a team may earn a prize for a specified level of performance on a quiz, or part of one student’s grade may be based on the improvement of someone else in the group.  Models of CL can be divided into three categories with respect to this issue:

(1) Some — notably the “student team learning” methods developed by Robert Slavin and his colleagues — are fundamentally and explicitly driven by the use of extrinsic motivators.  For Slavin, the very idea of a “group goal” means “working to earn certificates or other recognition, to receive a few minutes extra of recess, or to earn bonus points on their grades.”[49]

(2) Other approaches seem to reflect some ambivalence about the use of rewards, accepting them for promoting interdependence or boosting motivation, on the one hand, but expressing reservations, or at least not insisting upon their use, on the other.  The Johnsons endorse the strategic use of grades to ensure effective teamwork, calling this “one of the ways in which students are given the message, ‘We sink or swim together.'”  This, however, follows by a few pages their suggestion that extrinsic motivators “should probably be removed as soon as the intrinsic motivation inherent in cooperative learning groups becomes apparent.”[50]

(3) Finally, some models explicitly repudiate (or at least pointedly avoid) the use of extrinsic incentives on the grounds that they are manipulative, destructive, unnecessary, or all three.  Included in this category are:  the Group Investigation method devised by Shlomo Sharan and his associates,[51] the version of CL used in programs implemented by the Developmental Studies Center in California,[52] various models that grow out of a “constructivist” model of learning (see below) such as the collaborative learning approach favored in English-speaking countries other than the U.S.,[53] the work of Nancy Schniedewind and Mara Sapon-Shevin,[54] and my own writings.[55]

As I have already mentioned, a substantial amount of research in social psychology has established that when students see themselves as performing a task in order to receive a reward, their interest in that task is likely to decline, as is the quality of their performance over the long haul (particularly, but not exclusively, when creativity is required).  Recall that one of the chief reasons that competition tends to be counterproductive is precisely its status as an extrinsic motivator; to use other such motivators for groups would seem ill-advised, to say the least.

But let us put aside for a moment the relatively straightforward empirical question of what long-term consequences the use of classroom incentives will bring about.  The belief that incentives are necessary to promote interdependence and therefore learning — apart from relying on some very questionable assumptions[56] — reveals a rumbling under the surface of discussions about CL.  It is here that a great crevice between opposing approaches to learning itself is laid bare.  Because CL has such potential to transform education and society — the power to let a new generation experience firsthand the possibility of achieving one’s goals with others instead of against them — this division is worth exploring carefully.

On one side are those whose philosophy of learning is indistinguishable from that of mainstream educators in the U.S. except for the fact that traditional techniques and theories are applied at the level of the group instead of the individual.  For them, cooperation is really a package of isolatable behaviors that are taught through the use of reinforcements.  CL is a set of strategies that the teacher implements in prescribed ways[57] in order to improve children’s performance on whatever curriculum is already being taught.  (The curriculum often amounts to a collection of information bits that are transmitted by way of primers, basal readers, and worksheets; the effectiveness of that transmission is then assessed by way of standardized tests.)  It is taken for granted on this view, once the layers of assumption are peeled back, that human beings are inclined to learn and to care about each other only when they are given extrinsic incentives for doing so.  These beliefs and practices are separable, of course, but for convenience, I will refer to this whole constellation as the behaviorist position.

On the other side are those who see CL as part of a broader movement in education that represents a profound shift in how we construe and promote learning.  Here children are seen not as passive receptacles for facts but as beings who actively struggle to make sense of themselves and the world around them.  This they do, in large part, by talking to others.  Group discussion is not a “bull session” during which students react to what they have already learned; in large part, this is where the learning occurs.

The teacher’s role is to stimulate a child’s curiosity, to facilitate the process of playing with ideas and constructing meaning, and to aid in the development of intellectual and social skills.  The goal is to get the student to develop an intrinsic, enduring commitment to this process (and to working successfully with others), to take responsibility for her learning and her behavior.  For this reason, children are givenfar more responsibility, separately and as a class, for making decisions and solving problems than is the case in most U.S. schools.  This approach I will refer to as the constructivist position.

The basic distinction I am drawing has been around for decades, but rarely is it brought to bear on the question of CL.  It is not the only schema that can be used to distinguish various models in the field of cooperative learning but it is a division of monumental importance.  The fact is that some CL proponents differ from one another with respect to fundamental ideas about the purpose of schools.

In evaluating the behaviorist view, it must immediately be conceded that a sufficiently attractive reinforcement can get students to do almost anything in the short term.  (This concept fascinates children, in fact, judging by their tendency to entertain themselves by speculating how much they would have to be paid to perform various unappetizing feats.)  What we care about, though, is the effect on motivation in the long term, once reinforcements are no longer present.  Consider just one of many studies that speak to this issue:  Preschoolers who were praised or promised movie tickets for drinking an unfamiliar beverage did indeed drink more of it than those who expected nothing.  But a week later, when the latter group liked the beverage just as much as, if not more than, they did before, the children who had been rewarded for drinking no longer wanted to touch the stuff.[58]

Similarly, when children see schoolwork as a way they can individually orcollectively snag a good grade, avert a bad grade, receive a certificate, earn a popcorn party, win a contest,[59] or secure the approval of a teacher or parent, they have been given no reason to continue learning (or cooperating) when these rewards are no longer available.  In fact, the continuous presentation of rewards to students helps to explain the very absence of interest in learning that is then cited as the reason teachers have no choice but to use rewards.  Much like competing in order to feel good about oneself, offering students extrinsic motivators to get them interested in a task is like giving them salt water to quench their thirst.

The Three C’s of Cooperation

Someone who accepts the premises of behaviorism would argue that CL, like anything else, is dependent for its success on extrinsic motivators.  Speaking from within the constructivist tradition, let me indicate, by way of response, how CL can work effectively in the absence of these incentives.  I will focus on three interlocking domains to which we must attend in order to maximize the benefits of CL — or, for that matter, any learning environment.  Let us call them control, curriculum, and community.

CONTROL     Adults who are routinely told exactly what they have to do at work and how they have to do it are likely to become victims of what is known as “burn-out” — a vivid metaphor that suggests an extinguished candle or a dark bulb.  Some will become actively resentful while others will just go through the motions of working in order to pick up their paychecks.  Teachers, who know this syndrome well, do not always seem to realize that students are subject to it, too.  Conversely, when students (or other people) feel a sense of control over what they are engaged in, they are more likely to find it engaging.  Autonomy, then, is not simply an alternative to extrinsic motivators:  It is far more effective than any such inducement could be at producing interest in learning — cooperative or otherwise.  After all, it is the absence of felt control that, on some accounts, explains why rewards cause intrinsic interest to evaporate in the first place.

