spacespace space space space

Action Learning in a Nutshell

Robert L. Dilworth




There can be few tasks so daunting as to describe action learning in an abbreviated and concentrated way. Those who have experienced action learning know the wide variety of forms it can take. There can be vast differences of interpretation and application. The lack of a tightly defined framework can be a distraction, especially to those accustomed to curriculum design. On the other hand, the flexibility of action learning in promoting learning and elevating organizational performance can be highly attractive.

This article covers action learning in a contextual way, first, by relating it to two societal trends. Three examples are then provided that leave the reader free to interpret their significance and how they differ from traditional approaches to problem resolution. The remainder of the article then outlines action learning principles. The goal is an integrated view of action learning in application, including some contrasting beliefs.


Action learning is gaining in popularity as a way to improve performance, promote learning, and position organizations to adapt better in turbulent times. It is also seen as a way to develop the capabilities of individuals, teams, and overall organizations. An excellent report, Continuous Learning, published by the Canadian Centre for Management Development (1994), suggests that “some of the most interesting and promising innovations in management learning have taken the form of what is called action learning.”

The number of corporations that have or are now using action learning approaches is growing. They include TRW, Inc., General Electric (GE), Andersen Consulting, Conoco, Whirlpool, Ameritech, and GEC (England). Public sector organizations are now also represented, such as the Federal Deposit Insurance Corporation (FDIC) and the Defense Management Systems College (DMSC) of the U.S. Department of Defense.

Action learning, as a concept, dates back more than 50 years. It has, until recently, received more interest and attention outside the United States. Its roots can be traced to action research, a concept and term originated by the German psychologist, Kurt Lewin, in the 1940s (Weisbord, 1987, pp. 183-195). Reginald W. Revans, of England, originally an astrophysicist, pioneered the concepts related to action learning over more than 50 years ago. His effort is extensively documented and involved much in-depth research, including work in coal mines, hospitals, and with industry in Belgium.

Key Questions About Action Learning

Given the keen interest in action learning in the 1990s, the pivotal question becomes, “What is action learning?” There are associated questions as well: “Why is there so much interest in action learning today?” “How can action learning be applied?” “What are the perceived benefit values?” “How can action learning be related to performance improvement?”

This is when things get difficult because defining action learning is not easy to do. Action learning can take a variety of forms. In other cases, it can be closely interwoven with other organizational interventions. In such cases, a number of labels may apply, including organization development (OD), management development, team building, and transformative learning. Revans, even with all his writing on the subject, avoids defining “action learning”. He is more inclined to describe action learning in terms of what it is not. Revans, in effect, holds the view that to try and build finite structures around it, as is usually done with management concepts, only robs action learning of its power. It can be like trying to sail the Queen Elizabeth II in a bathtub. A highly definitive concept with narrow parameters simply does not fit the subject. The eclectic nature of action learning, drawing from various disciplines, also makes finely chiseled depictions difficult.

My strategy for explaining action learning in this article is to frame it in relation to societal trends and the prism of practice. This will allow me to isolate some of the most basic characteristics and underpinnings associated with action learning.

I will begin with two broad societal trends that are perceived to be influencing present day interest in action learning. Then I will turn to three actual examples of action learning, two from the United States and one from Europe. What contributed to the positive results realized in each instance? The concluding coverage in my article will map some of the major features of what is commonly associated with action learning. This will also encompass some of the issues that can arise in implementing an action learning program and how to go about putting such a program in place.

Two Societal Trends and Their Implications

The first societal trend relates to significant disillusionment with initiatives to improve the quality of work life and performance. It is frequently expressed in terms of marginal results and knee-jerk, short-lived initiatives. The disillusionment often stems from management approaches that do not prove out. This can take several forms:

  1. The initiative was launched without a full awareness of its implications, and the workforce may have been left in the dark. This can be particularly true in downsizing situations. In a survey of 547 companies that had downsized, the American Management Association (AMA) found that operating costs had improved for fewer than half, while 77 percent experienced a decline in employee morale after downsizing (Boyett & Boyett, 1996, p. 54).
  2. Top management said they supported the initiative but then became distracted by crises and daily work activity, and the initiative died on the vine. It can also be a case of management losing heart when they do not see immediate results. This can be true with regard to quality management initiatives, even when experience shows that you can go through a productivity dip for one to three years while the quality program is being brought fully on line. In other cases, top management turnover occurred, and the workforce was immediately asked to go a different direction. Cynicism ensued. The net result can be deep reluctance on the part of employees to mobilize behind management initiatives except on a superficial level.
  3. Consultants came in, applied (or recommended) solutions, and then left systems that continued a downward spiral. Then, secondary consultants were called in to fix residual problems after the first intervention. This can lead to confusion of objectives and a workforce essentially immobilized by uncertainty.

