House lore of the Technosophists · @technosophists
DangerousSecret Science Club
Is there a secret science that bridges all science? There was. It had a name, a book, a dining club in London, a decade of meetings in New York, and a control room in Santiago. Then it stopped being anyone's department.
Declassified finding: it was the funding
The ring is real. Bollen and colleagues (2009) built it from nearly a billion clicks between journals on scholarly portals, and Katy Börner put it in Atlas of Science (2010). The fields close into a loop around an empty middle. The "???" was added by the meme. This page fills it in.12What lives in the hole
One loop, every field
Before any theory, four machines. Each one measures something, compares it to what it wants, and pushes back on the difference. That push-back is the whole trick.
The Greek for steersman is kybernētēs. In 1948 Norbert Wiener borrowed it for a book about this loop in "the animal and the machine," and the field had a name: cybernetics.3 James Clerk Maxwell had already worked out the math of the steam governor in 1868.4 What was new was the claim that the same loop runs a nervous system, a factory, an economy and a gun director, so one science could study all of them.
The club, in order
Twenty-five loud years, then quiet
1943
Rosenblueth, Wiener and Bigelow publish "Behavior, Purpose and Teleology": purpose explained as negative feedback. McCulloch and Pitts publish the first model of a neuron as a logic gate.56
1946
First Macy conference in New York: "Feedback Mechanisms and Circular Causal Systems in Biological and Social Systems." Ten meetings follow through 1953.7
1948
Wiener's Cybernetics comes out. Ashby finishes the homeostat at Barnwood House. Grey Walter's robot tortoises Elmer and Elsie roam his lab in Bristol.389
1949
Time calls the homeostat "the closest thing to a synthetic brain so far designed by man." The Ratio Club starts dining in London, with Turing joining after the first meeting.810
1954
The Soviet Short Philosophical Dictionary defines cybernetics as "a reactionary pseudoscience." The next year Sobolev, Lyapunov and Kitov publish its rehabilitation.11
1956
Ashby's Introduction to Cybernetics states the law of requisite variety. At Dartmouth, John McCarthy launches a new field under a new name: "artificial intelligence."1213
1958
Heinz von Foerster opens the Biological Computer Laboratory in Illinois. The Ratio Club holds its last meetings.1014
1971
Stafford Beer starts Project Cybersyn for Allende's Chile: a telex network and an operations room to steer the economy in near real time.15
1973
The coup of 11 September ends Cybersyn. The military destroys the operations room.15
1974
The Biological Computer Laboratory closes for lack of research funds, after military money narrows to "militarily useful" projects.14
Four rooms where it happened
The meeting places
The Macy conferences
The Josiah Macy Jr. Foundation put mathematicians, neurologists, anthropologists and psychiatrists at one table, ten times, with Warren McCulloch in the chair. The guest list reads like a syllabus: Wiener, John von Neumann, Claude Shannon, Walter Pitts, Margaret Mead, Gregory Bateson, Heinz von Foerster. By 1950 the series title began with the word "Cybernetics."7
The point of the room: a phenomenon (circular causality) was the agenda, not a discipline. Everyone had to explain their field to everyone else.
Ashby's homeostat
Four ex-RAF bomb-control units, each wired to the other three. Knock one out of balance and the whole set hunts, switching its own wiring at random, until every needle settles again. Ashby described it in Design for a Brain (1952). His later law of requisite variety, "only variety can destroy variety," says a controller has to be at least as varied as the trouble it absorbs.812
The needles in the drawing settle the same way. Reload the page to knock them over.
The Ratio Club
A dining club founded by the neurologist John Bates for young British researchers already building thinking machines in their spare time: Ashby, Grey Walter, