"What's systemic thinking got to do with anything? Levels of learning inorganisations."
by Mark McKergow
"Learning", both by individuals and in organisations, is proving to be one of the key business topics of the 1990s. There is currently a lot of writing about the idea of the learning organisation, and about how good it will be when we have them. It seems safe to work with the assumption that people in organisations can learn, and that the organisation can therefore change. But what kinds of activity will actually be going on in a learning organisation? How might we make concrete steps towards this goal? I seek to put forward auseful framework to help understanding of the many processes which may go on in a learning organisation, and their relative purposes.
Mike Pedler, John Burgoyne and Tom Boydell have discussed the idea of a "learning company", and given glimpses of their ownideas of the sort of things such a company would do. They have defined a learning company as;
"a company which creates learning opportunities for all its members and is able to transform itself as a whole."
Peter Senge, in his book "The Fifth Discipline", has perhaps gone further in terms of describing what a learning organisation would do. He describes a learning organisation as;
"a place wherepeople are continuously discovering how they create their reality. And how they can change it."
Senge's "Fifth Discipline" is systemic thinking, which he says will be a vital part of the ways in which learning organisations will build their capabilities to achieve their own business goals. Combined with four other disciplines, namely mental models, personal mastery, shared vision and team learning, Senge says that learning organisations can be created.
These are undoubtedly fine words. But how do we actually do it? And where do other current aspects of organisations and development such as TQM and business process re-engineering (and the investment already made in them) fit in? Let us begin by considering a framework for different levels of learning in organisations.
To explain this framerwork for different levels of learning, I will use two examples, both from outside the normal world of "management". Having developed our model of levels of learning, we can then see how it applies to the world of business.
First, a sporting analogy:
Imagine a game of soccer in progress, with the attacking team taking a corner kick. The ball arches towards the goal, and is punched away by the goalkeeper. Immediately, the defence comes away from the goal line, and the ball bobbles around outside the penalty area. The attacking centre forward lines up a powerful shot - and the ball goes over the bar. He's missed!
Let us suppose that the attacking team wish to learn from their experience. They could reflect on the incident. They might re think their strategy for taking corners, bring up a big centre back to intimidate the goalkeeper, practice shooting at the goal, and many other things. That might happen during the game, or more likely back at the training ground on Monday morning. The idea will be that, when a similar corner arises in the future, they will be better equipped to score.
Next an analogy from the world of drama:
A local theatre group is putting on a romantic tragedy. On the first night they reach the point in the action where the hero has to leave his lover, because his family demand it. He turns to say farewell.... and forgets his lines. Prompt!
How might the actors learn from this? They might rehearse some more. The actor playing the hero might go and learn his lines. If he still can't remember them, they might be written on his sleeve to help him.... and soon.
Now we can start to put together a framework for learning. The framework consists of a number of levels. Let us put the actual events which occur this time (the bit initalics above) as actions on a certain level. We can call this the "doing" level. Now, imagine the various learning activities mentioned above, which are designed to improve what happens next time. These involve examining the possibilities, deciding on what to try next, and putting it into action.
We now have a "doing" level and a first "learning" level. The range of activities looks like this: (Figure 1)
It's interesting to reflect that learning and doing may be quite different activities. Simply "learning from experience" isn't as easy as it might be - we not only have to have experience, but, separately, use it as a basis for learning. The football team plays 90 minutes on Saturday - but trains for it all week. They do not simply play game after game of football, two in the morning, two in the afternoon and an hour off for lunch, in the hope of getting better at it. Similarly, the actors do not run through the whole play again and again - they work on parts of it, looking at different options and possibilities. In a learning organisation, we might expect to spend some time engaged in learning activities, focussing on how to improve, as well as doing activities. This could be imagined in our framework as a shifting of attention from the doing level to the first learning level.
We can now apply the framework shown in Figure 1 to organisations and businesses. On the "doing" level we have the activities that the customer actually sees or is aware of. These might include manufacturing, marketing, customer service and many others. This is equivalent to the football team's performance in front of the fans, or the actors' in front of their audience.
Now let us focus attentionto the first learning level. Here, the organisation carries out work to make the things on the doing level better. We find many familiar activities; problem-solving and quality improvement teams could be said to operate on this level. So too do programmes of cost reduction (such as IBM's current contraction), and training to do the job better. Process reviews, where groups or individuals reflect on "How did that go?", and "What could we do better next time?" are basic examples of this first level learning. Boeing's development and launch ofthe new 767 was a triumph of first level learning - existing practices were reviewed, built upon and learnt from. The 767 launch had all-time low defects and difficulties. The same framework as Figure 1 applies to business organisations too.
The experiential learning research of David Kolb, and of Peter Honey and Alan Mumford, shows this basic pattern. The actual activities carried out at the first learning level may be simple (like reflecting on how to do better) or complex (like TQM). However, the basic pattern remains.
