SoL Public Home
Kurt Lewin's Change Theory in the Field and in the Classroom: Notes Toward a Model of Managed Learning

Edgar H. Schein / Professor of Management Emeritus MIT Sloan School of Management

Table of Contents

I. "There is Nothing So Practical as a Good Theory:" Lewin's Change Model Elaborated
1. Disconfirmation
2. Induction of Guilt or Survival Anxiety
3. Creation of Psychological Safety or Overcoming of Learning Anxiety
4. Cognitive Redefinition
5. Imitation and Positive or Defensive Identification with a Role Model
6. Scanning: Insight or Trial and Error Learning
7. Personal and Relational Refreezing
II. "You Cannot Understand a System Until You Try to Change It:" Process Consultation and Clinical Research
III. Kurt Lewin in The Classroom: Teaching the Management of Planned Change
The MIT One Semester Course on Managing Planned Change
A. My Multiple Roles
Monitoring and Grading
Consulting and Coaching
The Empathy Walk
Project Reviews and Final Reports
B. The Conceptual Core of the Course: Diagnosis as Initial Intervention and Process Consultation as a Change Strategy
Summary and Conclusions

Few people have had as profound an impact on the theory and practice of social and organizational psychology as Kurt Lewin. Though I never knew him personally I was fortunate during my graduate school years at Harvard's Social Relations Dept. in 1949-50 to have been exposed to Alex Bavelas and Douglas McGregor, who, in my mind embodied Lewin's spirit totally. As I will try to show in this essay, Lewin's spirit and the assumptions that lay behind it are deeply embedded in my own work and that of many of my colleagues who practice the art of "Organization Development." This essay will attempt to spell out some of Lewin's basic dictums and show their influence in my own and others' contemporary work. [2] I will endeavor to show how my own thinking has evolved from theorizing about "planned change" to thinking about such processes more as "managed learning."

I. "There is Nothing So Practical as a Good Theory:" Lewin's Change Model Elaborated

The power of Lewin's theorizing lay not in a formal propositional kind of theory but in his ability to build "models" of processes that drew attention to the right kinds of variables that needed to be conceptualized and observed. In my opinion, the most powerful of these was his model of the change process in human systems. I found this model to be fundamentally necessary in trying to explain various phenomena I had observed, and I found that it lent itself very well to refinement and elaboration.

My own early work in clinical/social psychology dealt with the attitude changes that had occurred in military and civilian prisoners of the Chinese Communists during the Korean war (Schein, 1956,1961,1968). 1 found contemporary theories of attitude change to be trivial and superficial when applied to some of the profound changes that the prisoners had undergone, but I found Lewin's basic change model of unfreezing, changing, and refreezing to be a theoretical foundation upon which change theory could be built solidly. The key, of course, was to see that human change, whether at the individual or group level, was a profound psychological dynamic process that involved painful unlearning without loss of ego identity and difficult relearning as one cognitively attempted to restructure one's thoughts, perceptions, feelings, and attitudes.

Unfreezing as a concept entered the change literature early to highlight the observation that the stability of human behavior was based on "quasi- stationary equilibria" supported by a large force field of driving and restraining forces. For change to occur, this force field had to be altered under complex psychological conditions because, as was often noted, just adding a driving force toward change often produced an immediate counterforce to maintain the equilibrium. This observation led to the important insight that the equilibrium could more easily be moved if one could remove restraining forces since there were usually already driving forces in the system. Unfortunately restraining forces were harder to get at because they were often personal psychological defenses or group norms embedded in the organizational or community culture.

The full ramifications of such restraining forces were only understood after decades of frustrating encounters with resistance to change, and only then did we begin to pay attention to the work of cognitive psychologists on perceptual defenses, to what psychoanalysts and the Tavistock group were trying to show us with their work on denial, splitting and projection, and to Argyris's seminal work on defensive routines (e.g. Argyris, 1990; Hirschhorn, 1988). In trying to explain what happened to POWs I was led to the necessity to further "unpack" the concept of unfreezing and to highlight what really goes on there. Unfreezing is basically three processes, each of which has to be present to some degree for readiness and motivation to change to be generated.

1. Disconfirmation

It is my belief that all forms of learning and change start with some form of dissatisfaction or frustration generated by data that disconfirm our expectations or hopes. Whether we are talking about adaptation to some new environmental circumstances that thwart the satisfaction of some need, or whether we are talking about genuinely creative and generative learning of the kind Peter Senge focuses on, some disequilibrium based on disconfirming information is a pre-requisite (Senge, 1990). Disconfirmation, whatever its source, functions as a primary driving force in the quasi-stationary equilibrium.

Disconfirming information is not enough, however, because we can ignore the information, dismiss it as irrelevant, blame the undesired outcome on others or fate, or, as is most common, simply deny its validity. In order to become motivated to change, we must accept the information and connect it to something we care about. The disconfirmation must arouse what we can call "survival anxiety" or the feeling that if we do not change we will fail to meet our needs or fail to achieve some goals or ideals that we have set for ourselves ("survival guilt"). [3]

[Back to Table of Contents]

2. Induction of Guilt or Survival Anxiety

In order to feel survival anxiety or guilt, we must accept the disconfirming data as valid and relevant. What typically prevents us from doing so, what causes us to react defensively, is a second kind of anxiety which we can call "learning anxiety," or the feeling that if we allow ourselves to enter a learning or change process, if we admit to ourselves and others that something is wrong or imperfect, we will lose our effectiveness, our self-esteem and maybe even our identity. Most humans need to assume that they are doing their best at all times, and it may be a real loss of face to accept and even "embrace" errors (Michael, 1973, 1993). Adapting poorly or failing to meet our creative potential often looks more desirable than risking failure and loss of self-esteem in the learning process. Learning anxiety is the fundamental restraining force which can go up in direct proportion to the amount of disconfirmation, leading to the maintenance of the equilibrium by defensive avoidance of the disconfirming information. It is the dealing with learning anxiety, then, that is the key to producing change, and Lewin understood this better than anyone. His involving of workers on the pajama assembly line, his helping the housewives groups to identify their fear of being seen as less "good" in the community if they used the new proposed meats and his helping them to evolve new norms, was a direct attempt to deal with learning anxiety. This process can be conceptualized in its own right as creating for the learner some degree of "psychological safety."

