Third-Order Cybernetics | Neo-Cybernetics

First-order cybernetics studies observed systems. Second-order cybernetics studies observing systems. A curious student will naturally ask: is there a third order? Several people think so, though they cannot agree on what it is. Let us walk through the main proposals and see what each has to offer.

Why the question arises

Second-order cybernetics focused on an observer: a scientist, a therapist, a designer, a nervous system. But in societies, observers observe each other. Their theories become part of what other observers are observing. Institutions, markets, and now algorithms act as observers too. A cybernetics with a single observer began to feel too small for the world it was trying to describe.

Four proposals

Umpleby, a long-time figure at George Washington University and in the American Society for Cybernetics, described three stages: engineering cybernetics (control of machines), the biology of cognition (the observer), and social cybernetics, where the theories that people hold change the society they are theories about. Economics is the standard example: when traders adopt a model, the market changes, which can invalidate the model. Umpleby saw George Soros’s theory of reflexivity as a practical case.

2. Self-developing reflexive-active environments (Vladimir Lepskiy)

Lepskiy, at the Institute of Philosophy of the Russian Academy of Sciences, frames the history as a move through three kinds of scientific rationality: classical (an external observer of an observed system), non-classical (an observer built into the system), and post-non-classical (a distributed observer in a “self-developing reflexive-active environment”). His focus is how many subjects, each reflecting on the others, can steer a society toward development, with an emphasis on social responsibility and civil society.

3. Hayles’ third wave (virtuality)

The literary scholar N. Katherine Hayles described three waves of cybernetics in How We Became Posthuman (1999): homeostasis (1945–1960), reflexivity (1960–1980), and virtuality (1980 onward), where information is treated as separable from its material body and artificial life, simulation, and virtual worlds take centre stage. Her version is a cultural history more than a research programme, but it is widely cited.

4. Critical or ethical cybernetics

Within the American Society for Cybernetics, some describe a third order in which the system includes accounts of itself and asks who gets to define the problem and who benefits. Klaus Krippendorff’s “critical cybernetics” belongs here: it examines how cybernetic tools such as classification and feedback can become instruments of oppression when people are treated as input-output devices.

Comparison

Proposal	Observer	Main concern
Umpleby	Society theorising itself	Theories changing what they describe
Lepskiy	Distributed, many subjects	Self-development, social responsibility
Hayles	Information without a body	Virtuality, posthumanism
Critical	Those with power to define	Ethics, inclusion, who benefits

All four move from one observer to many interacting observers. All four put ethics and power closer to the centre. And all four were developed in a period when technical systems began to act as observers at scale: credit scores, search ranking, social feeds, and now machine learning models trained on human responses that they themselves shaped.

A note of caution

Some cyberneticians reject the idea of a third order. Their argument is that second-order cybernetics already includes the observer, so “observing the observers of observers” adds nothing new in principle. Others worry that numbering orders invites an endless climb. Whether “third order” settles into a stable term or remains a label for several different projects is still an open question, and perhaps that is no bad thing. Open questions are where the interesting work happens.