Teachers need to do more than minimize the use and salience of extrinsic motivators; they should affirmatively help students to become responsible for their own learning and relationships.  When children were allowed to pick their own materials in one study, they produced more creative art projects than did those who used exactly the same materials but had them handed to them.[60]  A child who can make (teacher-guided) choices about what happens in his or her classroom is a child who will be less likely to require artificial inducements to learn and more likely to get hooked on learning.

One model of CL that takes the idea of autonomy seriously is Group Investigation.  Here, students form inquiry groups based on what they want to know about a given topic and then make decisions together about how they will divide up the labor and conduct their investigation.  Each group collects information and analyzes it, then prepares and shares a final report or innovative presentation that reflects what has been learned.  Finally, each group contributes to the evaluation process, perhaps making up the questions on their unit that will be included in a classwide test, if there is to be one.  The point is to “incorporate the evaluation into the learning process.”[61]

At both the elementary and secondary levels, this model has been associated with a higher level of academic achievement (along with other advantages) precisely because “it gives students more control over their learning.”[62]  Elsewhere, Shlomo Sharan, its co-creator, has described CL itself as exemplifying the best of the constructivist tradition:

Cooperative learning…gives students an active role in deciding about, planning, directing and controlling the content and pace of their learning activities.  It changes the students’ role from recipients of information to seekers, analyzers and synthesizers of information.  It transforms pupils from listeners into talkers and doers, from powerless pawns into participant citizens empowered to influence decisions about what they must do in school.[63]

This is, to be precise, a description of what CL can be and should be, but not, alas, what it always is.

CURRICULUM    Many books have been filled with prescriptions for improving the curriculum used in various subjects at various age levels.  I would therefore make only three points:  First, when children are given assignments that stir their natural curiosity and are neither so difficult as to be anxiety-producing nor so easy as to be boring, they generally do not require extrinsic motivators in order to approach them.[64]  (The corollary to this principle is that if students are required to fill in the missing words in an endless series of unrelated sentences or memorize the kings and queens of England, they may have to be bribed to do so.)

Second, a meaningful curriculum is a necessary and perhaps even sufficient way to draw children into CL without the use of rewards.  “If the task is challenging and interesting, and if students are sufficiently prepared for skills in group process, students will experience the process of groupwork itself as highly rewarding,” according to Stanford University’s Elizabeth Cohen.[65]

Third, if we do not address curricular issues in the context of CL, we may doom it to failure — or, put differently, we may doom it to success according to the wholly inadequate educational standards we have increasingly come to accept.  Apart from the use of rewards, one of the most telling questions that anyone considering the use of CL can ask is what it is children are learning in their groups.  Some models are advertised as being adaptable to any curriculum.[66]  The soothing message here is not only that CL will work regardless of what is being taught, but that educators need not reflect on how challenging or valuable their material may be because students will benefit from learning whatever it is so long as they do it in groups.  Sapon-Shevin has wryly referred to this approach as the “hamburger helper” model of cooperative learning.[67]

If CL is not to become a technique for dressing up ground beef — and if schools are not to become outlets of a McInformation chain, as it were — teachers will need to provide groups with stimulating, personally relevant material that pushes them to engage together in higher-order thinking.  Even the partisans of some extrinsically based models of CL concede that these methods are most appropriate for dealing with problems that have one right answer.[68]  The question, then, is whether we resign ourselves to the continued reliance on such tasks or instead use the potential heuristic power of CL as an opportunity to find things more worthy of our children’s (and teachers’) time.

Using cooperative techniques to have students cover the same boring, inconsequential, or biased material or to have them ‘get through’ worksheets with more efficiency doesn’t demonstrate the approach’s full potential for changing what goes on in schools.[69]

Likewise, the potential of CL is lost if it is geared toward improving scores on standardized tests — or promoted and sold on that basis.  The worshipful regard for these tests — and if this were not discouraging enough, the proposal to have schoolscompete to see who can get more students to blacken the correct bubbles with their No. 2 pencils — comes chiefly from people who are uninformed about pedagogical or motivational issues.  Most educators, by contrast, know that these measures fail to capture what is meaningful about learning and have the effect of pressuring teachers to “teach to the tests,” foregoing potentially innovative lessons in order to prepare children for exams.  Anyone who wanted to destroy what is left of rich, creative teaching — or children’s love of learning — could do no better than to increase our reliance on standardized tests.[70]

Of course, all educators ought to insist on a curriculum of quality and attempt to resist and reverse the diminution of schooling.  But those who are associated with a reform as exciting as CL have a special responsibility to provide content worthy of the method.

COMMUNITY    The assumption that students will work together only on the basis of self-interest and, specifically, only in the hopes of receiving a reward for doing so — the assumption, that is, that no classroom environment could possibly develop norms leading naturally to cooperation — betrays a profoundly cynical view of “human nature.”  Happily, it is a view that enjoys very little empirical support.  Given the right circumstances, caring for others is no less natural than caring for oneself.[71]

A classroom that emphasizes and promotes the value of community — that has, in fact, been transformed into a caring community — allows positive interdependence to take hold.  CL can thus become successful without promising each student extra points for helping someone else.  This is one of the guiding premises of the Child Development Project in San Ramon, California, a comprehensive, long-term elementary-school program that has achieved significant success in helping children to become more caring and responsible.[72]  Drawing in part from the work of that Project, I have elsewhere sketched an approach that all teachers can use to promote a sense of community, the idea being to provide an environment that does not offer incentives for helping or sharing but that encourages children to ask, “What kind of classroom do we want to have?”[73]

Once again, an exploration of this topic takes us beyond its role as an alternative to the use of extrinsic motivators in CL.  In this case, it propels us clear beyond cooperative learning (at least in the way the concept is usually understood) and into the realm of what we might call the cooperative classroom.  Anyone who needs to be convinced to make such a shift should consider the predicament of the teacher who confessed “that she caught herself yelling at a group of students, ‘Stop helping each other; we’re not doing cooperative learning now!'”[74]

I am arguing that cooperation must come to be seen as a fundamental orientation toward other people rather than a set of techniques that are hauled out for specific lessons.  This means working to change the feel of the classroom itself.  As one second-grade teacher (affiliated with the Child Development Project) put it:

You can’t just teach Cooperative Learning as a separate subject that you just do for thirty minutes a day and then just treat the children differently the rest of the time.  The whole atmosphere of the classroom from the time the kids get to school until the time they go home is based on the idea that you expect them to get along, help each other, be cooperative, all the same guidelines in everything they do.[75]

Notice that this does not mean instituting a circumscribed program of CL and passively waiting for its benefits to seep out and improve the dynamics of other interactions and activities.  It means teachers must broaden their vision, rethinking the entire learning experience and working with students to fashion that experience into something that is cooperative inside and out.

