Volume 4, Number 2, REFLECTIONS
4
Teaching Smart PeopleHow to LearnChris Argyris
Any company that aspires to succeed in the tougher business environment of the 1990smust Žrst resolve a basic dilemma: success in the marketplace increasingly dependson learning, yet most people don’t know how to learn. What’s more, those members ofthe organization that many assume to be the best at learning are, in fact, not very goodat it. I am talking about the well-educated, high-powered, high-commitmentprofessionalswho occupy key leadership positions in the modern corporation.Most companies not only have tremendous difŽculty addressing this learning di-lemma; they aren’t even aware that it exists. The reason: they misunderstandwhat learn-ing is and how to bring it about. As a result, they tend to make two mistakes in theirefforts to become a learning organization.First, most people deŽne learning too narrowly as mere ‘‘problem solving,’’ so theyfocus on identifying and correcting errors in the external environment. Solving problemsis important. But if learning is to persist, managers and employees must also look inward.They need to reect critically on their own behavior, identify the ways they often inad-vertently contribute to the organization’s problems, and then change how they act. Inparticular, they must learn how the very way they go about deŽning and solving problemscan be a source of problems in its own right.I have coined the terms ‘‘single loop’’ and ‘‘double loop’’ learning to capture thiscrucial distinction. To give a simple analogy: a thermostat that automatically turns on theheat whenever the temperature in a room drops below 68 degrees is a good example ofsingle-loop learning. A thermostat that could ask, ‘‘Why am I set at 68 degrees?’’ and thenexplore whether or not some other temperaturemight more economically achieve the goalof heating the room would be engaging in double-loop learning.Highly skilled professionals are frequently very good at single-loop learning. After all,they have spent much of their lives acquiring academic credentials, mastering one or anumber of intellectual disciplines, and applying those disciplines to solve real-world prob-lems. But ironically, this very fact helps explain why professionals are often so bad atdouble-loop learning.Put simply, because many professionals are almost always successful at what theydo, they rarely experience failure. And because they have rarely failed, they have neverlearned how to learn from failure. So whenever their single-loop learning strategies gowrong, they become defensive, screen out criticism, and put the ‘‘blame’’ on anyone andeveryone but themselves. In short, their ability to learn shuts down precisely at the mo-ment they need it the most.The propensity among professionals to behave defensively helps shed light on thesecond mistake that companies make about learning. The common assumption is thatgetting people to learn is largely a matter of motivation. When people have the rightattitudes and commitment, learning automatically follows. So companies focus on creatingnew organizational structures—compensation programs, performance reviews, corporatecultures, and the like—that are designed to create motivated and committed employees.But effective double-loop learning is not simply a function of how people feel. It is areection of how they think—that is, the cognitive rules or reasoning they use to design© 1991 Harvard Business Review.Distributed by The New York TimesSpecial Features/Syndication Sales.
Chris ArgyrisJames Bryant Conant ProfessorHarvard Business School

REFLECTIONS, Volume 4, Number 2
Teachin
gSmart
People
Howto
Learn
·ARG
YRIS
5
and implement their actions. Think of these rules as a kind of ‘‘master program’’ storedin the brain, governing all behavior. Defensive reasoning can block learning even whenthe individual commitment to it is high, just as a computer program with hidden bugscan produce results exactly the opposite of what its designers had planned.Companies can learn how to resolve the learning dilemma. What it takes is to makethe ways managers and employees reason about their behavior a focus of organizationallearning and continuous improvement programs. Teaching people how to reason abouttheir behavior in new andmore effectiveways breaks down the defenses that block learning.All of the examples that follow involve a particular kind of professional: fast-trackconsultants at major management consulting companies. But the implications of my ar-gument go far beyond this speciŽc occupational group. The fact is, more and more jobs—no matter what the title—are taking on the contours of ‘‘knowledge work.’’ People at alllevels of the organization must combine the mastery of some highly specialized technicalexpertise with the ability to work effectively in teams, form productive relationships withclients and customers, and critically reect on and then change their own organizationalpractices. And the nuts and bolts of management—whether of high-powered consultantsor service representatives, senior managers or factory technicians—increasingly consistsof guiding and integrating the autonomous but interconnected work of highly skilledpeople.How Professionals Avoid LearningFor 15 years, I have been conducting in-depth studies of management consultants. I de-cided to study consultants for a few simple reasons. First, they are the epitome of thehighly educated professionals who play an increasingly central role in all organizations.Almost all of the consultants I’ve studied have MBAs from the top three or four U.S.business schools. They are also highly committed to theirwork. For instance, at one company, more than 90% of theconsultants responded in a survey that they were ‘‘highlysatisŽed’’ with their jobs and with the company.I also assumed that such professional consultantswouldbe good at learning. After all, the essence of their job is toteach others how to do things differently. I found, however,that these consultants embodied the learning dilemma. Themost enthusiastic about continuous improvement in theirown organizations, they were also often the biggest obstacleto its complete success.As long as efforts at learning and change focused on external organizational factors—job redesign, compensation programs, performance review, and leadership training—theprofessionals were enthusiastic participants. Indeed, creating new systems and structureswas precisely the kind of challenge that well-educated, highly motivated professionalsthrived on.And yet the moment the quest for continuous improvement turned to the profession-als’ own performance, something went wrong. It wasn’t a matter of bad attitude. Theprofessionals’ commitment to excellence was genuine, and the vision of the companywasclear. Nevertheless, continuous improvement did not persist. And the longer the contin-uous improvement efforts continued, the greater the likelihood that they would produceever-diminishing returns.What happened? The professionals began to feel embarrassed. They were threatenedby the prospect of critically examining their own role in the organization. Indeed, becausethey were so well paid (and generally believed that their employers were supportive andfair), the idea that their performance might not be at its best made them feel guilty.Far from being a catalyst for real change, such feelings caused most to react defen-sively. They projected the blame for any problems away from themselves and onto whatthey said were unclear goals, insensitive and unfair leaders, and stupid clients.Consider this example. At a premier management consulting company, the managerof a case team called a meeting to examine the team’s performance on a recent consultingproject. The client was largely satisŽed and had given the team relatively high marks, but
Professionals embody the learningdilemma: they are enthusiasticabout continuous improvement—and often the biggest obstacle toits success.