[Back to Table of Contents]

3. Creation of Psychological Safety or Overcoming of Learning Anxiety

My basic argument is that unless sufficient psychological safety is created, the disconfirming information will be denied or in other ways defended against, no survival anxiety will be felt, and, consequently, no change will take place. The key to effective change management, then, becomes the ability to balance the amount of threat produced by disconfirming data with enough psychological safety to allow the change target to accept the information, feel the survival anxiety, and become motivated to change.

The true artistry of change management lies in the various kinds of tactics that change agents employ to create psychological safety. For example, working in groups, creating parallel systems that allow some relief from day to day work pressures, providing practice fields in which errors are embraced rather than feared, providing positive visions to encourage the learner, breaking the learning process into manageable steps, providing on-line coaching and help all serve the function of reducing learning anxiety and thus creating genuine motivation to learn and change.

Unfortunately, motivation is not enough. A theory or model of change must also explain the actual learning and change mechanisms, and here Lewin's cognitive models were also very helpful in providing a theoretical base.

[Back to Table of Contents]

4. Cognitive Redefinition

By what means does a motivated learner learn something new when we are dealing with thought processes, feelings, values, and attitudes? Fundamentally it is a process of "cognitive restructuring," which has been labeled by many others as frame braking or reframing. It occurs by taking in new information that has one or more of the following impacts: 1 ) semantic redefinition--we learn that words can mean something different from what we had assumed; 2) cognitive broadening--we learn that a given concept can be much more broadly interpreted than what we had assumed; and 3) new standards of judgment or evaluation--we learn that the anchors we used for judgment and comparison are not absolute, and if we use a different anchor our scale of judgment shifts.

An example will make this clear. The concept of "teamwork" is today highly touted in organizational circles, yet the evidence for effective team work is at best minimal. The problem lies in the fact that in the U.S., the cultural assumption that society revolves around the individual and individual rights is so deeply embedded that when teamwork is advocated we pay lipservice but basically do not change our individualistic assumption. How then does change in this area come about? First, we would need to re-define teamwork as the coordination of individual activities for pragmatic ends, not the subordination of the individual to the group. If we define teamwork as individual subordination, as treating the group to be more important than the individual, we arouse all the defenses that lead to quips like camels being horses constructed by a committee, negative images of "group think," lynch mobs, etc.

Second, the redefinition of teamwork also allows one to redefine individualism in a way that preserves its primacy, not to "substitute" groupism for individualism. This process of redefinition in effect enlarges the concept of individualism to include the ability and obligation to work with others when the task demands it. In other words, helping a team to win is not inconsistent with individualism. And, third, one can change the standards by which individual performance is rewarded. Instead of rewarding "rugged individualism" or the competitive winning out over others (which makes collaborative behavior look "weak"), individuals can be increasingly rewarded for their ability to create, lead, and participate in teams (which makes collaborative behavior look "strong"). The best individual, then, is the one who can be an effective team player. What Lewin did with the housewives, was to help them to change their standard of what was an acceptable meat, so that kidneys, liver, etc. became cognitively redefined as acceptable to buy and serve. This process is fundamental to any change if one wants it to last.

The new information that makes any or all of these processes possible comes into us by one of two fundamental mechanisms--1 ) learning through positive or defensive identification with some available positive or negative role model, or 2) learning through a trial and error process based on scanning the environment for new concepts (Schein, 1968).

[Back to Table of Contents]

5. Imitation and Positive or Defensive Identification with a Role Model

Cognitive re-definition occurs when the learner has become unfrozen, i.e. motivated to change, and has, therefore opened him or herself up to new information. The next question to address, then. is how the new information comes to the learner. The most basic mechanism of acquiring new information that leads to cognitive restructuring is to discover in a conversational process that the interpretation that someone else puts on a concept is different from one's own. If one is motivated to change, i.e. if the factors described above have been operating, one may be able to "hear" or "see" something from a new perspective.

The best examples come from what has colloquially been labeled "brainwashing," where POWs who were judged "guilty" yet felt innocent, finally were able to admit their guilt when they could identify with their more advanced cell mates sufficiently to realize that the concepts of "crime" and "guilt" were defined differently by the Chinese communists. One was guilty because a crime was defined as "any action that could be harmful to the communists" even if no harm had occurred. A postcard to home, could conceivably contain information that would help the enemy, so sending the postcard was an act of espionage and the sender had to learn to appreciate and confess his or her guilt. Being born into the wrong social class was a crime because middle class attitudes could be very harmful to the communist cause. Semantic redefinition, cognitive broadening and changing standards of judgment were all present in this process.

Only by recognizing this potential for harm, confessing one's guilt, and acknowledging the incorrectness of one's social origins could one hope to learn how to be a good communist or to be released from jail. Once one had accepted the new cognitive frame of reference and learned the new definitions and standards, one could make rapid progress in re-education and remove the heavy disconfirming pressure. The key to the whole process, however, was to identify psychologically with other prisoners who had already made the cognitive shift and learning to see the world through their eyes.