Too much of American education consists of teachers spewing out facts and requiring students to spew them back; if the latter is done correctly, the transaction is pronounced successful.  This goes on in elementary school (“Europe is a what, boys and girls?  That’s right, a continent. Very good.”) and it goes on in college as instructors reward polysyllabic conformity by playing a game that might be called Guess What’s on My Legal Pad.[76]  The research and values that support the use of CL can be understood as an invitation to change all this — not to abandon the teaching of basic skills but to distinguish necessary skills from trivia and to embed those skills in the more substantive process of meaningful thinking.

There is good reason to think that CL can be enormously successful without falling back on what are, in effect, bribes for cooperating — so long as students are provided with a measure of control over their learning, a meaningful curriculum, and the opportunity to create a palpable sense of community in the classroom.

Research from at least three CL programs has shown that extrinsic motivators simply are not necessary to produce academic achievement.  First, as I noted earlier, the Group Investigation approach has achieved consistent advantages in terms of the quality of learning.  Second, the Child Development Project found that children in its program classrooms did significantly better than a carefully matched group of comparison students on an essay exam measuring higher-order reading comprehension, while suffering no disadvantage on standardized tests.[77]

The third and most recent example comes from a second-grade mathematics program based on constructivist principles and small-group work, and characterized by “a complete absence of extrinsic motivation or external reward systems, including praise.”  In place of textbook assignments, students were encouraged to take an active role in figuring out and discussing problems that occurred in concrete situations.  Compared to their peers who had been in traditional classrooms, these children developed more sophisticated higher-level reasoning skills without falling behind on basic computational tasks.[78]

To the extent that attention is paid to the issues of control, curriculum, and community, CL can flourish and, further, can be part of a radical transformation of American schooling.  To the extent that these things are absent, education itself — not only CL — is in trouble.

The Prospects for Cooperation

To what extent has CL made an impact on how our children are actually taught?  The answer one gives will depend on when one takes a reading (the assessment that follows was written in 1992) and where one happens to live (CL’s popularity in California may lead residents to assume that children everywhere are now cooperating).  It also may reflect one’s feelings about the desirability of CL:  Partisans, carried away by their enthusiasm, sometimes exaggerate the influence of the movement, while opponents, seeking to alarm the populace, may do the same thing.[79]

By now, many — perhaps most — teachers in the U.S. have at least heard the phrasecooperative learning.  Tens of thousands have been formally trained in one version or another.  CL has its own magazine, which is intended for researchers as well as classroom teachers.  Articles on the subject show up with some regularity in other education journals and magazines, and several bibliographies have been compiled, listing some of the hundreds of books and articles published to date.[80]  In March 1992, the newsletter of a major American education organization declared that, “unlike some other innovations in education, cooperative learning has not been a flash in the pan.  After years of attention, it remains a hot topic among educators.”[81]

Various mainstream organizations have come to recognize CL’s potential for enhancing the quality of learning, and some educators realize that CL is a logical alternative to tracking – that is, separating children by putative ability. [82]  The American Association for the Advancement of Science released a report in 1989 declaring that “the collaborative nature of scientific and technological work should be strongly reinforced by frequent group activity in the classroom.”[83] That same year, the National Council of Teachers of Mathematics also affirmed the value of small-group work.[84]  (Some universities are now beginning to use a collaborative approach for teaching higher-level mathematics.)[85]  The National Council of Teachers of English has published an anthology on the subject and has urged that secondary teachers offer “multiple opportunities for students to work together.”[86]  In Britain, meanwhile, “Her Majesty’s Inspectorate (HMI) have been actively promoting group work in schools in the belief that it increases children’s motivation, develops responsibility and initiative, and creates a context for effective learning.”[87]

The growing popularity of CL, however, is cause for concern among those who recognize its potential to reshape the educational landscape.  Any development that is sufficiently well-known and well-regarded may begin to take on the contours of a fad, which, in turn, means that it will be treated as a fad rather than with the necessary seriousness.  In practical terms, this means that principals and superintendents, trying to keep up with pedagogical fashion, may hire consultants who offer to provide brief (and therefore seductively inexpensive) training sessions for their faculties:  “Give me your teachers for two days of in-service and I’ll show them how to do cooperative learning.”  Some trainers even claim to be able to give teachers sufficient skills to become trainers themselves.

One of the central theses of this chapter is that CL at its best requires a significant reconceptualization of what learning involves and how the people who spend the day together in a classroom relate to each other.  Any attempt to implement CL will therefore raise a host of problems and questions, and many of them will not even occur to teachers until they have had the chance to work with the new arrangement for a time.  The fact is that the process of making this kind of change must be done over a period of months, if not years, with frequent observation and coaching as well as continuous, structured peer support.[88]  (Moreover, it must be done in a way that is respectful of teachers’ experiences and concerns.  Even good ideas should not — and, ultimately, cannot — be forced down unwilling throats.  In fact, we might even go so far as to say that policy makers, trainers, and administrators cannot change what goes on in classrooms.  All they can do is invite teachers to change what goes on in classrooms.)

One writer has estimated that only five to 10 percent of participants in a CL workshop will continue to use the cooperative approach over time if ongoing coaching and support are absent.[89]  If teachers receive quick-and-dirty training sessions, if the trainer is long gone by the time difficulties start to present themselves, then many educators will discard CL as “unworkable” or “unrealistic” and settle back to wait for next year’s fad.  To this extent, there is cause for worry about CL’s short-term popularity among those who discern its long-term potential.

On the other hand, the extent of that popularity should not be overstated.  If it is not quite accurate to call it a drop in the bucket of national educational practice, cooperative learning is still not much more than a puddle.  Walk into a random American classroom and you will still probably find each child at a separate desk facing the back of someone else’s head, those contests for attention and grades, that structurally enforced cruelty that Jules Henry described more than three decades ago (see p. 26).  Flip through the pages of debate about education reform and you will see scant attention to the question of how one student’s learning is related to another’s.

Clearly, we have a very long way to go.  The distance seems even greater, though, in light of the fact that some teachers who have been introduced to CL react much as a body does to the implantation of foreign tissue.  The extent of this rejection is difficult to gauge, but one observer wrote recently that “despite the academic vogue of cooperative learning and efforts at dissemination made by its proponents, it remains an instructional strategy seldom used in a systematic manner over the course of a school year or more.”[90]

It is difficult to verify this pessimistic assessment, but no one seeking to promote the use of CL should minimize the challenge that is involved in doing so successfully.

