This report examines how ideas related to Complexity Science are helping organisation to achieve optimum performance. It is based on presentations by Susanne Kelly and Christopher Davis at an LSE Complexity Group seminar, including material generated from their discussions with the audience.

Susanne Kelly is Vice President of Citigroup's Corporate Technology Office (CTO). She started her business career over thirty years ago as a programmer at AT&T and subsequently developed her interest in organisational behaviour. In 1991, she joined the CTO to work on organisational change in the Citibank technology community. Citibank, a subsidiary of Citigroup, has a presence in 100 countries, a staff of over 90,000 and relies heavily on technology to deliver global financial services. In Part 1, Kelly explains how Citibank, like other organisations, has been going through a major transition to an 'Information Age'. She describes how this change was facilitated at Citibank by introducing radical new ways of thinking and being to remain effective as an organisation and sustain a competitive edge. Kelly identifies four 'Simple Rules' derived from applying Complexity Science to her worldwide work with organisational change. She explains how these rules assist in surfacing generic patterns of functional versus dysfunctional group interactions. The patterns provide a sound framework for facing the key challenge of organisational improvement: self-diagnosis in one's own local environment, followed by effective intervention in self-organisation to create more productive emergent behaviour.

Christopher Davis is Managing Director of Business Design Associates (BDA) Germany, based in Frankfurt. BDA was formed in San Francisco in 1989. It has subsequently extended its operations to Latin America, Canada and, since 1997, a European operation spearheaded by BDA Germany. The approach adopted by BDA to organisational diagnosis and intervention is grounded in the study of 'speech acts' and a perception of human society operating through a self-organising network of commitments. This has been successfully applied in many companies, including Citibank. In Part 2, Davis shares some of that experience. BDA and its distinctive approach were originated by Fernando Flores. At the age of 29, Flores became the youngest cabinet minister in Salvador Allende's Chilean government in the early 1970s. He rose quickly to become Minister of Finance, but was jailed after the overthrow of President Allende's government. While in prison, he read the early works of Humberto Maturana on autopoiesis. When he was released and went to the University of California at Berkeley, he was exposed to work on Speech Act theory and to the writings of the German philosopher Martin Heidegger. In 1982, he hosted a series of workshops and started a business. Davis attended one of Flores' first workshops and started working with him in 1983.

This report was edited by London-based Editorial Consultant Malcolm Peltu.

Note that Tables are at the end of the report.

PART 1

EVOLVING THE ENTERPRISE

BY SUSANNE KELLY

Business in transition

Due to changes being experienced in the economic environment, people at Citibank have been interested in Complexity Science since the mid 80s. There has been a growing awareness that our enterprise, industry, competitive environment and economy are undergoing a major transition.

An increasing percentage of the GNP in many countries is shifting from activities aligned with Manufacturing to Information-based service. Although this represents a significant change of focus, we acknowledge that information was key to conducting business in the past, and that manufacturing activities will continue into the future. In fact, many organisations, including Citibank, seem to operate in a paradoxical environment with a foot in both camps, using repetitive process to deliver information-based service.

Table 1.1| illustrates how this transition can be broadly characterised as a move from what we call a 'Manufacturing Age', founded on machine-based models of organisation, to an 'Information Age', where more flexible and dynamic models are emerging.

While I refer to the Information Age, some people prefer to call it the 'Knowledge Age', as it highlights the importance of knowledge-based products and intellectual capital. Although Complexity Science recognises information and knowledge as phenomena embedded in neurological and organisational relationships, for our purposes, we'll stick to the distinctions I make in Table 1.1|. We will forego the more abstract debate about the nature of knowledge and information in complex adaptive systems.

Differences between the Manufacturing and Information Ages

From one perspective, Citibank can look at itself as a 'bulk processor': a key distinction of the Manufacturing Age (Table 1.1|). We employ what one might call assembly line procedures to sort customer cheques and render statements. The old manufacturing paradigm regards business essentially as a Newtonian closed system or machine using linear cause-and-effect control to produce commodity products. We also engage in activities concerned with the design and use of technology to deliver rapidly changing, knowledge-based products. This part of our environment is often characterised by uncertainty, ambiguity and non-linear, complex interactions.

Due to today's technology, communications and the nature of knowledge-based products, the Information Age summarised in Table 1.1| has expanded business competition to a global, rather than regional, context. Change has now become a way of life; it appears to be accelerating and at times becomes overwhelming.

Previously, things seemed to be more static. For instance, as a programmer with AT&T in 1965 (automating manual accounting procedures that had been in place for 10-15 years), we developed requirements for systems ¾ and they remained fixed while the system was implemented. Today, the requirements for most systems constantly change as they're implemented. Change in the business environment of the 60s was more digestible because it was periodic and occurred at a slower rate. This made it common from the 60s to early 80s for companies to develop and implement 5-year strategic plans. Citibank and many other enterprises now use scenario analysis rather than fixed long-term plans because we know something unexpected will happen the day after the plan is completed, and new opportunities will emerge as we coevolve with our market.

In the Machine Age, the leadership challenge was managing the events in the strategic plan to a target end-state. Now it is very different. We develop shorter-term scenarios, to understand the implications of moving in one direction or another and align with probable outcomes of different possible futures. A leader in the Information Age must help the enterprise navigate as the future unfolds, envisioning probable scenarios and coaching people in the right direction, at the right time. The challenge is to develop a capacity for change to augment the enterprise capacity to deliver.

Our current business transition also alters our key factors of production and the resources to which we attribute most value. Not only must we focus on material and financial capital, we must amass appropriate human, intellectual and social capital as well. We used to establish control systems to ensure we would not move too fast and get out of control; now, we focus on decision making and communication systems to ensure we do not move too slowly and fall out of the running. We must maintain control while operating at the edge with rapidly changing developments like the Internet and related technology-driven opportunities.

Towards a holistic view of the organisation

In the Manufacturing Age, we saw organisations as machines consisting of parts designed to interact, for end-to-end efficiency, in a sequence of standard steps, to reach a goal. Much of my work in process and systems since the mid 1960s has been looking at how to make the sequence of steps more efficient, more effective and more standardised. The emphasis was on optimising quality, lowering production costs and increasing productivity in repetitive day-to-day operations.

Today, experience in the Information Age tells us that the whole emerges from an organisation's parts interacting in a non-linear way. Hundreds of thousands of transactions take place daily in a large enterprise. People interact independently and we can't guarantee that resulting actions will occur in any particular sequence. In our fast paced world the primary aim is to be FIRST (and of course best if possible). The quest is for competitive edge through innovation.

In a non-linear, emergent world, the focus shifts from end-to-end efficiency to 'micro-to-macro integrity'. In my work with organisational improvement, I found that the main problem was not the lack of formal processes, but that people were not following them. The minute many processes were put on paper, they were already obsolete because someone had already found a better way to do it. This made me realise that the key is to keep the integrity of the way people actually behave equated with the processes we put down on paper.

We write processes on paper so that people can coordinate their interactions. These bits of paper are our coordination points. However, they don't coevolve with us, as our behaviour is changing rapidly. People learn and change naturally, but our papers don't; they stay in black and white as originally captured. So, we need to learn how to keep our artefacts coevolving with us at our speed. We also need to understand where we should have artefacts ¾ and where we should not because they get in our way, rather than help in coordination.

This ties to our need in the Information Age to focus on constant feedback, instead of standardisation. We don't have time to get everyone doing things the same way. By the time they are doing that, the way is obsolete. We do, however, need to know what other people are really doing to allow us to work together.

Citibank's interest in Complexity Science for a changing world

Citibank has explored Complexity Science because it offers new models for absorbing the changes we face in current business transition. Complexity Science is the interdisciplinary investigation into patterns and the macro phenomena which emerge from non-linear interactions of micro elements in an 'open adaptive system'. This is defined as:

1. a system in which the whole is manifested through the interaction of parts;

2. open, in that it exchanges energy, information and material with its environment;

3. adaptive, as it uses imported energy to self-organise and to maintain the stability of some internal elements, while changing others in response to changes in the environment

Citibank has been one of the funders of the Santa Fe Institute since 1984, when it was established to explore Complexity Science. Our initial interest was in using non-linear dynamics and concepts of emergence to help us better understand and predict our business, industry, market and economic environment.

In 1991, when I joined Citibank's Corporate Technology Office (CTO) to work on organisational behaviour change for 5,000 technologists around the world, many of them were acting like independent artists or craftsmen. We needed them to develop the discipline of an engineering team. This meant shedding some bad habits and learning new ones.

Many technologists had difficulty delivering joint results because they weren't always coordinating what they were doing and, worse yet, were set in loops of behaviour that kept them where they were. Some followed their own agendas and didn't share information. Others evolved their work as they went along, rather than engaging in coordinated planning.

Technologists need to display artistry and discipline. When I studied music composition as an undergraduate, I started by looking at great musicians almost in mathematical terms, in order to understand the patterns they used. A person can then apply their own creativity to the discipline of the masters and create something new. An organisation is like an orchestra, where all players must blend together. We had members of the orchestra who wanted to be independent artists and who had difficulty coming together for a collective performance. Eventually, models for Complexity Science helped us to get groups of technologists to work on conscious coevolution so that appropriate organisational behaviour could emerge.

