How to grasp an obscured yet fundamental truth?
Chapter 3 of Churchman’s “The design of inquiring systems” (1971) is entitled “On whole systems: the anatomy of goal seeking.” What this suggests is that the notions of ‘whole systems’ and ‘goal seeking’ are somehow related. It also suggests that the idea of ‘goal seeking’ can be disentangled by looking at its anatomy. In other words, the idea of ‘goal seeking’ needs to be dissected, or else we won’t understand what ‘whole systems’, and by extension ‘the systems approach’, mean or purport to achieve. Churchman himself admits that “the details are somewhat tedious”, which suggests that a non-genius like me won’t understand it, unless he really applies himself. With almost 40 pages the chapter is the longest in the book and I must admit: It was tough going. Hope you find my brief useful. It may help if you first read my previous post on the same chapter. You may also like to read Churchman’s 1962 working paper “On inquiring systems“, which corresponds by and large to Chapter 3 of the book published 9 years later.
Four main steps After having read the chapter “On whole systems”, I concluded that Churchman asks the reader to take four giant steps in order to really understand what he means by a system. None of these steps are likely to correspond to anything the reader has learned before, although he/she may be vaguely or intuitively aware of it. The reader must: (1) acknowledge the four main types of relationships in a human activity system; (2) conceptualize a factory as a simple, non-separable system, i.e. a system packed with interdependencies; (3) conceptualize how Churchman´s system categories can be used to learn about or (re-)design an actual system such as a factory; and finally (4) attempt to see the importance of a systems perspective of the world.
Explanatory models [step 1] The designer seeking to improve human activity systems has 4 key explanatory models to conceptualize alternative systems that are hopefully more effective than the one they seek to improve on: (i) cause-effect relationships (A leads to B); (ii) producer-product relationships (A may lead to B under certain conditions; (iii) functional relations (A influences system aspect B); and (iv) teleological entities, i.e. human activity systems, rank and choose functional entities to achieve a purpose.
Factory as a system [step 2] A factory is an example of a human activity system with a large number of interdependiencies, i.e. relationships that produce non-linear effects thus turning the factory into a non-separable system that is not overly complicated. The south-eastern quadrant of the concept map is sufficiently self-explanatory.
Non-separable systems [step 3] What turns a relatively simple non-separable system such as a factory into something highly complex is a combination of parts with multiple feedback loops and the presence of humans. In this case two humans (or roles of these) stand out: the designer and the decision-maker. In many cases it is best to conceive of these two as roles not persons or groups of persons, not in the last place because they affect one another. Another issue is that of handling ‘ the whole’, which is hard to define, even to the extent that some would argue that ‘the whole’ does not exist or cannot be proven to exist (as was demonstrated in a comparable case by Kant). If ‘the whole’ is a problematic concept, then it will be difficult to design well-integrated parts. Fateful options include: (i) ignore the whole (this sounds silly, but is actually quite common); (ii) make assumptions about the whole (not bad, especially if it involves modelling of system behaviour); (iii) assume temporary separability by adopting a strategy of incrementalism (also known as ‘muddling through’); (iv) use a system of going back-and-forth between the parts and a central decision-maker/designer until a seemingly satisfactory configuration of parts is achieved. Both, options (iii) and (iv) are typical of large organizations in business and government.
Monism [step 4] Churchman translates the main tenets of monism of past rationalist philosophers to a systems approach, all the while contrasting it with the pluralism of the empiricist. Scientific empiricism tends to dismiss or resist the notion of the existence or importance of whole systems and designs its inquiring system accordingly, i.e. by looking at the parts in segregation from the whole or other parts by imposing artificial separability. Monism, in contrast, uses non-separable inquiry, which is based on three monistic ideas: (i) whole performance encompasses part performance, i.e. the whole performance needs to be clarified in order to value and rank the relationships within and among part performance; (ii) the system as purposeful existence (because of the purposive individuals that operate in it) aimed at improving or optimizing a system or a problematic situation. Without this notion of optimization it is impossible to clarify the performance of the whole, i.e. the system effectiveness. Optimization is impossible without optimal inquiry, i.e. without an inquiring system meeting the nine conditions mentioned in my previous post (purpose, resources, implementation, client, decision-maker, designer, measure of performance, environment, and guarantor); (iii) reality is elusive and can only be approximated by means of optimization using multi-perspective inquiry to ensure the highest (possible) level of objectivity (which could be said to be inter-subjectivity in an effort to expand the Leibnizian fact nets as much as possible).