The reasons for that challenge are multiple.[91]  They include inadequate training in, and support for, the practice of CL and widespread misunderstanding about what CL does and does not involve.  But some obstacles to the effective adoption of cooperation run deeper and would likely cause problems even if trainings were perfection itself.

Specifically, we can identify four primary aspects of CL that may be unsettling to many educators.  The first is that it challenges our cultural commitment to the value of competition; the second is that it challenges our cultural commitment to the value of individualism.  I have already had my say on both of these issues.

The third attitudinal barrier is that CL demands attention to social goals.  Teachers can scarcely avoid noticing that children sometimes seem indifferent to suffering, unable to resolve conflicts fairly, and likely either to try to get their needs met by coercing others or, conversely, to be victimized by coercion.  But even those who recognize that the time spent together in the classroom could be used to attend to these problems may believe that this focus would be inappropriate because their charge is limited to providing instruction in the traditional academic subjects.

Whatever the reason, Goodlad’s study — probably the most comprehensive survey of American schools ever made — found that with respect to goals such as “developing productive and satisfying relations with others based on respect, trust, cooperation, and caring,” what goes on in the classroom at best contributes nothing and often actually impedes the promotion of such values.[92]

Even when attention is given to the development of children’s social skills and prosocial orientation, this enterprise is “frequently viewed through an instrumental prism of how [these skills] affect academic achievement rather than as schooling goals with inherent legitimacy,” according to educational researcher Yisrael Rich.[93]  The models of CL that call for explicit instruction in social skills, and especially those that recommend taking time to create a caring classroom community, would presumably be resisted by teachers who are concerned to help children become good learners but not necessarily good people.  As Rich observes,

If cooperative learning is perceived by teachers as primarily promoting pupils’ personal or social goals, we would not expect very many teachers to voluntarily participate.  And if they are required to participate in the workshop, few of them will arrive at the decision to adopt the new method, assuming they are allowed some choice in the matter, no matter how well the workshop is conducted.  And if they are required to adopt the new method, even fewer will implement it with a reasonable degree of fidelity.[94]

The final factor that may serve as an obstacle to the widespread use of CL is the fact that having children work in groups (at least in a way that is consistent with the constructivist approach to learning) reduces a teacher’s control over the classroom and the predictability of what happens each day.  Someone — it might have been me — once said that the traditional model of teaching amounts to a rehearsed solo performance by the instructor (with students being relegated to the role of audience), whereas CL not only offers instruments to everyone in the room but invites a jazz improvisation.  The analogy has its limits, but it captures two features of CL:  its demand that the teacher guide students in helping each other to learn (rather than being the only source of ideas and information in the room)[95] and its introduction of uncertainty in place of a predictable progression through a prepared lesson plan.  Some teachers have not bargained for either of these changes.

There is a certain pleasure to be taken from the role of king or queen, even if one’s subjects are very short.  To the extent that the process of schooling has been predicated on compelling students to follow directions, to absorb information and regurgitate it on command, to work silently on whatever task is presented, the profession may have attracted some people who thrive on autocracy. I regularly meet teachers who shine with generosity of spirit and an instinct for what children need to grow.  But others, let us frankly admit, are disinclined to embrace an approach that gives students an active role in their own learning and has them look to each other for help.  With CL the teacher has allies throughout the room — a scenario exciting and refreshing to some educators but highly disconcerting to those who, like trial attorneys during cross examination, never ask a question to which they do not already know the answer.[96]

One way around this barrier, of course, is simply to dilute CL until it is not so different from the present arrangement.  When the nature of the interchange among group members is highly circumscribed — more than the students’ age or the subject matter would seem to require — we might suspect that the teacher has compromised the process of cooperation more to maintain control than to maximize the heuristic value of the experience.  Some trainers are water-carriers, so to speak, for this dilution.

Interestingly, the versions of CL that seek to dictate to students each component of cooperation — thereby reducing their sense of autonomy as fellow meaning-creators and idea-explorers — are likely to be so structured and systematized that teachers, too, are deprived of authority.  So-called “teacher-proof” curricula, we ought to have realized by now, are not only disrespectful but chimerical:  They are the perpetual-motion machines of education.  Cookbook approaches to CL similarly attempt to specify in advance what cannot be specified in advance, to reduce learning to a series of discrete steps that renders the process sterile and excludes both teachers and students from the real work (and joy) of what happens in the classroom.

CL is not simply the status quo except in groups.  At its best, it is an entirely different way of approaching the act of learning.  Sharan warns that if we reduce the potentially liberating principles of CL to

a set of prepackaged procedures for managing the movements of warm bodies in the classroom…to get them to study, with just a little less boredom, the same material that would bore them more if it were studied in some other way…[then CL] will soon be discarded as another fraud, as will so many other packaged methods now loudly touted in the educational marketplace.[97]

In fact, the lack of predictability that CL can engender is not a necessary evil; it is a requirement for learning.  Neither CL nor education in general should allow itself to get too comfortable.  A century and a half ago, Kierkegaard, observing that most public figures pride themselves on figuring out how to make people’s lives easier in one way or the other, set for himself “out of love for mankind” the task of “creat[ing] difficulties everywhere.”[98]  In a sense, and with compassion, this is the teacher’s task:  to “design instructional activities that are likely to be problematic for children”[99] and to guard against relaxing when things are going too smoothly.  “Deeper learning would look somewhat ‘messier’ than what I am seeing in most of our [cooperative] classrooms,” is how Eric Schaps, the director of the Child Development Project, once put it.[100]


The prospects for CL to sweep the country, and for the country to stay swept, depend to some extent on the willingness of parents to demand that their children — that all children — be permitted the benefits of working together.  Many parents, looking back on their own years in school, may recall a superlative teacher here, a few satisfying friendships there, an occasional moment of excitement when a connection was made or an idea understood.  But the background to these pleasant memories may be something very like what Goodlad and Henry and scores of other observers have documented:  isolation, humiliation, self-doubt, rivalry, pointless tasks, boredom.  All of us for whom this account rings true have a choice to make.  We can set our jaws and harden our hearts, muttering, “Hey, if it was good enough for me….” — thereby dooming another generation to the same fate.  Or we can ask:  “What can be done to make sure our children get better than we got?”  Even to ask this question presupposes that we have recognized — and this may come as an epiphany — that, just as a game does not have to be a contest, so the stultifying aspects of our education are not inevitable features of going to school.