In 1995, Citibank began working with the Centre for Adaptive Systems Applications (CASA). A group of physicists from Los Alamos National Laboratory in New Mexico established CASA to build banking industry, market and economic models for us. They developed risk models and customer behaviour models that gave us insights into our outside context. Inspired by their work, I began to read books on Complexity Science. My interest was growing in it's potential for understanding patterns and emergent phenomena within the organisation. In my work, I had begun to see common organisational patterns, regardless of whether I was working with technologists in Japan, Singapore, Hong Kong, Brazil, Argentina, Chile, Belgium, England or Germany.

Complexity theory and organisational behaviour

As I learnt more about Complexity, it intrigued me that we had been using Complexity tools only to look at the outside world, but not inside the organisation. There seemed to have been a jump from an understanding that biological beings are complex adaptive systems to looking at our whole industry and the economy as complex adaptive systems. This leap had missed seeing the organisation itself as a complex adaptive system. Therefore, I raised my hand in 1997 and proposed to our CEO, John Reed, that we study our own enterprise as a complex adaptive system. He found this to be an intriguing thought, so gave the go-ahead for me to begin a Complexity Theory and Organisational Behaviour research project.

As I began to talk to managers, I realised the failure to apply Complexity to the internal environment arose because so many retain an illusion of control. It's easy for them to see the outside world as a set of nested systems that emerge as an uncontrollable web of activity. They see the industry and market as systems coevolving in step with the environment in which they are nested. However, they were leaving out an important piece in the nested web: the organisation itself. Managers could see that the external things fully fit the model of being open adaptive systems but still regarded the enterprise as a closed machine. We needed to bring the enterprise back into the picture and try to understand how Citibank operates as an open adaptive system.

Interlocking behaviour and stability

My interest in Complexity took me on my search for a few simple rules that might be applied locally at each level of an organisation to give rise to many diverse patterns of behaviour. While contemplating these simple rules, I also began to struggle with the question of stability. As complex open systems, we exchange energy and information with our environment. If we absorbed all available information from the environment, we would indeed display chaotic behaviour. So what phenomena gives rise to our stability? I followed two paths in searching for my answers.

My first insight came when I was at a seminar with Humberto Maturana who said that his 'light bulb' on understanding biological stability came in the form of a closed loop. His realisation was that DNA gives rise to a certain kind of protein, which gives rise to DNA, which gives rise to the protein, and so on to create a closed loop within the open system. Some people at the meeting said it would be more accurate to refer to 'enzymes' instead of proteins, but that was beside the point which had made my own light bulb go on. I realised that a system totally open to information, energy and material would be truly chaotic unless it was influenced by a 'closed' loop similar to the DNA-protein-DNA flow. That set me thinking about the nature of such closed loops in organisations.

Eventually, I saw that stability in human open systems was derived from our own behavioural loops: we do something that succeeds; we like success; we do more of it; we get better at it; we do it more; we succeed again; and so on until we develop a closed loop. This happens at the individual and organisational level; and when it does, we begin to close ourselves off from any information that would tell us to do other than what we are doing. IBM was successful in the business of mainframes and continued to focus on mainframes, liked it, did more of it, was successful, liked it, and did more of it to a point where a loop was created and closed off information saying: 'Hello! PCs are coming to do things in a new and better way'. This also explains the Peter Principle. People approach tasks in the way that has always worked for them in a given context while closing their eyes to the fact that the context around them is changing and they are no longer coevolving. Until, all of a sudden, they are no longer effective in that new environment.

My other breakthrough in looking at the question of stability in open systems came when I began to focus on the factors that drive our behavioural loops. I started thinking about what psychologists like Jung had said about human behaviour, when one day in my office I happened to catch sight of my Myers-Briggs reference book (Kroeger with Thuesen 1992). The Myers-Briggs Type Indicator is based on Jung's work with personality types and uses four indices to describe personality types: Introvert/Extrovert (I/E); Sensing/iNtuitive(S/N); Thinking/Feeling (T/F); and Perceiving/Judging (P/J). I began to rethink these indices in terms of how individuals behave as complex adaptive systems:

1. I/E is about how we get our energy;

2. S/N is how we get our information;

3. T/F is how we make choices about our environment;

4. P/J is about how we prefer to coevolve and whether we like to plan and coordinate our convolution with others ¾ or do so spontaneously.

One of the things I love about Complexity Science is that you often find the models you have applied usefully in the past gain new meaning when seen through a Complexity lens. Thinking of other models from a Complexity perspective may also explain why they don't work very well.

Group interlocking behaviour

This led me to think about organisations as complex adaptive systems. Groups of people exchange energy, share information, align choices and coevolve. I used my new Complexity interpretation of the Myers-Briggs indices to derive four natural group-behaviour elements of open human systems:

1. The exchange of energy (Competitive … Collaborative). When you interact with another person, there is automatically a relationship formed. That relationship can be anywhere from totally competitive to fully collaborative. Competition can provide the adrenalin necessary for a new level of performance. Or you could get a synergy and feeling of collaboration that enables you to do things in concert that you were unable to do before on your own.

2. Sharing 'information' to learn together (Limited … Openly and Fully): One person may prefer sensing information, while another may prefer getting information from an intuitive standpoint. But if we are able to discuss and share our perspectives on how we see things, we can get a fuller picture of the context we are in. We can share information for learning together either in a very limited way, or much more openly and fully, again along a continuum.

3. Aligning choices for interaction (Shallow .. Deep Commitment): You may make your choices based on feelings, while I may think mine through more logically. If we are making choices that will affect both of us, and the people we work with, we need to align choices through deep commitment to one another. In business, a deep commitment is indicated when someone says something like: 'Yes I will do that for you by this date and time' and you know they really mean it. In contrast, phrases like 'Let's talk about it sometime over lunch' or 'I'll call you' indicate a more shallow commitment.

4. Coevolution in language and action (On the Fly … With Coordination). This determines the degree to which people in a group prefer to communicate their intentions as they coevolve, along the spectrum from liking to act independently on the fly to seeking a more coordinated coevolution.

I believe these are the four natural elements of open human systems and suggest four simple rules for effective organisational behaviour. Analysing different patterns of these four elements offers some important understandings of why organisations behave in particular ways. I have been reassured that these patterns reflect what is really happening because so many people who have heard me describe them, recognise the phenomena and ask if I've been working in their organisation.

One extreme: a vicious cycle driven by fear

At one extreme (Figure 1.1, Patterns from 4 Elements: One Extreme…), an organisation's four natural open system elements could be characterised as:

1. Exchange of Competitive Energy

2. Limited Learning

3. Shallow Commitment

4. Coevolution on the Fly

In an environment where people compete for survival with others in the organisation, an energy of mutual fear is generated, which causes the people to engage in deception, which engenders mistrust, which causes people to compete with each other. This creates a feedback loop based on fear that propagates itself. Once an amplifying or positive feedback loop gets started and continues unchecked, it often leads to runaway.

Part of employing deception is to agree to anything, which becomes an element in a new closed loop of shallow commitment. One of the unwritten rules in business says that you are not supposed to cause conflict or be confrontational. You are supposed to be part of a team, so you pretend to be cooperative by agreeing to any request, whether or not you intend to fulfil it. You intend your best effort: 'I really will try to get to that, although I have so many things on my plate … but maybe I will be able to do it…'. You eventually deliver what is possible in the circumstances: 'I didn't get to that for you this week, but I'll have it next week.' This disappoints the requester, so you over-compensate by promising: 'Not only will I give it to you next week, but I will make it twice as good.' You intend best effort, deliver what's possible, etc…..

Fear and shallow commitment contribute to coevolution on the fly. When people intend their best effort to achieve a shallow promise, they usually follow their own personal agenda. They say things like: 'I committed this to somebody and this to somebody else. As I can't get them both done, I guess I'll do this instead.' We blame everybody else if, while pursuing our personal agenda, we are surprised in some way or something goes wrong. Then we tell our manager something like: 'You see I'm in this bind. I can't get it done because that person didn't have their part done for me. But tell me, do you want the project done ¾ or do you want the documentation?' This means we offer limited possibilities for consideration. As a result, a power outside of us is forced to make a decision, like a manager agreeing: 'OK, just give us the system and put off the documentation'. If that system crashes, you can then say: 'It's not my fault we don't have documentation because the boss told me not to do it!' In these circumstances, we react to whatever happens on the fly, and push responsibility to someone else.

An important consequence of people following personal agendas is that individuals gain unique experience, and often hold onto that unique capability to feed a Limited Learning loop. In that case, I don't tell anybody about what I have learnt because I am put one edge ahead and given power by knowing something which other people don't. I will then ward off change because it might devalue my power if I no longer have the unique capability and unique information. In protecting my domain, I will continue to get unique experience.

When warding off change in the Limited Learning loop in order to keep myself in power, I will consider only limited possibilities, which is also one of the elements in the Coevolve-on-the Fly closed loop. Similarly, whenever the Shallow Commitment loop disappoints a requester, further mistrust is engendered and fed back to the basic Fear loop. This illustrates how amplifying feedback loops can interact and interlock to create ingrained patterns of organisational behaviour. I have found it is not possible to adjust the behaviour of organisations to a better set of habits if these fear-engendered patterns underlie those existing habits. In such circumstances, organisational behaviour cannot be changed because behavioural patterns have become locked-in through the amplified feedback loops. Before the organisation can change, the core closed loops must be first unlocked.

A 'vicious cycle' eventually emerges from these fear-generated interlocking loops. This can be envisioned literally like a cycle, especially like a kids' mini-bike with little training wheels at the back to help keep the bike stable (Figure 1.2, A Vicious Cycle - Emerges). The main feature of the 'training wheels' of the vicious cycle are the need for 'heroes' who respond to irrational demands by working unreasonable hours. This generates errors, which requires massive rework, that creates further irrational demands, which generates the need for more 'heroes'. The result is another closed loop that adds to the cycle's momentum in propagating the behaviour patterns we are locked into.

