6.2 Experiencing complexity as mess or difficulty
In this section, I want to take the ideas of mess and difficulty and explore them in the context of complexity. I want to determine how these ideas are connected, how significant the connections are and what the differences illuminate. I shall draw on the ideas of three writers: Schön, whose central theme is practice (e.g. Schön, 1983; 1987); Ackoff, who explores the characteristics of mess; and Rosenhead, who shows how different approaches to practice may be contrasted in terms that illuminate the distinction between difficulty and mess.
When reflecting on his own professional experience of engaging with complex situations, Donald Schön, author of Educating the Reflective Practitioner (1987) had this to say:
In the swampy lowland, messy, confusing problems defy technical solution. The irony of this situation is that the problems of the high ground tend to be relatively unimportant to individuals or society at large, however great their technical interest may be, while in the swamp lie the problems of greatest human concern. The practitioner must choose. Shall he [sic] remain on the high ground where he can solve relatively unimportant problems according to prevailing standards of rigour, or shall he descend into the swamp of important problems? (p.28)
The metaphor of the swamp provides some useful images for this section, which is concerned with the problems and opportunities of the swamp. Schön argues that:
all professional practitioners experience a version of the dilemma of rigour and relevance and they respond to it in one of several ways. Some of them choose the swampy lowland, deliberately immersing themselves in confusing but critically important situations. When they are asked to describe their methods of inquiry they speak of experience, trial and error, intuition or muddling through. When teachers, social workers, or planners operate in this vein, they tend to be afflicted with a nagging sense of inferiority in relation to those who present themselves as models of technical rigor. When physicists or engineers do so, they tend to be troubled by the discrepancy between the technical rigor of the ‘hard’ zones of their practice and apparent sloppiness of the ‘soft’ ones. People tend to feel the dilemma of rigor or relevance with particular intensity when they reach the age of about 45. At this point they ask themselves: Am I going to continue to do the thing I was trained for, on which I base my claims to technical rigor and academic respectability? Or am I going to work on the problems – ill formed, vague, and messy – that I have discovered to be real around here? And depending on how people make this choice, their lives unfold differently (1995, p.28).
In my view the argument Schön presents is simple: there are many domains of human activity where professionals fail to take action in situations of uncertainty, complexity, uniqueness and conflict or where past actions have had unintended, sometimes surprising and catastrophic, consequences. It seems to be a common human experience, for example, that a well-meaning attempt to improve a complex and problematic situation has the effect of making the situation worse in quite unexpected ways. Such situations also arise when experience is at odds with intuition about how things should behave or should be. This class of experiences is described as counter-intuitive understanding. The idea is explored further in Box 2.
Box 2 Counter-intuitive understanding – an example
In many parts of the world since the end of the Second World War planners and individuals have begun to live with the common sense, and simple view (thanks to the so-called laws of supply and demand) that increasing the supply of something will lessen demand for it, i.e. they are inversely related. However, recent experience shows that increasing the supply – both in number and capacity – of roads creates its own demand, i.e. that increasing supply increases demand. This phenomenon, known as the Pigou-Knight-Downs Paradox, is one in which positive feedback operates, at least until basic resources are totally used up or skyrocketing costs block the positive feedback. In southeast England it is possible to speculate that the incidence of gridlock, pollution effects and social effects such as road rage, as well as increased fuel costs are beginning to block positive feedback. This provides some explanation for why political parties found it possible in the 1990s to curtail road-building programmes. On the other hand, the car lobby is powerful. They are major employers, contributing to economic growth, as currently measured, and it is in their interests to see the road-building programme continue as this also increases the demand for cars.
In the early 2000s English road building policies have, once more, been adopted as one solution to traffic congestion. Demand for new cars has also increased fuelled by a drop in their relative cost. One possible outcome is that pollution levels per car decrease, as newer cars are far more energy efficient, but that aggregate pollution may stay the same or increase due to the rising number of cars, increased number and duration of journeys – the latter exacerbated by increasing frequency of gridlock. Other forms of controlling demand such as the central London congestion charging have been introduced.
An intervention in a policy process designed to alleviate some critical need or process, may at first seem logical and intuitively correct, but may exacerbate the situation in the future. It is for this reason that representing systems of interest, particularly through some form of modelling which makes modes of thinking, particularly in terms of patterns of influence, or cause and effect, is at the core of most systems approaches for managing complexity.
