Soft systems thinking

 Soft  systems thinking

In the previous chapters we have introduced what a system is, developed various concepts and definitions that underpin systems thinking, and introduced a number of complementary, commonly cited systems thinking models and frameworks providing different perspectives to systems. In the previous chapter we elaborated further on the earlier discussion of hard and soft systems introduced in Chapter 2, and went on to develop hard systems thinking and introduce a number of techniques for modelling hard systems.

As a reminder, our earlier definitions of hard and soft systems were as follows:

.Hard system: a system consisting of high-integrity parts that are connected through well-understood interaction patterns producing predictable behaviours.

.Soft system: a system consisting of autonomous parts that are characterized by high variability and unpredictable behaviours and connected through a loosely defined dynamic web of relationships, power structures, shared or conflicting interests and values.

At the end of the last chapter, we discussed limitations of the hard systems modelling approach, primarily its inability to incorporate social dynamics within the system model, i.e., the soft aspects of systems. We also alluded to some soft systems modelling techniques, such as storytelling and role playing when we introduced innovations for overcoming the limitations of the hard systems modelling approach.

In this chapter we will introduce Soft Systems Methodology and several approaches to modelling soft systems including storytelling, role playing, rich pictures and causal loop diagrams. However, before we introduce the methods, we will start by outlining the characteristics of soft systems, which...

will further reinforce your understanding of the differences between hard and soft systems and the thinking that underpins these two paradigms. The last section will introduce causal loop diagrams as a means of capturing the social dynamics of a system, which will provide a segway to the next chapter, where we build upon causal diagrams to introduce causal mapping and its use in group decision making.

LEARNING OUTCOMES

.Understand soft systems and soft systems thinking

.Become familiar with Soft Systems Methodology including concepts such as root definition, CATWOE and trim-tab

.Understand the relative merits and limitations of different approaches to understanding and modelling soft systems, including

storytelling and roleplay

rich pictures

causal loop diagrams

.Be able to model, analyse and improve systems using soft systems thinking and approaches

6.1 Characteristics of soft systems and soft systems thinking

Earlier we defined a soft system as a system consisting of autonomous parts that are characterized by high variability and unpredictable behaviours, connected through a loosely defined dynamic web of relationships and power structures. From this we can deduce that the key characteristic that defines a soft system is the unpredictable behaviour of the system, which is caused by autonomy of each part together with loosely defined dynamic relationships between these parts.

Autonomy of each part means that each part may have its own worldview, which may be quite different from the worldviews of other parts of the system. It would also mean that as an autonomous agent each part would operate to maximize its own purpose, which may be related to financial outcomes such as profit, career outcomes such as promotion, or social outcomes.

such as recognition or acceptance. This autonomy creates a high degree of variation in the behaviour of each part.

To further complicate matters, in soft systems the relationship between each part of the system is loosely defined and highly dynamic. This is best explained in modern-day politics when we see nations who were considered allies not long ago become enemies. In soft systems, the nature of the relationship between two entities can change over time. Sometimes the change can occur over a long period of time; at other times it can happen almost instantaneously. Just think back over your relationships with family, friends and colleagues. Can you think of an instance where you had a close friend and suddenly, over something very minor, they appeared to have fallen out with you and you were left puzzled? Most of us have experienced this kind of change in relationships in our lifetimes, which can often be explained by different worldviews or values, i.e. something you would see as a minor issue maybe a very important issue for them.

In soft systems it is the combination of the autonomy of each part together with the loosely defined and dynamic relationships that makes the system behaviour unpredictable. For these reasons the behaviour of soft systems cannot be reliably explained through mathematical and/or statistical models. To add to the complexity of soft systems, just like any system, soft systems also suffer from entropy. If they are left to their own purposes, they will gradually or sometimes rapidly deteriorate into disorder. Whilst in hard systems the entropy is somewhat predictable, in soft systems entropy is a lot more difficult to predict. In most organizations it is therefore necessary to have constant change, renewal and improvement just to maintain the steady state.

With this level of unpredictability, in soft systems, feedback is used to compensate for deviation from expectations through social mechanisms such as trust, fear, influence, motivation, coercion, engagement and persuasion as well as through more technical mechanisms such as target setting, benchmarking, reward and discipline.

Based on this we can summarize the characteristics of soft systems as follows:

unpredictable behaviour;

autonomous parts with high levels of variation shaped by different worldviews;

loosely defined dynamic relationships that can change significantly in the short and long term;

.parts connected through a defined dynamic web of relationships, power structures, shared interests and values;

.feedback is used to compensate for deviation through social as well as technical mechanisms.

Soft systems thinking is a way of thinking about systems whereby the analyst/thinker is assuming that systems behave as soft systems as defined above. From this definition one can surmise that we would use soft systems thinking to analyse, understand and improve soft systems, while hard systems thinking is used to analyse, understand and improve hard systems. However, this is only partially true, as in principle, we can use hard and soft systems thinking to analyse, model and improve both hard and soft systems. This is best explained through an example.

Let’s consider a bicycle and its rider as a system. The bicycle itself is an engineered system. It clearly is a hard system and there is little benefit in trying to understand the bicycle from a soft systems perspective as each part of the bicycle is deterministic. A bicycle has no autonomy, feelings or sense of purpose. So, it would be safe to limit our analysis of a bicycle to a hard system modelling approach. In contrast, if we consider the bicycle and a rider as a system we have the human component, which makes the system much less predictable. In this case we can still take a view and model the bicycle with a rider from a hard systems perspective. Taking this approach, we would focus on the mechanistic aspects of the system as illustrated below.

BICYCLE AND RIDER AS A HARD SYSTEM

The rider sits on the seat, holds the handlebars with his hands and places his feet on the pedals. The rider then exerts downwards pressure on the pedals, alternating between the left and then right pedals. The pedals and the crank convert this force to rotational motion and transmit it to the back wheel through the chain and the gears on the back wheel. The wheel turns and moves the bicycle forwards. The rider moves the handlebars left and right to steer the bicycle. The rider can squeeze the brake levers that tighten the cables and engage the brake pads on to the wheel rim, slowing down or stopping the bicycle.

From previous trials and measurements of the rider’s strength and stamina, we know that the rider is capable of exerting 200Nm of torque onto the pedals constantly for two hours, which will generate about 660 watts of energy, giving the bicycle an average speed of 38km/h over the two-hour period on flat ground. Specifically, we can predict this system’s performance.

In the above example, the analysis considers the rider and the bicycle as deterministic parts with predictable capabilities that are likely to deliver predictable results. In this analysis we do not consider the rider’s mental state, health or motivation. If we take a soft systems approach, we can develop a very different picture of this system.

BICYCLE AND RIDER AS A SOFT SYSTEM

Prior to a race the next day, the rider has a sleepless night worrying about rising interest rates and energy prices. He is worried whether he will be able to continue funding his daughter’s private education.