Volume 4, Number 2, REFLECTIONS
Teachin
gSmart
People
Howto
Learn
·ARG
YRIS
6
the manager believed the team had not created the value added that it was capable of andthat the consulting company had promised. In the spirit of continuous improvement, hefelt that the team could do better. Indeed, so did some of the team members.The manager knew how difŽcult it was for people to reect critically on their ownwork performance, especially in the presence of their manager, so he took a number ofsteps to make possible a frank and open discussion. He invited to the meeting an outsideconsultant whom team members knew and trusted—‘‘just to keep me honest,’’ he said.He also agreed to have the entire meeting tape-recorded. That way, any subsequent con-fusions or disagreements about what went on at the meeting could be checked againstthe transcript. Finally, the manager opened the meeting by emphasizing that no subjectwas off limits—including his own behavior.‘‘I realize that you may believe you cannot confront me,’’ the manager said. ‘‘But Iencourage you to challenge me. You have a responsibility to tell me where you think theleadership made mistakes, just as I have the responsibility to identify any I believe youmade. And all of us must acknowledge our own mistakes. If we do not have an opendialogue, we will not learn.’’The professionals took the manager up on the Žrst half of his invitation but quietlyignored the second. When asked to pinpoint the key problems in the experience with theclient, they looked entirely outside themselves. The clients were uncooperative and ar-rogant. ‘‘They didn’t think we could help them.’’ The team’s own managers were un-available and poorly prepared. ‘‘At times, our managers were not up to speed before theywalked into the client meetings.’’ In effect, the professionals asserted that they were help-less to act differently—not because of any limitations of their own but because of thelimitations of others.The manager listened carefully to the team members and tried to respond to theircriticisms. He talked about the mistakes that he had made during the consulting process.For example, one professional objected to the way the manager had run the project meet-ings. ‘‘I see that the way I asked questions closed down discussions,’’ responded themanager. ‘‘I didn’t mean to do that, but I can see how you might have believed that I hadalready made up my mind.’’ Another team member complained that the manager hadcaved in to pressure from his superior to produce the project report far too quickly, con-sidering the team’s heavy work load. ‘‘I think that it was my responsibility to have saidno,’’ admitted the manager. ‘‘It was clear that we all had an immense amount of work.’’Finally, after some three hours of discussion about his own behavior, the managerbegan to ask the team members if there were any errors they might have made. ‘‘Afterall,’’ he said, ‘‘this client was not different from many others. How can we be moreeffective in the future?’’The professionals repeated that it was really the clients’ and their own managers’fault. As one put it, ‘‘They have to be open to change and want to learn.’’ The more themanager tried to get the team to examine its own responsibility for the outcome, the morethe professionals bypassed his concerns. The best one team member could suggest wasfor the case team to ‘‘promise less’’—implying that there was really no way for the groupto improve its performance.The case team members were reacting defensively to protect themselves, even thoughtheir manager was not acting in ways that an outsider would consider threatening. Evenif there were some truth to their charges—the clients may well have been arrogant andclosed, their own managers distant—the way they presented these claims was guaranteedto stop learning. With few exceptions, the professionals made attributions about the be-havior of the clients and the managers but never publicly tested their claims. For instance,they said that the clients weren’t motivated to learn but never really presented any evi-dence supporting that assertion. When their lack of concrete evidence was pointed out tothem, they simply repeated their criticisms more vehemently.If the professionals had felt so strongly about these issues, why had they never men-tioned them during the project? According to the professionals, even this was the fault ofothers. ‘‘We didn’t want to alienate the client,’’ argued one. ‘‘We didn’t want to be seenas whining,’’ said another.The professionals were using their criticisms of others to protect themselves from thepotential embarrassment of having to admit that perhaps they too had contributed to the

REFLECTIONS, Volume 4, Number 2
Teachin
gSmart
People
Howto
Learn
·ARG
YRIS
7
team’s less-than-perfect performance. What’s more, the fact that they kept repeating theirdefensive actions in the face of the manager’s efforts to turn the group’s attention to itsown role shows that this defensiveness had become a reexive routine. From the profes-sionals’ perspective, they weren’t resisting; they were focusing on the ‘‘real’’ causes.Indeed, they were to be respected, if not congratulated, for working as well as they didunder such difŽcult conditions.The end result was an unproductive parallel conversation. Both the manager and theprofessionals were candid; they expressed their views forcefully. But they talked past eachother, never Žnding a common language to describe what had happened with the client.The professionals kept insisting that the fault lay with others. The manager kept trying,unsuccessfully, to get the professionals to see how they contributed to the state of affairsthey were criticizing. The dialogue of this parallel conversation looks like this:Professionals: ‘‘The clients have to be open. They must want to change.’’Manager: ‘‘It’s our task to help them see that change is in their interest.’’Professionals: ‘‘But the clients didn’t agree with our analyses.’’Manager: ‘‘If they didn’t think our ideas were right, how might we have convincedthem?’’Professionals: ‘‘Maybe we need to have more meetings with the client.’’Manager: ‘‘If we aren’t adequately prepared and if the clients don’t think we’re cred-ible, how will more meetings help?’’Professionals: ‘‘There should be better communication between case team membersand management.’’Manager: ‘‘I agree. But professionals should take the initiative to educate themanagerabout the problems they are experiencing.’’Professionals: ‘‘Our leaders are unavailable and distant.’’Manager: ‘‘How do you expect us to know that if you don’t tell us?’’Conversations such as this one dramatically illustrate the learning dilemma. Theprob-lem with the professionals’ claims is not that they are wrong but that they aren’t useful.By constantly turning the focus away from their own behavior to that of others, the pro-fessionals bring learning to a grinding halt. The manager understands the trap but doesnot know how to get out of it. To learn how to do that requires going deeper into thedynamics of defensive reasoning—and into the special causes that make professionals soprone to it.