What have we learned The notion of teleological entities (i.e. purposive humans such as designers) is often ignored in inquiry for planning or science as if inquiry design is possible without considering inquiry as an integral part of system design or without a decision-maker deciding that inquiry or system optimization is in some way desirable. No matter how hard we try, we cannot design away the inquiring system and its design. The systems approach looks far beyond the physical interdependencies and feedback loops of classical system dynamics. Churchman writes “We never know what really exists, but at any time we do the best we can to construct an image of the world in which our observations, thinking, feeling and intuition will live as well together as possible.” Science, business and government are in themselves great enterprises, but fail to optimize this conviviality of our observations, thinking, feeling and intuition, exactly because we are endowed with all those different mental faculties. This applies to us individually (highly inquisitive or simply living our lives), as small or large groups, and internationally. The systems approach provides an inquiring system that can help us clarify why things don´t work the way they ought to and how we could improve value creation, both in the short and long run, and in all areas of work and life. It does so by helping us avoid common traps such as erroneous assumptions about the whole, the parts, and the environment among other things.
Nonseparability, the concept Let me ramble on a little bit. Nonseparability and holism are sometimes discussed together as in “Holism and Nonseparability in Physics” or “Heisenberg’s uncertainty principle” or Schrödinger’s cat, whose death and life turned out to be nonseparable in Schrödinger’s thought experiment. To monists (perhaps some systems thinkers consider themselves so) some of these principles or phenomena could be perceived as particular instances of systemic behaviour. I don’t think we need to go that far. Churchman’s nonseparability was probably just a clever way for circumventing all talk of holism, when there is no need for it. So, it is also a way of avoiding the conflation of systems thinking with holism, although ultimately, in some way, but perhaps only conceptually, the two may seem to converge. This last point is one of the great riddles to perplex us to all eternity. Schumacher makes a case for embracing this perplexity. I prefer Churchman´s stance, because talk of nonseparability may be more widely acceptable than talk of holism, thus increasing the likelihood that the world can benefit of the broad application of the systems approach. Moreover, nonseparability is fairly obvious, whereas holism (like God) is likely to remain subject to belief.
Nonseparability and system boundaries The concept of nonseparability is useful for understanding one other key systems aspect, that of the system boundary. It is fairly obvious to state that a system is defined by its boundary, simply because the boundary determines what is “in” and what is “out”. However, this ‘obviousness’ is very deceptive. In the case of the factory we may be inclined to think of the factory as a system, but problems start when we start considering the client (and his value ‘system’) as part of the system. Things become even more complicated when we look at the role of the designer, whose inquiring system is also part of the system, thus drawing in notions of the environment in which the factory operates. In the end we are left with a situation that it is better not to talk of factory as a system at all. Instead it is much better to debate (often carried out as ‘boundary critique’) the nonseparability (or relevance) of the inter-relationships in a (problematic) situation.
Apperception … is the last thing that remains to be discussed (see quadrant 4 above). There are quite a few definitions that do not differ a lot, whether they refer to Leibniz, Kant or William James, so let me just pick a general one that is readily understandable and seems good enough: “the process by which new experience is assimilated to and transformed by the residuum of past experience of an individual to form a new whole.” Now why do I want to bring this up? Just because it may clarify in a succinct way what the systems approach is all about, which is “forming a new whole,” but not just in any way that occurs to us. Apperception according to Churchman must respect as fully as possible the teleological character of this “forming a new whole.” As we can see in quadrant 4 of above concept map, this involves not only multiple perspectives, but also revaluing and reconfiguring existing inter-relationships. So, perhaps, we could simply say that the systems approach is nothing more than explicit teleological apperception. The difference between ‘normal’ apperception and the systems approach of apperception is that considerable thought has been given to how it can be done as well as possible.