The second trend relates to learning organizations and continuous learning. Interest in this area is now becoming pervasive, and there is reason to believe that it will be a lasting focus. There is an awareness that organizations as human systems must constantly learn to adapt if they are to survive. As the turbulence of the environments in which organizations exist accelerates, the learning must be continuous, as opposed to being anchored to intermittent formal training. This second trend also spawns intensive interest in workplace learning.

There are at least three questions that arise when considering the need to orient on learning organizations:

  • What constitutes a learning organization? (The concept is still evolving.)
  • How do you jump-start an organization to get it into a learning mode?
  • How do you go about shaping the “organizational DNA” to sustain a learning organization culture once it is created?

I personally believe that the answers to these questions can be linked to action learning, and the number of corporations turning to action learning tends to confirm this. Action learning is usually discussed in relation to organizational learning and creation of a learning organization. In some cases, the organizations decide to try action learning out of disillusionment with approaches that failed. There can also be a belief that action learning can be an important vehicle for transformation of organizational culture, increasing the learning capacity of the enterprise and empowering workers.

Three Examples of Action Learning in Practice

Rather than launch into an explanation of what constitutes a learning organization and action learning, let me start with three examples.

Example 1: Computer Technology

I headed a major organization with hundreds of computers organized in an Intranet. The organization was highly dependent on this communication mode for tens of thousands of daily transactions in conducting its highly complex and diversified business operations. Therefore, the efficiency and responsiveness of the computer capability directly influenced organizational performance.

We observed that the computers were slow to respond in moving between screen images. Such time delays collectively constituted a major drag weight on performance. My internal computer experts had promised prompt resolution but failed to deliver. The problem persisted even after the national headquarters of the major computer firm was activated (I personally wrote a letter to the Chief Executive Officer), and further troubleshooting was accomplished. The computer firm expressed the belief that the problem was unfixable and that a new multi-million computer system would be required.

Convinced that a solution was possible, even after the further “expert” initiatives proved unsuccessful, I called together my 22 management trainees from the fourteen directorates in my organization. All trainees had basic computer literacy, but only two were on a development track to becoming computer specialists. While serving in the same large organization, some of the trainees had never met before. Most had never worked together.

I informed the trainees of the problem, its significance to the organization, our unsuccessful efforts to solve it, my belief that it could be solved, the importance of avoiding the cost of a new computer system, and my belief that working to solve the computer problem could constitute a wonderful learning experience. They were asked if they would be willing to take on this problem as a group. After huddling briefly, they came back and accepted the challenge. There were comments circulated in the organization that the trainees must be extremely naïve to believe they could resolve such a problem. Others commented about the unfairness of asking trainees to take on such a challenge. The trainees never showed any evidence that the problem was beyond their depth. On the other hand, the trainees did view the problem as extremely challenging. My computer experts offered to attend the organization meeting called by the trainees, but the trainees declined, preferring to follow their own instincts and call up specific expertise as appropriate.

What happened? A month later, the computer problem had been fixed. The trainees briefed their problem solving approach to me, the in-house computer experts, representatives from the national computer firm, and other computer organizations. There was high interest in how a seemingly intractable problem had been overcome.

It became apparent in listening to the trainees that they had explored some avenues not explored by others. Their problem solving approach was of their own design. They found an array of causal factors rather than any single problem driver. Because our computer system interfaced with others in the overall organization, the trainees had gone outside our organization. They found sister organizations suffering with the same inefficiency. The trainees therefore adopted a total systems approach and set about solving the problem in the other organizations as well. The trainees statistically demonstrated that their time expenditure in doing the problem solving would be quickly amortized in terms of measurable improvements in organizational performance. They also gave considerable thought to what they had learned and classed it as one of the best learning experiences of their lives. A camaraderie grew out of the experience. They had bonded as a group and asked to be allowed to take on other complex troubleshooting projects.

What are the implications and lessons learned?

  • Because they did not know a great deal about computers, the trainees were induced to ask fresh questions.
  • The problem was real, and once they accepted responsibility for the project, they were expected to solve it.
  • Trainees did not know enough to start with customary troubleshooting techniques. They invented their own process.
  • They drew fully on the intellectual resources of their trainee group. There was no leader. They operated as equals in a trusting environment.
  • They were absolutely certain of top management support and backing.
  • They broke their overall group of 22 down into smaller groups of roughly six each to examine various aspects of the problem as they understood it.

Example 2: Florida Power and Light

Florida Power and Light (FPL) was the winner of the Deming Prize for Quality in 1990. This is a Japanese award named after an American, Edward Deming. FPL was the first U.S. corporation to earn the award. It was also the first major quality initiative related to the service sector, as opposed to manufacturing, in the United States.

At about the time that they earned the Deming Prize, FPL had experienced a difficulty with its power generation systems. In fact, the problem had persisted for several years, and various task force attempts to solve it had failed. FPL was experiencing an unacceptable inefficiency relating to the conversion of energy to electricity in its power generation systems. Electricity generated was significantly less than it should have been, based on energy input to power generator systems. They called together a team of people from different FPL areas to troubleshoot the problem. Team members had not worked together before as a team. They had a range of skills.

The team solved the problem, developed a strong team spirit (they asked the company to continue them as a team for other problem solving), and gave their presentation of results in shirts with their self-determined team name, “Drips,” embroidered on them. The team name related to some of the problem sources they had identified.

The "Drips" found that there were many individual problems, not one overriding circumstance that was causing the inefficiency. They found, among other things, that birds nesting on power lines across the state of Florida could lead to problems, one of them being shorting out of lines.

What are the lessons learned in this case?

  • They were not mired down in the past. They started a new line of inquiry.
  • They asked fresh questions. It represented unfamiliar territory to several of the team members. Therefore, they had to question assumptions and explore avenues that traditional troubleshooting approaches might have overlooked.
  • What they learned achieved some breakthroughs in thinking about problems of the type addressed. In other words, the established company knowledge in this area had to be updated and modified.
  • It was a real problem, and they were expected to solve it.

Example 3: Belgium

This example is drawn from the experience of Reginald Revans. Revans is recognized widely as the principal pioneer of action learning and has done much research in this area. The example involves the five universities of Belgium working with a number of larger companies, with the goal of elevating the economic performance of Belgium (Revans, 1982, p. 329). Begun in 1968, action learning was the technique of choice. Various companies identified pressing problems within their firms and an internal client organization dedicated to problem resolution. As this national program began, a group of five senior executives was brought together as a group (called an action learning set in Revans’ parlance) to deal with problems within the various companies.

The five executives had never worked together before, and each came from a different background. Each executive was matched with a company and their problem. In every case, the executives had a background different from the company and problem they were being asked to examine. One of the executives who came from a major bank was asked to deal with the problem surfaced by the largest steel company in Belgium. The problem concerned the inability of the steel company to produce adequate quantities of alloy steel. They had strong technological expertise, and their research and development was of the first order, yet the Japanese outpaced them in the area of production.

Each executive worked with the firm assigned, and then the five executives met as an action learning set at regular intervals over the course of a year in sharing their learning experience and pooling their knowledge.

The executive examining the problem with the steel company interviewed a number of personnel, including the chief executive officer during his fact finding. In the end, he discovered that the problem of alloy steel production had little to do with production capability. The problem stemmed from company policy related to its compensation system that went back many years. All company compensation policies, from entry level to corporate board room, were predicated on tonnage of steel shipped. Since alloy steel is light weight, there was no incentive to bring production to the required levels. Once that became known, it became possible to correct for this constraint. An action learning set within the specific client organization of the company was then activated to determine solution options. In the end, many action learning sets of executives were formed in Belgium to address major industrial problems using this model.

What were the lessons learned?

  • As was true of the other examples, working on unfamiliar problems triggers fresh questions.
  • The energy and synergy generated by working with people from dissimilar backgrounds (the group of executives, in this case) can stimulate learning.
  • The participating executives learned a great deal and developed their managerial expertise in the process.
Relating Theory to Practice

There are several fundamental aspects of action learning that relate to the examples given and the ability to stimulate continuous learning.