Defensive Reasoning and the Doom LoopWhat explains the professionals’ defensiveness? Not their attitudes about change or com-mitment to continuous improvement; they really wanted to workmore effectively. Rather,the key factor is the way they reasoned about their behavior and that of others.It is impossible to reason anew in every situation. If we had to think through all thepossible responses every time someone asked, ‘‘How are you?’’ the world would pass usby. Therefore, everyone develops a theory of action—a set of rules that individuals useto design and implement their own behavior as well as to understand the behavior ofothers. Usually, these theories of actions become so taken for granted that people don’teven realize they are using them.One of the paradoxes of human behavior, however, is that the master program peopleactually use is rarely the one they think they use. Ask people in an interview or question-naire to articulate the rules they use to govern their actions, and they will give you whatI call their ‘‘espoused’’ theory of action. But observe these same people’s behavior, andyou will quickly see that this espoused theory has very little to do with how they actuallybehave. For example, the professionals on the case team said they believed in continuousimprovement, and yet they consistently acted in ways thatmade improvement impossible.When you observe people’s behavior and try to come up with rules that would makesense of it, you discover a very different theory of action—what I call the individual’s‘‘theory-in-use.’’ Put simply, people consistently act inconsistently, unaware of the con-tradiction between their espoused theory and their theory-in-use, between the way theythink they are acting and the way they really act.

Volume 4, Number 2, REFLECTIONS
Teachin
gSmart
People
Howto
Learn
·ARG
YRIS
8
What’s more, most theories-in-use rest on the same set of governing values. Thereseems to be a universal human tendency to design one’s actions consistently accordingto four basic values:1. To remain in unilateral control;2. To maximize ‘‘winning’’ and minimize ‘‘losing’’;3. To suppress negative feelings; and4. To be as ‘‘rational’’ as possible—by which people mean deŽning clear objectives andevaluating their behavior in terms of whether or not they have achieved them.The purpose of all these values is to avoid embarrassment or threat, feeling vulnerableor incompetent. In this respect, the master program that most people use is profoundlydefensive. Defensive reasoning encourages individuals to keep private the premises, in-ferences, and conclusions that shape their behavior and to avoid testing them in a trulyindependent, objective fashion.Because the attributions that go into defensive reasoning are never really tested, it isa closed loop, remarkably impervious to conicting points of view. The inevitable re-sponse to the observation that somebody is reasoning defensively is yet more defensivereasoning. With the case team, for example, whenever anyone pointed out the profes-sionals’ defensive behavior to them, their initial reaction was to look for the cause insomebody else—clients who were so sensitive that they would have been alienated if theconsultants had criticized them or a manager so weak that he couldn’t have taken it hadthe consultants raised their concerns with him. In other words, the case team membersonce again denied their own responsibility by externalizing the problem and putting it onsomeone else.In such situations, the simple act of encouraging more open inquiry is often attackedby others as ‘‘intimidating.’’ Those who do the attacking deal with their feelings aboutpossibly being wrong by blaming the more open individual for arousing these feelingsand upsetting them.Needless to say, such a master program inevitably short-circuits learning. And for anumber of reasons unique to their psychology, well-educated professionals are especiallysusceptible to this.Nearly all the consultants I have studied have stellar academic records. Ironically,their very success at education helps explain the problems they havewith learning. Beforethey enter the world of work, their lives are primarily full of successes, so they have rarelyexperienced the embarrassment and sense of threat that comes with failure. As a result,their defensive reasoning has rarely been activated. People who rarely experience failure,however, end up not knowing how to deal with it effectively. And this serves to reinforcethe normal human tendency to reason defensively.In a survey of several hundred young consultants at the organizations I have beenstudying, these professionals describe themselves as driven internally by an unrealisticallyhigh ideal of performance: ‘‘Pressure on the job is self-imposed.’’ ‘‘I must not only do a good job; I must also bethe best.’’ ‘‘People around here are very bright and hard-working; they are highly motivated to do an outstandingjob.’’ ‘‘Most of us want not only to succeed but also to doso at maximum speed.’’These consultants are always comparing themselveswith the best around them and constantly trying to bettertheir own performance. And yet they do not appreciate being required to compete openlywith each other. They feel it is somehow inhumane. They prefer to be the individualcontributor—what might be termed a ‘‘productive loner.’’Behind this high aspiration for success is an equally high fear of failure and a pro-pensity to feel shame and guilt when they do fail to meet their high standards. ‘‘You mustavoid mistakes,’’ said one. ‘‘I hate making them. Many of us fear failure, whether weadmit it or not.’’To the extent that these consultants have experienced success in their lives, they havenot had to be concerned about failure and the attendant feelings of shame and guilt. Butto exactly the same extent, they also have never developed the tolerance for feelings offailure or the skills to deal with these feelings. This in turn has led them not only to fear
The very success of professionalsat education helps explain theproblems they have with learning.

