Return to Really Learning Home page

Briefing Paper Two: Organisational Learning and Learning Organisations

Again this is a lengthy account likely to be of greater interest to those teaching and consulting in this field than to health care practitioners. For a shorter more succinct account please see case study three of Developing Change Management Skills.

Organisational learning (OL) is a term introduced in the 1970s by Chris Argyris and Donald Schön. It draws on, among others, John Dewey’s Experience and Education (1938), one of the most influential texts on adult education, and Kurt Lewin’s (1946) development of action research. It is an academic field of study with a solid research base.

The Learning Organisation is a term that derives in part from OL, and includes a wide range of practical approaches advocated by consultants and practitioners, drawing on the fields of sociotechnical systems; organisational strategy; production; economic development; systems dynamics; human resources; and organisational culture. These approaches do not have the same rigour of research associated with them as do the core concepts of OL.

The three central concepts of OL:

  • Theory of action and theory-in-use
  • Model I and Model II
  • Single loop and double loop learning

Theory of action and theory-in-use

Argyris and Schön (1996) observed that within an organisational context individuals tend to promote one set of behaviours, and use another set. In explaining this disparity, Argyris and Schön  defined two kinds of theory of action: espoused theories and theories-in-use.

A theory of action has the following generic format: In situation S, if you want to achieve outcome O do activity A. A theory of action includes the values we attribute to O  that make us see it as desirable, as well as causal assumptions we bring that lead us to believe that A will lead to O.

Argyris and Schön suggest that we use espoused theories to explain or justify our actions. In practice, however, and especially when there is any risk of embarrassment or threat, we use a theory-in-use which is at variance with the espoused one.

Model I and Model II

Almost everyone participating in Argyris and Schön’s original research, when at risk of embarrassment or threat, could be seen as having adopted a theory-in-use that Argyris and Schön term Model I. This is a form of behaviours learnt early on in life and which is supported by a set of virtues widely held within society and within organisations. These virtues include:

  1. caring, help and support: give people approval and praise, tell people what you think will make them feel good about themselves, reduce their feelings of hurt – by saying how much you care, and if possible agreeing with them that other people have behaved improperly
  2. respect for others: defer to others when they are talking and do not confront their reasoning
  3. honesty: tell no lies, and/or tell others all you think and feel
  4. strength: advocate your own position and hold it in the face of attack from others. Feeling vulnerable is a sign of weakness.
  5. integrity: stick to your principles, values and beliefs.

 Behaviours associated with Model I include:




Define goals and try to achieve them

Design and manage the environment unilaterally - be persuasive, appeal to larger goals etc

Actor seen as defensive, inconsistent, controlling, fearful of being vulnerable, overly concerned about self and others, or under concerned about others

Maximise wining and minimise losing

Own and control the task -claim ownership of the task, be guardian of the definition and execution of the task

Defensive interpersonal and group relationship - depending on actor, little help to others

Minimise generating or expressing negative feelings

Unilaterally protect yourself -speak in inferred categories, accompanied by little or no directly observable date, be blind to the impact on others and to incongruity, use defensive actions such as blaming stereotyping, suppressing feelings, intellectualising

Defensive norms, mistrust, lack of risk taking, conformity, external commitment, emphasis on diplomacy, power centred competition and rivalry

Be rational

Unilaterally protect others from being hurt - withhold information, create rules to censor information and behaviour, hold private meetings


We apply Model I automatically because we become skilled at it from an early age. Because we are skilled at it, and the better at it we become the more averse to learning we are, Argyris and Schön call this skilled incompetence. If we are operating in Model I and are asked why we behaved in such-and-such a way, we tend to justify our actions by referring to our good intentions: the desire not to hurt other people’s feelings, the wish to advocate a position in accord with our values, and so on. If probed more deeply we will blame the situation we are in on other people, attributing to them negative attributes and motives, for example, their inability to handle the truth or readiness to play political games.

Although hidden from us, the disparity between our theories-in-use and our theories of action tend to be apparent to those we interact with. Do they draw our attention to this? Generally speaking, no. This is because they too have adopted a Model I theory-in-use: they too wish to avoid hurting our feelings and want us to save face. They in turn attribute to us an inability to handle honest feedback or a lack of willingness to work cooperatively with them.

In this situation both participants engage in what Argyris and Schön call bypass and cover up. Moreover, they make the bypass undiscussable, and they make that undiscussability itself undiscussable. This set of activities is known as an organisational defensive routine (ODR). Because ODRs are pervasive, individuals tend either not to notice them or to feel powerless to change them. They can see that they inhibit organisational effectiveness, yet challenging them requires the courage to risk making the situation more uncomfortable. Unchallenged ODRs lead to further fancy footwork, as Argyris and Schön call it, as people get to know the ‘way we do things round here’ and find ways around this. This in turn leads to what they call a state of organisational malaise whose general symptoms include hopelessness, cynicism, distancing and blaming others. Specific symptoms are:

  1. seeking and finding fault with the organisation, without accepting responsibility for correcting it
  2. accentuating the negative and de-emphasising the positive
  3. espousing values that everyone knows are not implementable but acting as if they are.

