Wicked assumptions and how to approach them
Part II of Challenge to Reason (CR: Churchman, 1968 ] is entitled The maximum loop. Deduction and possibly induction are examples of minimum loop inference. Deduction is the domain of logic, induction that of science. The American pragmatist Charles Saunders Peirce developed a third mode of inference, which he called abduction. It is arguable that, as Beer suggests (in Barton, 1999), management is effectively concerned with abductive processes. Churchman has been strongly influenced by American pragmatism along the line of Peirce (1839 – 1914)-James (1842-1910)-Singer (1873-1954)-Churchman (1913-2004). The systems approach could be considered a practical application of abduction.
Our narrow muddle At the end of chapter 13: Large models of systems, a reprint of an earlier presentation at the Second Stony Brook Conference on Advances in Computing, Stony Brook, New York, June, 1966, Churchman refers to John Dewey, another American pragmatist, “who showed that the quest for certainty is bound to fail, because certainty is an unattainable goal” (CR 167). Churchman also quotes Singer as having said that “When we reach a conclusion after having exposed our ideas to the most severe test we can imagine, then we have done the best that inquiry can possibly accomplish.” Churchman noticed that there was a snag: “How do we know we have exposed our ideas to the severest test? If we have bound ourselves by our thinking into one corner of reality, then we shall never expose ourselves to the really severe test. Instead, we shall wander aimlessly about in our own narrow muddle, thinking we are progressing, but getting nowhere at all.”
Large models of systems … is the title of the chapter summarized here (CR 155-167). During the 1940s and 1950s, Churchman was one of the founding fathers of operations research, also known as management science or decision science. One of the most common models used is the allocation model, which deals with the problem of distributing scarce resources among alternative activities. The idea is to maximize value according to measures of performance and subject to certain constraints. By the mid-1960s advances in computer technology made it possible to handle millions of variables and tens of thousands of constraints equations. It is no surprise that the well-known report “The Limits to growth” (Meadows et al. 1972) was initiated by the Club of Rome in the late 1960s. In fact, an earlier design or ‘prospectus’ of the report (“The predicament of mankind”) was based on Churchman’s systems approach (Christakis, 2014).
Flawed assumptions An obvious problem with large models is that only very few humans are able to program the computers that perform all the applied mathematics to handle the massive numbers of objective functions and constraint equations. This leaves most people out of the loop, which they will experience as threatening. The main flaw, however, is that large models assume that all the necessary correct information is available for specifying the functions and equations. This may or may not include the way people feel when they are left out of the loop. This situation is quite obviously the case in the European Union. The fundamental challenge to reason is that while trying our utmost best to be rational, we necessarily fail to do so. But there is a way out.
Sweeping in viewpoints Churchman’s chief interest is in human activity systems, also referred to as purposive systems or teleological systems. Examples include government activities, business operations, but also individual lives. Humans and their systems restlessly identify and pursue goals and try to use reason to do so more effectively. In the previous three posts we have explored the idea that we can decrease uncertainty by applying reason as an expansive concept, which is by sweeping in contrasting viewpoints. These viewpoints can portray reality, represent human experience, which in turn checks whether complete rationality has been attained by observing and experiencing that same experience. All this applies to the extent that it is relevant, provided we agree that relevance is a debatable, and therefore expanding, concept, too.
Churchmannian inquiry The point that Churchman emphasizes persistently is the problem that certain simple questions about subsystems cannot be answered without first having a full understanding of the whole system. He found that out while working on large models of systems from the 1940s to the 1960s. The wrong solution is that we easily bound ourselves into one corner of reality, where we can make all the wrong assumptions without being aware that they are not correct. In his books he gives many examples of a wide range of wrong assumptions. The right solution is to become aware of the deceptive mechanisms involved and the perceptive options available to us. These are: (1) critical awareness the limitations of inductive and deductive reasoning, no matter how seductive and convincing they may seem; (2) sweeping in contrasting viewpoints (examples in Churchman’s work include: (a) physicist vs. biologist or humanist vs. anthropologist; (b) management scientist vs. humanist vs. efficiency buff vs. anti-planner; and (c) the realist vs. idealist, see next post); and (3) understanding the key roles in decision processes in all their guises, including those of the client, the decision-maker and the planner (this resembles point 2.), in relation to the other 8 categories of the systems approach framework (see below and elsewhere in this blog).
Abduction and the maximum loop Peirce defined abduction as the cognitive process by which we are capable of isolating a relatively small number of plausible hypotheses to account for observable facts. It follows the same three steps as deduction and induction, but with a twist: (1) Statement B is true (result); (2) On the basis of my experience, my best guess is that A causes B (hypothesis/rule); (3) Therefore, A must be true. Now, clearly a guess does not seem a very good basis for certainty, but Peirce argued that abduction was the only form of inference that extends knowledge (as for innovation), while deduction simply develops logical results from hypotheses, and induction uses data to quantify arguments. One could say that deduction and induction follow a minimum loop, whereas abduction follows a much longer loop, of intermediate length, by ‘sweeping in’ experience. Churchman’s key contribution is to extend this model by positing that we must sweep in all of human experience: the maximum loop.
The categorical framework … of Churchman can be described in many ways. One way is by following Kant´s 3 x 4 tabular approach, which is in a way what Churchman is doing as well. There are four aspects for examining and judging a human activity or plan: value, activity, design, and approach. For each of this aspects there is a role, i.c. those of the client, decision-maker, planner and philosopher, respectively. The client enjoys the value, the decision-maker decides on the activity based on whether the plan can be justified, the planner designs the plan to maximize the value, and the (systems) philosopher decides on the (systems) approach for the plan to make sure it all works for the best. There are difficulties with all these roles in relation to each other. These difficulties are an important part of how and why we are so easily deceived when making our plans.
The systems approach of deception One of the problems is that of implementation. It is a well-known fact that many ‘wonderful’ plans end up not being implemented. This probably means that the planner has insufficiently taken into account how decision-makers make up their mind. One obvious aspect is that of the politics of a situation. Planners have their highly rational ideas about the feasibility of a plan in terms of effectiveness, but this may be quite different from feasibility in terms of acceptability, e.g. to the public. Would it then be reasonable to incorporate the politics into the plan or should more attention be devoted to how a plan can be ‘sold’. The trouble with these ‘solutions’ is that they work in tension with each other. The best plan is where an optimum concordance is struck between the twelve categories of the framework. In the dialectical process of doing so relevant viewpoints must be swept in and abduction must contribute innovative ideas. That, in short, is the systems approach as illustrated from the angle of large system models.
Barton, J. (1999). Pragmatism, systems thinking and system dynamics. In System Dynamics Conference (p. 17). Retrieved from Google Scholar.
Christakis, A. N. (2014). An Epic Learning Journey: From the Club of Rome to Dialogic Design Science and DEMOSOPHIA. In Social Systems and Design (pp. 37–70). Tokyo, Japan: Springer. DOC (June 7, 2017).
Churchman, C. W. (1968). Challenge to reason. McGraw-Hill New York. PDF.
Griffin, D. R., Cobb Jr, J. B., Ford, M. P., Gunter, P. A. Y., & Ochs, P. (1993). Founders of Constructive Postmodern Philosophy: Peirce, James, Bergson, Whitehead, and Hartshorne. Albany, NY: SUNY Press. Retrieved from Google Scholar. Partial preview.