Is cooperative learning out of step with our society’s values?  Some proponents deny this heatedly, pointing out that even a culture or an economic system as steeped in competition as our own requires cooperation to be successful.  To some extent this is true, but we might as well admit that CL does challenge the beliefs and practices that set people against each other.  Some curriculum guides augment this challenge by making cooperation and competition explicit topics for study,[101] but the very fact of learning by working together presents children with another model, a different experience, a sense of perspective on the competition they otherwise might have taken for granted.  Cooperative learning should not stand or fall depending on how well it serves our society’s institutions.  Rather, our institutions should stand or fall depending on how well they serve the sort of values represented by cooperative learning.



[For full citations, please see the Reference section of No Contest.]

1.For a cogent discussion of how competitive classroom games are both unnecessary and destructive — followed by suggestions for the kinds of games that might replace them — see Mara Sapon-Shevin, “Cooperative Instructional Games:  Alternatives to the Spelling Bee.”  One of Sapon-Shevin’s major points is that the anger and hurt feelings that so often attend these games do not reflect failings of the individual children but follow from the competitive structure of the games themselves.

2. Spencer Kagan quoted in Ron Brandt, “On Cooperative Learning:  A Conversation with Spencer Kagan,” p. 8.  Elsewhere, Kagan has observed that teachers who make public the achievement of their students set up the conditions for competition even if that is not their intent.  This would include posting either tests and papers that have been graded or charts with names and scores (Kagan et al., “Classroom Structural Bias,” p. 279).  We might add to this list the repugnant practice of announcing each student’s score on a test or paper as it is handed back.

3. The third edition of that book, Learning Together and Alone, was published in 1991.  Deutsch, in turn, had studied with Kurt Lewin, whose work in the 1930s and ’40s was instrumental in developing a systematic theory of how people behave in groups.  Lewin’s influence passed, on the one hand, to Deutsch and then to the Johnsons, and, on the other hand, to a second well-known student, Leon Festinger, who, in turn, was mentor to Elliot Aronson, developer of the Jigsaw model of cooperative learning.

4. Lilya Wagner, “Social and Historical Perspectives on Peer Teaching in Education.”  Also see Shlomo Sharan, “The Group Investigation Approach to Cooperative Learning,” p. 30.

5. For a discussion of the difference between cooperation and altruism, see my article, “Cooperation:  What It Means and Doesn’t Mean.”

6. Shlomo Sharan, however, has argued forcefully that “whole-class instruction should be retired as the primary mode of teaching, and, at best, should occupy a fraction of the time it presently occupies in the instructional repertoire of teachers.”  Eschewing the qualifiers that typically mute the impact of recommendations offered in academic essays, Sharan declares that whole-class teaching “often generates…social distance between peers in the classroom and between those from different ethnic groups in particular, insidious social comparison processes, more tightly knit cliques in classrooms, and many more students at the lower levels of achievement….The fact is that cooperative learning can be implemented only to the extent that traditional whole-class teaching is supplanted, not just altered!”  (“Cooperative Learning,” p. 298).

7. Robert Slavin, Cooperative Learning, p. 44.

8. David W. Johnson and Roger T. Johnson, Cooperation and Competition, p. 158.  These numbers reflect measures of overall self-esteem.  Of the studies that found a difference on measures of task or academic self-esteem in particular, 23 favored cooperation and one favored competition.  Of the studies that compared cooperative and individualistic arrangements on overall self-esteem, 39 found an advantage for the former and three for the latter.

9. “If we want to enhance self-esteem, we must first check to see whether the social environment is safe for the individual.  A debilitating environment is likely to squash fledgling self-confidence no matter how much we exhort the individual to persist….Moreover, suggesting that self-esteem can be preserved by developing ‘coping skills’ endorses the status quo….”  (James A. Beane, “Sorting Out the Self-Esteem Controversy,” p. 27).  Also see a thoughtful essay by Ellen Herman, “Toward a Politics of Self-Esteem?”

10. The work of Ruth C. Wylie is especially useful for understanding the peculiar characteristics and weaknesses of some of the many instruments used to measure self-esteem. [For more on the issue, see my later article “The Truth About Self-Esteem.”]

11. For an account of “the role education itself has played in causing students to fail” (p. xiii) and to feel like failures, see William Glasser, Schools Without Failure.

12. Johnson quoted in my article, “It’s Hard to Get Left Out of a Pair,” p. 54.

13. Johnson and Johnson, Cooperation and Competition, chap. 7; Slavin, Cooperative Learning, chap. 3.

14. Johnson and Johnson, Cooperation and Competition, p. 122.

15. Slavin, chap. 2.

16. Johnson and Johnson, Cooperation and Competition, pp. 40-41.  Those studies in which the subjects cooperated within their groups but the groups competed against each other found less of a benefit than those in which there was no intergroup competition.  In seven studies that directly compared cooperation with and without intergroup competition, no differences were found.  The Johnsons’ conclusion is that having groups compete against each other “does not enhance achievement and may decrease it” (p. 46).  (For a fuller discussion of the achievement effects of competition, see chapters 3 and 4 of their book.)

17. Johnson and Johnson, Learning Together and Alone, p. 40.

18. Richard J. Light,  The Harvard Assessment Seminars, pp. 6, 21, 64-65.  The science courses that students rated highest and lowest in overall quality had almost identical workloads.  What distinguished the lowest-rated courses was more intense competition for grades.

19. For more on the destructive effects on extrinsic motivators than the brief remarks on pp. 59-61, see my article, “Group Grade Grubbing vs. Cooperative Learning.”  Three indispensable books on the topic are The Hidden Costs of Rewards (edited by Mark R. Lepper and David Greene), Intrinsic Motivation and Self-Determination in Human Behavior (by Edward L. Deci and Richard M. Ryan), and The Competitive Ethos and Democratic Education (by John G. Nicholls). [Later, I wrote my own book:Punished by Rewards.]

20. See especially the work of Carole Ames and her colleagues listed in the reference section, some of which I described in earlier chapters.

21. This dynamic was noticed by a researcher as early as 1932:  “Generally, the usual classroom incentives call forth a response for maximum exertion only from the few very able pupils, while the majority of the pupils, knowing that their chances for excelling are limited, fail to be motivated to do their very best”  (Joseph Zubin, Some Effects of Incentives, p. 50).

22. R.E. Slavin, “Cooperative Learning and the Cooperative School,” p. 9.

23. David W. Johnson and Roger T. Johnson, “Motivational Processes in Cooperative, Competitive, and Individualistic Learning Situations,” p. 272.  Slavin, by contrast, reports mixed results from questionnaires that ask children whether they like class during CL sessions (Cooperative Learning, pp. 48-49).  He suggests that this may be due in part to some of the surveys having tapped students’ general feelings about school rather than their relative preference for different classroom approaches.  Here is another explanation for his ambiguous findings:  The version of CL that Slavin uses (and tends to rely on in his research reports) depends on rewards to foster interdependence.  We might expect, based on the documented consequences of using extrinsic motivators, that the enthusiasm generated by CL would be cancelled out by the reduced interest in tasks for which a goodie is offered.  If this is true, then models of CL that do not reward children to get them to work together would be expected to produce much more enthusiasm than other approaches.