The rider is also thinking about the race tomorrow where a good friend is competing against him, which may make things awkward. Furthermore, in the morning he finds that the batteries on the central heating controller have gone flat. When he gets up the house is cold and there is no hot water for a shower, so he leaves home feeling rather grumpy. He is tired, his mental state is not what it should be and he is not fully motivated to win his race today.

Given these conditions the rider is unlikely to be able to perform as previously predicted, but we cannot really say how these factors are going to combine to affect his performance. In fact, we may even be surprised. We may find that he is angry, and he takes his anger out by cycling a lot harder than he usually does, breaking his own record.

In this example, by analysing the bicycle and the rider as a soft system we have overlayed the social factors onto the more deterministic hard systems perspective to give us a more complete understanding of the factors that may contribute to the performance of the system. As in this example, even though we may still not be able to predict the behaviour of the system, we can at least use relationships between parts of the system to retrospectively explain the system’s behaviour. In short, hard and soft systems approaches are not alternative or competing ways of looking at organizational systems, but rather, in many cases, they provide complementary views to help us to develop a more complete understanding of the systems we live and work in.

In analysing and modelling soft systems the analyst tries to capture the inherent complexity from multiple perspectives to present an explanation of the system’s current (as-is) behaviour. In some ways, modelling soft systems is potentially a wicked problem, as it would be near impossible to incorporate everyone’s worldview in exactly the way they see the system. Thus, when modelling soft systems, the analyst is trying to integrate the views of different people whilst developing a credible and shared explanation of the system's behaviour. Therefore, building consensus and developing a shared understanding of the system is an integral part of the modelling challenge. Hence, we can think about what needs to change to improve the behavior of the system. Even then, due to the high degree of unpredictability, it would be prudent to think about any improvement or change proposition as a hypothesis and test this hypothesis before implementing the proposed change in a wider context, as unexpected and unpredicted side effects are likely to emerge. Such experiments can give us further insights into the way the system is working and responding to intervention.

In the following sections we will introduce Soft Systems Methodology together with various techniques for modelling soft systems, or modelling systems from a soft systems perspective. Once again, this is not intended to be an exhaustive list of systems modelling techniques and indeed there are many commercial tools available that use these techniques, or variations of them, to help model soft systems.

6.2 Soft Systems Methodology

Soft Systems Methodology (SSM) was developed during the early 1960s by Peter Checkland and his colleagues at Lancaster University in the UK to help in dealing with complex organizational and societal problems and situations that have divergent views about the problem and potential solution (Checkland, 1981; Checkland and Scholes, 1990 and 1999). As previously defined, when we are dealing with soft systems it may be difficult to agree on the actual problem to be addressed. SSM was developed specifically for dealing with situations where we are not sure what to do, why the problem exists, how urgent the situation is, who is involved, who the stakeholders are and what their views about the problem are. Essentially, SSM provides an organized way of thinking through these kinds of soft systems problems.

SSM comprises seven iterative steps as illustrated in Figure 6.1. In this figure the part above the black line is associated with real-world thinking and the part below the black line is associated with systems thinking. The first two steps of SSM focus on exploring and finding out about the problem in the real world. Steps 3 and 4 are about applying systems thinking by developing the root definitions for the relevant systems and building concepts.

models of the system. The final three steps are about exploring the models in the context of the real-world situation, experimenting with potential solutions, and taking actions to improve the situation. The methodology is not intended to be a linear process; rather, it is an iterative process between steps as it is often necessary to revisit the previous steps before progressing to the next steps. The overall process might also have to be repeated in cycles as it is often necessary to review the outcome of an action or an intervention to understand the unexpected emergent behaviours (i.e. side effects) to refine the interventions and continuously seek improvement of the system.

There are numerous examples of the successful use of SSM to resolve diverse problems, including ecological, environmental, political, economic, business, and military problems. In fact, in the 1990s it was the recommended complementary planning tool for the UK Government’s SSADM. This advice demonstrates the synergy and added value that a soft systems approach can bring to help us understand complex systems.

In terms of its limitations, SSM has been previously criticized as being too linear, too functionalist, and supporting the status quo and existing power structures. However, these claims are often rebutted by users arguing that these limitations are attributable to the users of the methodology rather than the methodology itself.

The overall methodology is supported by several techniques to help the users through the process. Below we summarize each step of the methodology. and include references to some of the commonly used techniques. In the following sections we will expand on some of these techniques to provide a complete understanding of the overall methodology.

Step 1: Problem situation – unstructured

This step is about engaging with various people concerned with the problem and gathering information about the situation that is considered to be problematic. Essentially, in this step the analyst is interviewing various people and getting them to explain what the problem may be, who the stakeholders may be, how the problems and potential solutions are being described by various people, and what the current performance measures and issues may be.

Although interviews are commonly used for data collection, we have observed that asking each person to compile short (five-minute) video diaries (daily or weekly) over a period of time describing what has happened and expressing their feelings and frustrations can be an alternative technique for collecting useful information about the current situation.

In this step it is important to interview people with different worldviews to ensure that we get as broad a picture of the situation as possible. For example, interviewing people new to the system (e.g., new employees) is as important as interviewing people who have some experience of working in the system. Similarly, interviewing customers, suppliers, and external users of a system will help to bring wider diversity in worldviews. This is important because, if people with similar views share slightly different overlapping interpretations of the problem, it’s less valuable than if overlap emerges from conversations with people with divergent backgrounds and worldviews.

Step 2: Problem situation – expressed

The purpose of this step is to capture multiple perceptions or views of the situation. As each person consulted in the previous step will have their own unique background, worldviews, and experiences of the situation, the challenge here is to identify common themes and develop different pictures of the situation. If we have interviewed 10 different people to understand a particular situation, it is unlikely that we will get 10 completely different.

views. It would be highly likely that a number of these perspectives would be similar to each other, and if we integrate these similar and complimentary views, we may end up with three or four quite different views of the situation.

An important outcome from this step is to note the differences in interpretation of the situation. As these differences are made explicit, further discussions and exploration of these differences enable the decision makers to arrive at accommodations/compromises or even a consensus over the situation.

To enable capturing and sharing the problem from different perspectives, Checkland and his colleagues developed the Rich Picture approach. The argument for the use of pictures is that words cannot adequately capture and describe a complex situation on their own. However, pictures used in conjunction with words (i.e., narratives) can help to develop a much richer picture of a complex situation, thus Rich Pictures. This approach has also proven to inspire people and help them talk more openly about problems and issues.

Step 3: Root definitions of relevant systems

In this step the objective is to capture the root definition or the purpose of the system that is relevant to the problem or situation at hand. A critical aspect of the root definition is the transformation that is performed by the system to deliver its purpose, which is captured by the verb in the root definition. For example, if we think of an education system we can formulate the purpose and therefore the root definition of the system in different ways as illustrated below. In this example, the transformation is highlighted in bold italics, and is significantly different in each root definition.

