Cybernetics began as engineering, but it carried a philosophy from the first day. Each generation pushed that philosophy further.
First order: purpose without ghosts
Before the 1940s, science had a problem with purpose. Physics explained things by prior causes. To say a cat moves in order to catch a mouse sounded like the outdated idea that the future pulls on the present (teleology). Biology tended either to avoid purpose or to explain it with vague “vital forces.”
The 1943 paper by Rosenblueth, Wiener, and Bigelow offered a way out. A system with negative feedback compares its state with a goal and acts to reduce the difference. Purpose becomes a structure you can build into a machine. No ghost is required.
Three philosophical claims came with that move:
- Function over substance. What matters is the organisation of a system, not what it is made of. A neuron and a relay can be described in the same terms.
- Circular causality. Causes run in loops. Effects return to modify their causes.
- Information as a physical quantity. Wiener wrote that “information is information, not matter or energy.” Communication became something science could measure.
These ideas fed directly into the mechanistic view of mind that later shaped cognitive science and AI.
Second order: the observer comes in
By the late 1960s some cyberneticians noticed a gap. They had described systems as if from outside. But every description is made by someone, with a nervous system, a language, and interests. If cybernetics was a theory of observing and regulating systems, it had to apply to the observer too.
Heinz von Foerster put it in a sentence: “Objectivity is the delusion that observations could be made without an observer.” Humberto Maturana added: “Anything said is said by an observer.” From here came several positions:
- Constructivism (von Foerster, Ernst von Glasersfeld): knowledge is built by the knower through action and does not copy an independent reality.
- Autopoiesis (Maturana and Varela): living systems produce and maintain themselves, and cognition is a feature of that self-production.
- Conversation (Gordon Pask): understanding is formed between participants, in exchanges that can be modelled.
- Ethics: von Foerster’s imperative, “Act always so as to increase the number of choices,” treats responsibility as something the observer cannot hand off to “the facts.”
The philosophical sources were wide: Kant on the mind’s role in shaping experience, Wittgenstein on language, Piaget on how children construct knowledge, phenomenology, and for Varela, Buddhist practice.
Third order: observers among observers
There is no single agreed “third order.” The proposals share one move: from an individual observer to many observers who shape one another, in societies, institutions, and technical systems.
- Stuart Umpleby described a progression from engineering cybernetics, to the biology of cognition, to social cybernetics, where the theories people hold change the society they theorise about.
- Vladimir Lepskiy, in Russia, describes third-order cybernetics as the study of self-developing reflexive-active environments, with a distributed observer.
- Others in the American Society for Cybernetics speak of a critical cybernetics that asks who defines a problem and who benefits from the definition.
The common thread is responsibility at scale. When algorithms observe us and our responses retrain the algorithms, observation and control are tangled together in a way no single observer can step outside of.
What stayed constant
Across all three orders, cybernetic philosophy keeps a few commitments: process over substance, loops over chains, relationships over things, and a suspicion of any account that hides who is doing the accounting.
Further reading
- Heinz von Foerster, Understanding Understanding (2003)
- Humberto Maturana and Francisco Varela, The Tree of Knowledge (1987)
- N. Katherine Hayles, How We Became Posthuman (1999)