Readers who are familiar with socialization processes in families, schools, companies, religious movements, and other organizational settings will readily recognize this mechanism as the key to apprenticeships, to "big brother" programs, to the concept of "mentoring" and to the various more formal group based indoctrination programs that organizations use. The mentor or big brother is often both a source of psychological safety and the role model to facilitate cognitive redefinition (Schein, 1968; Van Maanen & Schein, 1979)

Defensive identification is a rarer process that occurs when the learner is a captive in a hostile environment in which the most salient role models are the hostile captors, e.g. prison guards, authoritarian bosses or teachers, etc. The process was first described in relation to Nazi Concentration Camps where some prisoners took on the values and beliefs of the guards and maltreated fellow prisoners. In the face of severe survival anxiety, for some learners "identification with the aggressor" was the only solution (Bettelheim, 1943). Genuine new learning and change occurred, but, of course, in a direction deemed undesirable by others. In considering such outcomes one is reminded that unfreezing creates motivation to learn, but does not necessarily control or predict the direction of learning. If the only new information available is from salient and powerful role models, learning will occur in that direction. One of the key elements of a managed change process is, therefore, what kind of role models one makes available to the learners once they are unfrozen.

If either no good role models are available, or one wants the learning to be more genuinely creative one has to create the conditions for what I call "Scanning."

[Back to Table of Contents]

6. Scanning: Insight or Trial and Error Learning

A learner or change target can be highly motivated to learn something, yet have no role models nor initial feeling for where the answer or solution might lie. The learner then searches or scans by reading, traveling, talking to people, hiring consultants, entering therapy, going back to school, etc. to expose him or herself to a variety of new information that might reveal a solution to the problem. Alternatively, when the learner finally feels psychologically safe, he or she may experience spontaneously an insight that spells out the solution. Change agents such as process consultants or non-directive therapists count on such insights because of the assumption that the best and most stable solution will be one that the learner has invented for him or herself.

Once some cognitive redefinition has taken place, the new mental categories are tested with new behavior which leads to a period of trial and error and either reinforces the new categories or starts a new cycle of disconfirmation and search. Note that in the process of search, if role models are readily available, they will most likely be used. Identification is thus an efficient and fast process, but it may lead to solutions that do not stick because they do not fit the learner's total personality. If one wants to avoid that, one must create learning environments that do not display role models, thereby forcing the learner to scan and invent his or her own solutions.

It is this dynamic, to rely on identification with a role model, that explains why so many consultation processes go awry. The consultant, by design or unwittingly, becomes a role model and generates solutions and cognitive categories that do not really fit into the culture of the client organization and will therefore only be adopted temporarily. A similar result occurs when organizations attempt to check on their own performance by "benchmarking," i.e. comparing themselves to a reference group of organizations and attempting to identify "best practices." The speed and simplicity of that process is offset by two dangers. First, it may be that none of the organizations in the reference set have scanned for a good solution so the whole set continues to operate sub- optimally, or, second, that the identified best practice works only in certain kinds of organizational cultures and will fail in the particular organization that is trying to improve itself. In other words, learners can attempt to learn things that will not survive because they do not fit the personality or culture of the learning system. For change to remain more stable it must be "refrozen."

[Back to Table of Contents]

7. Personal and Relational Refreezing

The main point about refreezing is that new behavior must be to some degree congruent with the rest of the behavior and personality of the learner or it will simply set off new rounds of disconfirmation that often lead to unlearning the very thing one has learned. The classic case is the supervisory program that teaches individual supervisors how to empower employees and then sends them back into an organization where the culture supports only autocratic supervisory behavior. Or, in Lewin's classic studies, the attempt to change eating habits by using an educational program that teaches housewives how to use meats such as liver and kidneys and then sends them back into a community in which the norms are that only poor folks who can't afford good meat would use such poor meat.

The implication for change programs are clear. For personal refreezing to occur, it is best to avoid identification and encourage scanning so that the learner will pick solutions that fit him or her. For relational refreezing to occur, it is best to train the entire group that holds the norms that support the old behavior. It is only when housewives groups met and were encouraged to reveal their implicit norms that change was possible by changing the norms themselves, i.e. introducing collectively a new set of standards for judging what was"ok" meat.

In summary, what I have tried to show above is that Lewin's basic model of change leads to a whole range of insights and new concepts that enrich change theory and make change dynamics more understandable and manageable. It is a model upon which I have been able to build further because its fundamental concepts were anchored in empirical reality. Intellectual knowledge of the change process is not the same as the know-how or skills that are learned in actually producing change. In the next section I examine the implication of Lewin's thinking for the practice of change management.

[Back to Table of Contents]

II. "You Cannot Understand a System Until You Try to Change It:" Process Consultation and Clinical Research

The change and consulting literature is filled with the notion that one first diagnoses a system and then intervenes to change it. I learned early in my own consulting career that this basic model perpetuates a fundamental error in thinking, an error that Lewin learned to avoid in his own change projects and that led him to the seminal concept of "action research." The conceptual error is to separate the notion of diagnosis from the notion of intervention. That distinction comes to us from scientific endeavors where a greater separation exists between the researcher and the researched, particularly from medicine where the physical processes are assumed to be somewhat independent of the psychological processes (an assumption that is not even holding up in many parts of medicine).

