Systems Thinking: an introductory essay

1 Systems thinking : an introductory essay D. Wastell, March 2012. Design rules! Design must be at the heart of every business Design thinking and Systems thinking are one and the same. In great design, form and function come together seamlessly. Every part contributes to the whole in a way that seems inevitable. So too in a great system .(Peters 2005). Systems thinking is very much in vogue, in the public services especially (Wastell, 2011). The term embraces a plethora of specific meanings, methods and affiliated sects. Flood &. Jackson (1991) provide a handy overview of 8 distinctive approaches, including Soft Systems Methodology and Systems Dynamics, as well as their own Total Systems Intervention'.

2 In this essay , I provide an overview of two well-known approaches, Soft Systems Methodology (SSM) and the structural modelling technique of Peter Senge which focuses on the dynamic behaviour of Systems . Soft Systems Methodology SSM was developed by Peter Checkland of Lancaster University; it is known well in the UK. but also has a strong international following. It is described copiously in a series of books written by Checkland and various co-authors ( Checkland, 1981; Checkland & Scholes, 1990) and its pedigree is long. Checkland describes how SSM developed through 25 years of action research, aimed at dealing holistically with messy real-world situations using Systems thinking (Checkland & Holwell, 2005).

3 The empirical grounding of SSM is impressive. Checkland speaks of several hundred Systems studies carried out by Lancaster teams (ibid., ); the research, though based in the university, has always been conducted externally, working with managers of all kinds and at all levels try[ing] to cope with the complexity of life's rich pageant . Checkland makes an important distinction between hard' and soft' Systems . Hard Systems are typified by those dealt with by engineers, ranging across the spectrum of complexity from a domestic thermostat at one pole to a petrochemical plant at the other.

4 Such Systems can be rigorously defined and specified and can ultimately be realised as physical entities. Their design and optimisation is assisted by formal methods from mathematics and operations research; alternatives can be modelled and choices made on the basis of defined technical criteria. However, in many problematic situations faced by managers, such an engineering approach is not appropriate; indeed in such soft' situations it is the very inability to stipulate objectives, or even to define what the system really is, which caused the situation to be regarded as problematic in the first place (ibid.)

5 , ). Even situations which may seem to be simple, are not as simple as it may seem at first sight. Checkland gives the example of Concorde. Could the Concorde project simply be regarded as a system to create the first supersonic airliner? Or perhaps, as the name implies, it was really' a system to persuade the French that the British could be good European partners, nailing the myth of perfidious Albion. Or should it be seen as a system to develop UK aviation expertise? Of course, it was arguably all these things. What matters is that the project ( the system ) would be designed differently depending on the relative priorities.

6 If the second definition were the overwhelming priority, then there would be no ultimate need even to deliver a working product! Spurred by these considerations, Checkland developed a new approach, based on the fact that all real-world management problems have at least one thing in common: they contain people interested in taking purposeful action (ibid., ). Checkland coined the concept of the Human Activity system (HAS) to designate the web of activities linked together such that the whole set accomplishes some defined goal. Notations for producing diagrammatic representations of such soft Systems ' were developed.

7 These can be seen as blueprints for how the world might be', their role being to help in problem-setting, helping understand what the problem is and aiding in its solution or amelioration. Human activity Systems are not real, stresses Checkland, they are conceptual, existing only in people's heads, ways of making sense' of how things are and how they could be. SSM assumes a fluid social world, one which persists and changes, continuously socially created in never-ending social processes . Just as for Concorde, there will always be different interpretations and opinions as to what is really going on, or what should be done: as many different worlds as there are participants and spectators.

8 To accommodate such complexity, Checkland has a brilliantly simple solution: build a model for each relevant perspective! Checkland adopts the German word Weltanschauung (world-view) to refer to these different orientations. It is the immanent Weltanschauung which makes: a particular model meaningful, since the purposeful action which one observer perceives as freedom fighting will be perceived as terrorism by another observer with a different taken-for-granted image of the world (ibid., ). Checkland also adopts the term holon' from Arthur Koestler for abstract entities which are autonomous wholes and, like Russian dolls, potentially part of larger wholes.

9 An example is urgently needed, methinks! Figure 1a shows a simple HAS. The hypothetical context (visible in Figure 1b) is higher education and the problem being addressed is the need to produce students with qualifications suited to the needs of potential employers. The notation is simplicity itself; bubbles denote activities and links indicate logical dependency, if A is dependent on B, this means that without B, A could not be done. A set of activities is shown which certainly seem to be relevant to the issue at hand. But is this a system ? It might seem so, but it is not.

10 To see why, let me recapitulate the meaning of the ubiquitous term system ' in the technical sense in which Checkland uses it and in which it is used more broadly in the world of Systems thinking . Crudely, a system has the following characteristics: it is goal driven, purposefully transforming inputs to desired outputs, and feed-back loops are present (performance measures) to ensure that goals are achieved. Systems also operate in a defined environment, to which they are open'; they are encased by a permeable boundary and are capable of adaptation to changing circumstances, ensuring viability.