The following overall sequence:

skilled incompetence, ODRs of bypass and cover up, fancy footwork, organisational malaise, mediocre performance is called an organisational defensive pattern (ODP), and the reasoning at its heart is known as defensive reasoning. Argyris and Schön contrast this with productive reasoning, a way of thinking and talking which enables us to test the validity of our own and others’ theories. To engage in productive reasoning we need to adopt a new theory-in-use, called Model II.

Model II is not the converse of Model I, since we still need to be able to draw on the behaviours and virtues of Model I. We simply need to choose more judiciously and awarely when to use or be guided by them. As with Model I, there is a set of corresponding virtues that support Model II behaviours:

  1. help and support: increase others’ capacity to confront their own ideas, to face their unsurfaced assumptions, biases and fears, by acting in this way towards them.
  2. respect for others: attribute to other people a high capacity for self reflection and self examination – without becoming so upset they lose their effectiveness and sense of self respect and choice. 
  3. strength: combine advocacy with inquiry and self-reflection. Feeling vulnerable during inquiry is a sign of strength.
  4. honesty: encourage self and others to say what they know ( having tested assumptions and attributions) and yet fear to say.
  5. integrity: advocate principles, values and beliefs in a way that invites inquiry into them and encourages others to do the same.

 Model II behaviours include:




Valid information

Free and informed choice

Internal commitment to the choice and constant monitoring of its implementation

Design situations where participants can be origins of action and can experience high personal causation

Task is jointly controlled

Protection of self is a joint enterprise and oriented towards growth

Bilateral protection of others

Actor experienced as minimally defensive

Minimally defensive interpersonal relations and group dynamics

Learning oriented norms

High freedom of choice, internal commitment and risk taking

To be able to implement Model II we usually need to slow down our reasoning and increase our capacity for analysis and reflection, otherwise we unwittingly revert to Model I. One way of doing this is to use the approach introduced by Argyris and Schön, known as the Left hand column.

Think of a work situation you are concerned about.

Think of a conversation you either have had, or would like to have, with a colleague or someone else involved in that situation – someone you perceive as contributing to the problem.

Divide a piece of paper into two columns. In the right hand column write down what you actually said or did  (or would like to say and do).

In the left hand column write all the things you would be thinking in response to what the other person was (or would be) saying.

Review the left hand column entries to see how often you fell into Model I thinking. For example, consider asking yourself the following questions:

  • How often did I attribute negative motives or evaluations to the other person’s performance and yet not want to tell them?
  • How confident am I that I attributed those negative motives etc correctly? If I use the ladder of inference can I see that I have used data very selectively and added other beliefs that are not necessarily appropriate?
  • Did I advocate my own position firmly to the exclusion of the other person’s?
    Did I tell the other person that I care about his or her views while not truly being open to these?
  • Did I find a third party to blame for the overall situation, e.g. budget, the Government, and so on?

Model II leads us to be able to reason productively rather than defensively and to:

  1. strive to make premises and inferences explicit and clear
  2. develop conclusions that are publicly testable
  3. test them in ways that are independent of the logic used by the actor involved
  4. while taking action, reflect and be aware of own thoughts and feelings
  5. be clear about the position we are advocating and about any evaluations or attributions we make of others
  6. check constantly for unrecognised gaps or inconsistencies and encourage others to do the same
  7. combine taking the initiative with being open to any constructive confrontation of own views, evaluations and attributions.

Single loop and double loop learning

Sometimes the kind of learning that is needed in an organisation, as discovered through an organisational inquiry, is simple. If something has gone wrong and we put it right that is single loop learning. If we look deeper we may find that what went wrong did so because of the way systems were designed, and if we change the systems we can prevent it happening again, and this is double loop learning. If we reflect further still, about what prevented us from seeing that the systems needed changing, before something went wrong, we call this deutero learning.

A Model I theory-in-use is often appropriate where single loop learning will suffice. But for double loop learning Model II is necessary.

A learning organisation is one in which there is a lot of double loop and deutero learning going on at individual and organisational level.

The view of Learning Organisation practitioners from the field of Organisational Learning

Argyris argues that most advice from academic researchers, management consultants and executive development programmes, including those advising on Learning Organisations reinforces ODRs, and indeed that these advisors are often guilty of defensive routines themselves.