A virtuous hypercycle powered by trust

At the other extreme (Figure 1.3, ……To Another Extreme: Four Simple 'Rules'), you could have a pattern of natural open systems elements consisting of:

1. Exchange of Collaborative Energy

2. Shared Learning

3. Deep Commitments

4. Coordinated Coevolution

In an environment characterised by collaboration between people, a basic closed loop of trust is established. Collaboration generates mutual respect, so that people engage honestly with each other to engender trust, which in turn allows us to collaborate with others to generate mutual respect, and so on to create an amplifying feedback loop. Engaging honestly implies that people in the organisation will agree only to mutually meaningful and doable requests. This means saying 'No' if you can't do something, and having people believe that you are saying 'No' because you really cannot do it.

I like to illustrate this by reminding people who 'jog' for exercise, that they know they can run a given distance in a certain amount of time. If I said to them 'Tomorrow I want you to run it twice as fast', they would say they couldn't do it. But in business there is a tendency to say 'Yes' to everything, as I explained in relation to the fear-based vicious cycle of behavioural lock-in. When you engage honestly and agree to meaningful and doable request, you ensure fulfilment and satisfy the requestor, which makes what you say credible. You are also believed when you tell someone you can't do something.

When we commit deeply by promising to fulfil a request in a trust-based environment, we are really agreeing to a course of action for performing a joint task. Once we have agreed on our course of action, we will check what we have done by jointly reviewing the consequences against what we initially intended. If we haven't fulfilled all promises, we act jointly again to explore our new space of possibilities for jointly accomplishing new intentions and improving what we do. Out of this exploration, we can agree on a new plan for a future course of action. This loop of Coordinated Coevolution is essentially the Shewart cycle used in Deming quality management, another old model that works even better in a Complexity context.

When we jointly perform something in the way agreed during Coordinated Coevolution, we can also gain from tracking our joint capability as defined by our successes and failures. Sharing lessons learnt, for example, includes informing people about our mistakes in order to warn others not to make the same ones. This joint tracking of capability and sharing of lessons feed into a Shared Learning loop that also involves building, or changing, joint skills and approaches. This helps us to create a new space of possibilities for ourselves. The gaining of joint experience enables us to begin to change our world through a Coordinated Evolution amplified feedback loop in which we can ensure fulfilment and satisfy the requestor, thereby engendering more trust. This set of positive feedback loops enable the emergence of a wholly new set of behavioural patterns and a totally different dynamic to the vicious cycle. We call this new way of behaving a 'virtuous hypercycle' (Figure 1.4, A Hypercycle - Emerges).

At Citibank, we are looking to establish an enterprise hypercycle that is self-perpetuating. This must have a proper tension between our commitments, trust and learning, generated by a professional team focused on the terrain ahead. Everyone needs to be out scanning the environment so they can sense information and contribute insights about changing context. The hypercycle requires the development of new skills that enable a team to understand the dynamics of their environment and to be aware of the organisation they are creating. The team's fitness will improve as its people explore new routes and clear new paths to jointly open new spaces of possibility.

Changing from a vicious to a virtuous cycle

The behavioural loops I have been discussing give rise to the underlying structure of the organisation, which may be different from the structure we see on formal organisational charts. We create this structure ourselves because it emerges from the interlocking loops of our behaviour.

Like others who have worked in both types of organisation, I know that a trust-based, coordinated coevolutionary environment is so much more exciting to work in. It generates a wholly different energy to what happens in fear-generated interlocking loops, where everyone is unhappy and feel they are victims. In this vicious cycle, managers say: 'I don't know what's going on' and their people say: 'The manager is always committing me to things I can't do'. The passing of blame to everyone else shows that people don't understand their own personal responsibility for causing and perpetuating the dysfunctional set of loops.

Complexity Science has helped us to understand that the main reason why such a dysfunctional and uncomfortable environment is perpetuated is that people find it as stable as the virtuous cycle. It may be a horrible state of stability, but that at least gives people a context in which they know what to expect and how to behave, like saying 'Yes' to everything. In one organisation, a command-and-control type of manager felt the only way he could get things done was by making everyone fearful. He would demand something be done by an employee in a given time, even when that might be unrealistic; threatening he would find someone else to do it if that employee couldn't. Naturally, people agreed. Then they played a game of 'chicken', waiting for a sacrificial victim to emerge when one of the unattainable milestones wasn't met. Others could excuse their own failures by saying their problems were caused by the first 'chicken game' loser.

I saw a new manager come into this environment seeking to change it to a trust-based cycle. He told hundreds of people in his group that he trusted them and they could trust him. But their fear-based dysfunctional stability had led to such deeply ingrained behavioural patterns that one manager could not dismantle the interlocked cycles on his own. As they were involved in an open system, people were filtering out information that didn't accord with their closed loops of stability. So they didn't really hear what the manager was saying or detect the signs that they could trust him, which meant they continued to propagate old habits. We were eventually able to change behaviour to a sustainable hypercycle with the help of BDA (see Part 2).

Growing organisational change like a virus

Over the past few years, we have learnt a great deal at Citibank about the ways in which change towards a virtuous hypercycle can be achieved. The nature of the structures and behaviour I have been describing highlights the importance of seeing both the whole and the parts, and of intervening in both. Individuals and groups must see how they are contributing to the whole. You must work with people to change their individual habits, as well as intervening at organisational levels. And leaders must establish an environment for this to happen.

What excites me about understanding change in terms of complex adaptive systems is the challenge it provides in helping new behaviour to emerge. In the same way that people do not create laser beams, but an environment where those light beams can self organise to be coherent, you cannot create change in an organisation. You can only establish an environment in which people can come to a new coherence (Kelly and Allison 1999). This process must move through a number of layers of evolution, starting small. One of nine rules in Kevin Kelly's (1995) book Out of Control talks about this requirement. You start small, get that working, then build on this success through connections and relationships.

I typically start by catalysing a small team to help them make the changes necessary to be successful, which stimulates other teams to follow their example. The process of change and adaptation will then grow, like a virus. You must proceed bottom-up, starting with close colleagues, like a family. First, you must get a team that is functioning well in a coordinated plan by trusting each other and focusing on interactive, rather than reactive, decision making. Only then can you begin to string together teams and people to create an organisation that behaves in a hypercycle.

Five key elements in creating a hypercycle

We have identified five conditions which need to be in place to create the right kind of environment for hypercycle behaviour (Kelly and Allison 1999):

1. You must have a leader who wants the change to happen. Although leaders can't do it on their own, they carry more weight than others in the group. Despite this weighted position, leaders are still just one autonomous agent among others and must show the way forward by example through their own behaviour.

2. There must be a catalyst role filled by a person who knows how to help the right behaviours emerge. Just like a chemical catalyst helps change to occur more rapidly, you also need human catalysts to facilitate organisational change. That is the kind of role I play at Citibank.

3. The players must also want things to be different. When I started working on organisational change, I concentrated on telling managers how much more productive the organisation would be if they changed the way they did things. But then I began to realise that it is insufficient to get just managers to buy into this story. Everybody needs a convincing reason for investing their energy in changing personal habits. This means getting to the kind of personal impetus which makes people change a habit like smoking ¾ only after they have survived a heart attack. A concept like 'improving productivity' is too abstract to achieve that personal impetus for change. So, you must spend time talking to people about how their working environment affects their lives. For example, where people are in a vicious cycle involving many heroes who work all hours of the day and night, I ask whether they feel this is providing the quality of a life they want, and ask how their family feels about it. I get them to talk about what they feel when they have to cancel out frequently on doing things with their kids because they must rush to the office to sort out a crisis. I then tell them stories which open the possibility of an alternative, but emphasise that it will only emerge if they want to make it happen themselves. It cannot be imposed from the outside.

4. You need boundaries within which change can happen. Boundaries act both to keep things in or to keep things out of a certain space. For organisational change to occur, the right level of energy and focus of information needs to be maintained within a group. Boundaries can be established by geographic distance or management attention. For example, a systems group located in Florida was so remote that it wasn't distracted by extraneous information and rumours which abound in our New York headquarters. The location of this systems group allowed its members to concentrate on changing their own environment for themselves. On the other hand, when we needed to make required change in a corporate headquarters function, the unit's manager created the boundary. He reinforced the messages and information leading to change and acted as a buffer to keep out distractions that could have taken the group's focus off their change programme.

5. It's easier to coevolve with, rather than try to work against, the environment. The change in the Florida group was facilitated because their client organisation was also interested in improving the quality of mutual relationships, process and the products they produced.

Effective leadership to support change in organisations

In targeting change, it is important to establish policies and timely consequences that encourage desired goals, actions and ongoing performance. If a leader wants more of a particular kind of behaviour to emerge, that behaviour must in turn trigger immediate consequences of importance to the people being asked to change. Unfortunately, there is a tendency to rely on end-of-year performance rewards to encourage a new kind of behaviour needed tomorrow.

When change is needed now, immediate feedback and perceptible consequences are needed 'now' as well. This requires helping people to envision their organisation in the desired state with consequences favourable to all. The most effective actions for a leader to take are those which demonstrate that the change has a tangible and beneficial impact on everyone. Also, to engage people in behaviour that is more trusting and deeply committed, a leader must set new standards of behaviour by personal example, or for instance by stopping on the spot to tell someone who is propagating fear that things aren't being done that way anymore.