Higher education system...

...a system to drive economic growth through innovation and enterprise

...a system to make money through international student fees

...a system to develop new knowledge and global impact through leading-edge R&D

To help us ensure that the root definition is appropriate, Checkland and colleagues have developed a completeness test based on the mnemonic CATWOE, which stands for Customer, Actor, Transformation, Weltanschauung (Worldview), Owners and Environmental constraints. Below, each of the components is described in more detail:

.Customers are people or organizations that receive and are impacted (positively or negatively) by the output from the transformation. More recently this concept has been further developed by splitting the customer into two categories: beneficiaries and victims of the system.

Beneficiaries of the system are those customers experiencing a positive impact from the outputs of the system, and victims of the system are those customers experiencing a negative impact from these outputs. With this distinction, sometimes you may encounter the mnemonic BATWOE appearing instead of CATWOE.

At this point, it is also important to point out the difference between customers and clients. Clients are not the same as customers, rather they are actors (see below). In SSM, many systems that serve one group (clients) will often impact a different group (customers).

.Actors are those individuals, groups or organizations that make the system work. They carry out the transformation. According to this definition, actors would include clients, suppliers, and business partners.

Sometimes actors could also be customers or even owners (see below) of the system, but more often they are not. A good test to consider if an individual or a group of individuals are also customers is to think about whether they are positively or negatively impacted by the system. Similarly, considering if the actors also have the power to change the system or disband it would help us decide whether an actor is also an owner.

.Transformation in its simplest form is the purposeful activity that transforms inputs to outputs. However, in complex systems the input-output view is often too complex to express simply in one sentence, thus it is often better expressed as a system to deliver a purpose through some function; for example, an education system is a system to drive economic growth (purpose) through innovation and enterprise (function).

.Weltanschauung is the German word that means ‘worldview’ and it has somehow become part of the terminology often used in systems thinking. It captures the underlying and often uncommunicated beliefs that give meaning to the root definition. In this book, we chose to use worldview.

.Owner(s) is/are the key decision maker(s) who are concerned with the performance of the system and who can change the performance measures of the system. As previously explained, an owner can also be a customer or an actor.

.Environmental constraints recognize that all systems exist as parts of larger systems; thus, their operations and performance are constrained by factors outside the system’s boundary. Thus, environmental constraints are those constraints that come from outside the boundaries of a system and are significant to the system’s operation and performance. Therefore, the ‘root definition’ should include anything special in the environment of the system that is likely to affect the transformation.

Based on the CATWOE test as described above, a complete root definition of the system could be structured as follows:

A system owned by ‘O’ in which actors ‘A’ perform transformation ‘T’ for customers ‘C’ within the worldviews ‘W’ and the environmental constraints ‘E’.

Thus, our earlier definition of the higher education system could be developed into a complete root definition as follows:

A system owned by ‘O’ in which actors ‘A’ perform transformation ‘T’ for customers ‘C’ within the worldviews ‘W’ and the environmental constraints ‘E’.

Thus our earlier definition of the higher education system could be developed into a complete root definition as follows:

A system owned by the government (O) in which universities (A) in collaboration with schools, further education colleges, and enterprises (A) drive economic growth (T) by creating and growing enterprises for the benefit of the wider society (C) underpinned by the values of responsible and sustainable business (W) and within technological capabilities (E).

In general, there are no particular rules about how to develop root definitions as long as they conform to the CATWOE test. Some users prefer to identify all the CATWOE elements first and then start compiling a root definition; others may prefer compiling alternative root definitions first and then using the CATWOE test to refine them.

Whichever approach is taken, it is important at this stage to note that as part of this step (Step 3) it is likely that we end up with a number of competing root definitions due to different views about the purpose of a system. Such an outcome was illustrated with the example we provided earlier in this section about the purpose of the higher education system (i.e., drive economic growth vs. make money vs. develop new knowledge).

At this stage it might be appropriate to conduct the first cycle of the iterative process. If you are faced with multiple root definitions, it would be appropriate to reduce these to a smaller number by scrutinizing each of the definitions and through further discussions with key stakeholders. In some cases, it is also possible that the different root definitions could co-exist within the same system but as different subsystems within the larger system as discussed in Step 4.

Step 4: Building conceptual models

Once we have agreed on one or several root definitions, we can move on to building conceptual models of the system and its behaviour. If after the discussion with key stakeholders you have not reached a compromise on the common view, it may be that you have more than one root definition representing different distinct subsystems within the wider system. If this is the case, you would need to develop a separate conceptual model for each one of the subsystems.

At this stage it is worth noting that it is not uncommon to discover a number of distinct subsystems within an organization with different root definitions. Often some of the problems and issues we are trying to address can be attributable to the internal conflicts between these subsystems. Therefore, recognizing these as different subsystems with different root definitions can be a key step towards addressing these problems. To illustrate this, we have provided two examples below.

EXAMPLE 1: PUMP MANUFACTURING

A large engineering company is engaged in the design, manufacture, and installation of pumps for their customers in the mining, oil, and gas sectors. On the one hand, they design, manufacture, and sell standard pumps. These standard pumps are made to stock based on a forecast. The customers buy them from a product catalogue; they expect short delivery lead times and make buying decisions primarily based on the price.

They also design, manufacture, and install custom-engineered pumping solutions that have been engineered and manufactured as one-off products to.

meet specific customers’ requirements. The main order-winning criterion is the organization’s capability to design, develop, manufacture, deliver and install a pumping solution fit for purpose, to budget and to timescale. Thus, these products are significantly more complex and uncertain when compared to the standard products.

Clearly, here the company has two different systems, each with a different root definition:

A system where marketing, engineering, manufacturing and sales departments collaborate to create value by designing, manufacturing and selling market-leading standard pumping solutions.

A system where clients, engineering, manufacturing and projects departments collaborate to create value by designing, manufacturing and selling customized pumping solutions to address the needs of each client.

EXAMPLE 2: EDUCATION

A UK-based research-intensive university has undergraduate and postgraduate programmes as well as being active in local innovation and enterprise development activities. In this context, their undergraduates are mostly UK-based students who undertake five-year engineering degrees. The university actively coaches these students to develop innovations that could be transferred to the private sector or developed into start-up enterprises. In contrast, the postgraduate programmes are mainly popular among fee-paying overseas students undertaking standard one-year master’s programmes. The university also has an active research and development culture and is one of the leading UK universities in terms of research intensiveness.

In this case it is conceivable that the university has three distinctive systems, each with a separate root definition:

 A system to drive economic growth through innovation and enterprise.

 A system to fund R&D and economic growth initiatives through international student fees.

 A system to develop new knowledge and global impact through leading-edge R&D.