REFLECTIONS, Volume 4, Number 2
Teachin
gSmart
People
Howto
Learn
·ARG
YRIS
9
failure but also fear the fear of failure itself. For they know that they will not cope withit superlatively—their usual level of aspiration.The consultants use two intriguing metaphors to describe this phenomenon. Theytalk about the ‘‘doom loop’’ and ‘‘doom zoom.’’ Often, consultants will perform well onthe case team, but because they don’t do the jobs perfectly or receive accolades from theirmanagers, they go into a doom loop of despair. And they don’t ease into the doom loop,they zoom into it.As a result, many professionals have extremely ‘‘brittle’’ personalities. When sud-denly faced with a situation they cannot immediately handle, they tend to fall apart. Theycover up their distress in front of the client. They talk about it constantly with their fellowcase team members. Interestingly, these conversations commonly take the form of bad-mouthing clients.Such brittleness leads to an inappropriately high sense of despondency or even de-spair when people don’t achieve the high levels of performance they aspire to. Suchdespondency is rarely psychologically devastating, but when combined with defensivereasoning, it can result in a formidable predisposition against learning.There is no better example of how this brittleness can disrupt an organization thanperformance evaluations. Because it represents the onemoment when a professional mustmeasure his or her own behavior against some formal standard, a performance evaluationis almost tailor-made to push a professional into the doom loop. Indeed a poor evaluationcan reverberate far beyond the particular individual involvedto spark defensive reasoning throughout an entire organi-zation.At one consulting company, management established anew performance-evaluation process that was designed tomake evaluations both more objective and more useful tothose being evaluated. The consultants participated in thedesign of the new system and in general were enthusiastic because it corresponded totheir espoused values of objectivity and fairness. A brief two years into the new process,however, it had become the object of dissatisfaction. The catalyst for this about-face wasthe Žrst unsatisfactory rating.Senior managers had identiŽed six consultants whose performance they consideredbelow standard. In keeping with the new evaluation process, they did all they could tocommunicate their concerns to the six and to help them improve. Managers met witheach individual separately for as long and as often as the professional requested to explainthe reasons behind the rating and to discuss what needed to be done to improve—but tono avail. Performance continued at the same low level and, eventually, the six were let go.When word of the dismissal spread through the company, people responded withconfusion and anxiety. After about a dozen consultants angrily complained to manage-ment, the CEO held two lengthy meetings where employees could air their concerns.At the meetings, the professionals made a variety of claims. Some said theperformance-evaluation process was unfair because judgments were subjective and bi-ased and the criteria for minimum performance unclear. Others suspected that the realcause for the dismissals was economic and that the performance-evaluation procedurewas just a Žg leaf to hide the fact that the company was in trouble. Still others arguedthat the evaluation process was antilearning. If the company were truly a learning orga-nization, as it claimed, then people performing below the minimum standard should betaught how to reach it. As one professional put it: ‘‘We were told that the company didnot have an up-or-out policy. Up-or-out is inconsistent with learning. You misled us.’’The CEO tried to explain the logic behind management’s decision by grounding it inthe facts of the case and by asking the professionals for any evidence thatmight contradictthese facts.Is there subjectivity and bias in the evaluation process? Yes, responded the CEO, but‘‘we strive hard to reduce them. We are constantly trying to improve the process. If youhave any ideas, please tell us. If you know of someone treated unfairly, please bring itup. If any of you feel that you have been treated unfairly, let’s discuss it now or, if youwish, privately.’’Is the level of minimum competence too vague? ‘‘We are working to deŽne min-imum competence more clearly,’’ he answered. ‘‘In the case of the six, however, their
Performance evaluation is tailor-made to push professionals intothe doom loop.