Given that in many situations, unexpected and potentially disastrous events may occur, it makes sense to think about doing some things differently. Doing things differently requires changes in thinking and in the actions that result from thinking. Being prepared for, minimizing, or even avoiding unintended potentially disastrous consequences means engaging with complexity. The effect of not engaging with Schön's swamp is to run the risk of unintended consequences of unknown seriousness, even if the intervention seems the right thing to do. We risk doing the wrong things with greater and greater efficiency rather than establishing what is the right thing to be doing. Russell Ackoff (1995) claims that it is better to do the right thing imperfectly than to keep doing the wrong thing better and better.
The experiences that have led to claims that different ways of thinking and acting are required for managing complexity have been derived in many domains. Examples include:
The computer press is littered with examples of […] information technology fiascos or near disasters. An example is the computer-aided despatch system introduced into the London Ambulance Service in 1992. The £1.5 million system was brought into full use at 07:00 hours on 26 October and almost immediately began to ‘lose’ ambulances.[…] the system reverted to […] manual methods on 4 November when the system locked up altogether.
(Fortune and Peters, 1995, p. 33)
One of the striking things about public policy […] is that so many of the most pressing problems are ones that cut across departments, cut across disciplines; issues like social exclusion, the environment, the family. [so ] My fifth point is about thinking systemically. (Geoff Mulgan; ex Demos, Director PIU, UK Cabinet Office)
… one of the more remarkable aspects of British debate is how little analysis is made in […] systemic terms. (Will Hutton, journalist and former editor of The Observer, a London Sunday newspaper)
I felt that a concern for and systematic study of the social and environmental aspects of technology was essential. Certainly environmental problems were approachable only by means of systemic and interdisciplinary methods and I felt convinced that any Faculty of Technology that did not concern itself with such problems could not claim to be either modern or responsible, whether socially or academically. (Geoff Holister, founding dean, Faculty of Technology, The Open University)
(Holister, 1974, pp.149–152)
Education for sustainability is the continual refinement of the knowledge and skills that lead to informed citizenry that is committed to responsible individuals and collaborative actions that will result in an ecologically sound, economically prosperous, and equitable society for present and future generations. The principles underlying education for sustainability include, but are not limited to, strong core academics, understanding the relationships between disciplines, systems thinking, lifelong learning, hands-on experiential learning, community-based learning, technology, partnerships, family involvement, and personal responsibility. (President's Council on Sustainable Development, USA, under the Clinton administration)
(President's Council on Sustainable Development, 1996)
These quotations are used in relation to at least four different domains. These are situations associated with:
The use of information and communication technology to develop information systems;
Organizational arrangements and associated policies and programmes;
Approaches to learning in technology education and in education for sustainable development;
Practice – whether in conducting an analysis or being professional.
There are a number of responses available to Schön's invitation to descend into the swamp of messy, confusing problems. Russell Ackoff uses the term messes to refer to the swamp, and difficulties to refer to the high ground. You should already have encountered Ackoff s terms in your earlier study of Systems.
Try to describe three features a practitioner might use to distinguish a mess from a difficulty. Is any one of these distinguishing features more significant than the others?
The three main features a practitioner might use to distinguish a difficulty from a mess are:
Messes are made up from a network of problems and opportunities that will be described differently by different people engaged in the situation. By contrast a difficulty will be described much the same, even from a diversity of perspectives.
The improvement in a mess is not just the sum of the improvements in its component parts. The improvements in a difficulty are easier to identify and describe and it is easier to identify how they came about.
Because a mess is a set of external conditions that causes dissatisfaction, a judgement about whether or not it has been improved, and by how much, will depend upon the perspective of the observer. The improvement in a difficulty will be generally agreed upon by observers from any perspective.
To deal with messes requires a holistic or systems approach, therefore it makes little sense to distinguish one feature as more important than another. A core concept at the heart of the idea of mess is, however, that of emergence, meaning the whole is greater than the sum of its parts.
Activity 19 should take about 15 minutes.
Refresh your understanding of messes and difficulties.
Read Ackoff's points about messes and difficulties in Box 3 below. Relate these points, and your previous understanding of messes and difficulties to the child support case study. From your perspective on the case study, are there aspects that appear to be difficulties and others that appear to be messes?
Make notes on these.
Box 3 Some features of messes and difficulties
A problem or an opportunity is an ultimate element abstracted from a mess. Ultimate elements are necessarily abstractions that cannot be observed.