24. This reluctance to leave could also, in theory, be explained by the sort of food that awaited them.  However, other reports corroborate the excitement that students demonstrate for their work when engaged in CL, groaning when the bell rings and displaying other signs of being engrossed that many teachers thought they would never live long enough to witness.  CL sometimes makes learning so engaging, according to one researcher and trainer, that students will sometimes “be prepared to remain in the classroom and continue working on learning tasks rather than taking a recess and going out to play” (Sharan, p. 288).

25. John I. Goodlad, A Place Called School, pp. 233, 232.

26. Goodlad, pp. 229, 233.

27. “The subject of student boredom in school in general, and under varying instructional conditions in particular, has yet to be studied systematically.  It is almost as if this topic was purposefully shunned by educational researchers!”  (Sharan, p. 287).

28. David W. Johnson, Roger T. Johnson, and Edythe Johnson Holubec, Circles of Learning, p. 65.

29. In addition to the sources cited below, see the discussions of this question in Bonnie K. Nastasi and Douglas H. Clements, “Research on Cooperative Learning,” pp. 117-120, 124-126; and Sharan, pp. 289-291.

30. See Noreen M. Webb, “Student Interaction and Learning in Small Groups” and virtually any of her other articles and chapters.

31. Hugh C. Foot et al., “Theoretical Issues in Peer Tutoring,” p. 72.  Tutoring also can enhance the self-esteem, social skills, and motivation of the tutor.

32. Carl A. Benware and Edward L. Deci, “Quality of Learning with an Active Versus Passive Motivational Set.”

33. Gerald W. Foster and John E. Penick, “Creativity in a Cooperative Group Setting.”

34. Johnson and Johnson, Cooperation and Competition, p. 48.

35. This notion of learning through encounters with conflicting points of view is at the core of Jean Piaget’s approach to cognitive development and, following his work, the model of learning known as constructivism.  For a recent elaboration of this theory from another Swiss researcher, see Willem Doise, “The Development of Individual Competencies through Social Interaction.”

36. Michael Marland, a British educator, is quoted in Joan Green and John Myers, “Conversations,” p. 330.

37. Judy Clarke, “The Hidden Treasure of Co-operative Learning,” p. 3.

38. Robert Bellah et al., The Good Society, pp. 172, 176.

39. For one discussion on this topic, see Johnson and Johnson, Learning Together and Alone, p. 64.

40. Webb, “Student Interaction and Learning in Small Groups,” pp. 165-167.

41. For some concrete suggestions on random-assignment procedures, see Dee Dishon and Pat Wilson O’Leary, “Tips for Heterogeneous Group Selection.”

42. For a thorough discussion of different kinds of learning groups, see Judy Clarke et al., Together We Learn, chaps. 3 and 4.

43. Johnson and Johnson, Learning Together and Alone, p. 146.  Also see their Circles of Learning, pp. 80-85; Judy Clarke et al., Together We Learn, chap. 5; and two papers by Nancy B. Graves and Theodore D. Graves: “Creating a Cooperative Learning Environment” and “Should We Teach Cooperative Skills as a Part of Each Cooperative Lesson?”

44. Selma Wassermann, “Children Working in Groups?  It Doesn’t Work!”, pp. 203-204.

45. See Johnson and Johnson, Learning Together and Alone, pp. 58-59.  For an example of research supporting the value of group processing, see Stuart Yager et al., “The Impact of Group Processing on Achievement in Cooperative Learning Groups.”

46. Kipling D. Williams and Steven J. Karau, “Social Loafing and Social Compensation,” p. 570.

47. Besides asking “accountable to whom?” we might ask “accountable with respect to what?”  For one group of researchers, individual accountability is defined “in terms of operating in accordance with the norms of a classroom community…rather than in terms of performance on achievement tests”  (Erna Yackel et al., “Small Group Interactions as a Source of Learning Opportunities in Second-Grade Mathematics,” p. 398).

48. Thomas L. Good et al., “Using Work-Groups in Mathematics Instruction,” p. 60.  I am indebted to Judy Clarke for her incisive analysis of role assignments.

49. Slavin, “Cooperative Learning and Student Achievement,” p. 31.

50. Johnson and Johnson, Learning Together and Alone, pp. 134, 143.

51. See Yael Sharan and Shlomo Sharan, “Group Investigation Expands Cooperative Learning” and virtually anything else written by either of them.

52. The best known of these programs is the Child Development Project, for information about which see Daniel Solomon et al., “Cooperative Learning as Part of a Comprehensive Program Designed to Promote Prosocial Development”; chapter 6 of my book, The Brighter Side of Human Nature; or my article, “The ABC’s of Caring.”

53. See, for example, Mark Brubacher et al., eds., Perspectives on Small Group Learning, the contributors to which repeatedly show how children learn by talking, and the work of Judy Clarke in Canada, Helen Cowie and Jean Rudduck in England, and Joan Dalton in Australia, among many others.

54. Together, in Mara Sapon-Shevin and Nancy Schniedewind, “Selling Cooperative Learning Without Selling It Short”; separately, in Schniedewind and Ellen Davidson’s excellent guidebook for teachers,Cooperative Learning, Cooperative Lives, and in Sapon-Shevin’s “Cooperative Learning, Cooperative Visions.”

55. See my article “Group Grade Grubbing versus Cooperative Learning” and a forthcoming book titledPunished by Rewards.

56. Nastasi and Clements point out that in all the studies purporting to show that rewards are necessary for learning to occur, CL was basically glued on to “traditional classroom structures which are individualistic and competitive in orientation” (p. 123).  For additional criticisms of the claim that CL yields achievement gains only if extrinsic motivators are present, see my article, “Group Grade Grubbing versus Cooperative Learning.”

57. Spencer Kagan notes that his “structural” methods of CL shares with Slavin’s approach “an emphasis on specific behaviors among teachers” as opposed to “general principles” of cooperation or collaboration  (Brandt, “On Cooperative Learning,” p. 10).  While I am not aware of any CL theorist or researcher who makes the connection explicitly, anyone who relies on extrinsic incentives to get children to cooperate might very well rely on extrinsic incentives to get teachers to use CL; concepts like merit pay are quite consistent with this paradigm.

