What is a Systems Approach?
In Bertalanffy's introduction to his 1968 General System Theory (glossary) (GST) book, he characterizes a systems approach as: “A certain objective is given; to find ways and means for its realization requires the system specialist (or team of specialists) to consider alternative solutions and to choose those promising optimization at maximum efficiency and minimum cost in a tremendously complex network of interactions”. (Bertalanffy 1968, 4)
He goes on to list as possible elements of a systems approach: “classical” systems theory (differential equations), computerization and simulation, compartment theory, set theory, graph theory, net theory, cybernetics, information theory, theory of automata, game theory, decision theory, queuing theory, and models in ordinary language.
This description is similar to what Warren Weaver identified as the methods used successfully by “mixed teams” during World War II (WWII) on “problems of organized complexity”. However, some conditions that had contributed to success during wartime did not hold after the war, such as a clear focus on well-defined common goals that motivated participants to work across disciplinary boundaries.
By the early 1970’s, there was growing disillusionment with the promise that a systems approach would provide easy solutions for all complex problems. There was particular criticism from some, including pioneers of Operations Research and Management Science (ORMS) like Ackoff and Churchman, that reliance on rote mathematical methods to identify optimal solutions among fixed alternatives had become just as inflexible and unimaginative an approach to complex problems as whatever it had replaced. Interest grew in examining and comparing methods and methodologies to better understand what could help ensure the best thinking and learning in terms of systems in systems approaches to practice.
Issues in Systems Approaches
A systems approach is strongly associated with systems thinking and how it helps to guides systems practice. In What is Systems Thinking? the key ideas of considering a system holistically, setting a boundary for a problem/ solution of interest, and considering the resulting system-of-interest from outside its boundary are identified (Senge, 2006), (Churchman 1979), Meadows (2010).
A systems approach can view a system as a “holon” – an entity that is itself a “whole system” that interacts with a mosaic of other holons in its wider environment (Hybertson, 2009), while also being made up of interacting parts.
We can use this model recursively – each part of the system may be a system in its own right, and can itself be viewed both as an entity as seen from outside, and as a set of interacting parts. This model also applies in upwards recursion, so the original “system-of-interest” is an interacting part of one or more wider systems.
This means that an important skill in a systems approach is to identify the “natural holons” in the problem situation and solution systems and to make the partitioning of responsibilities match the “natural holons”, so as to minimize the coupling between parallel activities when applying a solution. This is the “cohesive/loose coupling” heuristic that has been around for a long time in many design disciplines.
Another consequence of the holistic nature of a systems approach is that it considers not only a problem situation and a solution system but also the system created and deployed to apply one to the other. A systems approach must consider both the boundary of the system of concern as well as the boundary of the system inquiry (or model). Real systems are always open, i.e., they interact with their environment or supersystem(s). On the other hand, real models are always “closed” due to resource constraints — a fixed boundary of consideration must be set. So there is an ongoing negotiation to relate the two in systems practice and the judgment to do so is greatly helped by an appreciation of the difference between them.
Thus, a systems approach can be characterized by how it considers problems, solutions and the problem resolution process itself:
• Consider problems holistically, setting problem boundaries though understanding of natural system relationships and trying to avoid unwanted consequences.
• Create solutions based on sound system principles, in particular creating system structures which reduce organized complexity and unwanted emergent properties.
• Use understanding, judgment and models in both problem understanding and solution creation, while understanding the limitations of such views and models.
Systems Methodologies
One topic that has received significant attention in the systems science community is the analysis and comparison of methodologies which implement a systems approach. A methodology is a body of tools, procedures, and methods applied to a problem situation, ideally derived from a theoretical framework. These describe structured approaches to problem understanding and/or resolution making use of some of the concepts of systems thinking. These methodologies are generally associated with a particular system paradigm or way of thinking, which has a strong influence on the three aspects of a systems approach described above.
The most widely used groups of methodologies are as follows, see also History of Systems Science:
• Hard System (glossary) methodologies (Checkland 1978) set out to select an efficient means to achieve a predefined and agreed end.
• Soft System (glossary) methodologies (Checkland 1999) are interactive and participatory approaches to assist groups of diverse participants to alleviate a complex, problematic situation of common interest.
• Critical Systems Thinking (glossary) methodologies (Jackson 1985) attempts to provide a framework in which appropriate hard and soft methods can be applied as appropriate to the situation under investigation.