Must these first level learning processes be carried out separately from the actual "doing" of the jobs concerned? There is some evidence that they must, even if the shift of attention is momentary and informal. To actually work out what you do, precisely as you do it, and decide how to do it better, can be really very difficult. Try explaining to someone exactly what you do when driving a car, as you do it, to experience these difficulties first hand!
Peter Senge refers to "team learning" as a key activity in a learning organisation. This equates to the kind of practice games that football teams have between each other, or rehearsals by actors, to see how their learning is coming along. Would you not only support, but bet your livelihood on a football team which didn't practice? No - but that's exactly what many organisations do with their management and working teams. Practice is another kind of first level learning.
Is first level learning the end of the story? No - there is another fundamentally different way in which organisations can learn to do things differently.
Let's go back to the football match described above. The attacking side are trying to get the ball into the defence's goal. The defence might be expected to get between the attackers and the goal to prevent them. Yet, after the ball was punched clear, the defence ran away from the goal. Why?
Similarly, in the play, our hero is about to leave the woman he loves. Surely, he might tell his family to get stuffed, and stay with her, but he doesn't. Why? The reason for both these rather odd situations is the rules which are underlying the actions.
In football, the offside rule allows the defence to regain possession if they can leave any attacker between all the defence (except the keeper)and the goal mouth, and the attacker then receives the ball.Hence the rather bizarre behaviour (on the face of it) of leavingthe goal undefended at the crucial moment. Indeed, if we didn't understand this particular aspect of the rules of football, we would be nonplussed (as indeed are many close relatives of fans at the offside rule!).
In our theatrical example, the rule concerned may be that the script says that this is what happens- so the actors do it. The audience may be rooting for the hero to stay with the heroine - but if the play says he goes, then he goes. Sad, but normally true.
So, the rules of the situation also govern what actually happens. What would happen if we changed them? That would depend on what we changed them to. If we do change the rules, then what happens to the actual actions? Football without the offside rule would be a rather different game. The play in which the hero succumbs to temptation would be a different play. Maybe better, maybe worse, but different. In both cases, however, further first level learning would be required to improve the actual performances, under the new rules.
So, we can change what actually happens in two ways. We can learn to do things within the rules (first level learning), and we can change the rules. We can call the latter option "second level learning". This kind of distinction has its origins in cybernetics, and was originally introduced into management thinking by Chris Argyris and DonaldSchön in the late 1970s.
They referred to "single loop" and "double loop" learning rather than first and secondlevel learning. Argyris and Schön wrote about the difference between changing things within a framework (single loop/firstlevel) and changing the framework itself (double loop/ secondlevel). In order to go double loop learning, it was necessary to find the "governing variable" of the system under consideration, and change that. However, not much practical use has been made of the ideas until more recently. Changing the "governing variable" is the same as changing the "rules" of the system. The framework of learning levels including second level learning is shown in Figure 2.
To achieve second level learning, we have to change the rules. But not just any old rules, and not just in any old way. We first have to decide what we want to achieve after the changes. It's no longer enough to want "fewer defects" or "costs down" - these are ways of doing the same things, only better. Instead we must decide what we want more of. We must then seek out the rule which governs those things we want to change, and adjust it for a different rule which produces those things we want instead.
This is where systemic thinking comes in. In order to find the particular rule which governs the bits we want to change, we need to consider the system from a viewpoint outside it. By starting to see our situation to terms of loops of events rather than simple cause and effect relationships, we can find out the rule which will actually allow us out of that system and into a preferable one. Sometimes, this rule will be a long way from the area where the changes are wanted, and can appear rather perverse or counter-intuitive.
Let's take a simple example. There was an account of an emergency rescue on television recently. A young man was canoeing in a canal. He capsized, got out of his canoe and was swept over a weir. The swirling water at the base of the weir buffeted him around, keeping his head underwater for long periods and threatening to drown him. He did the obvious thing -struggled for air, and kept trying unsuccessfully to flounder to the surface. As he was on the verge of losing consciousness, he had an idea for one last effort to save himself. He undid his life jacket, and sank into the water! Within seconds, he had been pulled clear of the swirling currents, and appeared a few yards downstream.
On the face of it, for the man to remove his lifejacket was suicidal. However, it saved his life. The buoyancy of the jacket was the very thing that held him in the dangerous currents close to the weir. The surrounding water simply flowed under the swirl, and continued downstream. The "obvious" thing - keeping the life jacket and keep trying to surface - would have been fatal. The kind of thinking that led this young man to remove his life jacket is precisely the systemic thinking which Senge refers to, and the change in the situation's rules which he made saved him.