The classical model is that the doctor makes an examination, runs certain tests, decides what is wrong, and writes a prescription which includes recommendations for therapy or, if necessary, for other interventions such as surgery. The consulting industry has perpetuated this model by proposing as a major part of most projects a diagnostic phase in which large numbers of interviews, questionnaires, and observations are made the basis of a set of recommendations given to the client. Consultants differ on whether they feel they should also be accountable for the implementation of the recommendations, but they tend to agree that there is a discrete billable period in any project that is basically considered necessary--namely a diagnosis of the problem--and that the consultant's basic job is done with a set of recommendations "for future intervention." If interviews or surveys are done, the attempt is made to be as scientifically objective as possible in gathering the data and to interfere minimally during this phase with the operation of the organization. What is wrong with this picture?

If Lewin was correct that one cannot understand an organization without trying to change it, how is it possible to make an adequate diagnosis without intervening? So either consultants using the classical model are getting an incorrect picture of the organization, or they are intervening but are denying it by labeling it "just diagnosis." Isn't a better initial model of work with organizations something like the stress test that the cardiologist performs by putting the heart under pressure to see how it will perform, even knowing that there are some risks and that some people have been hurt during the test itself? This risk forces the diagnostician to think about the nature of the "diagnostic intervention" and to apply clinical criteria for what is safe, rather than purely scientific criteria of what would seemingly give the most definitive answer.

It is my contention that Lewin was correct and that we must all approach our consulting work from a clinical perspective that starts with the assumption that everything we do with a client system is an intervention, and that, unless we intervene, we will not learn what some of the essential dynamics of the system really are. If we start from that assumption, we need to develop criteria that balance the amount of information gained from an intervention with the amount of risk to the client from making that intervention. In other words, if the consultant is going to interview all the members of top management, he or she must ask whether the amount of information gained will be worth the risk of perturbing the system by interviewing everybody, and, if the answer is "yes," must make a further determination of what is to be learned from the reactions of the management to being interviewed. That is, the interview process itself will change the system and the nature of that change will provide some of the most important data about how the system works, i.e. will respondents be paranoid and mistrusting, open and helpful, supportive of each other or hostile in their comments about each other, cooperative or aloof, and so on. The best information about the dynamics of the organization will be how the organization deals with the consultant, because his or her very presence is de facto an intervention.

Yet the focus in many traditional consultation models is on the "objective data obtained in the interview" with nary a reference to how the interviewer felt about the process and what could be inferred from the way he or she was received. The irony in all of this is that Lewin was by training a physicist and knew very well the rules of scientific inquiry and objectivity. For him to have discovered that human systems cannot be treated with that level of objectivity is, therefore, an important insight that is all too often ignored in our change and consultation literature.

In actual practice what most change agents have learned from their own experience is that "diagnostic" activities such as observations, interviews, and questionnaires are already powerful interventions and that the process of learning about a system and changing that system are, in fact, one and the same. This insight has many ramifications, particularly for the ethics of research and consulting. Too many researchers and consultants assume that they can "objectively" gather data and arrive at a diagnosis without having already changed the system. In fact, the very method of gathering data influences the system and, therefore, must be considered carefully. For example, asking someone in a questionnaire how they feel about their boss gets the respondent thinking about an issue that he or she might not have focused on previously and it might get them talking to others about the question in a way that would create a common attitude that was not there before.

The concept of process consultation as a mode of inquiry grew out of my insight that to be helpful one had to learn enough about the system to understand where it needed help and that this required a period of very low key inquiry oriented diagnostic interventions designed to have a minimal impact on the processes being inquired about (Schein, 1969,1987,1988). Process consultation as a philosophy acknowledges that the consultant is not an expert on anything but how to be helpful, and starts with total ignorance of what is actually going on in the client system. One of the skills, then, of process consulting is to "access one's ignorance," to let go of the expert or doctor role, and get attuned to the client system as much as possible. Only when one has genuinely understood the problem and what kind of help is needed, can one even begin to think about recommendations and prescriptions, and even then it is likely that they will not fit the client system's culture and will, therefore, not be refrozen even if initially adopted. Instead, a better model of help is to start out with the intention of creating in insider/outsider team that is responsible for diagnostic interventions and all subsequent interventions. When the consultant and the client have joint ownership of the change process, both the validity of the diagnostic interventions and the subsequent change interventions will be greatly enhanced.

The flow of a change or managed learning process then is one of continuous diagnosis as one is continuously intervening. The consultant must become highly attuned to his or her own insights into what is going on and his or her own impact on the client system. Stage models which emphasize up front contracting do not deal adequately with the reality that the psychological contract is a constantly evolving one and that the degree to which it needs to be formalized depends very much on the culture of the organization.

In summary, Lewin's concept of action research is absolutely fundamental to any model of working with human systems, and such action research must be viewed from a clinical perspective as a set of interventions that must be guided primarily by their presumed impact on the client system. The immediate implication of this is that in training consultants and change agents one should put much more emphasis on the clinical criteria of how different interventions will affect client systems than on the canons of how to gather scientifically valid information. Graduate students should be sent into field internships as participant observers and helpers before they are taught all the canons of how to gather and analyze data. Both are necessary, but the order of priority is backward in most training programs.

What can be done to enhance an understanding of these models and to begin to build the necessary skills to implement them? We turn next to an experimental course that attempts to teach "the management of planned change."

[Back to Table of Contents]

III. Kurt Lewin in The Classroom: Teaching the Management of Planned Change

The idea for a "planned change workshop" goes back to the mid 1960's when Richard Beckhard and I designed a program on "planned change" for the National Training Labs. The essence of our program was that participants should be involved in real projects which could be of one or two years duration, and that the time spent together should be devoted initially to learning diagnostic intervention tools and models and, thereafter, to reporting progress to each other. That program started with a one week workshop and was followed by quarterly meetings of three days duration. Participants were organized into teams geographically and were expected to meet regularly with each other to share problems and progress.

