Soft Systems Analysis

Soft systems analysis (SSA) is a method for investigating problems and planning changes in complex systems. It can also be used to design new systems. It was developed initially by Peter Checkland from Lancaster University (Checkland, 1981). The method can best be seen as a practical working tool, although it has been widely used in applied research. A core issue surrounds the meaning of “soft” as opposed to “hard” systems. The central idea in soft systems thinking is that people see and interpret the world differently. Discrepancies in the views held by individuals are not sources of invalidity or “noise” in the data; rather, differentiation reflects the nature of reality. Pluralism is the norm. In complex systems, individuals or groups are likely to construct quite different views of how the system works, what may be wrong with it, and how it should be improved.

Although recognizing that SSA can be used in a variety of ways, it is easiest to describe in its most straightforward format (Naughton, 1984). The method is organized in a series of relatively formal and well structured stages through which its users work. In practice, considerable iteration can take place between the stages. The analyst initially gathers data about the problem situation (Stage 1), and this is represented in pictorial form (a “rich picture”) (Stage 2). The users of the method (i.e., the analyst and the system participants) then explicitly try looking at the system in a number of different ways, searching for views that add some new light (a “relevant system”) (Stage 3). They [Page 1047]select a new perspective on the problem situation and develop a conceptual model of what the system would logically have to do to meet the requirements of this new view (Stage 4). This model is then compared with the existing problem situation to see if there are any lessons for change (Stage 5). These are then discussed by the participants, who decide which should be implemented (Stage 6). Changes that are agreed to are then implemented (Stage 7). If the new view (from Stage 3) does not appear to offer help to the participants, another perspective is tried (i.e., the process returns to Stage 3).

An example helps illustrate key parts of the method. In this case, Symon and Clegg (1991) studied the implementation of a Computer Aided Design Computer Aided Manufacturing (CADCAM) system in a large aerospace company in the United Kingdom. The study was undertaken between 1988 and 1990, with a follow-up in 1996. After a period of intensive data gathering (Stages 1 and 2), the researchers argued that the company was utilizing an inappropriate (and very limited) view (or relevant system) of the implementation process, and offered a different one (Stage 3). The old relevant system saw the process as a straightforward technology implementation project, whereas the new one emphasized organizational change. A new conceptual model was generated, one required to meet the needs of this new view of the CADCAM implementation (Stage 4). The new model stressed the need to resource, plan, participate, educate, support, and evaluate. These needs were then matched against the actual implementation activities, and this helped identify some major gaps in practice. This led to the generation of a series of recommendations for action (Stage 5). After discussion (Stage 6), some of these were implemented by the company (Stage 7).

...