Problems, even as abstract mental constructs, do not exist in isolation, although it is possible to isolate them conceptually. The same is true of opportunities. A mess may comprise both problems and opportunities. What is a problem for one person may be an opportunity for another – thus a problem can be an opportunity from another perspective.
The improvement to a mess – whatever it may be – is not the simple sum of the solutions to the problems or opportunities that are or can be extracted from it. No mess can be solved by solving each of its component problems/opportunities independently of the others because no mess can be decomposed into independent components.
Simple situations do exist that can be improved by extracting one problem from them and solving it. These are called difficulties and they are seen as exceptions rather than the norm in terms of decisions that are needed in environmental, organizational and other information-related contexts.
The attempt to deal with a system of problems and opportunities as a system – synthetically, as a whole – is an essential skill of a systems practitioner.
Russell Ackoff first coined the term ‘mess’ in 1974. He did so in response to the insights of two eminent American philosophers, William James and John Dewey. These philosophers recognized that problems are taken up by, not given to, decision-makers and that problems are extracted from unstructured states of confusion. Ackoff (1974a,b) argued, in proposing his notion of mess that:
What decision-makers deal with, I maintain, are messes not problems. This is hardly illuminating, however, unless I make more explicit what I mean by a mess. A mess is a set of external conditions that produces dissatisfaction. It can be conceptualized as a system of problems in the same sense in which a physical body can be conceptualized as a system of atoms.
From this definition of mess, Ackoff recognized a number of features of messes and difficulties (Box 3) that, if one is aware of them, affect the way a practitioner engages with a ‘real-world’ situation (see Figure 3 again).
When you have refreshed your understanding of messes and difficulties and re-read Box 3 spend about 15 minutes on the next activity.
Explain some implications of treating a situation as a difficulty.
You have been asked by the relevant government minister to prepare five quick-fix actions he can take to improve the child support situation. As a systems practitioner you are reluctant to take this approach. Write a few paragraphs briefing the minister about the possible implications of treating the situation as if it were a difficulty rather than a mess.
I find it interesting that Schön and Ackoff both have a professional background in planning. It is not surprising therefore that they have made similar distinctions when describing, or accounting for, their experiences in the messy business of planning. For me, they exemplify the aware practitioner juggling all the balls I described in Figure 4. What these planners have in common is they recognize that if the situation is engaged with as a difficulty there will be an outcome that will be different than if the situation is engaged with as a mess. They also agree that the traditional problem-solving methods, which are often associated with fields such as operations (or in the UK, operational) research (OR), or ‘scientific management,’ become useable only after the most important decisions have already been made. In other words, a difficulty is first abstracted from the mess and then the difficulty is treated using a traditional problem-solving approach.
I have summarized some of the characteristics associated with traditional OR in Table 3. Characteristics of an alternative, ideal, approach to OR, envisaged by Rosenhead (1989, pp. 1–20) in the early 1980s, are included in the table. Surveys had shown a low level of satisfaction on the part of managers with OR and management science projects at the time Rosenhead suggested his alternatives.
Table 3 Characteristics of doing traditional operations research in comparison to alternatives that were suggested in the early 1980s
|Characteristics of traditional OR||Alternative characteristics for OR|
|1 Problems and opportunities are formulated in terms of a single objective that can be optimized. Trade-offs are made by reducing variables to a common scale||1 Does not seek to optimize. Done by seeking alternative solutions that are acceptable on different dimensions without trade-offs|
|2 Has overwhelming data demands, which leads to problems of distortion, data availability and data credibility||2 Has reduced data demands because of integrating qualitative and quantitative data with social judgements|
|3 Subjected to demands of science (scientization), assumed to be depoliticized and that consensus exists||3 Strives for transparency and simplicity so as to clarify terms of conflict|
|4 People are treated as passive objects||4 People are regarded as active subjects|
|5 Assumes a single decision maker with abstract objectives from which concrete actions can be deduced for implementation through a hierarchical chain of command||5 Facilitates planning from the bottom up|
|6 Attempts to abolish future uncertainty and pre-take future decisions||6 Accepts uncertainty, and aims to keep options open for later resolution|
One way of interpreting Table 3, is that Rosenhead regarded the traditional OR approach of staying on the high ground, of treating the ‘real world’ situations with which many practitioners engage, as made up of difficulties to be solved rather than messes to be improved. I find many similarities with the ideas in Rosenhead's table with the following observation attributed to Richard Dawkins (Plsek, 2001):
If I hold a rock, but want it to change, to be over there, I can simply throw it. Knowing the weight of the rock, the speed at which it leaves my hand, and a few other variables, I can reliably predict both the path and the landing place of a rock. But what happens if I substitute a [live] bird? Knowing the weight of a bird and the speed of launch tells me nothing really about where the bird will land. No matter how much analysis I do in developing the launch plan … the bird will follow the path it chooses and land where it wants.
