In his paper, “The Human Use of Human Beings”, Weiner asserts that the only true way of understanding a society (or system) is to study the messages and communication facilities that belong to it. Booker interprets this claim in his piece, “How to be Universal” by arguing cybernetics as being universal, and proposes a new economy of sciences in order to leverage what he calls a “universal language”. In order to understand the pitfalls of this method of distilling a system down the the information exchanged within it, we must first analyse the claim to universality.
According to Bowker, at the heart of cybernetic theory lies abstraction. By quoting Bateson and Lettvin et al, he claims abstraction is separate from the conscious mind, and is a language that is universal, and already organized and interpreted. Thus he elevates abstraction from mere being an ability of the evolved mind, to a physical fact of nature.
The other important concept at the core of cybernetics is entropy. Entropy is defined as the “measure of disorder in a system”. This disorder gets in the way of effective communication by creating inconsistencies and incoherence in the exchange of information. Booker argues that cybernetics is negentropic, that is, it resists this natural disorder.
As a result, he positions cybernetics — and by extension abstraction — as a method of communication which leads to the least loss of information, and therefore universal.
This is where I think the flaw lies in the argument. Can something be truly negentropic? The second law of thermodynamics states that the total entropy of any system never decreases, because systems will always tend towards, and rest at, a state of equilibrium, where entropy is maximum.
Does abstraction truly provide the crystal clear exchange of information that makes is “universal”? Bowker does not address the concept of noise in this model. Noise, as defined in Shannon and Weaver’s model of communication theory, which Fiske outlined in chapter one of his book “Introduction to Communiation Studies”, is an unintentional signal added in between the source and the receiver that distorts the message. As long as noise exists in the signal, is it possible to have a truly negentropic method of communication? Does abstraction truly provide the tools to communicate without any noise? To assume that abstraction and cybernetics provide a method that overrides a basic law of the universe is, to me, somewhat pompous.
Furthermore, cybernetics defines all behavior in the form of feedback loops. Both Bowker and Weiner argue that feedback loops make behaviors, in both humans and machines, purposive. To clarify this, Bowker gives the example of reaching your hand out to attempt to lift an object, and then course correcting until you’ve successfully performed the action. They define any behavior not in terms of what is, but in terms of what should be. The feedback-control method only focusses on what should be, and adjusts behavior in order to achieve that result. This sounds an awful lot like Alexander’s method of problem solving by defining the “bad fit”, but cyberneticians apply this model to human behavior. I do not feel that defining something in negatives gives the whole picture. By describing what something isn’t, you cannot always successfully and accurately describe what something is. This sounds like a gross oversimplification of the nature of human beings, and the wide range of behaviors they exhibit.
John Fiske, “Communication Theory,” in Introduction to Communication Studies (London: Routledge, 1990)
Norbert Wiener, “Cybernetics in History” and other excerpts, The Human Use of Human Beings (London: Free Association Books, 1950)
Geof Bowker, “How to Be Universal: Some Cybernetic Strategies, 1943–70,” Social Studies of Science 23(1993)