Change Versus Learning

Change now tends to outdistance our ability to learn. Established knowledge can quickly be outdistanced by the pace of new challenges and events. This happened in all three examples provided. Knowledge on hand tended to misdirect inquiry rather than facilitate problem resolution. It was only when new lines of inquiry and fresh questions were introduced that problems were overcome. In each instance, the organization learned new ways of coping with problems. Those working in the groups (sets) learned the value of pooling their intellectual capital. The learning that occurred had greater value strategically for the organization than the immediate tactical advantage of early problem correction. The learning capacity of the participants and organizations involved was enhanced. This enhancement made the organizations better able to deal with change dynamics.

As Revans points out in all of his writing, you must capacitize organizations to learn apace with (ideally, faster than) change if you expect to survive. This has applicability to both individuals and organizations. If you fail to keep current from a competency point of view, as an individual, the degree of success you can expect to experience in your career will tend to be limited. The risks to an organization are equally severe as we enter the 21st century. You either become a learning organization or lose your competitive edge.

A way to explain the need for synchronization of change and learning can be related to the example of a small aircraft that encounters severe air turbulence aloft (analogous to the velocity of change in the environment surrounding organizations). The pilot, as the story goes, has three choices. First, she or he can slow down the aircraft. However, when the law of physics is taken into account, this will only add to aircraft instability. Second, the pilot can simply maintain speed. However, with the external environment moving at an accelerated pace, this also tends to produce instability. The answer lies in the need to increase the speed of the aircraft (in terms of organizations, this is the learning component). The point of the story is that the learning capacity of organizations tends to lag behind the dynamics of what confronts them as we enter the 21st century. The answer lies in improving the intellectual capital and learning properties of organizations.

Learning Comes First

The most important link to action learning is that you bring people together for reasons other than problem resolution. You want a problem solved, but the primary value is in the learning that occurs. You want to build the learning capacity of the organization as a way to boost organizational performance. Therefore, the employment of action learning is strategic rather than tactical. You need fresh thinking if you are to avoid responding to yesterday’s challenges as today’s problems and tomorrow’s opportunities engulf you. The goal becomes dynamic equilibrium, with learning and change intertwined.

The Learning Equation

Revans suggests that learning is derived two ways–both through programmed instruction (which he calls “P”) and questioning insight (the “Q”). The “P” comes through textbooks, lectures, case studies, computer-based instruction, simulations, and the like. This is an important source of learning but carries with it an embedded caution flag: the “P” is all based in the past. Therefore, it is unlikely to match precisely with today’s needs. In the worst case, it fogs your lens and can lead you astray. That happened in each of the three examples given.

In every example provided, the pitfall of imperfectly constructed past knowledge was avoided by beginning with questioning insight (“Q”) rather than by using past knowledge as the first reference point. There is great power in this conception. By going through the “Q” step first, you are able to begin gauging whether the information available is adequate to your needs. It will lead you to a determination of the relevancy of information. It will also point to areas that will require creation of new knowledge (new “P”). The key is to start with fresh questions, not with constructs from the past.

Problem versus Setting

As Revans reminds us, there are two kinds of problems and two kinds of settings where they are found. The problem is either familiar or unfamiliar. The same applies to the setting. Most problem solving to be accomplished in organizations is done in natural work teams in a familiar work area. This tends to make us comfortable with what we know and with the established way of thinking.

When we are asked to examine a familiar problem in an unfamiliar setting, it can open up our eyes to new alternatives. We may also find ourselves addressing an unfamiliar problem in a familiar setting. This was true to a large extent in the computer responsiveness problem I outlined. This kind of situation also causes us to ask fresh questions and reexamine our basic assumptions.

The Belgium case involved an unfamiliar problem and unfamiliar setting for the business executives involved. In addition, each executive was surrounded by unfamiliar colleagues. It promoted new ways of going about learning and problem solving. Revans considers this the most potent form of the action learning experience.

In action learning, you deal with real problems, not problems that have been prefabricated for instructional purposes. The best way to promote fresh questions and new thinking is by causing the problem, setting, and colleagues to be unfamiliar.