Volume 4, Number 2, REFLECTIONS
Teachin
gSmart
People
Howto
Learn
·ARG
YRIS
10
performance was so poor that it wasn’t difŽcult to reach a decision.’’ Most of the six hadreceived timely feedback about their problems. And in the two cases where people hadnot, the reason was that they had never taken the responsibility to seek out evaluations—and, indeed, had actively avoided them. ‘‘If you have any data to the contrary,’’ the CEOadded, ‘‘let’s talk about it.’’Were the six asked to leave for economic reasons? No, said the CEO. ‘‘We havemorework than we can do, and letting professionals go is extremely costly for us. Do any ofyou have any information to the contrary?’’As to the company being antilearning, in fact, the entire evaluation process was de-signed to encourage learning. When a professional is performing below the minimumlevel, the CEO explained, ‘‘we jointly design remedial experiences with the individual.Then we look for signs of improvement. In these cases, either the professionals werereluctant to take on such assignments or they repeatedly failed when they did. Again, ifyou have information or evidence to the contrary, I’d like to hear about it.’’The CEO concluded: ‘‘It’s regrettable, but sometimes we make mistakes and hire thewrong people. If individuals don’t produce and repeatedly prove themselves unable toimprove, we don’t know what else to do except dismiss them. It’s just not fair to keeppoorly performing individuals in the company. They earn an unfair share of the Žnancialrewards.’’Instead of responding with data of their own, the professionals simply repeated theiraccusations but in ways that consistently contradicted their claims. They said that a gen-uinely fair evaluation process would contain clear and documentable data about perfor-mance—but they were unable to provide Žrsthand examples of the unfairness that theyimplied colored the evaluation of the six dismissed employees. They argued that peopleshouldn’t be judged by inferences unconnected to their actual performance—but theyjudge management in precisely this way. They insisted that management deŽne clear,objective, and unambiguous performance standards—but they argued that any humanesystem would take into account that the performance of a professional cannot be preciselymeasured. Finally, they presented themselves as champions of learning—but they neverproposed any criteria for assessing whether an individual might be unable to learn.In short, the professionals seemed to hold management to a different level of perfor-mance than they held themselves. In their conversation at the meetings, they used manyof the features of ineffective evaluation that they condemned—the absence of concretedata, for example, and the dependence on a circular logic of ‘‘heads we win, tails youlose.’’ It is as if they were saying, ‘‘Here are the features of a fair performance-evaluationsystem. You should abide by them. But we don’t have to when we are evaluating you.’’Indeed, if we were to explain the professionals’ behavior by articulating rules thatwould have to be in their heads in order for them to act the way they did, the rules wouldlook something like this:1. When criticizing the company, state your criticism in ways that you believe arevalid—but also in ways that prevent others from deciding for themselves whetheryour claim to validity is correct.2. When asked to illustrate your criticisms, don’t include any data that others could useto decide for themselves whether the illustrations are valid.3. State your conclusions in ways that disguise their logical implications. If others pointout those implications to you, deny them.Of course, when such rules were described to the professionals, they found themabhorrent. It was inconceivable that these rules might explain their actions. And yet indefending themselves against this observation, they almost always inadvertently con-Žrmed the rules.Learning How to Reason ProductivelyIf defensive reasoning is as widespread as I believe, then focusing on an individual’sattitudes or commitment is never enough to produce real change. And as the previousexample illustrates, neither is creating new organizational structuresor systems. Theprob-lem is that even when people are genuinely committed to improving their performanceand management has changed its structures in order to encourage the ‘‘right’’ kind of

REFLECTIONS, Volume 4, Number 2
Teachin
gSmart
People
Howto
Learn
·ARG
YRIS
11
behavior, people still remain locked in defensive reasoning. Either they remain unawareof this fact, or if they do become aware of it, they blame others.There is, however, reason to believe that organizations can break out of this viciouscircle. Despite the strength of defensive reasoning, people genuinely strive to producewhat they intend. They value acting competently. Their self-esteem is intimately tied upwith behaving consistently and performing effectively. Companies can use these universalhuman tendencies to teach people how to reason in a new way—in effect, to change themaster programs in their heads and thus reshape their behavior.People can be taught how to recognize the reasoning they use when they design andimplement their actions. They can begin to identify the inconsistencies between theirespoused and actual theories of action. They can face up tothe fact that they unconsciously design and implement ac-tions that they do not intend. Finally, people can learn howto identify what individuals and groups do to create organi-zational defenses and how these defenses contribute to anorganization’s problems.Once companies embark on this learning process, theywill discover that the kind of reasoning necessary to reduceand overcome organizational defenses is the same kind of‘‘tough reasoning’’ that underlies the effective use of ideas in strategy, Žnance,marketing,manufacturing, and other management disciplines. Any sophisticated strategic analysis,for example, depends on collecting valid data, analyzing it carefully, and constantly testingthe inferences drawn from the data. The toughest tests are reserved for the conclusions.Good strategists make sure that their conclusions can withstand all kinds of critical ques-tioning.So too with productive reasoning about human behavior. The standard of analysis isjust as high. Human resource programs no longer need to be based on ‘‘soft’’ reasoningbut should be as analytical and as data-driven as any other management discipline.Of course, that is not the kind of reasoning the consultants used when they encoun-tered problems that were embarrassing or threatening. The data they collected was hardlyobjective. The inferences they made rarely became explicit. The conclusions they reachedwere largely self-serving, impossible for others to test, and as a result, ‘‘self-sealing,’’impervious to change.How can an organization begin to turn this situation around, to teach its membershow to reason productively? The Žrst step is for managers at the top to examine criticallyand change their own theories-in-use. Until senior managers become aware of how theyreason defensively and the counterproductive consequences that result, there will be littlereal progress. Any change activity is likely to be just a fad.Change has to start at the top because otherwise defensive senior managers are likelyto disown any transformation in reasoning patterns coming from below. If professionalsor middle managers begin to change the way they reason and act, such changes are likelyto appear strange—if not actually dangerous—to those at the top. The result is anunstablesituation where senior managers still believe that it is a sign of caring and sensitivity tobypass and cover up difŽcult issues, while their subordinates see the very same actionsas defensive.The key to any educational experience designed to teach senior managers how toreason productively is to connect the program to real business problems. The best dem-onstration of the usefulness of productive reasoning is for busy managers to see how itcan make a direct difference in their own performance and in that of the organization.This will not happen overnight. Managers need plenty of opportunity to practice the newskills. But once they grasp the powerful impact that productive reasoning can have onactual performance, they will have a strong incentive to reason productively not just in atraining session but in all their work relationships.One simple approach I have used to get this process started is to have participantsproduce a kind of rudimentary case study. The subject is a real business problem that themanager either wants to deal with or has tried unsuccessfully to address in the past.Writing the actual case usually takes less than an hour. But then the case becomes thefocal point of an extended analysis.For example, a CEO at a large organizational-development consulting company waspreoccupied with the problems caused by the intense competition among the various
Until senior managers becomeaware of the ways they reasondefensively, any change activityis likely to be just a fad.