What Beckhard and I learned from this program is 1) to learn about managing change one must be involved in a real project, and 2) one of the most powerful sources of motivation to work through all the frustrations involved in managing change is to have to report regularly on progress to "team mates" and to the faculty. All of the participants noted during and after the program how important it had been to give quarterly progress reports, to have a chance at those times to rediagnose, to recalibrate their own situation and to share war stories and frustrations with others who were in the same boat.

Criteria for choosing the initial project were 1 ) something that the workshop participant was personally involved in and cared about; 2) something that would make a real contribution to the organization from which the participant came; and 3) something that was realistic in terms of being doable in the time allocated to the workshop, i.e. one or two years. We considered the workshop a success and felt we had learned what the essential components of such a learning experience had to be. But it was not until two decades later that I found a way to implement my own learning in the more traditional classroom environment.

[Back to Table of Contents]

1. The MIT One Semester Course on Managing Planned Change

In 1987 I decided to experiment with a version of the Beckhard/Schein model in the regular Masters curriculum of the MIT Sloan School. I offered a mini-course that ran for 10 weeks, three hours per week. Eventually it was expanded to a full 14 week long semester elective course for full academic credit. Enrollment in the first three years averaged around 25 students, but in the last year or so it caught on so I ended up in 1994 with three sections of 30 to 35 students each.

In the first session I emphasized that the core of the course was not the class time or reading, but two actual change projects--one personal and one focused on an organization and carried out by a group. The personal project asked each student to pick some personal change goal that he or she wanted to work on for the next 14 weeks. The first week's paper had to spell out the goals and the method that would be used to achieve them, including some system for appraising progress week by week. Each week a one page progress report had to be handed in to me detailing outcomes and any reactions or thoughts about the change process. These reports were private between me and each student and provided me an opportunity to react and coach, typically by asking questions and making suggestions. Reading 100 one page papers was time consuming but very engaging because each student was wrestling with real and personally meaningful issues--stopping smoking, losing weight, overcoming shyness, learning to talk more in large classes, improving relationship with spouse or a child, increasing reading speed, developing a more healthy balanced life style, overcoming chronic lateness, and so on.

The group projects were to be realistic efforts to make an organizational change somewhere in the MIT environment. At the opening session I collected data from the class on possible organizational change projects they might wish to undertake in small teams. Given that the project had to be completed in 14 weeks, we focused on organizations to which students had access already, which meant de facto that most of the projects were located in and around the MIT Sloan School.

We started with a brainstorming session on all kinds of things that could and/or should be changed around the school, followed by a joint critical analysis of what was feasible and worthwhile. My role in this was to provide a "sanity" or "reality" check on the ideas that were brought up. When we had a list of feasible projects we duplicated it and then, in the second class session, did a straw vote to see how many people were interested in which, to reduce the number down to roughly one-fourth the size of the class so that each team could consist of four or five students. Final choice of projects and signing on to the teams was the last step, usually accomplished by the third or fourth class session.

In the end I only required that each team had at least two people and no more than seven or eight. It was essential that each student picked a project that he or she was genuinely motivated to complete. This process stood in sharp contrast to what most other classes were offering as projects where students selected from pre-arranged topics, sites, or problems instead of having to wrestle with what they would personally actually commit themselves to. Lewin's insight about the importance of involving the learner were not lost here.

Once the teams were formed, they met weekly during and after the class sessions and were required to submit a weekly progress report on specific goals selected, diagnostic thinking about the project, action steps taken, and results. Sample projects that were undertaken were to revise the particular curriculum of a key course on strategy to make it more international, to resurrect the European Club and to improve its process of helping students find jobs in Europe, to improve the responsiveness of the career development office, to reduce the bureaucracy of the MIT housing office, to fix a leak in the bridge between two buildings that had been left alone for the past three years, to develop a student lounge, to redesign the form on which students gave feedback to faculty on their teaching, to increase the interaction between first and second year masters students, to increase the range of food offerings in the local student cafeteria, to create a lecture series that would expose students to some of the more prominent faculty at MIT, and so on.

[Back to Table of Contents]

My Multiple Roles

I served as the animator, teacher, monitor, coach and consultant. In the initial three hour session I provided the structure, the tasks, the rules, and the challenge. The bulk of the time in class was devoted to explaining how things would work, convincing the class that these projects were for real and that at our last session we would all share what was actually accomplished. Students were so overtrained to be passive that animating them to get involved was, in fact, the first challenge. The most important element of that process was to convince students that I meant it--that they actually had to choose their own projects and commit to them.

[Back to Table of Contents]

Teaching. Starting with the second class I played a teacher role in providing various diagnostic models for the students to use in analyzing their individual and team projects. I suggested a number of books and asked people to read as much as possible early in the 14 week period since all of the diagnostic material was relevant up front. At the same time I gave weekly reading assignments to focus us on relevant materials during the first half of the semester. Diagnostic models such as the Beckhard/Harris change map, force field analysis, role network analyses, and the Lewin/Schein stages of change were presented in the early weeks and rediscussed at later sessions so that the groups would have all of the tools available early on but could revisit them as they became more relevant.

A major chunk of time was devoted initially to the concept of process consultation because the change teams would have to operate without formal position power. I argued that their best chance of forming into effective teams vis-s-vis each other and their change targets, was to define themselves initially as internal process consultants who would have to develop some kind of access and a constructive relationship with their selected change targets. I also pointed out that this way of defining planned change was virtually synonymous with how one might define the process of management itself, except that one did not have formal position power. In this context I also reminded students that most managers report that having position power is not enough to make planned change happen.