Let's be clear about what we mean by "rules" here. We don't just mean the written rules, the laws which apply, the corporate policies or whatever. What I mean by the rules is that set of things which defines the activity in question, and makes it the way that it is. We not only have to examine the formal rules. We also have to include the "rules" we all carry around in our heads - the collection of presuppositions, assumptions and previous learnings by which we each make sense of the world every day, and of which we are often actually unaware.
How can organisations use second level learning methods? For a start, they can realise that there's more to learning and changing than first level learning! The first level learning activities described above all take place within a framework or set of assumptions. It's interesting to wonder what might happen if those rules were changed in a suitable way.
There are interesting examples of second level learning at strategic levels in organisations.
Some realignments of the marketplace come into this category. The upsurge in interest and design (not to mention prices!) of training shoes over the last few years is one. Originally, the market for trainers was simply that - shoes to play sport in. Gradually, young people started to wear the shoes as leisure footwear. On the streets the trainer started to become not just a shoe, but a fashion accessory, and a vital one at that. The number of designs and colours mushroomed, as did prices. In other words, the rules changed.
Businesses can change the rules too. The original British motorcycle industry was world number one for many years - playing by their own rules. The Japanese manufacturers changed those rules by introducing new fangled rider-friendly aspects like reliability, electric starter and indicators. The UK manufacturers continued to play by their own rules, and were finished - because there weren't as many customers getting pleasure from tinkering with their machines as those wanting to ride them!
An example of a change to the formal rules of a system is the introduction of unleaded petrol to the UK market. For some time, unleaded appeared to be offered by a few filling stations as an interesting extra for eco-freaks to put into their vehicles. Then, the Chancellor of the Exchequer reduced the tax on unleaded, compared to leaded fuel. Within months, unleaded was available at many more filling stations, and hundreds of thousands of motorists were getting their cars converted. On this occasion at least, the Chancellor boxed smart. He didn't put energy into advertising unleaded, or trying to sell it to forecourt managers - he changed the rule that held the market in position, and at a stroke changed that position.
The above examples are all very well if you're a company executive. But how about the rest of the workforce - can they use second order learning to change things too? They most certainly can. But because the formal rules are not usually accessible at these levels in organisations, the methods tend to look at the "informal" rules - the mental models of the people involved. And these informal rules influence chairmen like everyone else.
One way is by the process called "Socratic dialogue", where the participants (with the aid of a facilitator) suspend their assumptions and therefore start to notice differences between what they had assumed and what actually is. Another way is to describe in great and observable detail what they would like to do, irrespective of what might be "possible" or "reasonable", before getting to grips with what they need to do, and think, to achieve it. In other words, they will be able to define the rules to fit the end result, rather than taking the result which they get from the (existing) rules.
When people become skilled at second level learning methods, they will have learnt how to not only play the game better, but to change the rules as well. However, this ability brings with it a new set of demands. Once we have learnt to change the rules, we can in principle play any game we like. To return to our footballing analogy, it could be possible in principle to change the rules from 11 men a side, a green field and a dummy pig's bladder to two men and two ladies, a strung bat each and a shuttlecock....if that's what we want. In the theatre, we can write new plays, which perhaps get more acclaim or better reviews - if that's what we want. We have to actively choose what we want before setting out. Imagine changing the rules of football randomly, to see if we get a better game. Then imagine the chaos. Then remember not to do it without being clear about what you mean by "better".
Second level learning methods can focus on the desired outcome, and seek to avoid the traps posed by existing rules, formal and informal. Sometimes desired outcomes are described in great detail. The more detail in the description, the less chance for our mental models to start interpreting and inventing things for us - and the less we fall back into playing by our existing rules.
Some guidelines for actually starting to apply second level learning:
* Become aware of your own formal and informal rules, your assumptions and preconceptions. It's difficult to be entirely sure sometimes about what your own personal rules are - you might do well to get some help from someone with appropriate skills.
* Don't simply substitute someone else's rules for your own. Get outside everyone's rules, and focus on what's actually happening.
* Describe in detail the outcomes you want. The more detail, and the more specific, the better.
* Find the particular rule(s) which will make the most difference. Go for the change with the most leverage on your outcome.
* Change them in a way which will produce the results you described.
* Check that it worked. If there's still a way to go, you haven't yet found the key rule.
Let's finish by reviewing the question posed at the start of this piece. What sort of activities will be going on in a learning organisation? There will be activity at three levels:
The "doing" level; where customers are delighted.
The "firstlearning" level; where we learn how to do the doingactivities better.
The "secondlearning" level; where we learn how to change the formal and informal rules which govern what happens on both the other levels.
We will also be learning to do the activities on the two learning levels differently and better - but that's another article!
"What's systemic thinking got to do with anything? Levels of learning in organisations.",Organisations and People, 1, No 1, pp 16 - 20 (1994)
To sign up for our tips and news please fill in your details below