When trying to catalyse change, how we behave and interact day-to-day is more important than what kind of reward might be given in a year's time.

Moving up the maturity ladder

In the IS groups I worked with at Citibank, we were driving change using the quality management model developed by the Software Engineering Institute (SEI) at Carnegie Mellon University. This model classifies software development environments into five levels of 'process maturity'. Its lowest 'Level 1' rung is known as ad hoc because of little adherence to formal procedures. The ladder moves progressively through successively higher levels, where processes are classified as 'repeatable' (level 2) through 'defined' and 'managed' to 'optimised' at Level 5. When acting as a catalyst to help groups self-assess their maturity level, they began to realise why they were at the lowest level.

It helped during self-assessment to ask everyone about their quality of life and what they felt was going wrong or going well. Often, groups felt things were going terribly because the manager didn't communicate with them to keep them informed. Often, managers agreed the situation was terrible, but felt it was because his people were deceptive and didn't keep him in the loop of what was really happening. People would claim that their manager made ridiculous and stupid decisions that affected them badly. This was matched by the manager's belief that he was being forced to make decisions without proper information because he wasn't being told the truth.

When we shared this kind of feedback openly as part of one self-assessment process, the manager became defensive. He was a reasonable, well-intentioned person who thought he was communicating and was upset to find that people had seen him as perhaps being inept or malicious. But when he realised his people genuinely felt he wasn't telling them things, he asked how he could improve his communication. In situations such as this, I find people tend to provide feedback in generalities. To avoid this, I asked members of the group what specific actions the manager could take to better communicate. I asked them to envisage what it would be like if he were communicating to their satisfaction and to explain what behaviours would be involved in achieving that.

In response, the manager committed to holding a meeting with his senior managers every Monday morning. By the afternoon, the unabridged minutes of that meeting were put on Lotus Notes for everyone to see. People no longer felt he was failing to keep them informed. The group began sharing information and experiences as well. The manager established a catalyst in the organisation: a person people would go to when they found new ways of doing things. In addition to incorporating these ideas into an online database of procedures available to the group, she broadcast an email to everyone, alerting them of new approaches that might be useful. Simple recognition like this can make a difference - information sharing grew rapidly.

The group worked their way up the SEI maturity ladder to level 2 where processes are more visible, enabling people to repeat things which worked for them ¾ and avoid repeating approaches which didn't work. In November 1998 they are being assessed for 'defined' level 3 process maturity. Even more important than this formal recognition has been the fact that everyone is so much happier, as I found out when I worked with them on their level 2 assessment. At first they were concerned that improving processes would lead to more bureaucracy and unnecessary documentation. However, I pointed out this would happen only if they made it work that way.

The group actually ended up with more documentation than I would have produced. But they were happy with it because they felt everything they did was useful to them. People in the group began to say the manager was so much better because he was making the right decisions; and he was saying his people were so much better because they were creating real, doable plans. It wasn't that everyone had become so much more competent within just 18 months; by coordinating they were able to leverage their competence. If you ask them how they did it, they say by talking and listening to each other, then progressing slowly.

This is an example of how self organisation and selection for fitness can generate coevolution which can move an enterprise from near extinction to becoming a new breed with the drive and capacity for change needed to evolve successfully today, and to influence tomorrow (see Figure 1.5, Coevolution). A sustainable enterprise works through self-organisation to counter the pressures of environmental selection. It becomes a dominant species by successfully adapting to the environment as well as triggering new changes within that environment.

Hierarchies, webs and 'ecotechs'

Kelly was asked how the ideas she discussed in her talk related to the observations of Oshry (1996) on the profound influence of hierarchy in organisations. Oshry was said to have highlighted the existence of very different worlds in the 'tops', 'middles' and 'bottoms' in any organisation. Those at the bottom level were said to be likely to establish a coherence, empathy and sense of community among themselves very quickly, for example through trade unions; the middles, such as middle managers, find it very difficult to find any sort of commonality as they each exist in the middle of different worlds; and the tops live in a world where they are often told very little about what is happening elsewhere. Kelly responded:

I believe that organisations consist of a diverse range of people and groups who relate to each other in a variety of ways, through a web of nested connections. People may have different positions in terms of 'hierarchies' within different groups, perhaps being a manager in one, but operating as lower level member in others. The one-to-one relationships between people are as important to focus on as the artificial hierarchies they are in.

Nevertheless, organisations often need to designate a communication hierarchy on top of web relationships for purposes of rapid information dissemination or timeliness and accountability in decision making. Power hierarchies need to be kept to a minimum so as not to constrain that web of relationships, through which things tend to get done. In a dysfunctional power hierarchy, people play to the interests of powerful individuals rather than the good of the enterprise. Business organisations are the only ecology where the being with power, the manager, can be thought of as not only having a prey but also able to control the food chain of that prey. People are afraid to do things because they could 'be eaten for lunch by the manager who also controls all the resources they need to survive'.

In addition to designated hierarchies, natural hierarchies come and go within the web of relationships. I refer to these flexible hierarchies as earned rather than designated. For instance, in this seminar I could be said to be at the top of the hierarchy for the moment as I have certain knowledge that others seek. I wouldn't be at the top tomorrow if we met to discuss physics or economics. We coevolve in communities of people depending on the timing and our needs, which suggests that hierarchies should be fluid and change with the situation. If someone rises to the top of a hierarchy because of the influence, knowledge or expertise they offer in one situation, they might not earn that position in different contexts. So, we must recognise and separate the earned, designated, and power hierarchies which have gotten very mixed up in many organisations.

'Black holes' in the middle of organisations

I have also gained important insights into the nature of the designated top, middle and bottom worlds when working on organisational change. I have found people at the top spend their time focusing on strategic trends in external market, industry and economic coevolution over the longer term. This often leads them to develop a vision that highlights the need for internal coevolution or change as well. They say 'We need change'. People at the bottom are on the firing line, struggling to keep up with their changing context. Therefore, they are also saying 'We need change'. Middle managers, on the other hand, seem to be buffered on both sides. They do not have the same focus on strategic needs nor the same reality of facing day-to-day tactical obsolescence.

Information is needed at all levels to counter the forces propagating stability. Some psychological research shows middle managers are the group most under stress. Regardless, I find many of them in dysfunctional and stressful environments still give a high priority to stability because of a fear that things could get even worse. It's a case of 'better the devil you know'. The middle then forms a black hole as messages in favour of change from the top and bottom go nowhere. People at both ends of the spectrum might see the need for change while propagating behavioural loops that mitigate against the change.

This highlights why it is necessary to focus on micro-to-macro integrity and feedback. It is also another reason why a catalyst for change must figure out how to take the power hierarchy apart. The aim of doing this is to facilitate the development of a flexible decision-making hierarchy and create communication paths that get clear messages efficiently to everyone, everywhere in the organisation.

Who you know and how you navigate through relationships with others lies at the heart of how an organisation actually works. Leveraging this web within the formal organisational system is crucial. Focusing on this natural web, in Kelly and Allison (1999) we emphasise the significance of important roles in organisations, such as being an autonomous agent, catalyst or leader. We include the crucial new role we call ecology technicians, or 'ecotechs'. They are typified by the experts at CASA, who have the skills to help us influence our organisational ecology by constructing models that provide insight into the current ecology, the non-linear nature of our human systems and trends in environmental coevolution. Regardless of what they are called, it is important to have such people to help navigate effectively in today's rapidly changing global economy.

PART 2

LISTENING, LANGUAGE AND ACTION

BY CHRISTOPHER DAVIS

Flores' unique cocktail: the 'mobilisation of change' approach

We at Business Design Associates (BDA) offer to work with our clients to produce a competitive edge by transforming their organisation along the lines of the perspective discussed in this report. This offer has grown over the past twenty years from Fernando Flores' work, which he developed in collaboration with colleagues, including myself, at BDA. There are three original sources to this approach:

1. the work of Humberto Maturana on complex adaptive systems and autopoiesis (see Maturana and Varela 1987);

2. Speech Act Theory, originally developed at Oxford University by the philosopher J. L. Austin (1962) and later refined by the philosopher John Searle at the University of California, Berkeley (see primarily Searle 1979);

3. the work of German philosopher Martin Heidegger, who is known for his practice-based or existential account of the understanding of being (see Heidegger 1962; 1971).

Fernando Flores is remarkable as a philosopher for his combination of the Anglo-American tradition's Speech Act Theory with a continental German hermeneutical approach to philosophic problems. Even more unusual has been the way he has brought these generally incompatible traditions together while adding the apparently incompatible thinking of Maturana about biology-basic cognitive investigations. Flores added his own unique insights as he began to blend these different flavours into a unique cocktail which has formed the foundations of the methodology used by BDA.

We have never adopted an explicit Complexity Theory viewpoint in BDA's approach. We became aware of how our approach related to Complexity only when we met Susanne Kelly while working at Citibank (see Part 1). Since she said what we were doing is at the heart of what she believes needs to be done through Complexity, she asked us to take that perspective seriously. I believe the approach I describe here lines up very well with the new paradigm of understanding also provided by Complexity. I find it very encouraging that the new thinking indicated by the growing interest in Complexity is challenging certain traditional 'common sense' notions which have held dominion since the 1980s. We have been pushing against these outdated attitudes for some time, so we welcome our allies from the Complexity universe.

Three kinds of processes

There are many places where I could begin to explain our approach to the mobilisation of change. One, which business people can relate themselves to easily, concerns our analysis of the processes of businesses. Unlike others, we maintain that business takes place in three dimensions (Table 2.1|). Switching the emphasis from one to the next corresponds to the movement from the Manufacturing to Information Ages and beyond as discussed in Part 1:

1. Materiel, the materials and equipment of an organisation: Logistical and other processes that move and transform physical goods. Chemical engineering and manufacturing activities are found in this sphere. Materiel processes have been studied well for over a hundred years, going back to early time-and-motion studies. Major breakthroughs are rarely made here anymore, although occasionally some important advances occur.

2. Information: This dimension is about moving data around, comparing it, changing it, storing it, retrieving it, and so on. These information processes can be done with or without computer-based systems, although with the start of the computer age about thirty years ago information processes became a major area of concern for all companies.

3. Commitment: Here we are concerned with the space containing the network of personal relationships that comprise a company, which encompasses questions about how those relationships can be made to function well. Commitment processes are connected to the concern for producing change in organisations.

These three dimensions should not be considered as developing linearly with hard-and-fast stops. Referring to them as 'ages' or 'eras' may, therefore, not be strictly correct. The different processes have things in common, in as much as they are all concerned with some kind of recurrent 'process' activity. They also all fit together. A simple example is shown in the 'basic capitalism' game played by kids in America by setting up a lemonade stand. The materiel process in the game is the actual production of the lemonade. The information processes include keeping track of the number of lemons, of whose turn it is to make lemonade and whose to take money from the customers, making change, and so forth. Executing these processes is included in the commitment process where the child agrees to make you a cup of lemonade of a certain standard, for a certain price. This illustrates how all three dimensions are part of every process, with a continuum between them.

BDA has focused on the commitment process. We do not disparage materiel and information processes because they are also necessary. However, we believe a sharp line can be drawn between what could be called the materiel/information paradigm and the dimension of commitment processes. This is highlighted by the way the verbs used in Table 2.1| to describe materiel and information processes are similar or identical. They define actions dealing with the movement, transformation, and assembly of certain products. There is a very different set of verbs for commitment processes: declare, offer, request, promise, assess, assert. If you try to manage people in the commitment paradigm with the same rules used for materiel and information processes, you are therefore likely to get into trouble. The verbs are different because the domains are radically different. Many process-based companies have learnt this distinction at a very high cost, especially when they have thought of people as if they were another information-processing, goods-moving machine. That mindset will lead to a fundamentally badly designed organisation because it will leave out the power of commitments to move people and flexibility. It will also leave out the business' relationship to its customers.

Waste down through the ages

As commerce has moved from industrial-based to information-based, and beyond, the nature of business process has changed. This change is also mapped by our three dimensions as depicted in Table 2.1|:

1. Waste in materiel/industrial processes has been considered mainly in terms of unnecessary time, motion, resources, and physical movement activity that do not contribute to product value.

2. In the Information era, waste is seen primarily as relating to inaccurate, untimely, insufficient, or under-utilised information.

3. In the commitment/relationship dimension, incompetence for listening, mistrust and anything else that reduces the capacity to maintain relationships are sources of waste.

A manager's emphasis on commitment and relationship has generally been seen as being very 'soft', unmeasurable and lacking any clear evidence that it adds to the bottom line, particularly as seen by the very traditionally trained 'bean counters'. More and more, however, the growth of theoretical stances like Complexity is leading to a recognition that a company should be dealt with as a holistic system. This new holism is showing why some interventions that seem to make sense in a linear way, like lay-offs, actually have a very real negative impact on the value of the company as a whole. Some research is just catching up to demonstrate this.

Contrasts between materiel/information and commitment

Table 2.2| summarises important differences between the materiel/information and commitment 'paradigms' or relationships. These differences again highlight concerns about linear and holistic thinking. For example, the classical Cartesian view, which the theories of materiel and information processes follow, is of a 'world' that is universal and objective, and so is available in one form to everyone. In contrast, we argue that the commitment paradigm is based on a holism and pluralism of many 'worlds' or 'sub-worlds' brought forth by our pluralistic cultural traditions. Each set of practices or world discloses a different set of possibilities. This claim becomes obvious when you look into big companies, where the world of marketing people is not the same as the world of engineers, which is not the same as the world of accountants. Even when people use the same words in their different worlds, they are often not talking about the same things. This produces enormous mis-coordination and bad listening in many ways. Yet we continuously see efforts in companies which persistently assume that all these people are working in the same world.

Perception in the classical, universal, objective world is considered to be a matter of passive receipt of data by the senses, which usually perceive the world 'as it is'. At BDA, we emphasise that this is not the case. We point to the commitment relationships we all in fact live in, and in which what we perceive is understood differently according to the social, historical and biological background practices of the world. People trained as engineers see people as setting standards and making predications. People in the marketing tradition see people as collections of preferences. My ability to move into any situation and cope with it should then be seen as being directly the product of having the appropriate background practices. A corollary of this is that organisations and people are inevitably 'blind', in the sense that that they are disposed to see only certain aspects based on the patterns they have followed in the past. This is neither bad nor good, but it means we must take this inevitable blindness into full account when designing what people and organisations do. This blindness appears must clearly when examining how people listen to others with different background practices.

Listening within the typical materiel/information process is about gathering information or requirements: 'If we could just get the requirements from our customers, we would know what to do …' You will know you are not in the presence of a great listener if you ever a meet someone newly minted from a sales course who looks you in the eyes and says: 'Just tell me what your requirements are.' That is the sign of someone caught in a procedure and, literally, going through a checklist. In contrast, building on the work of Heidegger in the 1950s, listening in the commitment process is always about interpreting, and listening to the interpretation of others. That is what skilful listeners do. It is a skill that is particularly underdeveloped in the typical engineer.

Communication in materiel/information processes is considered to be essentially about the transmission of information between a sender and receiver. From the perspective of commitment processes, communication is about the successful coordination of action. That is why offers, requests, and promises play such a large role. We can then know that successful communication has taken place if all parties assess their actions are inter-coordinated. For instance, we have evidence that some communication acts took place concerning this seminar, because we all arrived at the right room at the LSE by 2pm. That required offers and requests to speakers and attendees, promises from speakers and those in charge of the rooms, and so forth. So far as it required information, the information was in the context of an offer, request, declaration or promise.

Finally, materiel/information action is always concerned with activities relating to physical entities. What BDA calls 'human action' is always connected to commitment. In this paradigm, the movement of things is considered as being action only if there is a background structure of commitments which make sense of what is being done. If we had randomly wandered to the LSE today at 2pm, we would not say we had engaged in any particular human action by showing up. Animals and leaves can just randomly appear here or there. However, it is clear we had a successfully coordinated action because our appearance came from a background in which we had all received and accepted an invitation and arrived at the appropriate time. Such action is always connected to an interpretation based on a commitment, using a finite and universal taxonomy: a 'grammar of action' which transcends all cultures and languages.

The grammar of action

Table 2.3| provides our overview of speech acts, which mirror the language acts summarised in Table 2.1|. This grammar was originally derived from the philosopher John Austin in the early 1960s, who observed that language does not simply describe the world; for a good part of the time we are inventing the world through language. When I invite someone to join me for lunch, I am not describing anything but opening a new possibility. According to Austin, this kind of speech act is the most interesting thing which happens in language because it is making something happen which didn't exist before. Other things we say do not invent anything.

If someone asks you the time and you say it is 4:25, you are not inventing something. We would say you are making an assertion. Such assertions are the least interesting things we do with language. Much more interesting are the speech acts where you make offers and promises, then fulfil them; where you utter new declarations that change the space in which people are moving; or make interesting assessments that provoke action in someone. This is the kind of activity people get paid for.

We agree with Austin's basic claims and have drawn on Searle to come to our own approach. This has been distilled into our 'grammar of action' involving all the basic 'atoms' of speech acts:

1. Declarations establishes new networks of relationships in which we can take action. Declarations require that the speaker be recognised as having the appropriate social authority. So, when a minister in a church pronounces that a couple have become 'husband and wife', at that moment it becomes reality in part because we have given the minister authority to do so. Similarly, a CEO has the authority to make particular decisions about the future of a firm, while the Board of Directors has the authority to remove the CEO by pure declaration. The decision of a jury in a court of law in many countries is accepted purely by the declaration about the accused's innocence or guilt. This acceptance is not a matter of evidence, or of the 'truth' about whether the accused was actually innocent or guilty. It is solely a matter of social authority and declarative action on the basis of that authority.

2. Offers and requests elicit mutual commitment and coordinated action. They are about making a move with someone else. When you make an offer to someone, you are making it clear that you will fulfil certain conditions if the person accepts. With a request, you ask someone to do the same. This covers a broad class of speech acts. An order from a general to a lieutenant is a kind of request; so is a beggar putting his hand out. The fundamental act is always a request, although a wide range of force and other background dimensions can be applied to back it.

3. Assessments evaluate progress and help us to navigate in our projects and worlds. Assessments always involve value judgements: 'This is better than that'; 'She's a better manager than he'; 'This plan makes more sense than that one'.

4. Assertions build confidence in our judgements and our ability to undertake consistent, reliable coordination. They are the speech acts which describe the world and where you say something is true and I am prepared to provide evidence to that effect. Facts are assertions assessed as being true; lies are assertions assessed as being intentionally false.