Once we have finalized our root definition(s) we can move on to build conceptual models of the system. To build the model we start by identifying the key activities that underpin the root definition. Often these activities can be ascertained from the verbs contained in the root definition, which help start building the models. However, in complex systems it is not always possible to ascertain all the key activities from the root definition. For example, in the pump manufacturing example above, activities such as collaborating, designing, manufacturing and selling are part of the root definition and would provide a useful starting point for building a conceptual model. In contrast, however, the root definitions for the education system are less helpful in this regard. Whilst 'driving economic growth', 'innovating' and 'enterprise' are not useful to enable us to start building conceptual models, we can always go back to the data from stories, interviews and rich pictures in Step 2 to establish how these outcomes or functions are achieved.

Building a convincing conceptual model involves capturing the key activities and other elements that shape the behaviour of the system together with the logical links that connect the elements together. For example, if the outcomes of the system are leading to poor customer experience, there must be logical links from different parts of the system that lead to poor customer satisfaction; this could be product quality, late delivery or even unhelpful staff behaviour. Taking the latter as an example, when we look for the cause of this unhelpful staff behaviour, we may find that it is mostly attributable to staff morale and motivation, which in turn may be caused by other things.

Although using storytelling and rich pictures can help us to develop conceptual models, from experience we find causal loop diagrams more useful as a formalized way of developing these conceptual models. Indeed, they also enable us to explore solutions more easily in the real world before taking action (i.e., Steps 5, 6, and 7).

In this section we will refrain from going into further detail on storytelling, rich pictures and causal loop diagrams, as we will discuss these in greater detail later in this chapter. However, at this point it would be appropriate to include a health warning. In the literature and wider internet there are some examples of soft systems models using rich pictures and causal loop diagrams that are rather simplistic and arguably focus on modelling the hard aspect of the system. The real value of soft systems modelling is that it helps us understand and model the social dimensions of the system, which helps us develop a profound understanding of why complex systems behave the way they do. Thus, when reading other sources on systems thinking, 

The final but important part of building a conceptual model is to identify and agree on the measures of performance for the system. According to SSM there are three measures of performance: effectiveness, efficiency, and efficacy (which we discussed in Chapter 3). It is thus important that before proceeding to Step 5 we agree on the performance measures that enable us to evaluate and compare the current state of the system to any future improved state considering efficiency, effectiveness, and efficacy.

Step 5: Comparing the conceptual model with the problem as expressed

In this step the main concern is to identify the potential changes that can be made to the system to resolve a particular problem/issue or improve the performance of the system. A simple way of doing this is to focus on the key elements of the system (from the root definition) and compare real-life issues with what could be done in the future. This is illustrated in the example in Table 6.1, which is based on the root definition of the higher education system introduced earlier. As a reminder, the root definition was:

A system owned by the government (O) in which universities (A) in collaboration with schools, further education colleges and enterprises (A) drive economic growth (T) by creating and growing enterprises for the benefit of the wider society (C) underpinned by the values of responsible and sustainable business (W) and within technological capabilities (E).

Commonly our initial exploration of the real world (Steps 1 and 2) would have surfaced assumptions about the underlying causes of the problem and what needs to be done to resolve the problem. This approach enables us to test these assumptions, which may be ill-founded. It is these differences between what happens in reality and the logical model that raise the questions that will ultimately lead to change.

Step 6: Feasible changes or interventions

This step is primarily concerned with identifying the feasible changes or interventions that would improve the behaviour of the system and consequently its performance. A table such as 6.1 is likely to contain several.

Root definition Issues in the real world What can be done?

… universities (A) in collaboration with schools, further education colleges and enterprises Poor alignment between schools, colleges and universities, little or no collaboration with enterprises Review the purpose and objective of schools, colleges, and universities

… drive economic growth In the fourth quartile for developed (OECD) economies Create measures and incentives

… creating and growing enterprises Enterprise death rate is higher than birth rate Increase financial support for start-ups

… benefit wider society Not visible. No clear benefits Fund enterprise education

Owned by government Primarily concerned with measuring and reporting outputs Advisory support for growth

Measure the increase/decrease jobs versus higher value jobs

Focus on measuring and growing underlying capabilities

possible interventions that would bring the real world closer to our conceptual model. However, there are a number of pertinent points here that need to be at the forefront of our minds when deciding which changes to implement.

In line with the fundamentals of systems thinking discussed earlier in this book, the first point is that once we make a change to the system, the system is changed, and it would need to be restudied and reanalyzed in order for us to understand the new system and the new constraints. To give you a simple example, if we take a length of chain, how can we improve the strength of the chain? The simple answer is that we find the weakest link in the chain and strengthen it. If we need to improve the strength of the chain further, we would need to study the chain again, find the next weakest link and improve that link and so on. In summary, identifying and making several changes to the system may not deliver the expected results. In fact, it is likely to be counterproductive, with greater likelihood of delivering unexpected and undesirable side effects.

The added benefit of this approach is that by focusing on one change/intervention at a time we are more likely to get the change done, do it well and do it quickly. We must recognize that all change initiatives consume resources. One scarce resource is people's time. People have to work on their 'day job', i.e. designing products, serving customers, operating machines, loading trucks, etc., and in addition, they have to work on 'the project'. So, in terms of people's time, the change project is already in competition with the day job. If we introduce several projects simultaneously, there is a good chance that they will be competing for the same resources.

The second point is that the change and intervention should focus on identifying the most significant constraint that is preventing the system from getting one step closer to its purpose or delivering its performance measures. That is, the change or the intervention should be subordinate to the purpose and performance measures of the system.

The third point is that we should focus on finding the one simple change we can make that will have a profound effect on the behaviour of the overall system. 'Call me trim-tab' is the phrase engraved on the gravestone of Buckminster Fuller (1895–1983). Fuller, an American architect, systems theorist, author, designer and inventor, is frequently quoted for his use of trim-tabs as a metaphor. In 1972 Fuller said:

"Something hit me very hard once, thinking about what one little man could do. Think of the Queen Mary [the ocean liner] – the whole ship goes by and then comes the rudder. And there’s a tiny thing at the edge of the rudder called a trim-tab. It’s a miniature rudder. Just moving the little trim-tab builds a low pressure that pulls the rudder around which brings the whole ship around. [It] takes almost no effort at all...!" (Kowalski, nd)

This trim-tab analogy, which goes by different names such as levers or pressure points in the system, captures an important feature of a systems approach. That is, one small change in an almost insignificant part of a complex system could end up changing the behaviour of the whole organization. The trick is knowing where or what the trim-tab is, or even how to find it. We will further discuss the techniques for finding the trim-tab when we look at causal loop mapping in detail later in this chapter. We will then discuss different types of trim-tabs or levers in the system in Chapter 9 after introducing system dynamics as a method for simulating the future behaviour of a system. The fourth point is that people are not always motivated to implement change, particularly when people involved in the potential change have. 