58. Leann Lipps Birch et al., “Eating as the ‘Means’ Activity in a Contingency.”  The experimenters did not expect to find that verbal rewards would have precisely the same motivation-killing effects as tangible rewards.

59. Sadly, some models of CL require groups of students to compete against each other.  Like team sports, the central message this conveys is that the reason to work together in a group is to defeat another group; triumphing over a common enemy is established as the ultimate goal.  This not only sends children conflicting messages, but takes away with one hand everything that the other has given by making CL available.  “In STAD or TGT [two of the student team learning methods developed at Johns Hopkins University], learning and cooperating are the means; the goal is winning,” observed Spencer Kagan (“Co-op Co-op,” p. 439).  The third edition of a book describing these methods (Robert E. Slavin, Using Student Team Learning) begins with the announcement that “competition between teams is no longer recommended” (p. 1).  This is particularly remarkable for two reasons:  First, the reason given for this shift is not a change of heart or mind on the part of the author but a growing resistance to the technique on the part of educators themselves:  “The same teachers who are attracted to cooperative learning are often repelled by moving competition up to the team level” (p. 1).  Second, notwithstanding this comment in the introduction, the manual proceeds to set out the rules for how “students compete” in the tournaments (p. 24; see also Slavin, Cooperative Learning, chap. 4).

60. Teresa M. Amabile and Judith Gitomer, “Children’s Artistic Creativity.”  For more evidence of, and a theoretical basic for, the auspicious effects of autonomy, see the work of Edward L. Deci, including his book with Richard M. Ryan, Intrinsic Motivation and Self-Determination in Human Behavior.

61. Shlomo Sharan, “Cooperative Learning and Helping Behaviour in the Multi-ethnic Classroom,” p. 158.

62. Yael Sharan and Shlomo Sharan, “Group Investigation Expands Cooperative Learning,” p. 20.

63. These remarks were part of a speech delivered at the second annual cooperative learning conference in 1986.  See Sharan, “Cooperative Learning:  Problems and Promise,” p. 4.  Striking a similar theme, Mara Sapon-Shevin writes that instead of wondering, “What cooperative learning method will increase student test scores?” we might ask, “What kinds of cooperative learning methods and practice best allow students to experience control over their own learning and learn to make meaningful decisions related to their own education and that of others?”  (Sapon-Shevin, “Cooperative Learning, Cooperative Visions,” pp. 26-27).

64. For discussions of task characteristics that tap and promote intrinsic interest, see Thomas W. Malone and Mark R. Lepper, “Making Learning Fun”; and Raymond J. Wlodkowski and Judith H. Jaynes, Eager to Learn.

65. Elizabeth G. Cohen, Designing Groupwork, p. 69.  Cohen emphasizes that the curriculum, besides being interesting, must be structured so that different kinds of abilities are necessary for success.  This allows everyone to make a contribution, which in turn reduces the differences in status that students represent in each other’s eyes.  (See also Cohen, “Continuing to Cooperate.”)

66. For example, Spencer Kagan has remarked that his “structural” version of CL “shares with David and Roger Johnson’s approach the idea of giving teachers new methods so they can teach whatever they want to teach more successfully.  It’s curriculum free; the choice of a structure does not involve choice of any particular curriculum or curriculum materials…”  (Ron Brandt, “On Cooperative Learning,” p. 10).

67. Mara Sapon-Shevin, “Cooperative Learning:  Liberatory Praxis or Hamburger Helper?”

68. “I’m often depressed, however, to see these methods [Student Teams-Achievement Divisions and Teams-Games-Tournaments] applied to subjects that lend themselves more to discussion and controversy”  (Robert E. Slavin, “Here to Stay — Or Gone Tomorrow?”, p. 5).

69. Mara Sapon-Shevin and Nancy Schniedewind, “Selling Cooperative Learning Without Selling It Short.”  Also see Eric Schaps and Catherine Lewis, “Extrinsic Rewards Are Education’s Past, Not Its Future.”

70.  For one of many criticisms, see D. Monty Neill and Noe J. Medina, “Standardized Testing:  Harmful to Educational Health.”  For reactions to the prospect of expanding such tests, see Susan Chira, “Prominent Educators Oppose National Tests.”  Some criticize these exams for various methodological flaws that limit their usefulness at doing even what they claim to do.  I am more concerned with their inherent limits, the misconceived purposes for which they were designed.  The only legitimate reason to evaluate students, I would argue, is to help them learn more effectively.  Tests that sort children like so many potatoes — or that are intended to serve as extrinsic motivators for studying (or teaching) — ultimately interfere with a meaningful education.

71. See my book, The Brighter Side of Human Nature, and some of the more than 600 sources cited there.

72. For that matter, the Project also places special emphasis on the quality of the curriculum and on giving students more control over their learning — all of which have moved the Project’s designers away from relying on punishments or rewards.  For details, see the sources listed in note 52, above.

73. See my article, “Caring Kids:  The Role of the Schools.”

74. Sapon-Shevin and Schniedewind, p. 64.

75. Nan and Ted Graves, “Sue Smith:  The Child Development Project in Action,” p. 12.

76. One observer argues that this attenuated model of learning in higher education helps to explain why CL has rarely taken root in the college classroom.  “The traditional role of the university professor as an expert who initiates novices into disciplinary ways of knowing and thinking is itself an impediment to creating genuinely cooperative learning environments,” according to Steven A. Gelb, who teaches at the University of San Diego.  The other impediments he mentions are the fact that many students expect to defer to authority rather than learning cooperatively, and the extraordinarily competitive milieu in which faculty members work, with cooperation construed as a hindrance to professional success.  (Gelb, “On Being Cooperative in Noncooperative Places,” pp. 1-2).

77. Developmental Studies Center, “Evaluation of the Child Development Project.”

78. See Paul Cobb et al., “Assessment of a Problem-Centered Second-Grade Mathematics Project”; and Erna Yackel et al., “Small-Group Interactions as a Source of Learning Opportunities in Second-Grade Mathematics.”  (The quotation is from the latter article, p. 394).  Children in the project classroom were also less motivated to be superior to their peers — further corroboration of the argument that excellence and competition tend to pull in opposite directions.  In third grade, project students were plunged back into the conventional textbook-based approach to learning math.  Nevertheless, those who had spent the preceding year in constructivist, cooperative classrooms continued to score higher than the others on conceptually challenging tasks (Paul Cobb et al., “A Follow-Up Assessment of a Second-Grade Problem Centered Mathematics Project”).