Part of each class during the remainder of the course was devoted to short lectures on whatever seemed relevant at the time, war stories from my own experience, war stories that students told from their experience, and dealing with student questions on their projects. In dealing with questions I shifted my role increasingly to being a process consultant to the class and to the projects to highlight the importance of this role.

[Back to Table of Contents]

Monitoring and Grading. The monitoring role was most salient in how I dealt with the papers. For example, if a paper stated a goal of losing 30 pounds by the end of the semester, I might ask whether or not that was realistic, how much weight loss that would mean per week or per day, and how the person would monitor his or her own progress. If the goal was to overcome shyness I might ask the person to translate that into something concrete and measurable such as how many new contacts were made per week at parties, etc. I gave relatively few hints or suggestions unless the person specifically requested that kind of help, but concentrated on "process" monitoring: "How will you measure your progress toward your goal?" "Have you thought about how you will know at the end of the week whether you have made any progress?" "What will this mean for your daily behavior?" etc. Suggestions were always couched as questions: "Have you done a force field analysis relative to your change target?" "Who are the people in your role set and how will they react?" "Have you thought of involving your spouse in your project?" etc. If the logic of what was in the paper did not hold up I would question it or point out inconsistencies or lack of realism.

I made it clear at the outset that I expected everyone to do all the work, attend all of the classes, submit all of the papers, and that would result in a grade of A for every student. The only way to get a poor grade would be to shirk on the work or to put in obviously substandard papers. If students were absent or did not hand in papers two weeks running, I put notes in their boxes reminding them of their commitment. My goal was to create a climate where everyone would learn to the maximum of their own potential and would, therefore, merit the grade of A. I did not require that every project had to meet its change targets, but I did require that every project maximize its own learning.

[Back to Table of Contents]

Consulting and Coaching. These roles came up most often when I was asked questions about "what to do if....," usually in relationship to some "impossible" situation that the class member had experienced. Implicit in these questions was the assumption that since I was an expert on change I would be able to advise anyone on anything having to do with change. It is on these occasions that I found myself having to subtly shift my role to that of process consultant by asking inquiry types of questions to learn more about the reason for the question, the context, and what the questioner had already thought of. Sometimes I discussed the process directly by noting that the question was putting me into an expert role that I was not prepared to fulfill.

If team members asked me what do in relation to some aspect of their specific project, I attempted to get them to think it out with my help rather than giving them an "expert" answer. Or I would provide a number of alternatives instead of a single solution if it was clear that I had to provide some level of expertise. The best way to get this across was to think of myself as a "coach" who would help with the projects but could not do the actual work.

The best setting for coaching was when one group was asked to consult to another group, an activity that I started midway into the course. Sometimes I would role play the consultant before asking class members to do it, but the best learning actually arose when groups consulted with each other. Inevitably the consultants would make ineffective comments, or ask confrontive questions, or in some other way create a tense rather than a helping relationship. Once this happened I had two choices. I could let the interaction run its course and then get a reconstruction. A more effective intervention was to jump in immediately when something happened that seemed not to be optimally effective and provide an alternative or actually "role model" the alternative. This was direct coaching and was deemed by class members to be the situation in which they learned the most. In these settings I became the "process expert" because we were working on real situations in which I did indeed have more experience.

[Back to Table of Contents]

Dialogue. During the last two years I changed the structure of the class sessions by arranging us all in a circle, introducing the concept of dialogue, and starting each class with a "check-in" which involved asking each student in turn to say something about "where you are at right now" at the beginning of each class (Bohm, 1989; Isaacs, 1993; Schein, 1993). Though this was at times cumbersome because it took quite a while for 30 people to check in, the ritual itself became very meaningful and important to the class. The circle format and the dialogue assumptions made each session much more interactive and comfortable. It allowed me from time to time to also ask for a check out by going around the room near the end of class to see where people were at. If we were short of time we used a truncated version of check in by asking each person just to say two or three words such as "anxious but motivated," "tired and sleepy," "comfortable and eager," "distracted" and so on.

The Check-ln guaranteed that everyone would have a voice without having to raise their hand or figure out how to get in, a process that was especially important for the foreign students with language problems. One could see week by week how they become more comfortable during the check in and how this generalized to comfort in the remainder of the class session. Check-ln also revealed the class mood, things that were going on in the students' lives that were a distraction, fatigue levels and other factors that enabled us all to start class work on a more "realistic" level. It reinforced the dictums I had espoused--"always deal with the reality as you find it" and "go with the flow."

[Back to Table of Contents]

The Empathy Walk. At roughly eight to nine weeks into the semester I asked each class to form itself into pairs and to do the following exercise developed by Richard Walton and me at a workshop in the 1960's:

1 ) Talk with your partner to identify someone in the greater Boston area whom the two of you consider to be most different from the two of you. This will require you to think about how you are similar and along what dimensions someone would be really different.

2) Locate someone who fits your definition of someone most different and establish a relationship with that person so that you can spend a few hours getting into that person's world.

3) Be prepared to report back to the class what you learned.

We typically devoted one whole class session to the "war stories" students brought back and pulled out insights about the process of developing empathy. In addition each student wrote up their individual experience in the weekly paper that week.

Post class feedback consistently confirms that this is one of the most potent exercises of the semester because it forces confrontation of self and others at multiple levels. I assigned readings from Erving Goffman (1959, 1967) during these weeks to provide some conceptual handles. The ingenuity and cleverness of students that this exercise releases is dramatic. Students have found and built relationships with homeless people, street musicians, prostitutes, go-go dancers, trappist monks, convicted murderers, blind people, dying aids patients! successful celebrities, fishermen, hare krishnas, and so on. They discover, among other things, that the difference between them and their target is often less that their difference from each other. They realize how insulated their lives are from many real world problems, and how narrow their own perspectives are. They come face to face with social status and the dilemmas of having a privileged position in society, usually in the form of anxiety and guilt when they contemplate how one approaches a homeless person without "talking down to them." The discovery that some of these people have had or still have rich lives comes as a shock. In every case it opens the student up to becoming more inquiring and more sensitive to others, an essential step in becoming a successful change agent or manager.