"conflicting views or even see themselves as the victims of the change—the logic of the conceptual model is undeniable. Therefore, when selecting which change to implement, it is important to understand the resistance to change. In this context, the good thing is that during the potential stages of SSM we should have understood different worldviews of people and this will give us a good idea as to how the change will affect the customers of the system, and whether they will become beneficiaries or victims. Needless to say, where a change is going to produce more victims than beneficiaries, it is going to be harder to implement. In this context, a simple change that might produce a lot of resistance is not as simple as it seems, and thus we must either find ways to mitigate resistance to change or seek alternative interventions that would be simpler to implement.

Step 7: Action to improve the problem

Once we have identified the change that we would consider feasible, then take action to implement this change. As stated above, this change will rest in a new system that may even affect the wider system within which our system exists, leading to more problems and opportunities. Thus, the process starts again from Step 1."

REFLECTIVE EXERCISE

Think about a system you know well. This could be a business or a department you work in or a university, school or department within which you are studying. Use the CATWOE model and the examples from earlier in this chapter to develop a root definition for this system.

Then talk to some friends and colleagues who also work or study in the same system to see how they conceptualize the system, but at this stage don’t share your root definition with them. Also, you do not need to educate them about systems thinking and all the concepts associated with it to do this. Just ask them a few questions around what they think the purpose of the system (business/department/club/school) is, who (people and organizations) they think are the key players in the system, and who organizes the system, measures its performance and is ultimately responsible for the system.

Compare their answers to the root definition you developed earlier. Do they agree with your root definition? Are they similar or wildly different? How can you further develop your root definition to incorporate their views?

"Having introduced Soft Systems Methodology in some detail, in the following section we will introduce different but complementary techniques for modelling soft systems."

6.3 Storytelling and Roleplay

Just like playing is the most elevated form of learning and investigation, storytelling is also an effective way of communicating and learning that has been developed and refined by different cultures over centuries. Storytelling is a social activity of sharing stories, sometimes including props, artefacts, pictures or roleplay/theatrics. Every culture has its own stories or narratives, used as a means of communication for the purposes of cultural preservation, education or entertainment. Therefore, it is considered a useful way of capturing and sharing information about the real world as well as sharing and communicating information about the conceptual model.

Essentially, a story is a narrative that is structured around a plot and characters, as we already observed at the end of Chapter 5 when we discussed how one organization overcame the limitations of hard systems modelling (SSADM) by roleplaying and storytelling along the process. In this particular case, when finding out about the real-world problems (Steps 1 and 2) the team captured various individual’s stories about the company (the system) from different perspectives. These stories covered numerous themes explaining what worked, what did not, how and why, from different perspectives. Some were factual, some were sad and others very funny. These stories became infinitely valuable when identifying different root definitions for the system.

In a similar vein, vignettes are brief stories of events that may describe a particular situation. Typically, they are 800–1,000 words, but they can be just a few sentences long, as exemplified by the following episode demonstrating the (lack of) value placed on annual performance reviews in an organization:

… in our organization we do annual performance reviews, but we just pay lip service to them, we do not really value them. Last year my manager told me that we will do my performance review next week on the flight when we are travelling to Milan, which I was not happy about…

Vignettes are commonly used in research to understand a particular phenomenon. Collecting numerous stories about the phenomenon from a diverse  

"range of people with different perspectives enables the investigator to analyze the vignettes for patterns to identify themes and help them build theories to explain this phenomenon.

In similar ways, storytelling and roleplaying can be used to share feedback and build consensus about the current state of a system, helping to pave the way towards identifying a common solution that is acceptable to all. There are no hard and fast rules to storytelling and roleplay as it is considered an art form, and requires a certain degree of creativity, vision, skill, and practice. Having said that, the basic skills of storytelling are not beyond the capabilities of most people, they may just need some good examples of some advice and coaching to point them in the right direction. The internet is full of advice and guidance about how to tell effective stories, most of which would be useful for developing basic storytelling skills required to capture and tell stories about systems we are studying.

Roleplaying is an effective way of bringing a story to life. Similarly, roleplaying is also an artform that could be developed by most people over time. It comes more naturally to some people than to others. In most situations there are people who have some experience of roleplaying either from school or a local theatre group. Tapping into people who have some experience of roleplaying or who are eager to participate is a good way of getting people engaged in the story and communicating the current situation within the system."

6.4. Rich pictures

The rich picture technique was developed as an action learning mechanism to help learn about complex and ill-defined systems. There are no formal rules, nomenclature or syntax to follow; instead, they rely purely on the user’s creativity in using symbols, cartoons, and sketches to help them tell a story. The real value of the technique is that it forces the creator out of their comfort zone and makes them think more deeply about the system or the problem. Combined with the narrative that accompanies the illustration, they provide a powerful way of telling a story about the behavior of a system, as exemplified in Figure 6.2.

The following is his narration of the picture, with our comments inserted in brackets:

As a management team, we are trying to create this tranquil environment [oasis with palm trees] where academics, students, administrative and technical staff have some stability and can perform. However, we have a number of forces that we have to battle with on a daily basis. On the one side we have kids [school pupils] wanting to come and study with us, on the other hand we have industry wanting more and more from us and our students. We also have the faculty, university, funding councils and professional bodies throwing more and more demands on us [illustrated as people firing arrows]. Quite frankly I cannot help but think that we are sitting on a time bomb [note a clock and dynamite at the bottom of the picture].

From the above example we can see that the picture alone is of limited use to communicate the message behind the picture. However, in this case the technique has enabled this person to express his frustrations with the current system and the situation as he sees it. It becomes much clearer when the picture is accompanied by his narrative.

Liver disease

"Earlier in this book we introduced the concept of rich pictures, albeit with out referring to them as rich pictures. The pictures in Figure 2.1 illustrate how different students conceptualize a university. Effectively these are rich pictures that could be useful, particularly if accompanied by a narrative for communicating and sharing each student’s conceptualization of a university from their own viewpoint. In fact, rich pictures are commonly used to help management teams in an organization to understand each other’s viewpoints, which can help ease some of the underlying frictions with the organization. The example, Figure 6.3 illustrates rich pictures drawn by three different senior managers from the same organization. You will notice that the three stories have certain similarities and differences. Although free-form pictures as illustrated in Figure 6.3 are useful to understand how to conceptualize a particular system or problem, the"

"more structure and process a rich picture contains the more useful it becomes to help us model the system later. As there are no hard and fast rules for developing rich pictures, we can often use the basic concept of a system, i.e., a system is a collection of interacting parts in which the interactions result in system-level properties and behaviours not attributable to the sum of individual parts. Based on this definition we can start building rich pictures that help us describe the parts and how they are connected to one another. In Figure 6.4 we have developed a rich picture, using standard PowerPoint icons, of the situation we described earlier in this chapter. As a reminder, the earlier story is as follows: With the recent increase in interest rates and energy prices the rider is worried about his family’s finances and whether they will be able to continue funding their daughter’s education in the private school. The rider has a sleepless night worrying about finances. The rider is also thinking about the race where a good friend is competing, which is going to make things awkward. Furthermore, in the morning he finds that the batteries of the central heating controller have gone flat. When he gets up the house is cold and there is no hot water for a shower. So, he leaves home feeling rather grumpy. He is tired; his mental state is not what it should be, and he is not fully motivated to win this race today."