79. Most of the opposition to CL comes from individuals put off by the very idea of teaching children to cooperate.  One academic psychologist told a reporter that cooperative learning should be opposed because it “goes against the American grain….Education should prepare kids for life in a particular culture.  In reality the name of the game is dog eat dog.”  (Barbara Foorman quoted in William J. Warren, “New Movement Seeks to Replace Rivalry in Class With Team Spirit.)  The closest thing to organized opposition comes from advocates for special programs for “gifted” children, who bolster what they see as their underdog status by wildly overstating the popularity of CL.  “In the [current] climate, marshalling an argument against cooperative learning is an unpopular, if not risky position,” is how one writer prefaces her attack (Ann Robinson, “Cooperation or Exploitation?”, p. 9).

80. See, for example, Samuel Totten et al., Cooperative Learning: A Guide to Research; and the September 1990 issue of Cooperative Learning magazine.  The March 1987 issue of the IASCE Newsletter (precursor to the magazine) is devoted to the academic roots of CL, with articles on the work of Lewin, Deutsch, Dewey, and others.

81. Scott Willis, “Coop. Learning Shows Staying Power,” p. 1.  The article appeared on the front page of the newsletter of the Association for Supervision and Curriculum Development.

82. Considerable research has demonstrated the academic and psychological peril in this widespread practice:  Ability grouping does not respond to differences in what children can learn so much as itcreates differences in what they will learn.  (The indispensable source on this topic is Jeannie Oakes’sKeeping Track.  For a summary of other empirical research, see my book, You Know What They Say…, pp. 163-166, 223.)  The sensible response to tracking, however, is not simply to dump students of widely varying achievement levels into one classroom but to provide the opportunity for them to engage in structured heterogeneous interaction.

83. AAAS report excerpt quoted in Susan F. Wooley et al., “BSCS Cooperative Learning and Science Program,” p. 32.

84. “Small groups provide a forum for asking questions, discussing ideas, making mistakes, learning to listen to others’ ideas, offering constructive criticism, and summarizing discoveries in writing”  (National Council of Teachers of Mathematics, Curriculum and Evaluation Standards for School Mathematics, p. 87).

85. See Elizabeth Culotta, “The Calculus of Education Reform.”  In a cooperative and conceptual approach to teaching calculus to students at Duke University and elsewhere, “much of their grade depends on co-authored written reports,” which means that “students have to collaborate during lab and outside class.  And so they befriend each other” (p. 1061).

86. The recommendation quoted here is taken from a pamphlet entitled “The English Coalition Conference:  Assumptions, Aims, and Recommendations of the Secondary Strand,” distributed by the NCTE.  The 1988 anthology, Focus on Collaborative Learning, was edited by Jeff Golub and the National Council of Teachers of English Committee on Classroom Practices.

87. Helen Cowie and Jean Rudduck, “Learning from One Another,” p. 236.

88. For a sense of what is involved in helping teachers to become skilled in CL, see Susan S. Ellis, “Introducing Cooperative Learning,” and the entire Winter 1991/92 issue of Cooperative Learningmagazine.  Peer support, in fact, is an essential ingredient in the creation of effective schools.  The higher self-esteem, more positive relationships, and enhanced performance that attend cooperation among students also attend cooperation among teachers.  Conversely, “if teachers spend five to seven hours a day advocating a competitive, individualistic approach — telling students, ‘Do your own work….Try to be better [than your neighbor],’ those are the values the teachers are going to have in their relationships with colleagues and their administrators” (Ron Brandt, “On Cooperation in Schools:  A Conversation with David and Roger Johnson,” p. 15.  For more on the subject, see the Johnsons’ book, Leading the Cooperative School.) Complicating matters further is the fear of making oneself vulnerable by seeking help from a colleague about a problem in the classroom.  But the administrator who wants to create a cooperative school must do more than exhort teachers to exchange ideas; carefully designed structures and adequate time must be provided for collaboration.

89. Mary Male, “Cooperative Learning and Staff Development,” p. 4.

90. Yisrael Rich, “Ideological Impediments to Instructional Innovation,” p. 83.

91. I am borrowing, in the discussion that follows, from my article, “Resistance to Cooperative Learning:  Making Sense of Its Deletion and Dilution.”

92. Goodlad, pp. 239-242.

93. Rich, p. 83.

94. Rich, p. 89.

95. “The teacher in the cooperative classroom context will be more a facilitator than a director of learning….Children make sense of the learning events together, within a mutually constructed experience.  The teacher participates in this process, and as appropriate, models strategies and in other ways provides scaffolding to support children’s thinking until they can function independently” (Nastasi and Clements, p. 126).

96. “Placing more emphasis on students’ explanations necessarily requires teachers to relinquish some control over the direction the lesson will take.  This can be a frightening prospect to a teacher who is unprepared to evaluate the validity of a novel idea that students inevitably propose”  (James W. Stigler and Harold W. Stevenson, “How Asian Teachers Polish Each Lesson to Perfection,” p. 44).  To the extent that this process of empowering students explains much of the resistance, it may be this rather than CL, per se, that is hard to do.  Sharan has argued that successful cooperative learning is no more complex or difficult to bring off well than whole-class instruction; the fact that whole-class instruction is used widely does not mean it is being done well (see Sharan, “Cooperative Learning:  New Horizons, Old Threats,” p. 5).

97. Sharan, “Cooperative Learning:  New Horizons, Old Threats,” p. 5.  Likewise, a British educator rites:  “Unless we are prepared to recognize that what’s involved here are radical changes in our expectations of pupils, our own role in the classroom, and our method of classroom management, small group work is likely to remain just another teaching technique, rather than, as it could be, a road toward new patterns of thinking and feeling for the pupils, or new ways for them to make sense of their world” (Alan Howe, “A Climate for Small Group Talk,” pp. 115-116).

98. Soren Kierkegaard, Concluding Unscientific Postscript, pp. 165-166.

99. Erna Yackel et al., p. 396.

100. Eric Schaps, “Cooperative Learning:  The Challenge in the ’90s,” p. 8.  Schaps describes his growing sense that something may be wrong even in smoothly functioning cooperative classrooms where children are “on task” and interacting with each other.  “They aren’t as careful, as serious, as probing as one would like.  They are too casual and quick, and too accommodating, about deciding on their answers.  And in the processing afterwards, they tend day after day to give the same, short, almost ‘canned’ answers:  ‘We shared and helped each other, and everyone got along.'”  Many teachers, he continues, “seem satisfied with easy or predictable answers.  Their questions do not often probe or challenge; their comments are often routine and formulaic.”  If teachers and students were really exploring ideas wholeheartedly, there would be “more conflict, more frustration” (p. 8).

101. For example, see Nancy Schniedewind and Ellen Davidson, Cooperative Learning, Cooperative Lives; and Ebba Hierta, Building Cooperative Societies.


Copyright © 1992 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. We can be reached through the Contact Us page.