Cybernetics travelled fast because its core was mathematical. A feedback loop works the same way in Boston, Moscow, Santiago and Beijing. What changed from country to country was the question people asked it to answer, and as we shall see, the questions tell us a great deal about the people asking.
What they shared
All four traditions agreed on the basics:
- Systems can be described by information flows and feedback loops.
- Purposeful behaviour can be explained and engineered.
- Computers would make large-scale regulation possible.
- The same principles apply to machines, organisms, and organisations.
They also shared a belief, strong in the 1950s and 1960s, that science could make social decisions more rational.
Where they differed
| United States | USSR | Chile (1971–73) | China | |
|---|---|---|---|---|
| Main question | How do brains, machines, and markets regulate themselves? | How can computers perfect central planning? | How can information support worker self-management? | How can engineering guide the nation? |
| Patron | Military, foundations, universities | Academy of Sciences, Party | Socialist government | Defence establishment, state |
| View of control | Distributed, self-regulating | Hierarchical, optimising | Recursive autonomy | Top-down systems engineering |
| Key figures | Wiener, McCulloch, von Foerster | Kitov, Berg, Glushkov | Beer, Flores | Qian Xuesen, Song Jian |
| Signature project | Macy Conferences, BCL | OGAS (unbuilt) | Cybersyn | Missile programme, population policy |
| Fate | Absorbed into AI and computing | Bureaucratised, faded with the USSR | Ended by a coup | Integrated into governance |
The politics of feedback
The deepest divide was about where regulation should happen.
American cybernetics, especially in its later counterculture form, leaned toward decentralisation: systems regulate themselves, so leave them alone. That fitted neatly with market liberalism, and some of its vocabulary later fed into free-market arguments about “spontaneous order” (Friedrich Hayek drew on cybernetic ideas in his later work).
Soviet cybernetics leaned toward centralisation: with enough data and computing power, planners could steer the whole economy. Reformers sometimes used cybernetics to argue for more autonomy at enterprise level, but the official framing stayed hierarchical.
Stafford Beer’s Chilean design tried a third way. His Viable System Model is recursive: every unit is itself a viable system with its own autonomy, and the centre intervenes only when lower levels cannot cope. Beer described this as maximum autonomy consistent with the cohesion of the whole.
Chinese cybernetics, rooted in Qian Xuesen’s engineering, treated society as a large system to be designed and optimised by experts. It has had the most direct and lasting influence on state policy.
The irony
Benjamin Peters’ observation about networks applies more widely. The capitalist US built its computer network (ARPANET) through state funding and cooperation among researchers. The socialist USSR failed to build one because ministries competed to protect their own information. Each system ended up behaving the way its official ideology said the other would. It is hard not to smile at that.
What the comparison teaches
The same cybernetic tools can support very different political projects. Feedback can serve self-management or surveillance, markets or plans, autonomy or control. The choice always rests with the people who set the goals, draw the boundaries and decide who gets to see the data. This is one of the strongest arguments second-order cybernetics makes for including the observer, and their purposes, in any description of a system.
Further reading
- Eden Medina, Cybernetic Revolutionaries (2011)
- Slava Gerovitch, From Newspeak to Cyberspeak (2002)
- Benjamin Peters, How Not to Network a Nation (2016)
- Ronald Kline, The Cybernetics Moment (2015)