Action Learning Sets

Usually, action learning is accomplished in action learning sets of five or six to provide a group size that promotes ease of communication. A facilitator can be used. Some of us believe that facilitation should be used primarily at the start of an action learning experience and then quickly fade out. Revans believes that learners are their own best facilitators. There can be an important role for facilitators in helping the learners slow down their process in order to allow sufficient time to reflect on learning. Reflection is an important part of an action learning experience. Executives, in their impatience to move forward and solve problems, can find it extremely difficult to take time to reflect on what has been learned. When they are encouraged to do so, reflection on learning can represent a positive executive development and individual growth experience. A facilitator can help participants learn this skill. Members of the set, as part of the norming process, can also promote use of reflection in the learning process.

A Summary of Action Learning Fundamentals

  • Questioning insight is always the start point.
  • The problem must be real.
  • The problem to be solved can be tactical or strategic, but the learning is strategic. You are striving to continuously upgrade the intellectual capital of the firm in terms of adapting to change and sustaining a competitive edge. This is a key ingredient of the learning organization.
  • Reflection is as important as action. Learning journals/logs are a good way to induce greater reflection and enhance the learning that occurs.
  • Three basic questions commonly begin the action learning process in addressing a real problem. First, what should be happening? Second, what is stopping us from doing it? Third, what can we do? This is reminiscent of Kurt Lewin’s Force Field Analysis (Driving Forces and Restraining Forces) (Weisbord, 1987, pp. 77-79).
  • Learning is the primary goal, even though the problem solving is real and important. Learning is facilitated, to include breaking out of well-established mind sets by having the setting, the problem, and colleagues to some degree unfamiliar.
Why is Action Learning Gaining Attention?

What is driving organizations in terms of rapid environmental change requires rapid real-time adaptation. Action learning is well suited to this type of adaptation. Work complexity and organizational design also make it a natural intervention. Work complexity is moving downward in organizations as the need for timely action and multiple competencies grows. In its early use, action learning tended to focus on executives, as occurred in Belgium. Today, the realities of the workplace and growing competency base make it a model that can be used at all levels in an organization. In fact, in Sweden, action learning has even been employed in the K-12 classroom environment, with parents, teachers, and children participating in the experience.

The need for organizational integration in responding to change also fits well with action learning. The concept of action learning tends to be very egalitarian, with action learning sets operating without a designated leader. Any mantle of authority is usually left at the door. This makes action learning a good fit with self-directed work teams (SDWT). Such teams are rapidly becoming a feature of U.S. business organizations.

There is also a growing awareness of the learning challenge now confronting organizations. You see this in the choice of terms now becoming common place: organizational learning, continuous learning, life-long learning, deutero-learning (learning to learn), boundarylessness, self-ordering systems (chaos theory), integration with environment, learning organizations, building (and sustaining) intellectual capital, self-directed learning, the emerging concept of the chief learning officer (CLO), and virtual organizations. Because of the significance assigned to real problem solving and unfamiliar venue, action learning meshes well with the need for spontaneity and speed of adaptation. TRW has a program designed to underwrite its emphasis on globalization. It involves taking executives away from their corporate site to work on such issues in action learning sets. TRW then may send them overseas in order to further induce fresh thinking and elevate the degree of learning taking place.

Action Learning As Distinct Phases

Let me tie the concept down as a progression of phases. Each phase sets the stage for the next phase. It is a cascading effort.

  • By placing individuals in unfamiliar settings and having them work on unfamiliar problems, you induce fresh questions. When we are out of our comfort zone, of necessity we must look at things through a different lens.
  • When you ask fresh questions, you begin to unfreeze and reshape your underlying assumptions. It can be transformative learning.
  • As assumptions come into question, they are either confirmed, modified, or rejected. When we end up changing the texture of our assumptions, we then begin to create new mental models.
  • The new mental models, together with the shifts in underlying assumptions that prompted them, cause reassessment of the programmed knowledge (“P”) at our disposal. This causes us to reject some of the “P” available and replace it with new “P”.
  • The learning capacities and performance levels of the enterprise are enhanced by the renewal accompanying generation of new knowledge and questioning insight (“Q”).
Implementing Action Learning

Broadly conceptualizing action learning is one thing, but how do you go about the actual process of implementation? The concept can be strongly resisted. The training community can see such an experiential approach as threatening to long-established training programs. It can even be viewed as placing traditional training jobs in jeopardy. Top management can frown on it because it does not fit the established mold. It can be a desire for quantitative results. Action learning can and has produced this, but it can be difficult to measure results in traditional ways and in the short term. How do you go about gauging the increase in learning capacity of the organization? While some tools for doing this are emerging, it is still in its early stages. Top executives can also resist on the basis that it seems nonsensical to take executives out of their primary area of expertise and assign them responsibilities for solving a problem outside the bounds of that expertise.