[Back to Table of Contents]

Project Reviews and Final Reports. Toward the latter third of the course I began a series of project reviews by inviting any groups that wanted some help to present their issues and have other groups or individual students be consultants. After a half hour or so of the group and their helpers operating in a fish bowl I would open it up to the floor to get other comments. As unhelpful comments were made such as unsolicited advice or even punishment for mistakes that the group was perceived to have made, I would intervene in a coaching mode to examine what was happening. As pointed out above, these turned out to be some of the most salient learning experiences.

During the last two class sessions, usually accompanied by cookies and drinks, each group reported its final outcomes, salient points about their process, and the major things they had learned from doing the project. It was at this point that many students revealed the importance of doing both a personal and group change project because their struggles with themselves in the personal project gave them real insights into the problems of resistance to change in the group projects. Different groups reported different kinds of learning but a common theme that ran through all of them was the importance of making a commitment to the change, having an audience in the form of faculty and fellow team members, and having weekly reports that forced constant planning and replanning, and provided opportunities to get feedback.

The real payoff to the students is to discover that they can actually produce changes that have an impact. To see the Sloan School adopt a new faculty feedback form, to see actual changes in the student cafeteria menu offerings, to be thanked by the MIT Housing Office for improving the system of dealing with applicants, to create a new physical space and student lounge, to create events that increase the interaction between faculty and students and have those events become regular annual events, and, most importantly, to hear the Dean's office make reference to future student projects as a positive force for change is the best feedback possible. My own assessment is that student teams well training in planned change methods can accomplish more than powerful committees of faculty and administrators who do not understand how change can and should be managed. Finally, what surprises us all most is that change can happen fairly rapidly. Fourteen weeks is enough to make fairly substantial changes happen. But the conceptual core must be the right one.

[Back to Table of Contents]

The Conceptual Core of the Course: Diagnosis as Initial Intervention and Process Consultation as a Change Strategy.

The most important and most difficult concept to get across early in the course is that diagnosis is intervention and, in fact, that everything that involves the target system in any way is intervention. The discovery by students that diagnosis is intervention is paradoxical. In order to figure out what we need to change and discover where there is already some motivation to change that we can link with, we have to find out things about the present state of the system that we cannot know without inquiring. In order to gather such information we have talk to people in the system and ask them questions or conduct surveys. What is especially important to discover is where there is already motivation to change, where there is already survival anxiety that can be harnessed, because for many kinds of projects, students are not likely to be able to disconfirm or induce survival anxiety or guilt. On the other hand, if the change project involves organizational structures where the students are the recipients, they can often marshal potent disconfirming data and induce considerable survival anxiety.

The mental model at this stage that they are "just gathering preliminary diagnostic data" overlooks that the very people whom they have involved in the question asking may later be the prime targets whom they are ultimately trying to change. And, by asking those people various kinds of questions, they have 1 ) influenced their thinking by raising certain issues; 2) created an image in their minds of our own style and approach; and 3) created a degree of awareness and self- consciousness (possibly even defensiveness) because the targets now know that "there is a game afoot" and they are in some unknown way part of it.

Furthermore, as change agents, students often assume that they must remain fairly private about just exactly what they are trying to do, so they ask very broad inquiry type of questions, never once considering that the very vagueness of their questions may produce tension and anxiety in the interviewee precisely because he or she does not know what the change agents are after. How then do we gather the data necessary to determine what the present state of the system is without creating anxiety, misrepresenting ourselves, and unduly influencing the interviewee prematurely?

The answer lies in working from several assumptions that underlie process consultation (Schein, 1987,1988) and what has more recently been called appreciative inquiry (Cooperrider & Srivastva, 1987; Shani & Pitman, 1991). From process consultation one derives the assumption that one must always work in the present reality and must understand the ebb and flow of that reality moment to moment, shifting roles as necessary. If a student is going to gather data from a faculty member, the student must understand that there are already strong role expectations on both sides and one must work initially within that set of expectations. For example, some amount of deference is expected and must initially be honored. The faculty member would expect to be asked questions that draw on his or her field of expertise and the student would be expected to listen politely.

On the other hand if the student knows that the faculty member knows that the student is part of a team that has been set up to redesign portions of the curriculum, the student can assume that the faculty member would be curious, possibly anxious, and would prefer to find out first from the student what this was all about before revealing his or her own information. In that case the student might open the discussion by volunteering a description of the project in terms that are informative and minimally threatening.

Alternatively, the faculty interviewee might seize the initiative and ask a bunch of questions about the project. In those preliminary questions, the student would have to assess how much anxiety is present and vary his or her tactics accordingly. It is in the design of those tactics where "appreciative inquiry" plays a role. One of the core assumptions of appreciative inquiry is to focus initially on what is working well and avoid criticism or problem foci. The interview might well start with what the faculty member is most proud of or what works best in the curriculum. If the interviewer focuses on success and what works well, he or she is creating psychological safety that will make it easier for both parties later in the interview to discuss problem areas, difficulties, things that need improvement. The prime data that the interviewer needs and wants is where the faculty member sees problems or has motivation to change, but the initial assumption has to be that he or she will not be ready to talk about problems until they feel safe with the interviewer, and they will only feel safe if the interviewer displays appreciation of what works well.