"In Figure 6.4, we can see that we can capture richer information which we did not identify from our earlier narration. This is because the act of drawing the picture makes us think more deeply about the situation and identify further linkages that had not thought about. Particularly, there are some self-reinforcing loops identified, such as the cyclist being angry at himself for not checking the battery in the central heating controller. Also, his awareness of little sleep, not having a clear head and lacking motivation are causing further anxiety. With these forces it would be safe to predict that this race will not be this cyclist’s best performance.

In summary, rich pictures provide a flexible and unconstrained platform for communicating experiences and viewpoints of a given system or problem situation. Thus, it is a valuable technique for capturing and communicating how people see a situation individually and collectively. It is also worth noting that it is possible to produce rich pictures as a group by having everybody contribute to its creation. This approach can also help to develop and reinforce a shared understanding of the system or problem situation."

"6.5 Causal loop diagrams

We have already introduced the concepts of causality and feedback loops earlier in this book (see section 3.7). In this section we will build upon the concept of causality and feedback loops as well as other things we have learned about systems and systems thinking so far, to develop an insight into how we can start modelling and understanding complex systems.

Essentially, a causal loop diagram illustrates the causal relationships between the entities within a system. As in rich pictures, they help visualize how different factors in a system are causally interacting to shape the behaviour of the system. In essence, one could argue that the rich picture we constructed in Figure 6.4 is an informal causal loop diagram. Whilst in rich pictures there are no hard and fast rules, in causal loop diagrams we have entities described in words and relationships between entities described by arrows. Furthermore, the relationships between entities can be characterized as positive (+) or negative (-) links. A positive causal link means that the two entities change in the same direction, e.g., both improving or both deteriorating. A negative causal link means that the two entities change in opposite directions, e.g., an improvement in one will cause deterioration in the other.

Let us develop these relationships through a simple example:

"When you feel heavy and lethargic you feel more fed up, and you eat more cake, which makes you feel worst (Figure 6.5b). This is a feedback loop, which is self-reinforcing, i.e., a positive feedback loop which will continue getting worse if we do not intervene. This kind of feedback loop is known as a reinforcing loop.

Instead, when you are fed up if you drink water, exercise and eat fruit, you will still get the energy from the sugar in the fruit but without feeling heavy and lethargic. When you feel less heavy and lethargic (Figure 6.5c). Now we have a negative feedback loop, which balances the effects of the reinforcing loop. These negative feedback loops are known as balancing feedback loops.

Understanding these feedback loops is important, as they are essential features of causal loop diagrams. It is important point to note that the arrows.

"In causal loop diagrams represent causal relationships between entities, which are variables with values that can increase or decrease. Do not confuse this with data flow diagrams where the arrows represent material, information or even people flows.

Another important point to remember when constructing causal loop diagrams is the language we use when explaining the variables as this can impact on the nature (positive or negative) of the relationship between entities. When naming entities, you should refrain from qualifying them with phrases such as increasing, decreasing, improving, deteriorating or similar, as they can significantly change the nature of these relationships, from positive to negative or vice versa. This advice is the opposite of what is recommended in formulating statements in causal mapping discussed in the next chapter, whereby we are looking for the statements to be actionable. We can clearly observe the effects of such qualifiers in the example illustrated in Figure 6.6.

In Figure 6.6a we can observe each entity without any qualifiers and their relationships. In contrast, in Figure 6.6b we can observe the impact of using..."

"qualifiers on the same relationship. For example, when we qualify Fed Up with Less Fed Up, and Eat Cake with Eat More Cake the original positive relationship between Fed Up and Eat Cake reverses. Nevertheless, in practice we have found that some people find it useful to use qualifiers with entities when building causal loop diagrams as it enables them to keep all relationships either positive or negative to tell the story. This approach also enables the reader to follow the story more easily. For this reason, it is acceptable practice to build causal loop models using qualifiers. Naturally the model can always be edited to remove the qualifiers and correct the nature of the relationships. We will illustrate this in the example that follows.

Furthermore, to make causal loop diagrams comprehensible and easier to follow, diagrams are usually accompanied by a narrative that explains what is happening. To accurately relate the narrative to the diagram, sometimes the entities, relationships and feedback loops may be labelled, e.g. E1, E2 and E3 to represent entities and R1, R2 and R3 to represent relationships. This kind of labelling can also help you to tell the story of the model in a consistent manner."

"In terms of Soft Systems Methodology, development of the causal loop diagram is understanding the system and developing a conceptual model that represents how the system behaves and the forces that make the system be the way it does. Once we have completed building the causal loop diagram the next step is to analyse it to see what can be changed to improve the current situation and explore different solutions offered by different people (customers, owners, actors) in the system; we illustrate how this can be done in the case study which is included at the end of this section."

"In searching for a feasible change to improve the system we go back to the causal loop diagram and look for a number of patterns:

We explore the balancing and reinforcing loops to see what changes could be made to improve the system’s behaviour.

We look for busy entities with a lot of inputs and outputs as busy entities are influenced and are potential influencers of the entire system.

We look for entities with a lot of outputs as they tend to influence a lot of other entities and can be the main influencer.

Finally, finding the changes to make to the system is not a straightforward task. Even after considerable modelling and analysis the changes we come up with may not deliver the expected outcomes. Even then, if we treat these as experiments, we can learn more about the system and eventually find the right intervention that will deliver the improvement we seek.

In the following case study, we illustrate this phenomenon together with various aspects of Soft Systems Methodology as well as the use of causal loop diagrams in understanding, modelling, analysing and then improving a complex system."

"CASE STUDY

Systems thinking case study

Understanding and modelling the system

We were invited by a national utility provider to help them resolve what they perceived as a wicked problem. To develop a causal loop model of the current situation we have progressed through steps 1 to 4 of Soft Systems Methodology. In summary:

Step 1 – we interviewed several people from the organization including senior management from corporate headquarters, middle managers managing the operations, customer service supervisors and agents, and even some customers. We summarized each interview into a short story reflecting their position in the system. In line with the CATWOE framework, we identified Customers – home residents and businesses who use the service and might call the call centre, Actors – customer service agents, supervisors and middle managers, and Owners – senior management from Headquarters, and described their worldviews and the environmental constraints (e.g. media and the wider society)."