Which of these metaphors (the rock or the bird) do you think best describes the process of launching change in the Child Support Agency case study?
There are differences as well as similarities in the explanations the two planners (Schön and Ackoff) provide when they reflect on their experiences. Schön in particular, chose to focus on the characteristics of the practitioner. I referred to some of these characteristics in the section above on being a systems practitioner. Schön's ideas, among others, have already informed the approach taken in Parts 1 and 2 of this course.
Plsek, a change consultant based in the USA, used the ‘rock-bird’ story in an address to a UK National Health Service (NHS) Conference entitled: ‘Why Won't the NHS Do As It Is Told?’ The UK NHS is the world's third biggest employer after the Chinese Red Army and Indian Railways. Understandably many people involved in the NHS experience it as complex. In his presentation Plsek evokes different metaphors as means for the audience to make new distinctions. He contrasts the machine metaphor (as characterized by traditional OR in Table 3 and scientific management) with an alternative metaphor of complex adaptive systems (CAS) as exemplified by the bird in the rock-bird story (CAS is explained in Section 6).
Ackoff, in his definition of a mess as a system of problems and opportunities chose systems thinking as his strategy to make sense of the mess of the swamp. His strategy was to look for system within a mess as a means to do something about it. Please note I am not referring here to ‘discovering’ the system or a system but the process of distinguishing one or many systems of interest in a context. The end product of the process of finding system within a mess is called formulating a system of interest.
Let me consider now what I think Ackoff was doing in terms of a practitioner juggling the E ball. In my terms, Ackoff was a systems practitioner (Ps) engaging with a ‘real-world’ situation that he could choose to recognize as either a mess or a difficulty using a systems approach (As). This leads me to ask a fundamental question: are the characteristics of a mess part of the situation or a function of the choice the practitioner makes? I will answer this question by grounding it in my responses to the case study.
If I use everyday speech to describe my initial experience of the child-support issue I say ‘it is a mess’, or ‘it is really complex’, or ‘I find it hard to understand it all’. You will notice I have used the word ‘it’ each time, which suggests the existence of something, an entity, a ‘real-world’ situation with which I have engaged. The structures of the language I use tie me into a linguistic trap – the naming of an ‘it’ that is independent of my act of distinction. Getting out of this trap means finding a language that avoids the implication there is a pre-existing ‘it’ waiting to be noticed by me. As someone once said every noun obscures a verb!
The same could be said of thinking that there is a NHS which is a machine or a complex adaptive system. I can get out of this trap by claiming a mess or a difficulty arises in the distinctions that a practitioner makes in a particular situation. If this is the case, a mess or a difficulty is not a property of the situation but arises as a distinction made by a systems practitioner – someone aware of the conceptual distinctions between seeing a mess and experiencing a difficulty – engaging with a particular ‘real-world’ situation (see Figure 16).
If, on the other hand, my experience had led me to say, ‘Oh, I know what the problem is – we just have to do X and that will fix things’, then I would be implicitly seeing the issue as comprising a difficulty.
Which of the following statements conform to the idea of formulating a system of interest?
I am fascinated by the solar system.
When I engaged with the issues surrounding child support, I thought it might be helpful to consider it as a system from a number of perspectives. For example:
As a system to reduce the social security budget;
As a system to secure the best future for children in lone-parent families;
As a system to ensure the non-resident parent contributes equitably to the raising of their children.
I am interested in making computer systems function more effectively.
Statement (2) best conforms to the idea of formulating a system of interest because it considers systems not as things out there in the ‘real world’ – as in answers (1) and (3) but as a useful way of thinking about – engaging with – complexity from different perspectives. Statements (1) and (3) have little to say about context therefore it is more difficult to consider how these might be reformulated as systems of interest.
Consider your own practices.
Consider your own practices in some recent situation(s) in the light of Figure 16 and the question I posed above:
Are the characteristics of a mess [or a difficulty] part of the situation or a function of the choice the practitioner makes?
You might benefit from writing your answer in your notebook and returning to it as your own systems practice develops.