Understanding the context of speech acts

All speech acts are context dependent. The same sentences can be quite different speech acts in different contexts. Cases at law famously provide good examples for this ambiguity. When a spectator in the 'peanut gallery' says 'He's guilty!', that is a person without authority making a value judgement. It may be interesting, but it remains only an assessment. The prosecutor might say 'Ladies and gentlemen of the jury, he is guilty. Find him that way.' That would be a request for their vote in this regard. The jury will debate assertions, going back and forth, when they have been sequestered. But in the end they have just one job to do: come back with a declaration of 'Guilty' or 'Not Guilty'. Their declaration does not change whatever actually happened when the alleged crime took place, and it does not mean that, with another standard of proof another verdict might have been reached. Nevertheless, by the rules of our legal game and social engagement, we must accept the declaration of a verdict as the basis on which subsequent actions are coordinated. The jury has declared what the facts of the case add up to. In the moment that the jury's foreman answers 'Guilty' in responses to the judge's query 'How do you find him', we deal with him as being guilty of the crime he was accused of. In normal cases, the judge then declares what this guilt means in law by saying how the court finds the defendant. That declaration invents a social reality, but is not concerned with assigning absolute truth or false values. The court journalist who rushes out to report the verdict is not inventing anything, just reporting a reality that has already come into existence. When the journalist says, 'He is guilty', that is an assertion.

It is never words alone that matter. It is the contexts in which they are used which affects our future coordination. Words expressing a descriptive assertion in one context can become a declaration creating a new space of possibilities in another, and they do that purely by convention and authority.

The conversation for action

The speech acts in Tables 2.1 and 2.3 constitute the fundamental atoms of human coordination. In the discipline of coordination, they are analogous in significance to the periodic table for chemistry. Both the periodic table and those provided here identify the discrete elements that combine to produce the matter studied by the two disciplines: chemical compounds in chemistry and coordinated action in organisational change management.

One of the many ways we work with speech acts in managing organisational change concerns the basic 'dance' that takes place between people in order to get into action. We call this the 'conversation for action'. It consists of a loop with four quadrants (see Figure 2.1, The Conversation for Action), representing the structure of any conversation aiming to coordinate action between two people. The four quadrants are:

1. Preparation (resulting in a request or offer). The conversation for action starts with a period involving all preparations relating to the concerns or activities that happen before the specific moment when a request or offer is made. If this moment consists of a request, it means the would-be customer is asking for something. If it is an offer, the would-be performer is offering something to the would-be customer. As soon as the offer or request moment happens, a new phase begins: the period of negotiation.

2. Negotiation (to gain acceptance of two mutual promises). Of the many kinds of conversation that can happen in the negotiation process, only a few maintain a state of commitment between the two parties: accepting the request; declining it; counter-offering and reaching an acceptable conclusion; or committing to reach a resolution at a future date. Any other kind of response breaks the sincere commitment between the parties. The dynamics of 'shallow' versus 'deep' commitments discussed in Part 1 become manifest in this phase, for example through observing whether people maintain a state of commitment in their conversations, or allow them to be watered down and dissipated. Actions move to the performance quadrant if the customer accepts the performer's offer and, in turn, promises some form of payment.

3. Performance (to a declaration of completion). The performance activity aims to generate what BDA calls the 'conditions of satisfaction' to fulfil what was agreed on as a result of the previous preparation and negotiation phases of the loop. Effective performance involves understanding the conditions of satisfaction and what needs to be done to bring the performer to the moment when he or she declares completion and actions move into the next and final phase.

4. Assurance (to declaration of satisfaction). This phase is needed to maintain a state of coherence of coordination and interpretation, where the customer assesses what has been delivered and says what is good, what is not good and what is missing. The conversation-for-action loop is closed when the customer declares satisfaction because what has been delivered is what was wanted.

These loops move along very nicely when coordination works well. Lots of sub-loops are bound to occur within each of the phases. Our aim in explaining this is not to teach people new terms or a new method. BDA's intent is to reveal what is already going on when we coordinate action. Whatever words are used to describe it, people are naturally involved in all kinds of networks of conversational loops in any organisation. You automatically make assessments whenever you accept a promise. In an organisation where shallow promises are accepted, the kind of vicious cycle referred to in Part 1 emerges, producing redundancy, re-work backups, and an enormous amount of waste.

The conversation-for-action loops are being made, broken, and re-created over and over again. They form networks of many loops which serve a single central loop that organises an overall promise such as: 'I will build you a new dam in this place.' Out of that main loop, a multitude of major and minor sub-loops will develop, many of which will have their own sub-loops. Ultimately, millions of conversations for action will happen before the project delivers the dam wanted by the customer. The coherence and integrity of these conversation loops provides firm grounds for judging the success of an organisation in terms of whether the loops are well designed, link together well, are rigorously maintained, rarely break down, and maintain a level of commitment throughout.

Mapping commitment processes

When we at BDA work with an organisation, we ask its people to observe their commitment flows. In doing this we are requiring something new because the standard process maps of most organisations collapse together information, materiel, and commitment processes. For instance, a process map for mortgage lending would show how main actors, like the client, loan committee, appraiser, real estate broker, and loan officer communicate via documents which the IT areas service (Figure 2.2, Mortgage Lending: People, Objects and Information Flows). We therefore ask people to produce a map which separates out the commitment process involved by focusing on the major conversation-for-action loop between customer and performer.

In the mortgage loan case, the central commitment loop is that between the client and the bank, and its aim is to come to agreement on a loan (Figure 2.3, Mortgage Lending: Commitment Process). Each quadrant has key sub-loops, each of which has a customer and performer. So, for instance, a loan officer might be the customer who requests approval for credit within the bank while the loan executive is the performer who gives final approval. But that same loan offer might be the performer in dealing with the bank customer who requests credit in the first place. Each sub-loop has its own cycle time and associated other events, creating further sub-loops. The map can be enlarged to describe any level of detail sought. The advantage of this kind of map is that it can show how well the commitment processes in any organisation are designed, and how well the information systems support the flow of commitments.

A commitment map can reveal whether these processes really allow each person to get his or her job done in the easiest way possible at each step. It can also show when process are in a horrible muddle, as happened when we mapped the purchasing commitment process for a mining company in which nobody had any idea of what the whole process looked like. The map showed the commitment process was a bureaucratic nightmare of specifications, signatures and consulting sub-loops that had to be gone through before anything requested was delivered. Performers did not know the specific customers for whom they were working. Rather, they were performing to meet standards. Consequently, there was no negotiation over non-standard requests. Many performers simply had the job of approving what others had done without knowing why their approval was necessary. Such commitment structures indicated that the process was never designed, but was the product of drift over time. Processes were added according to the opportunity or breakdown of the moment and just remained in place. The result was a deeply resigned group of people who felt terrible about not getting things done, and they could not figure out why given their long hours and hard work. The map removed the mystery by showing there was no design and no real concern for moving the purchase through the organisation to satisfy requests.

Commitment process maps can be done very quickly because they are very close to the way work really happens. In one utility company, a group of clerks took just an afternoon to find, diagnose and fix the source a major complaint that had been an enormous source of aggravation to them. The map showed that when the clerks sent a lineman to do a repair, the promise they made to the customer was not coordinated with the promise the lineman had made to them. Preparing the commitment map which helped them to solve this problem did not require the kind of business process design team that has to be immersed away from the business for many weeks before producing any results.

It is important to emphasise that the commitment mapping process must not be treated mechanically. For example, some people in an engineering company we are working with initially saw the mapping process as a 'good and logical' way of 'fixing the people' by knowing what everyone should be doing. That fails to understand the significance of the background context in determining how well or badly the commitment process works. The process itself must always be considered alongside dimensions such as: the degree of trust that is present; the overall mood in the organisation (whether it is one of resentment towards management or an ambition towards what's possible); and the competence available for listening and appreciating the other worlds that exist within and outside the organisation.

Trust and the rigorous interpretation of 'soft' process issues

Using speech act analysis enables one to make rigorous interpretations of certain phenomena which are often thought of as being too soft to interpret properly. Trust is something we all think we know about. You either feel it or you do not, but it can seem difficult, too soft to interact with effectively. Seeing trust this way, whatever its legitimacy in some domains, is plainly harmful when you are seeking to make an effective business intervention. Consequently, we find it useful to deal with trust as something always connected to an assessment about whether the other person will fulfil the promises he or she makes. In these terms, trust always has three dimensions:

1. Sincerity: the degree to which people mean what they say and whether their promises are shallow or deep.

2. Competence: whether a person has the capability to fulfil a promise made sincerely. Managers are always making such assessments, for example in considering if a person can keep up with the speed at which the organisation has to move.

3. Involvement or care: how well a person can appreciate and care about someone else's concerns and, in particular, one's own concerns. For instance, someone may be sincere and may even have the competence to do something for me, but may not have taken the time to truly understand what I consider important and how my business is changing. If this is the case, a sincere competent person might well deliver something that is ultimately harmful to me. There was a consulting company that had such a gung-ho style that it was said the company would deliver on its contract even if they alienated the client's workforce and destroyed his business.

Without trust, poor listening, poor coordination, cost, and complexity all grow. Without trust, strategic and even tactical change cannot be managed. Therefore, the distrust in an organisation has a direct impact on the bottom line. Every interaction in which I cannot trust the other person to produce on what he or she promises will require me to build a redundant request to someone else to standby, in case of the failure of the first. One, two, all three, or none of these might actually satisfy the promise. In any case, it is wasteful and shows why distrust has a very real impact.