Volume 4, Number 2, REFLECTIONS
Teachin
gSmart
People
Howto
Learn
·ARG
YRIS
12
business functions represented by his four direct reports. Not only was he tired of havingthe problems dumped in his lap, but he was also worried about the impact the interfunc-tional conicts were having on the organization’s exibility. He had even calculated thatthe money being spent to iron out disagreements amounted to hundreds of thousands ofdollars every year. And the more Žghts there were, the more defensive people became,which only increased the costs to the organization.In a paragraph or so, the CEO described a meeting he intended to have with his directreports to address the problem. Next, he divided the paper in half, and on the right-handside of the page, he wrote a scenario for the meeting—much like the script for a movieor play—describing what he would say and how his subordinates would likely respond.On the left-hand side of the page, he wrote down any thoughts and feelings that he wouldbe likely to have during the meeting but that he wouldn’t express for fear they wouldderail the discussion.But instead of holding the meeting, the CEO analyzed this scenario with his directreports. The case became the catalyst for a discussion in which the CEO learned severalthings about the way he acted with his management team.He discovered that his four direct reports often perceived his conversations as coun-terproductive. In the guise of being ‘‘diplomatic,’’ he would pretend that a consensusabout the problem existed, when in fact none existed The unintended result: instead offeeling reassured, his subordinates felt wary and tried to Žgure out ‘‘what is he reallygetting at.’’The CEO also realized that the way he dealt with the competitiveness among depart-ment heads was completely contradictory. On the one hand, he kept urging them to ‘‘thinkof the organization as a whole.’’ On the other, he kept calling for actions—departmentbudget cuts, for example—that placed them directly in competition with each other.Finally, the CEO discovered that many of the tacit evaluations and attributions hehad listed turned out to be wrong. Since he had never expressed these assumptions, hehad never found out just how wrong they were. What’s more, he learned that much ofwhat he thought he was hiding came through to his subordinates anyway—but with theadded message that the boss was covering up.The CEO’s colleagues also learned about their own ineffective behavior. They learnedby examining their own behavior as they tried to help the CEO analyze his case. Theyalso learned by writing and analyzing cases of their own. They began to see that they tootended to bypass and cover up the real issues and that the CEO was often aware of it butdid not say so. They too made inaccurate attributions and evaluations that they did notexpress. Moreover, the belief that they had to hide important ideas and feelings from theCEO and from each other in order not to upset anyone turned out to be mistaken. In thecontext of the case discussions, the entire senior management team was quite willing todiscuss what had always been undiscussable.In effect, the case study exercise legitimizes talking about issues that people havenever been able to address before. Such a discussion can be emotional—even painful.But for managers with the courage to persist, the payoff is great: management teams andentire organizations work more openly and more effectively and have greater options forbehaving exibly and adapting to particular situations.When senior managers are trained in new reasoningskills, they can have a big impact on the performance ofthe entire organization—even when other employees arestill reasoning defensively. The CEO who led themeetingson the performance-evaluation procedure was able to de-fuse dissatisfaction because he didn’t respond to profes-sionals’ criticisms in kind but instead gave a clearpresentation of relevant data. Indeed, most participantstook the CEO’s behavior to be a sign that the companyreally acted on the values of participation and employeeinvolvement that it espoused.Of course, the ideal is for all the members of an or-ganization to learn how to reason productively. This hashappened at the company where the case team meetingtook place. Consultants and their managers are now able©EmilySpe
r

REFLECTIONS, Volume 4, Number 2
Teachin
gSmart
People
Howto
Learn
·ARG
YRIS
13
to confront some of the most difŽcult issues of the consultant-client relationship. To geta sense of the difference productive reasoning can make, imagine how the original con-versation between the manager and case team might have gone had everyone engaged ineffective reasoning. (The following dialogue is based on actual sessions I have attendedwith other case teams at the same company since the training has been completed.)First, the consultants would have demonstrated their commitment to continuous im-provement by being willing to examine their own role in the difŽculties that arose duringthe consulting project. No doubt they would have identiŽed their managers and the clientsas part of the problem, but they would have gone on to admit that they had contributedto it as well. More important, they would have agreed with the manager that as theyexplored the various roles of clients, managers, and professionals, they would make sureto test any evaluations or attributions they might make against the data. Each individualwould have encouraged the others to question his or her reasoning. Indeed, they wouldhave insisted on it. And in turn, everyone would have understood that act of questioningnot as a sign of mistrust or an invasion of privacy but as a valuable opportunity forlearning.The conversation about the manager’s unwillingness to say no might look somethinglike this:Professional #1: ‘‘One of the biggest problems I had with the way you managed thiscase was that you seemed to be unable to say no when either the client or your superiormade unfair demands.’’ [Gives an example.]Professional #2: ‘‘I have another example to add. [Describes a second example.] ButI’d also like to say that we never really told you how we felt about this. Behind your backwe were bad-mouthing you—you know, ‘he’s being such a wimp’—but we never cameright out and said it.’’Manager: ‘‘It certainly would have been helpful if you had said something. Was thereanything I said or did that gave you the idea that you had better not raise this with me?’’Professional #3: ‘‘Not really. I think we didn’t want to sound like we were whining.’’Manager: ‘‘Well, I certainly don’t think you sound like you’re whining. But twothoughts come to mind. If I understand you correctly, you were complaining, but thecomplaining about me and my inability to say no was covered up. Second, if we haddiscussed this, I might have gotten the data I needed to be able to say no.’’Notice that when the second professional describes how the consultants had coveredup their complaints, the manager doesn’t criticize her. Rather, he rewards her for beingopen by responding in kind. He focuses on the ways that he too may have contributed tothe cover-up. Reecting undefensively about his own role in the problem then makes itpossible for the professionals to talk about their fears of appearing to be whining. Themanager then agrees with the professionals that they shouldn’t become complainers. Atthe same time, he points out the counterproductive consequences of covering up theircomplaints.Another unresolved issue in the case teammeeting concerned the supposed arroganceof the clients. A more productive conversation about that problem might go like this:Manager: ‘‘You said that the clients were arrogant and uncooperative. What did theysay and do?’’Professional #1: ‘‘One asked me if I had ever met a payroll. Another asked how longI’ve been out of school.’’Professional #2: ‘‘One even asked me how old I was!’’Professional #3: ‘‘That’s nothing. The worst is when they say that all we do is inter-view people, write a report based on what they tell us, and then collect our fees.’’Manager: ‘‘The fact that we tend to be so young is a real problem for many of ourclients. They get very defensive about it. But I’d like to explore whether there is a wayfor them to freely express their views without our getting defensive.‘‘What troubled me about your original responses was that you assumed you wereright in calling the clients stupid. One thing I’ve noticed about consultants—in this com-pany and others—is that we tend to defend ourselves by bad-mouthing the client.’’Professional #1: ‘‘Right. After all, if they are genuinely stupid, then it’s obviously notour fault that they aren’t getting it!’’Professional #2: ‘‘Of course, that stance is antilearning and overprotective. By assum-ing that they can’t learn, we absolve ourselves from having to.’’

Volume 4, Number 2, REFLECTIONS
Teachin
gSmart
People
Howto
Learn
·ARG
YRIS
14
Professional #3: ‘‘And the more we all go along with the bad-mouthing, the more wereinforce each other’s defensiveness.’’Manager: ‘‘So what’s the alternative? How can we encourage our clients to expresstheir defensiveness and at the same time constructively build on it?’’Professional #1: ‘‘We all know that the real issue isn’t our age; it’s whether or not weare able to add value to the client’s organization. They should judge us by what weproduce. And if we aren’t adding value, they should get rid of us—no matter how youngor old we happen to be.’’Manager: ‘‘Perhaps that is exactly what we should tell them.’’In both these examples, the consultants and their manager are doing real work. Theyare learning about their own group dynamics and addressing some generic problems inclient-consultant relationships. The insights they gain will allow them to act more effec-tively in the future—both as individuals and as a team. They are not just solving problemsbut developing a far deeper and more textured understanding of their role as members ofthe organization. They are laying the groundwork for continuous improvement that istruly continuous. They are learning how to learn.Commentaryby Chris ArgyrisWe are inundated with examples of the defensive reasoning and organizational defensive routinesthat this article describes. There is Enron, Arthur Andersen, the CIA, the FBI, the Catholic hierarchy,and the administration of school systems.Several questions come to mind that I think are relevant. Enron and Andersen have receivedawards for leadership and enlightened human resource programs. In both organizations, the topmanagement genuinely championed these programs. Yet the same top executives violated the ten-ets of the old programs. Do we have theories that will not only explain the ip-ops, but predictwhen they will occur and how to prevent them?How is it that the church hierarchies, which espouse trust and honesty, skillfully producecover-ups and cover-up of the cover-ups? How do we explain that educational leaders cover upteacher incompetence by giving answers to students so that they can pass proŽciency tests?Finally, how do we explain that the ‘‘local’’ levels of the FBI feel free to assign responsibilityfor problems, when later they admit that they create the same problems at the local level andcovered up that this is the case?Another question focuses on the fact that most organizational change programs are based onemulating best practices. The Achilles heel of this strategy is that what creates the best practicescan also harbor processes that eventually bring them down. For example, 3M has for years beentouted as an innovative company. We now learn that this is no longer the case. How did this hap-pen? Could the deterioration have been avoided?Finally, there is the question about our competence to produce the claims that we espouse. Ina recent inquiry, I found that many professionals, when being challenged, respond with the samebehaviors for which they criticized line managers. Moreover, many of the programs were not im-plementable, and their creators appeared skillfully unaware of the inconsistencies (Argyris, 2000).ReferenceArgyris, C. Flawed Advice and the Management Trap (New York: Oxford University Press, 2000).Commentaryby Haridimos TsoukasVulnerability, Moral Responsibility, and Reexive ThinkingWhen Chris Argyris published his now classic article, the terms ‘‘knowledge work’’ and the‘‘knowledge-based organization’’ had not yet fully entered public discourse to the extent that theyhave today. Argyris, however, was prescient enough to realize that the kind of ‘‘smart people’’ hewas writing about were not the exception: in the advanced economies, they were becoming the