In order to minimise the risk of advisors exacerbating the situation he suggests that if advice is to be implementable it must have three essential parts:

  1. A causal theory (if you do x, then y will happen)
  2. It must illustrate x at two levels: the action strategy – for example,  ‘combine advocacy with inquiry’ – and actual examples that illustrate the saying and doing
  3. It should articulate the values that govern the suggestion – for example, if the governing value is ‘to win’ this should be expressed. If these governing values are made explicit then the advice can be awarely accepted or rejected by client organisations.

 He points out that even when people know what they should do there are often situations in which they are unable to and suggests that everyone is warned that ‘it is unlikely that you will be able to produce this advice when you are dealing with issues that contain components of embarrassment or threat. You are likely to be unaware of this fact while you are implementing this advice. Or if you are aware you will tend to blame factors other than yourself’.

The Learning Organisation

As we have seen, the Learning Organisation (LO) is a term often used to encompass a variety of approaches. These include the Learning Company (Pedler et al., 1991), ‘learning systems’ (Nevis et al., 1995) and ‘learning organisations’ (plural) (Davies and Nutly, 2000). One aim these approaches have in common is to show how organisations should be designed and managed to promote effective learning. Several approaches build upon the work of Argyris and Schön discussed above as well as other, related concepts such as ‘defensive’ and ‘offensive’ adjustment (Hedberg, 1981).

As well as texts that address specifically the learning organisation there are many that take a holistic approach to the design of organisations and these can be just as useful in considering how to promote learning. The seven S framework, for example, introduced by Tom Peters et al, can be used for just this purpose.

Perhaps the most widely known of the books on Learning Organisations is Peter Senge’s The Fifth Discipline (1990).

Senge draws on the work of Argyris and Schön  as well as the work of Jay Forrester and other systems scientists. In it he describes a LO as one

‘where people continually expand their capacities to create the results they truly desire, where new and expansive patterns of thinking are nurtured, where collective aspiration is set free, and where people are continually learning how to learn together’.

Senge suggests that learning is a journey which has no final destination; learning is never -ending and it is the journey of discovery itself which counts. Moreover, the more we learn, the more we become aware of our ignorance. He also asserts that the ability to learn faster than competitors may be the ‘only sustainable competitive advantage’ . However, he does not explain why this would be any more sustainable than other competitive advantages that involve new disciplines or technologies, and the overall tone of his work is inspirational and optimistic, in contrast to the more measured and sceptical tone of Argyris and Schön. For instance, Argyris and Schön point out that Model II theory-in-use is an ideal and have conceded that they are unaware of any organisation that has fully implemented a double loop learning system. A similar caveat needs to be given with claims made for LOs. For practising managers, therefore, Senge’s messages are likely to prove valuable when combined with a thorough understanding of the work of Argyris and Schön.

Senge lists five disciplines which he suggests will lead to the kind of LO defined above. These are:

Systems thinking – seeing processes rather than events, wholes rather than parts, dynamic rather than detail complexity

Personal mastery – the discipline of personal growth and learning, continually striving to clarify what is important, to be clear aboutthe vision we are aiming for, and at the same time being ruthlessly clear about the current reality

Mental models – the tacit models we use to interpret and interact with the world. This directly refers to Argyris and Schön’s Model I theory-in-use, skilled incompetence and ODRs

Shared vision – the vision that encompasses the personal visions of all those working within the organisation

Team learning – the process of aligning the personal visions and developing the capacity of the team to work together to achieve the results they are after. Here Senge promotes the virtues of Model II.

Key messages arising out of the five disciplines include:

Systems thinking
The essence of systems thinking, suggests Senge, lies in seeing inter-relationships rather than linear cause-effect chains and seeing processes of change rather than snapshots. It starts with an understanding of the principle of feedback and builds to learning to recognise types of recurrent structures. Click here for more information  on systems thinking.

Personal mastery

People with a high sense of personal mastery, says Senge, have a clear purpose, see current reality as an ally, feel connected to others and never arrive on their learning-in -progress journey. They see failure as an opportunity for learning, for example, about inaccurate pictures of reality, strategies which did not yield the results expected, and clarity of vision. They do not see failure as a sign of unworthiness or powerlessness. Above all, people with high personal mastery have a commitment to truth, to reality, to really understanding situations, including their own behaviour.

Mental models

To challenge mental models Senge draws on Argyris and Schön to recommend the use of the left hand column exercise referred to above and the ladder of inference, to recognise leaps of abstraction which often lead us to jump from initial data to the wrong conclusions. He similarly describes using reflection and inquiry skills to identify ODRs.

Shared vision

Senge indicates that without a shared vision it is not possible to have a learning organisation. This claim is probably aspirational. Visions of the people at the top of organisations are, he says,  usually just that, they may not be shared except at the top. Senge argues that to foster a learning climate in which individuals are encouraged to have personal visions means not imposing one top down. He also suggests that there must be a shared appreciation of the current reality, and of the gap between the vision and the reality and the work needed to reduce the gap.