Coordination pathologies

BDA has found that different organisations demonstrate more or less evidence of what we call 'coordination pathologies':

1. Losing the customer's trust. A customer's trust can be lost at any moment in the conversation-for-action loop. In the beginning, poor listening with low care causes breakdown. Typically, the performer hears only his own concern and the customer does not want to do business with him. In the second phase, trust is lost when negotiations are not sincere, or again when there is insufficient care. Here, no clear and committed promise comes out of the negotiation phase. In the third phase, performing poorly on quality and time promises destroys confidence in the performer's competence or sincerity. Finally, failure to solicit and take seriously the customer's assessments and declaration of satisfaction will reveal a lack of care for the relationship. The networks people work with in all organisations are networks of trust. An individual's network is based on knowing the people you can rely on to listen when you ask for something and deliver on agreed promises.

2. The executive's complaint. This breakdown occurs when a high-level person in a company repeatedly makes requests which seemingly disappear into black holes without the executive knowing why. In such cases, the senior managers do not see that their actions are failing to produce good coordination. They typically end up sitting at the front of meetings saying: 'I want someone to tackle this important unresolved problem.' But everyone looks away and nobody leaves the manager with a promise to do anything to deal with the problem.

3. The 'I'll do my best' syndrome. This happens where the response to requests coming in is to work through procedures in which there is neither a promise nor real concern for producing satisfaction. This is usually found in organisations where people are in a mood of feeling overwhelmed, so the most they can offer is to try their 'tops'.

4. Assembly-line work. In this situation, people perform routine tasks without any sense of requests, promises, or concern for customers' satisfaction. Work in this environment has become simply an activity that is being fed along as if it were on an automatic production line. Not surprisingly, this kind of environment is an unsatisfying one to work in because people generally get their satisfaction and enjoyment at work by closing the loop that lets you know you have produced value for a person requesting something.

Typical cultural difficulties

We at BDA have classified organisations stuck in the vicious cycle of decline discussed in Part 1 in terms of the following typical cultural dysfunctions:

1. Mistrust and poor listening. People feel isolated. There is a great deal of heroic behaviour, poor coordination, extra loops, re-work, and blame.

2. No one makes offers. Nobody speaks up to invent new ways of doing things. Innovation, then, becomes the task of top people only. The other members of the organisation are passively awaiting requests.

3. No one declines requests. This dysfunction results in shallow commitment to all requests because people say 'Yes' to everything. People and processes then become overwhelmed.

4. No one makes negative assessments. People, then, interact with great insincerity. But the insincerity is covered over by a cordial hypocrisy that follows the adage: 'If you can't say something nice, don't say anything at all.' Weaknesses and suffering will then persist. In order to build good teams and organisations, it is necessary to be able to give a direct and honest assessment in the appropriate context, for instance to say when someone is not performing well or has acted irresponsibly. Without such assessments, the level of care for the organisation declines. This care manifested in honest assessments is a key reason why some sports teams become champions, while others remain losers.

5. Over-the-wall execution. People become focused on their own local 'chimney' and seek to pass responsibility for the end result on to the next person. Consequently, in organisations with this dysfunction, there will be no concern for the unity of the organisation and what it is delivering. As a result, 'mysterious' failures, dissatisfied customers, a lack of innovation, and a culture of blame will be prevalent.

We at BDA have developed a number of practices for overcoming these dysfunctions in coordination and in corporate culture. As competition increases with globalisation and the lowering of barriers of entry provided by today's new technologies, we believe that the practices we have developed are becoming competitive imperatives.

Competitive imperatives and critical competencies

In our practice of mobilisation for change, we at BDA have identified a number of key areas where actions are needed if an organisation is to move from the negative business cycle of increasing distrust, shallow promises, and growing fear to a more positive cycle that also increases competitiveness. To achieve the competencies needed to acquire and retain customers, design and produce innovations, build alliances and maintain flexibility, a company must become good at:

1. Listening and innovation;

2. Coordination and commitment;

3. Mobilising new practices; and

4. Employing the tools for coordination and redesign of business processes.

These four competencies are discussed in the next four sections.

Listening in order to understand other worlds

A culture must be developed in which 'listening' is understood as seeking an interpretation of the 'worlds' other people live and work in. This requires coaching people out of the notion that they are just 'gathering requirements' from customers, whether internal or external. Key people in a company must be trained to listen to the concerns and commitments of others in the context of their lives and work habits. It is also important to realise that innovation can be cultivated. For this reason, BDA encourages people to cultivate 'gardens for innovation.' In such gardens, the practices of customers are explored alongside the practices for purchasing and using products anticipated by the business. The differences between the two practices lead to innovations, from enhancements to reconceptualising the nature of the business.

Establishing a 'culture of commitment'

Impeccable coordination should be made the standard for all activity in the organisation. That standard will not mean that everyone will keep his or her promises all the time, but it does mean that any lack of fulfilment will be properly managed so that incomplete fulfilment never becomes accepted as standard. An interpretation of what it means to make a commitment will be shared from the bottom to the top of the organisation. If a promise is not going to be kept, it will be renegotiated or revoked. This standard of operations with promises needs to be built through the cultivation of a 'culture of commitment'. In order to have a culture of commitment, the basic commercial speech acts must be respected:

1. Declarations must be made under valid conditions, so that what people with authority say is always respected and anything that gets in its way is eliminated.

2. Innovative offers should be prized and celebrated.

3. Requests should be well articulated, with clear and complete conditions of satisfaction, and entail the commitment of the customer to make assessments as a customer.

4. Assessments should be given when grounds can be supplied on request.

5. Assertions should be made with a clear understanding of the relevant evidence and a readiness to provide it in a clear way.

As pointed out in Part 1, one of the main sources of bad moods in companies is the swirling sea of ungrounded assessments. People will say such things as: 'Nobody ever tells us anything'. If you hear this kind of comment, you should ask 'Who doesn't tell you what?' Only then can you begin to define something clear and actionable, by turning the ungrounded assessment into a request.

Mobilising new practices for change

In today's market environment, many enterprises regularly seek to make organisational changes that run the gamut of supporting simple changes in customer satisfaction (new products and services) to strategic shifts that change the nature of the industry. However, I have seen over and over again that improvement initiatives end up in producing lots of nice new written procedures, pretty maps and great strategies that go nowhere.

In order to create real change, where new practices are fully brought in and sustained, organisations must address four key dimensions:

1. Trust must be built or rebuilt by listening to other people's concerns and taking action to address their needs. This requires the evolution of a tradition of 'taking care'. Without trust, organisational change will fall prey to fears and suspicions about each individual's welfare.

2. Responsibility must be the basis of moving in the organisation. There must be an end to any prevalent 'culture of victimisation', where the blame for any problem is always said to be the fault of someone else. The blame culture must be killed by example from the top. In its place, an environment must be created where everyone assumes they have a responsibility for the reason why things are not working and have a role in improving the situation, even if other people are also involved. Without shared responsibility, any change initiative looks like the punishment of all because of someone else's misdeeds. The changes then get undermined by resentment.

3. A story about the future gives purpose and meaning for today's actions. It is a failure of leadership if people feel their work in the company gives no meaning to their lives beyond being breadwinners. No interesting companies today have employees who work for the money. However, just coming out with trite slogans does not give meaning to people in their work. Leadership is needed to create a narrative that shows why the company is making a better shared life for all stakeholders.

4. New work practices must give employees the opportunity to fulfil commitments. New practices are not mobilised by disseminating procedures, rules and other information. People must get caught up with routines naturally, and this natural adoption occurs if the new practices enable people to make and fulfil commitment.

Tools for coordination

Reengineering the organisation by designing coordination practices around new IT tools gets mobilisation backwards. Instead, processes, practices and infrastructure should be built around coordination focused on the timely completion of commitments. Coordination around the commitment structure of the organisation should become what the organisation's IT service aims to support. We at BDA have found that organising a company around the flow of commitments makes IT support far simpler. Competences and practices must also be established for achieving shockingly quick development and intimate design partnerships. Such shocking quick redesign is particularly important for IT departments, where the dinosaur notion that you can spend years going out to gather requirements must be annihilated.

IT systems also need to be targeted at supporting the capacity to mobilise new procedures. For example, in some organisations e-mail is used to send requests and thereby avoid the negotiation necessary for people to make genuine commitments. BDA has a sister company founded by Flores, called Action Technologies, which produces software that can organise e-mail in terms of the commitments people are making and receiving. When this system is used, a request does not come back as a promise until the other person negotiates and commits. Simply sending a request does not mean you have a commitment. The performer must also promise. The way this software prompts and reminds users can help to get people out of irresponsible 'management by firing off email requests'. (Action Technologies provides a range of Web-based workflow and work management software for knowledge workers. It is based in Alameda, California, tel: 1-800-967-5356; fax: (510) 769-0596; Web:

www.actiontech.com|

How to implement change actions successfully

From our experience, we have identified the common elements of successful organisational change.

First, as highlighted in Part 1, you must have commitment from the top. We will not take on a project unless at least the senior executives within the business unit are fully committed to change. It makes absolutely no sense to move the organisation unless its leaders and managers take responsibility for that move. When we start working with an organisation, managers frequently complain about the mood of their people. We point out that they have everything to do with the mood and must take local responsibility for improving trust and transforming the whole mood into a positive, future-oriented mood, such as one of excitement or of caring for others.

Second, the change team must consist of those managers from different areas of the company who are both influential and who can listen to each other and the concerns of all stakeholders.

Third, this change team must provide a very focused intervention to increase the level of trust between people in the organisation and to enhance their efficiency and coordination. To do this, change teams must bring about the change the organisation seeks by also enhancing coordination and focusing on commitments.