REFLECTIONS, Volume 4, Number 2
Teachin
gSmart
People
Howto
Learn
·ARG
YRIS
15
norm. Knowledge workers, or symbolic analysts, have been the fastest growing group of employees(at least, in the US; Barley, 1996).Argyris’s argument about the inherent difŽculty smart people have in engaging in double-loop(or reexive) learning is particularly relevant today, as we are moving from the classic Weberianbureaucracy to post-bureaucratic forms of organization or, to use a different language, from themodern to the post-modern Žrm (Heckscher and Donnellon, 1994). Here is, I think, how Argyris’sinsights enrich those of other organizational psychologists, such as Larry Hirschhorn (1997), and ofsociologists, such as Shoshana Zuboff (1988) and Anthony Giddens (1991). For Hirschhorn, what isparticularly distinctive in the shift to the post-modern organization is a change in the very conceptof personhood: individuals increasingly rely less on internalizing organizational authority as repre-sented by the boss; instead, they rely more on internalized images of themselves, on their ownpersonal authority.What this means is that, in the post-modern organization, individuals bring more of them-selves (their ideas, their feelings) to their work. In Hirschhorn’s (1997: 9) terms, ‘‘they are morepsychologically present.’’ A post-modern Žrm—a Žrm that is rich in information and relies heavilyon the daily choices of its knowledgeable employees—is a place that, unlike the modern Žrm,thrives on doubt and challenge. As organizational ethnographers, such as Julian Orr (1996) andEtienne Wenger (1998), have shown, daily work in information-rich companies is more decisionintensive—more loci for decision making by employees are created. The more ‘‘informated’’ (to useZuboff’s term) a workplace is, the more decisions employees will have to make. Or, to put it differ-ently, the more informated a workplace is, the more reexive the organization is capable of be-coming (what Giddens calls ‘‘institutional reexivity’’); it has the opportunity to feed back, andreect on, the information about its modus operandi and the outcomes it brings about.In such organizations, individuals need to be able to ask critical questions of others and ofthemselves if they are to be effective in fully reaping the potential beneŽts reexivity brings about.Individuals, therefore, no longer need to uphold the ‘‘masculine ideal’’—that is, to suppress doubtand ambivalence. On the contrary, doubt, debate, and reexivity are the very qualities needed topromote learning. A knowledge-intensive workplace thrives on the exchange of ideas and experi-ences in the interest of enhancing the collective pool of knowledge and of generating new ideas.But, as we know from academic life, for ideas to ourish, debate is needed; hence, the importanceof criticism, learning, and reexivity.Throughout his work, Argyris has pointed out the difŽculties practitioners have in engaging inreexive thinking—in his terms, in ‘‘double-loop thinking.’’ This is particularly so in the case ofknowledge workers because, to the extent they are more psychologically present at work, they ex-pose more of themselves to others; hence, they are more vulnerable. Argyris documents this vul-nerability in his article, showing the defensive reasoning it brings out in knowledge workers. Morethan that, however, he shows what individuals need to do in order to stop being defensive whenthe spotlight is turned on themselves—how to engage in productive reasoning. The message Ar-gyris is getting across, it seems to me, is not only how productive reasoning may be achieved but,also, the importance of constantly challenging yourself, of expanding your horizons, of ‘‘knowingthyself.’’In other words, Argyris invites knowledge workers to undertake a primarily moral, not justtechnical, task: to be open to criticism, to be willing to test their claims publicly against evidence,to accept that they too are partly responsible for the problems they are confronted with. The clientmay or may not be ‘‘stupid,’’ but the real question, if a consultant is really keen on learning, is‘‘what can I do to improve the relationship with the client (or my boss, or anyone else)?’’ It allcomes down to individual responsibility, and this is, essentially, a moral issue. In that sense, as wellas being an inuential organizational psychologist and an implicit moral philosopher, Argyris is asystemic theorist, not too different from his own hero Gregory Bateson (1979): we partly createthe problems we face, he says, and we have a responsibility for this. An excellent point.ReferencesBarley, S. The New World of Work (London: British-North American Committee, 1996): 59.Bateson, G. Mind and Nature (Toronto: Bantam, 1979).Giddens, A. Modernity and Self-Identity (Cambridge, UK: Polity Press, 1991).Heckscher, C. and A. Donnellon. The Post-Bureaucratic Organization, (London: Sage, 1994).Hirschhorn, L. Reworking Authority (Cambridge, MA: MIT Press, 1997).Orr, J. Talking About Machines (Ithaca, NY: ILR Press/Cornell University Press, 1996).Wenger, E. Communities of Practice (Cambridge: Cambridge University Press, 1998).Zuboff, S. In the Age of the Smart Machine (Oxford: Heinemann, 1988).
Haridimos TsoukasGeorge D. Mavros ResearchProfessor in Organization andManagementALBA, GreeceProfessor of Organization Theoryand BehaviorUniversity of StrathclydeGraduate School of Business, UKhtsoukas@alba.edu.gr