"Step 2 – we consolidated different stories into a single story by discussing the problem situation and different views and making appropriate accommodations and compromises. By the end of this stage, we had most of the story developed, discussed and refined in the form of the causal loop model.

Step 3 – we explored various root definitions for the system and finally agreed it to be a system owned by the senior management (owners), in which customer service agents, supervisors and managers (actors) work to serve the people at home and at work (reflecting household customers and business customers) to resolve service problems and address enquiries (transformation) efficiently and effectively (worldview) within the technical capabilities of our equipment (environmental constraints).

Step 4 – we developed the causal loop diagram (Figure 6.7) and its accompanying narrative represented our conceptual model of the system as it existed at the time. The following is the abridged narrative for the causal loop diagram illustrated in Figure 6.7. We start our narrative at entity E1. The narrative follows the entity numbers in sequence."

"Utility is a national company providing services to households in the UK. At the time of conducting this analysis it was criticized on a weekly basis on national television (E1) for poor customer service (E2) and poor customer satisfaction (E3). Combined, these three factors cause more customers to call the company, increasing call volumes (E4). To deal with increasing call volumes and to address poor customer satisfaction the company hires more people, increasing headcount (E5) and national criticism. The increase in call volumes and headcount leads to reduced productivity (E7). This together with increasing costs attracts attention from corporate management (E8), which leads to senior management micromanaging (E9) with daily monitoring of costs (E10) and increased emphasis on productivity management (E11), and this in turn leads to deterioration of management morale (E12). To improve productivity and costs, call targets (E13) are introduced, mandating that agents deal with each customer’s call within eight minutes. To help manage this, call target egg timers (E14) are introduced on every customer services agent’s computer screen. These egg timers change colour to amber at six minutes and to red at eight minutes to signal to the agent that they are near or at call target. Daily performance reviews (E15) are used to review the average performance of each agent to the eight-minute target and agents with lower performance are referred to the performance improvement programme (PIP), which essentially retrains the agents on how to serve customers and which they had already completed when first joining the company. Furthermore, the PIP is conducted out of hours so that it does not negatively impact productivity, and conducting PIP is seen as a punishment (E16). With the threat of PIP, agents defer eight-minute target, daily performance reviews and the threat

"...include, ‘Sorry the computer systems are down, can you please call later?’ and ‘You have come to the wrong department, I will put you through to the correct department’, and the customer is put back to the same telephone queue for someone else to deal with. This dynamic leads to deteriorating employee morale (E18), which together with deteriorating management morale (E12) leads to increasing levels of absenteeism and attrition (E19). This contributes to the deterioration of employee engagement (E20), which in turn reinforces dysfunctional behaviours (E21) such as deferral of customers (E17). This leads to poor customer service (E2) and satisfaction (E3). Moreover, whilst increasing absenteeism (E19) impacts directly on productivity (E7), increasing attrition (E19) leads to erosion of knowledge and experience, which further impacts on productivity (E7).

"In contrast to the causal loop model illustrated in Figure 6.7, the model illustrated in Figure 6.8 is the same causal loop model with all the qualifiers removed and relationships expressed as either positive or negative relationships. Although this model better complies with the principles of causal loop mapping it is somewhat more difficult to follow compared to the causal loop model illustrated in Figure 6.7. In our experience in analysing complex systems, creating causal loop models using qualifiers and a narrative is sufficient to tell a story and start identifying opportunities for improving and innovating the system. While causal loop models with no qualifiers and positive and negative relationships become more useful if we are going to pursue a more quantitative modelling and analysis approach, such as systems dynamics modelling and simulation, which we cover in further detail in Chapter 8."

"CASE STUDY ... Continuing the systems thinking case study

Analysing, improving and innovating the system

Earlier in this chapter we introduced the concept of the trim-tab when we introduced Step 6 of Soft Systems Methodology. In short, trim-tab can be defined as the single small change we can make to the system that would end up changing the behaviour of the whole system. In this section we explain how we can analyse our conceptual model (i.e. the causal loop model in Figure 6.7) to identify the changes we can make to change the behaviour of the system.

Step 5 – we analysed the conceptual model to see what can be changed to improve the current situation and explored different solutions offered by different people (customers, owners, actors) in the system. Table 5.2 illustrates an abridged version of the potential solutions proposed by various people. We have organized them against the root definition and the issues observed in the real world."

Root definition

... in which customer service agents, supervisors and managers work

Issues in the real world

Deteriorating morale, poor engagement, dysfunctional behaviour

What can be done?

Better technology support

Standard scripts

More team-building activities for staff

More social events for staff

Make everyone redundant and hire back the right people

Remove egg timers and 8-minute call target

Introduce self-managing work teams

Root definition

... resolve service problems and address enquiries

Issues in the real world

Rarely solved at first call, average no. of calls unknown

What can be done?

Improve pay to bring in more committed people with customer service experience

Focus on the customer

Root definition Issues in the real world What can be done?

... efficiently and effectively Efficiency is managed on a daily basis, what about effectiveness? - Increase focus on effectiveness

- First call resolution

Owned by senior management Increasing micromanagement - Put systems in to remove the need for micromanagement

- Trust your people and get out of the way

- Introduce self-managing work teams

Step 6 – in searching for a feasible solution we went back to the causal loop diagram and looked for:

Balancing and reinforcing loops. However, sometimes these loops are not as simple and obvious. In the causal loop model (Figure 6.7) we can see that the whole model is a reinforcing loop which starts and finishes with increasing call volumes (E4). Simply put:

increasing call volumes (E4) lead to increased headcount (E5), increasing costs (E6) and reduced productivity (E7);

this in turn leads to increased management attention (E8), micromanagement (E9) and emphasis on productivity management (E11, E13, E14, E15 and E16);

these elements all serve to create problems with employee morale (E12 and E18), engagement (E20) and behaviour (E17 and E21);

resulting in poorer customer service (E2) and customer satisfaction (E3), which drive call volumes (E4).

Busy entities with a lot of inputs and outputs. If we follow the above logic, we can observe that entities such as call volumes (E4), deteriorating productivity (E7), increasing emphasis on productivity (E11), daily performance reviews (E15), deteriorating employee morale (E18) and customers being deferred (E17) are all busy nodes.

Entities with a lot of outputs. In our example productivity management (E11) has comparatively more outputs and appears to be a key influencer for the entire system.

Based on this analysis together with the following evidence we can conclude that this system there is an overemphasis on productivity management which creates the reinforcing loop we are seeing here without a balancing loop that reinforces the 'effectiveness' of the system. In this case, it was agreed that the first-call resolution...

"would be introduced as a performance measure to balance the emphasis on productivity that is being driven by the eight-minute call target and the egg timers on the screens of the agents. Although this change resulted in some improvements, it did not solve the problem.