Fourth, the change team must also begin the process of breaking up the previous 'common-sense', linear, Cartesian ways of approaching work.

Fifth, specialised intervention projects also play a valuable role in moving things forward. For example, in one company with extremely high quality costs, we were able to show that poor coordination produced much of the wasteful expenses. Indeed, we identified the costs as coming specifically from poor preparation of offers and poor negotiation, rather than execution, which was where the company had been focusing its attention. Hence, the redesign for change revealed itself as one that would go a long way toward solving nagging problems that were not the focus of the change initiative. After we had helped to overcome these quality inefficiencies, the organisation could move on to redesign other key processes and produce new practices with greater trust and equanimity.

Evidence of successful change

A BDA intervention is regarded as successful only when we see evidence that a new practice has taken hold and a new style of working has come into existence, and that this new style is allowing the company to produce financial results that it could not produce before. A simple example of such a success occurs with a team adopts a new kind of meeting with new participants and with rigorous attention to producing the results for which the meeting was designed. If the meeting is for achieving resolution, then the roles that enable this must be employed. But the use of roles such as champion, investor, and so forth is not enough. We must also see that the meeting produces increased trust, openness to innovation, responsibility for the organisation, and actions that advance the competitiveness of the organisation.

Sometimes change is a tough process, particularly at the senior ranks. Much time must then be spent getting people to listen more carefully. Careful listening to concerns that only make sense in other worlds of practice can be especially difficult with groups, like engineers, who have been trained in a very Cartesian interpretation of listening. Beginning to show them that there is a different way of listening is a first step forward.

One of the conversations I have found effective in training senior managers to listen focuses on distinguishing a 'procedure' from a 'sensibility'. I point out that the best cooks have all developed a sensibility about cooking that transcends the following of pre-determined recipes. They are able to taste the food, become inspired by something that's missing, then do something new and imaginative. Procedures can be very useful in many circumstances such as going through pre-flight check-lists. But procedures should never be confused with sensibility, which is the competence a person needs to feel at home and innovate in a domain.

We carry out a variety of exercises to get people to understand that listening is not a procedure, but a sensibility that can be cultivated to ever higher levels like a martial art. For instance, we send people out to do structured interviews with their internal customers and performers. We encourage them to listen carefully to the people being interviewed. They always come back startled, with comments like 'I had no idea those guys did it that way!' or 'What they asked for makes a lot more sense than what we are doing now' and 'They didn't know that doing things their way caused us two weeks extra work'. We also ask them to listen to suppliers and customers. This exercise almost always gives managers another surprise and is usually felt to be a sharp wake-up call.

A natural and important by-product of developing a new sensibility from listening to each other is the increased trust and improvement in the mood of the organisation. As a result of careful listening, previous areas of conflict, say, between marketing and engineering people, are transformed as conciliation increases with understanding and attempts to explore ways of improving coordination.

We have classified common moods in organisations and people into positive and negative categories (see Table 2.4|). The positive moods, like ambition, serenity, trust, acceptance, wonder, resolution, and confidence generate a surplus of energy. In contrast, energy is sapped in such negative moods as resignation, boredom, despair, distrust, resentment, confusion, and panic. Talk of energy here is no metaphor. You can literally see the energy level drop in a person who slumps into a mood of resignation, in which that the person believes nothing will ever change. Medical studies support these observations with claims that the negative moods just listed have a negative impact on one's immune system. In the positive mood of ambition, work will be exciting and fun. And health will increase.

Bibliography

Austin J L (Urmson J O and Sbisa Marina, eds) "How To Do Things With Words" Harvard University Press, 1962.

Denning Peter J and Metcalfe Robert M, eds "Beyond Calculation" Copernicus-Springer-Verlag, 1997.

Dreyfus Hubert L "Being-in-the-World: A Commentary on Heidegger's Being and Time" The MIT Press, 1991.

Flores Fernando "The Leaders of the Future" in Denning and. Metcalfe, 1997, 175-192.

Heidegger Martin "Being and Time" Trans. John Macquarrie and Edward Robinson, Harper & Row, 1962.

Heidegger Martin "On the Way to Language", Trans. Peter D Hartz, Harper & Row, 1971.

Kelly Kevin "Out of Control: The New Biology of Machines, Social Systems and the Economic World" Addison Wesley (1995).

Kelly Susanne and Allison Mary Ann "The Complexity Advantage" McGraw-Hill (1999).

Kierkegaard Søren "Fear and Trembling", Trans. Hannay Alastair, Penguin, 1985.

Kroeger Otto with Thuesen Janet M "Type Talk at Work" Dell Publishing (1992).

Maturana Humberto R and Varela Francisco J "Tree of Knowledge: Biological Roots of Human Understanding" Shambhala, 1988

Nietzsche Friedrich "The Birth of Tragedy and The Genealogy of Morals" Trans. Francis Golffing, Anchor-Doubleday, 1956.

Oshry Barry "Seeing Systems: Unlocking the Mysteries of Organisational Life" Berrett Koehler (1996).

Searle John "Expression and Meaning" Cambridge University Press, 1979.

Searle John "Intentionality" Cambridge University Press, 1983.

Spinosa Charles, Flores Fernando and Dreyfus Hubert L "Disclosing New Worlds" The MIT Press, 1997.

Winograd Terry and Flores Fernando "Understanding Computers and Cognition" Addison-Wesley, 1986.

Table 1.1: Business in transition

Table 1.1
	Manufacturing Age	Information Age
THE GAME	Bulk-material manufacturing	Design and use of technology
GOAL	Commodity products	Knowledge-based product
DOMAIN	Regional	Global
FUTURE	Somewhat predictable, deterministic	Uncertain, probable, possible
CHANGE	Periodic nuance, Steady Rate Digestible	Way of life, Acelerating, Overwhelming
RULES	Linear cause and effect	Non-linear complex interaction
GAME PLAN	Five year strategic plans Three year probability scenarios
LEADER	Manages strategic plan to end-state	Envisions and coaches on direction
CHALLENGE	Demand versus capacity to deliver	Demand versus capacity for change
RESOURCE	Material and Capital	Human, Social, Intellectual Capital
RISK	Moving too fast (out of control)	Moving too slowly (out of the running)
APPROACH	Quality, low cost of production Branding, emergent price standard	Be 1st (best if can), high cost R&D Market lock in, high margin
ROLW OF TEAM	Optimise quality/productivity Repetitive day-to-day operation Processing of resource	Quality=Productivity=Adaptability Quest for innovation Processing of information
PROCESS VIEW	Parts interact in sequence of steps	Whole emerges from interacting part
PROCESS	End-to-end efficiency	Micro-to-macro integrity
FOCUS	Standardisation	Feedback

Table 2.1: Three kinds of process

Table 2.1
Materiel	Information	Commitment
Move	Communicate	Declare
Assemble	Assemble	Offer
Transform	Transform	Request
Store/retrieve	Store/retrieve	Assess
Compare	Display	Assert
Things	Data	Relationship

Table 2.2: Contrasts between materiel/information and commitment paradigms

Table 2.2
		Commitment Paradigm
Aspects	Materiel/Information Paradigm	composed of many 'worlds' or sub-universes, each brought forth by historical practices
The 'World' is ...	universal and objective	the product of a social/historical/biological observer (which means 'blindness' is inevitable)
Perception is ...	the passive receipt of data by the senses
Listening is ...	hearing information or gathering requirements	interpreting and tuning to the interpretations of other
Communication is ...	the transmission of information	the successful coordination of action
Action is ...	physical activity	basedon commitments (for which there is a small generative set of speech acts: declare, offer, request, assess, assert)

Table 2.3: Overview of speech acts

Table 2.3
Speech Acts	What is the action?	For example?	What does it produce?
Declare	A speaker declares a new world of possibilities for action in a community.	'We are founding company X to provide Y to customers.' 'We are going to cut our costs by 10%.' 'A firm is a network of commitments.	Leadership and a new context for action to take care of the concerns of the community that listens to the declaration and makes it effective
Offer/Accept	A speaker offers to take care of something of concern to the listener.By accepting, the listener turns the offer into a promise.	'Would you like some coffee or dessert?' 'I propose we meet and discuss that.' 'I'll prepare a report on that by Wednesday.	Mutual commitment to action
Request/Promise	A speaker asks a listener to take care of something that the speaker is concerned about.	'Can you get me on a flight to Boston in time for my meeting?' 'Why did the power go off? (Questions are requests for language-action.)	Mutual commitment to action
Assess	A speakers assesses how some action or thing relates to specific concerns or commitments.	'We are in a mature industry.' 'Our users are happy.' 'IT is revolutionising our work.' 'Our costs are increasing.	Preparation for action: orientations, interpretations and attitudes towards actions or situations
Assert	A speaker asserts (i.e. reports) facts pertinent to the concerns at hand.	'It is 4pm GMT'. 'The drawer is open'. It is a robin'. 'The gauge reads 200 kg'. 'Our sales were $4.2 million this quarter'	Confidence that we share a reliable and observable interpretation of the situation in which we will coordinate action

Table 2.4: Some fundamental moods

Table 2.4
Positive	Negative
* Ambition, Persistence	* Resignation, Boredom
* Serenity, Peacefulness, Joy	* Despair
* Trust, Prudence	* Distrust, Skepticism
* Acceptance	* Resentment, Anger
* Wonder	* Confusion
* Resolution, Speculation, Urgence	* Panic, Worry, Anxiety
	* Arrogance