As the interview or interaction proceeds, the change agent must be constantly alert for changes in mood or feeling on the part of the interviewee, being especially sensitive to issues that may be threatening to the interviewee leading to a shutting down of the flow of information. It is in that ongoing interaction that the tactical use of inquiry questions, diagnostic questions, action oriented questions, and confrontive questions comes into play (Schein, 1987, p. 1 46).

The goal should be to create an interaction that will provide information to the change agent, begin to build trust with the potential change target, and begin to get the change target to think diagnostically and positively about the change project such that he or she will welcome another interview or interaction because their curiosity or their own energy for change has been aroused. In a sense the concept of "change target" has to become transformed in the change agent's mind into a "client" who seeks some help or into a "learner." The change agent has to become a facilitator of the learning process and the desired change has to be embedded in a "helping process" that makes sense to the learner.

In thinking this way we have come full circle once again to Lewin's original concept of involving the change target in the change process, but I have tried to elaborate and deepen our understanding of the issues involved in making that happen, especially when the change agent operates from a position of low status and minimal formal power.

[Back to Table of Contents]

Summary and Conclusions

As I reflect on the material in this essay I am struck once again by the depth of Lewin's insight and the seminal nature of his concepts and methods. I have only reflected on some aspects of Lewin's theory, but even those few aspects have deeply enriched our understanding of how change happens and what role change agents can and must play if they are to be successful. Lewin probably saw such issues more clearly because he was able to view U.S. culture from a European perspective. Important changes inevitably involve deep cultural and sub-cultural assumptions. The ability to perceive and appreciate the meaning of such tacit cultural assumptions is enhanced by working across several cultures. If we want to enrich our understanding of these dynamics further, we also should become cross- cultural learners, to expose ourselves to different cultures and begin to reflect on what it means to try to change cultural assumptions. We may then discover why "change" is better defined as "learning," why cultures change through enlarging and broadening not through destruction of elements, and why the involvement of the learner is so crucial to any kind of planned change or, as we might better conceptualize it-- "managed learning."

[Back to Table of Contents]


Argyris, C. (1990) Overcoming Organizational Defenses. Boston: Allyn & Bacon.

Bettelheim, B. (1943) "Individual and Mass Behavior in Extreme Situations," Journal of Abnormal and Social Psychology 38, 417-452.

Bohm, D. (1989) On Dialogue. Ojai, CA.: Ojai Seminars.

Bushe, G. & Pitman, T. (1991 ) Appreciative Process: A Method for Transformational Change. OD Practitioner. September, 1-4.

Cooperrider, D. & Srivastva, S. (1987) "Appreciative Inquiry." In R. W. Woodman & W. A. Pasmore (eds.) Research in Organizational Change and Development. Vol. 1 . N.Y.: JAI Press, pp. 129-169.

Goffman, E. (1959) The Presentation of Self in Everyday Life. N. Y.: Doubleday.

Goffman, E. (1967) Interaction Ritual. N. Y.: Aldine.

Hirschhorn, L. (1988) The Workplace Within. Cambridge, MA: MIT Press.

Isaacs, W. N. (1993) Taking Flight: Dialogue, Collective Thinking, and Organizational Learning. Organizational Dynamics, Winter, 24-39.

Michael, D. N. (1973) On Learning to Plan and Planning to Learn. San Francisco, CA.: Jossey Bass.

Michael, D. N. (1992) "Governing by Learning in an Information Society." In Rosell, S. A. (Ed.) Governing in an Information Society. Montreal, Quebec: Inst. for Research on Public Policy.

Schein, E. H. (1956) The Chinese Indoctrination Program for Prisoners of War. Psychiatry, 19, 149-172.

Schein, E. H. with Inge Schneier and C. H. Barker. (1961) Coercive Persuasion. N. Y.: Norton.

Schein, E. H. (1968) "Personal Change Through Interpersonal Relationships." In Bennis, W. G., Schein, E. H., Steele, F. 1. & Berlew, D. E. (Eds.) Interpersonal Dynamics. Rev. Ed. Homewood, 111.: Dosey Press.

Schein, E. H. (1987) Process Consultation. Vol. 2. Reading, MA.: Addison- Wesley.

Schein, E. H. (1988) Process Consultation. Vol. 1 (Rev. Ed.!. Reading, Ma.: Addison-Wesley.

Schein, E. H. (1992) Organizational Culture and Leadership. 2d. Ed. San Francisco, CA.: Jossey Bass.

Schein, E. H. (1993) How Can Organizations Learn Faster? The Challenge of Entering the Green Room. Sloan Management Review, 34, 85-92.

Schein, E. H. (!993) On Dialogue, Culture, and Organizational Learning. Organizational Dynamics, Winter, 40-51.

Senge, P. (1990) The Fifth Discipline. N.Y.: Doubleday, 1990.

Van Maanen, J. & Schein, E. H. (1979) "Toward a Theory of Organizational Socialization." In M. B. Staw & L. L. Cummings (Eds.) Research in Organizational Behavior. Vol. 1. Greenwich, Conn.: JAI Press.

[Back to Table of Contents]


1. Invited paper for a special issue of Systems Practice edited by Susan Wheelan, March, 1995. [Back]
2. I have deliberately avoided giving specific references to Lewin's work because it is his basic philosophy and concepts that have influenced me, and these run through all of his work as well as the work of so many others who have founded to field of group dynamics and organization development.[Back]
3. I am indebted to Colleen Lannon Kim tor these terms. I had originally used Anxiety 1 and Anxiety 2 (Schein, 1993). She helpfully put some useful labels on them.[Back]

[Back to Table of Contents]

Untitled Document