Team learning

Team learning is a process of aligning the individual within the team and developing the capacity of the team to achieve the results it is seeking. To facilitate team learning Senge advocates distinguishing between dialogue and discussion, and moving awarely from one to the other to explore different topics. His description of dialogue coincides with Argyris and Schön’s Model II. He also describes it as being a free, creative exploration of complex and subtle issues which involves deep listening, suspending one’s own views, being aware of one’s own thinking processes and of others’. Discussion is the presenting of views and defending them in a search for the best argument. This will usually be appropriate for more straightforward issues. Unless the decision to use one rather than the other is explicit, Senge suggests, the result will be an ineffective amalgam. Dialogue, he observes, will probably always need the involvement of a facilitator of some sort.

Introducing organisational learning

Argyris and Schon have found that defensive routines are alterable, indeed that given a commitment to change these defensive routines then change can be made relatively easily . Furthermore this change of this kind is realistic and not utopian.

They suggest that it is not necessary to think in terms of massive change programmes – that you can start small, especially if you start at the top. Also that you don’t need advisors who are highly skilled in changing defensive routines before they start, they can learn as they go along. They do say that people aiming to tackle these defensive routines should be able to:

  • use publicly compelling and testable reasoning when dealing with defensive routines minimise the use of their own defences to protect themselves or the client translate any error into an opportunity for learning, and
  • design programmes within their competences (so their anxiety levels are not raised).

They will be able to be effective before they are completely skilled themselves, if they work with a professional who is more skilled than they are in a ‘supervisory’ (as in clinical supervision) capacity.

 Argyris’ methods for teaching Model two all involve action and he suggests that this is why there is no forgetting curve, why people who have become familiar, through practice, with Model two will continue to use it in the same way that people who have learned to can always ride a bicycle.

A programme for encouraging organisational learning

As people’s theories-in-use are tacit, hidden from the person using them, they cannot be ascertained by interview or reflection, they can only be constructed through observation. Therefore, the first step is observation by a skilled outsider, able to draw up a map of the theories-in-use, the values that seem to be governing them, the action strategies that people are adopting, and the consequences of these. Often there are first-, second- and third-order consequences (the first set lead to the second and so on). This map is then fed back to the participants who then discuss all the elements and the feedback and feed -through processes within it, with the observer and each other. The aim of the discussion is that everyone agrees that these ODRs exist, and that they want to tackle them.

According to Argyris and Schön people most usually do recognise ODRs, and are keen to address something they have felt powerless to change. They stress that this ‘aha’ is not sufficient, that skilled incompetence is so great, and the interventions below are necessary. These can be thought of as  four main phases, as follows.

  1. At a first meeting the concepts of Model I and Model II are introduced. Often participants are fired with enthusiasm and believe they can change their behaviours instantly as a result of the ‘Aha!’ of discovery they have just experienced. Argyris and Schön have observed that instant change is not possible, because of the skill with which people use Model I and the way it is embedded into our habitual ways of responding.
  2. Participants are then asked to undertake the left hand column exercise described above, thinking about an issue of particular concern to them and imagining a conversation they would like to have with the person they perceive to be causing the greatest problem.
  3. They bring their imaginary scripts and the associated thoughts and feelings with them to a meeting at which they discuss their scripts and the left hand columns with the others. This process allows them to slow their thinking down sufficiently to be able to use Model II concepts. Argyris and Schön suggest these discussions are taped and that the tape is given to the person whose issue is being discussed. They find that people often use Model I thinking during the discussion to attribute words or phrases to others which are not in evidence when the tape is played back.
  4. It is useful then for participants to be able to have further one-to-one work with the facilitator in which they work on real issues, often involving unproductive relationships with each other. It can be useful to gather the learning experiences together in a further reflective seminar.

The work can then be expanded to the rest of the organisation in a similar way. If an organisation is serious about doing so, Argyris and Schön report some success with thedevelopment of a cadre of internal change agents and educators who foster a particular expertise in working with Model II thinking, in addition to their day jobs.


Chris Argyris and Donald Schon,  Organisational Learning II, Theory, Method and Practice, 1996, Addison Wesley OD series
Chris Argyris, Overcoming Organisational Defenses, 1994. Prentice Hall
Peter Senge , The Fifth Discipline , 1990, Century Business
Senge et al, The Fifth Discipline Field Book, 1994, Nicholas Brealey Publishing
Mike Pedlar, The Learning Company
Brian Quinn and Henry Mintzberg, The Strategy Process, Prentice Hall
For more about the literature and background of the Learning Organisation, see Iles and Sutherland: Organisational Change (2001), page 64.