An Introduction to ‘Systems’ By Jake Chapman Professor, Energy Systems, at the Open University 1978–2001, previously Lecturer and Senior Lecturer in Physics (1970–1978), now retired. This text aims to describe what is meant by systems thinking and systems practice in a way that is accessible to professionals who have not previously engaged with systems. The text does not therefore aim for theoretical rigour but rather aims to assist people to comprehend the core ideas. For simplicity the term ‘Systems’ is used to indicate the amalgamation of theory and practice that comprises the approach. One of the reasons why people find Systems difficult to comprehend is that as an intellectual discipline it is not defined by the subjects or issues to which its ideas may be applied. Subjects such as chemistry, economics and literature all have both a defined area of application as well as characteristic ways of thinking about and analysing that area. Systems is more like history or philosophy – it is an intellectual approach to issues that can range across the whole of human experience. This does not mean that Systems has some sort of universal application, any more than history or philosophy. Systems is useful for tackling issues that are embedded in complexity created by human activity. One way to understand Systems is by contrast with the reductionist approach to tackling complexity. Reductionist thinking has been remarkably successful, particularly in developing successful theories and models of the inanimate world when combined with scientific procedures. The essential characteristic of a reductionist approach is that the complexity is simplified by dividing the problem into sub-problems or components. The process of sub-division is continued until the resulting bits are simple enough to be analysed and understood. The operation of the original complex entity is then constructed from the operation of the components. But herein lies a potential problem. What if essential features of the complexity were not embedded in the components but in their interconnectedness? The very act of simplifying by sub-divisions loses the interconnections and therefore cannot tackle this aspect of complexity. Systems has an alternative strategy for simplifying complexity, namely going up a level of abstraction. Increasing levels of abstraction lose detail, and it is the loss of detail that provides the simplification1 . But the interconnection of the components is largely maintained in this process. For example a core systems idea is feedback, both positive (or self-reinforcing) and negative (or self-correcting), and one of the reasons why complexity can often appear mysterious is due to a rich set of feedback loops between the components2 . Because it retains connections and avoids breaking things down, Systems is a ‘holistic’ approach to understanding and managing complexity. 1 For example, when people talk about the behaviour of organizations they are eliminating the rich detail of how individuals or groups within that organization function. The organization is at a higher level of abstraction than the departments or individuals within it. 2 Human cognition appears to be limited in two ways. First, it is hard for people to comprehend the operation of more than about seven variables at once. Second, it is even harder for a person to understand a situation in which there are more than just a few feedback loops operating. Broadly, reductionist approaches reduce the number of variables by eliminating feedback and systems approaches reduce the number of variables and the feedback loops by eliminating detail. 2 of 3 It is now possible to say more about the domain of application of Systems. In general, Systems provides a useful approach to situations where reductionist approaches have failed. These are characteristically situations that involve human activity in which there is a significant level of ambiguity about what the problem is, what a solution would look like, and about the scale, in both time and people, of the problem or issue. Such situations are called ‘messes’. Systems sets out to provide a framework for improving messes, not solving or resolving them. There is another reason why Systems, as an intellectual approach, is hard to comprehend, namely that in everyday speech ‘system’ is used as a noun, as for example the solar system, the legal system and the eco-system. There is a link between the everyday use of the noun and the intellectual approach, but it is not a simple one. Within Systems, a system is usually defined in terms of “a purposeful assembly of components (or sub-systems) such that the behaviour of the components is influenced by being in the system”. The stuff around the system is referred to as its environment and the separation between the system and its environment is referred to as the system boundary. According to these definitions the solar system is not one of the entities that would be studied using Systems. The legal system is an entity that would be studied, though exactly what is regarded as within the legal system will depend upon the context and perspective of the study. This latter point causes difficulties for people previously trained in subjects where the definition of the entities of study is fixed and not dependant upon context or perspective. Within Systems theory it is formally acknowledged that the perception of a system is a subjective3 process and depends upon the purpose of the Systems practitioner. An example of this subjectivity of defining a system may make this clearer. Within the domain of energy policy the energy system is considered to comprise of the sources of supply of fuel (oil wells, coal mines, etc.), the fuel conversion plants (oilrefineries, power stations, etc.) and the ways in which fuel is distributed to end-users (gas pipelines, electricity grid, etc.) and all the associated organizations (fuel supply companies, regulators, etc.). However, if the focus of the study were ‘to investigate the impact of energy efficiency improvements’ then the system would be expanded to include the equipment used by the so-called end-users (fridges, boilers, cars, etc.). Whether the end-user equipment is inside or outside the energy system is determined by the purpose of the study, not any absolute definition. This emphasis on the perception and purpose of the analyst is a characteristic of systems thinking. For example, a powerful idea in Systems is that when someone finds an issue particularly difficult it is likely to be due to a ‘trap’ in their way of thinking about the issue. The ‘trap’ exists in the mental constructs that they are using to comprehend the situation – some of which may be the way that they are defining the system(s) involved. The solution to this category of difficulty is for the individual involved to make a change in their way of thinking about the issue – easily said, but extremely difficult for most people to accomplish. A range of Systems tools and methods are constructed in order to facilitate this shift in thinking. Once the shift has occurred then the individual may or may not find that, from the new perspective, the previously difficult issue is less intractable. Systems thinking values different perspectives precisely because it encourages insights and new approaches to complex issues. 3 Usually the system is agreed by negotiation between all those involved in a particular exercise. 3 of 3 Another feature of Systems tools and approaches is that they make use of a range of different types of diagrams. There are two reasons for this. First, diagrams are an economical way of representing interconnections between components and thus facilitate the process of abstraction used to simplify and represent complexity. Second, most professionals are more familiar with the use of language and mathematics, so using diagrams is strange and forces the adoption of a slightly different perspective. In short diagrams can assist in the process of changing the way an individual perceives and thinks about the issue. There are a number of formal Systems methods that have been developed to facilitate tackling different types of complex issues. One characteristic of the methods is that they provide a vehicle or framework within which different actors or stakeholders in a situation can communicate and share perspectives and understanding. This is a key component of tackling messes since part of what usually makes them intractable is that participants are focused on promoting their solutions rather than understanding the differences in perception of the problem(s). Another characteristic of many of these formal Systems methods is that at some stage the representation of the system under examination is compared to an ‘ideal’ system with the same or similar purpose. By identifying differences or missing components between the ideal and observed systems recommendations for improvement can be generated. Recommendations range from ways of improving control (for example by providing better feedback) to increasing autonomy by fostering self-governing sub-systems (as in learning organizations). Summary Systems provides a set of ideas, tools and methods for engaging with and improving complex situations, referred to as messes. It is a holistic approach that emphasizes the connections between issues and components in the mess and simplifies the complexity by thinking at a greater level of abstraction or generality. Systems fosters a multiple perspective approach to complexity and assumes that insights and ways of improving situations will be generated by facilitating stakeholders and participants to shift their established way of thinking about the mess. Systems is likely to be most useful in contexts where reductionist, single perspective or command and control processes have failed to provide adequate management of the complexity involved. Systems complements other methods of analysis, often by providing an overarching context within which more detailed analyses can be situated