What follows is an abbreviated sequential look at steps in implementing an action learning program.

Gaining Early Understanding and Support

As already outlined, both the training community and top management can resist an action learning approach. Therefore, it is important that they understand the rationale behind it and how the concept is being used elsewhere. It can be useful to expose both trainers and top executives to the concept, including the opportunity to meet with those who have used the action learning process. Trainers need to understand that the introduction of action learning does not cause the human resource development (HRD) function to go away. It simply changes the focus and kinds of skills HRD professionals need to possess.

This early acquisition of support is critical. Without such support, only a limited action learning program is usually possible.

Cultural Receptivity

If the culture of the organization is lacking in trust, this can be a major impediment to action learning. It operates best in a wholesome organizational climate. If the organization is heavily authoritative or status-conscious, this can also impose a constraint. Fear and uncertainty also work against action learning as an intervention. If top management is really on board, action learning can be employed to transform organizational culture in arriving at a more trusting and productive work environment. It can be an important force in opening up communication flows. But you do need very strong leadership at the top to bring about such a change. This is really true of any major shift of organizational direction, not just that specific to action learning.

Selecting a Problem

Because action learning is strategic and primarily designed to promote learning, the problem needs to be selected carefully. This also relates to how you will use your small groups (action learning sets). One option is to have the set take on a shared problem as was true in the computer problem case example. Another approach is to have each member of the five or six member set work on a different problem but learn from each other (the Belgium model).

Problem selection (and how you use the action learning set) can also serve strategic interests of the organization. For example, a problem might be selected for the set to work on that causes integration of various processes across the organization. If each member of the set operates on a separate problem, the set members individually need to be capable of dealing with the complexity involved, because while they can draw on the views of other set members, they will be left to deal with their own focus area and much of the fact finding and analysis.

Determining Set Composition

The determination of who will be in an action learning set is again strategic. It is not something that should be done randomly. There can be value in bringing people together from different organizations to work on a problem. Beyond the potential value of an unfamiliar problem for learning purposes, such a cross-section can help build organizational integration through creation of networks that did not exist before.

There can be an issue around voluntary versus involuntary membership in a set. My own view is that volunteers are desirable, but set assignment itself needs to fit strategic interests. If set selection is voluntary, you can end up with a clique of friends or people with compatible views. Your goal is to build in diversity. The problem solving and learning properties tend to be elevated through incorporation of a variety of perspectives.

Learning styles are another important factor in set selection. If you have all activists in a set and no reflective types, you can end up with a rush to judgment. Presence of reflectors will help slow the process sufficiently to better consider available options. It should also allow greater focus on the learning taking place. The Learning Style Questionnaire (LSQ) developed by Peter Honey and Alan Mumford (based on David Kolb’s model) specifies four learning styles (activist, reflector, theorist, and pragmatist) (Mumford, 1993, pp. 53-57). It is not a case of being one or the other. There are cases of an individual being strong or weak in all four types. It represents how we communicate, view our world, and what we emphasize in our life experience. It can be a gauge to how well we listen and what we hear. Strong activists can be disinclined to spend time in a reflective mode. While no instrument is fail safe, the LSQ can be helpful in balancing set composition. In addition to learning styles, it is also useful to balance gender, age, and ethnicity in sets. This adds diversity and richness.

The Facilitator Role

It is common to employ a facilitator with a set. As mentioned previously, there are wide differences of view on this. A good start point is to consider the role of the facilitator. It can be an important role at the outset in helping the set norm itself in terms of interpersonal dynamics. If you are dealing with mature learners, the role becomes more problematic thereafter. The facilitator is not a member of the set, and the mere presence of a facilitator can influence (even damper) what occurs. As a reality of group dynamics, there is no such thing as a benign presence. I have seen facilitators rejected in several instances because of perceived interference with set operation or simply unwanted presence. The presence of a facilitator, even when they remain silent for the most part, can influence interaction within the set.

My own belief is that the facilitator can play a key role in “jump starting” the set activity and orienting the set members on basic fundamentals of action learning (e.g., all set members are equal; the importance of listening to the views of others). After that, I believe the facilitator needs to fade back to what can be called mentor status. The facilitator is available as needed and will provide perspectives related to action learning but never tell the set how to deal with an issue. How the set deals with its problem is left to the learners. In my experience, it can be wise to make the presence of a facilitator at a set meeting “by invitation only.” Some advocates of action learning take a diametrically opposed view and believe that a facilitator presence at set meetings is necessary. My position is that people do not “learn how to learn” through the guidance of others so much as they do from creating their own meaning.

The Problem and Learning Balance

There can be a great urge to race forward to problem solutions in a set–and that is one of the goals. The learning can receive insufficient emphasis in such a situation. As a result, the action learning experience can become little more than another task-force-related excursion.

One way to avoid this common phenomenon is through proper priming of set activity at the start. This is something a good facilitator will do. It is also extremely useful to have set members keep a daily learning log, and then develop a reflective essay on the learning experience at program end. It can be beneficial to have participants think of several critical incident categories in making entries in their learning log.

  • When was I most engaged?
  • When was I most distanced?
  • When was I most affirmed?
  • When was I most puzzled?
  • What was the single most important thing I learned today? Why? How will I capitalize on it in the future?

Self-Norming of the Set

Set members need to learn about each other early in the action learning experience. You need to make the time to do this. The facilitator can assist. What background and skills are at the table? How can they potentially contribute to problem resolution? During this early state, norms also need to be determined. A recent set I was associated with came up with these norms:

  • We will meet only when all members can meet.
  • We will debate earnestly but never attack one another.
  • All will carry their share of the responsibility.
  • We will listen to one another.

The norming process is particularly significant because there is no designated leader. Therefore, interpersonal relationships become critical. In my recent experience, sets have operated throughout the action learning process without a single leader emerging. All tend to share the role. This is, of course, also a product of careful set member selection so there is relative parity in intellectual level and how personalities mesh. When the norming process is short-circuited because of pressure to get on with problem solving, it can become necessary to reopen the norming process. I watched a set in 1996 cease problem solving activities for half a day until the group could come to a meeting of the minds on what had turned out to be unfinished business in the norming area. In this instance, one of the set members had displayed excessive dominance, and the set had to talk it through.

Because action learning sets operate as self-directed teams, they can be an ideal bridge for those organizations moving in the direction of self-directed work teams. Action learning promotes empowerment and self-reliance. I have also seen it significantly elevate the self-confidence of participants.

Dealing With The Client

Working with a client can be new ground for set members. The facilitator can assist as an agent in the process but should never act as the representative of the set. I recently saw a facilitator overstep the role in speaking for a set with the client–and with the set present. The facilitator did not enjoy the full respect of the set again. One or more set members may end up selected by their fellows to be liaison with the client. Such an arrangement can prove necessary, but all set members should meet with the client concurrently whenever possible.

Alpha, Beta, Gamma

Revans speaks of Alpha, Beta, and Gamma in relation to action learning. In order to provide an integrated overview of action learning, let me address these three indices briefly.

System Alpha, in Revans’ definition, falls between the subjective and objective. It moves from personal values to external circumstances and internal resources. They were alluded to early in this article as the start point when an action learning set begins work. There are three questions that get asked. What should be happening? What is stopping it from happening? What can I do to remove the blockage? (Revans, 1982, p. 14)

System Beta relates to a five-stage process, endlessly repeated (as Revans describes it) as participants (set members) move toward their goals.

  1. Survey or observation.
  2. Hypothesis or theory (in formulating courses of feasible action).
  3. Experiment or test.
  4. Audit or evaluation (what happened?).
  5. Review or ratification (comparisons between expectation and experience) (Revans, 1982, p. 725).

Here you see the cycle repeating. As is done in action research, in effect, you research, act, review results, and repeat the cycle. It is trial and error, following the logic of the scientific method.

System Gamma is the interaction between the manager (or set member) and the situation she or he is trying to influence. (Revans, 1982, p. 348). This cannot be discounted. We are who we are and relate to events from our own unique framework of knowing. This is another reason why set members need to know other set members at a deep level. In the process, we also come to know ourselves better. This, by itself, can be an important learning outcome. System Gamma can end up being transformative of the individual learner and the organization involved.


What action learning can provide is elevated levels of discernment and understanding through the interweave of action and reflection. In a time of rapid change, it can be an intervention of choice. GE uses it throughout its global operations. In some cases, it takes the form of what is called the work-out, i.e., group work on real problems in real time, with key executives expected to provide on-the-spot decisions to employees as proposed solutions to problems are offered. At the higher levels of the company, it can take the form of change acceleration programs. Here you see the clear linkage between change and learning.

Action learning takes careful thought in execution. It can run cross-grain with established ways of doing business. While growing use of cross-functional (and global teams) is symbiotic with action learning principles, some corporations still think almost exclusively in terms of formal training. Since action learning can provide learning experiences outside the bounds of formal training programs, it can be viewed as a threat. The suggestion that people be considered for problem solving activity outside their expertise can also be viewed as a strange proposition.

One thing is clear. More and more corporations are turning to action learning because it is viewed as a way of transforming the culture and providing continuous learning. Some view it as the gateway to learning organizations. Most importantly, they have found it an excellent tonic for driving performance improvement.


Barker, A. (1997). Of…action learning, humanity, management and morality. Unpublished manuscript.

Boyett, J.H. & Boyett, J.T. (1996). Beyond workplace 2000. New York: A Plume Book.

Canadian Centre for Management Development (1994). Continuous learning (CCMD Report No. 1). Ottawa, Canada: author.

Chawla, S. & Renesch, J. (Eds.) (1995). Learning organizations: Developing cultures for tomorrow’s workplace. Portland, OR: Productivity Press.

Dilworth, R.L. (1996). Action learning: Bridging academic and workplace domains. The Journal of Workplace Learning 8(6), 48-56.

Dixon, N. (1994). The organizational learning cycle. Berkshire, England: McGraw-Hill Book Company Europe.

Marsick, V. (1987). Learning in the workplace. London: Croom Helm.

Mumford, A. (1993). How managers can develop managers. Brookfield, VT: Gower.

Mumford, A. (Ed.). (1994). Authors and authorities in action learning. England: MCB University Press Ltd.

Pedler, M. (1996). Action learning for managers. London: Lemos & Crane.

Pedler, M., Burgoyne, J. & Boydell, T. (1991). The learning company. New York: McGraw-Hill Book Company.

Pedler, M. (1991). Action learning in practice. Vermont: Gower Publishing Company.

Revans, R. (1983). ABC of action learning. Kent, England: Chartwell-Bratt Ltd.

Revans, R. (1982). The origins and growth of action learning. Lund, Sweden: Student Litteratur.

Revans, R. (1980). Action learning: New techniques for management. London: Blond & Briggs, Ltd.

Revans, R. (1971). Developing effective managers. New York: Praeger Publishers.

Weinstein, K. (1995). Action learning: A journey in discovery and development. London: Harper Collins Publishers.

Weisbord, M. (1987). Productive workplaces. San Francisco: Jossey-Bass.

The late Robert L. (Lex) Dilworth was an Associate Professor Emeritus of Virginia Commonwealth University in Richmond, Virginia, USA. He received his doctorate from Columbia University in Adult and Continuing Education. His specialties included human resource development (HRD), action learning, and organization development (OD). He spent a number of years involved with action learning internationally, including time spent in extensive collaboration with Reg Revans and Albert Barker of England, as well as Verna Willis at Georgia State University. Before he entered his educational career, he was a Regular Army Brigadier General in the U.S. Army. His military assignments included service as The Adjutant General (TAG) of the U.S. Army.

This article originally appeared in Performance Improvement Quarterly, 1998, Volume 11, Number 1, pp. 28-43.

ITAP Global Resource Team Offices

Seoul, South KoreaSingapore / China / PhilippinesMunich, GermanyPrinceton, New Jersey, USA

ITAP Asia Pacific

OD Corea / ITAP Asia Pacific
Seoul, South Korea

Contact now →

ITAP Singapore / ITAP China / ITAP Philippines

ODE Consulting /

ODE Consulting (China) Co., Ltd. / 
Shanghai, China

CMC Business Solutions / 
Manila, Philippines

Contact now →

ITAP Europe

TWIST Consulting / ITAP Europe
Munich, Germany

Contact now →

ITAP Americas

ITAP Americas
Princeton, NJ, USA

Contact now →