Clearly, having changed the system we have a new system. Theoretically, to analyse the new systems we would need to go through the SSM starting from Step 1. However, in this assignment we did not go through the second full iteration because we could clearly see that although first-call resolution was introduced as a measure, due to the learned behaviours at all levels productivity was being managed more actively. This was reflected in the following quotes by the customer service agents: 'I hate the egg-timer, it distracts me from getting the job done', 'Performance reviews are embarrassing, particularly if I spent ages trying to help a customer', 'The eight-minute target and PIPs force me to do things I do not like... I’d rather work elsewhere'.

From this it was clear that the balancing loop we created to resolve the problem was not strong enough to overcome the existing reinforcing loop. Furthermore, the data revealed that 68 per cent of the calls coming to this customer service centre were failure calls, i.e. because of not being able to deal with a customer’s problem/enquiry at their first call. To continue refining the solution, we went back to the root definition and modified it as follows:"

"A system owned by the senior management (owners), in which customer service agents, supervisors and managers (actors) work to serve the people at home and at work (reflecting household customers and business customers) to resolve service problems and address enquiries (transformation) on the customers' first call (worldview) within the technical capabilities of our equipment (environmental constraints).

The new root definition changed the worldview of the organization, which previously valued efficiency and productivity over effectiveness, i.e. first-call resolution. As a consequence, all productivity-related measures, such as the eight-minute target, the egg timer on the screen and individual productivity-related KPIs at operational level, were removed and the company operated using a single operational measure: first-call resolution. This resulted in a complete transformation in the way the system behaved. Customer service and satisfaction improved, call volumes dropped significantly, employee morale and satisfaction levels improved and the company no longer featured in the national media for their poor customer service. Furthermore, productivity improved significantly, despite management’s concern that removing productivity measures would result in deterioration of productivity. This is because they were resolving more issues at the first call, which resulted in the number of failed calls plummeting."

"At the start of this section, we started talking about the ‘trim-tab’, i.e., the one small change we could make to a system that would change the behaviour of the whole system. In the case study we demonstrated the use of causal loop diagrams for modelling, analysing, improving and innovating systems. We found the trim-tab to be a small change in the performance measures used. We have also observed that the introduction of first-call resolution measures was amongst the potential solutions that emerged from the analysis. However, modelling the system, understanding the reinforcing and balancing loops, and experimenting with the system enabled us to find the trim-tab. Even then, what worked was not just the introduction of the first-call resolution measures, it was a combination of this and the removal of the productivity measures."

6.6 Summary

In this chapter our objective was to reinforce the differences between hard and soft systems, introduce you to soft systems thinking and Soft Systems Methodology (SSM), as well as providing you with a working understanding of the approaches to modelling soft systems.

We started the chapter by looking at the differences between hard systems and soft systems in greater depth to reinforce your understanding of soft systems. We then introduced Soft Systems Methodology and the seven-step approach to understanding, modelling and changing soft systems that underpins SSM. We introduced storytelling, roleplaying, rich pictures and causal loop diagrams as techniques that enable us to navigate through the seven steps of the SSM. We illustrated the use of the seven steps and causal loop diagrams to model, analyse and improve a problematic system in a national utilities provider. Throughout the example we also highlighted the importance of finding the one small change we can make to the system that would change the behaviour of the whole system, i.e., the trim-tab.

In the next chapter we will build upon the causal loop diagrams introduced in this chapter. We will focus specifically on the use of causal loop diagrams to enable group decision making in complex systems where decision making is often a messy or wicked problem as there is always someone the decision does not suit.

REFLECTIVE EXERCISE

Think about a system you are interested in and try using all frameworks and models we discussed in this chapter to gain a better understanding of the system by answering the following questions:

What is it about the system that would characterize it as a soft system? Are there people involved with different agendas and worldviews? Are the parts of the system autonomous decision makers?

Reflect on the earlier exercises from the end of section 6.2, where you used the CATWOE model to develop a root definition for your system. Did other people in the system agree with your root definition? Were there similarities or significant differences? Could you explain these differences?

Reflecting on pump manufacturing and education examples provided in Step 4 of Soft Systems Methodology, does your system have clear subsystems, potentially with different root definitions? If so, what are they? Do other people see it the same way?

Try drawing a rich picture of your system. This does not need to be anything sophisticated; you can use just pen and paper if you like. Some people find it useful to sit in front of a screen using a drawing tool, as the act of drawing something makes them think about the system and the picture emerges after several iterations. You may wish to try this approach as it makes it easy to move things around, erase, redraw, cut and paste, etc. Once you have finished drawing the rich picture, try narrating your story, i.e., explain the story behind the picture. You will probably find that as you are narrating the story you will want to go back and make some changes to the picture. Once finished, sit back and reflect on the picture and the story. Is this what you intended when you started or is it different from what you imagined at the start?

5 Try converting your rich picture into a causal loop diagram. How easy or difficult did you find to do this? Are there any parts you were not sure about? Did it make you think that you need to go back and find out more information about the system?

TEAM EXERCISE

Rich pictures – Take a system that all participants may be familiar with and ask them to draw a rich picture of the system. If the participants are from the same organization, it is usually easier to ask them to draw a rich picture of their organization. Try not to qualify what the picture should do, such as explain how.

"the organization works or highlight the issues and problems in the organization. Let them work that out for themselves. Usually, giving an example of the sort of thing you are looking for helps. We usually use the example from Figure 6.2. Also, this exercise works best if each of the participants draws their pictures on a flipchart using markers. They will need about 15 minutes to complete this exercise.

Next, ask each participant to show their picture to others and talk through the picture (i.e., the narrative). During this process, note the similarities and differences in the pictures and stories. Are the different perspectives and worldviews of different people coming through? Also watch for participants’ reactions to each other’s models – often there are surprises and a-ha moments that you can observe from their faces and body language.

Next, facilitate a discussion about the differences and similarities and ask each participant what they have learned from the exercise. Often, your observations of participants’ faces and body language (from the previous stage) are useful material to facilitate the discussion."

Causal loop diagrams – The above exercise can be continued by asking the group to consolidate their rich pictures into a single causal loop diagram. This exercise takes a bit longer and depending on the complexity and number of people involved you will need to give at least one hour to enable people to complete it. Also, doing this exercise on a large whiteboard with several coloured pens works best. This allows the model to emerge through several iterations (rubbing off bits and redrawing). Using different colours to represent different parts of the system, e.g., different people's views from the rich pictures, will also aid further analysis.

To finish the exercise, ask the group what they think of the final model that emerged from this. Does it reflect reality from their perspective? Does it help to explain some of the behaviours? And what have they learned?

Further reading

Kowalski, K (nd) Call me Trim Tab – Buckminster Fuller and the impact of an individual on society, Sloww, www.sloww.co/trim-tab-buckminster-fuller/ (archived at https://perma.cc/W39V-Q9SM)

Bell, S and Morse, S (2013) How people use rich pictures to help them think and act, Systemic Practice and Action Research, 26, pp. 331–48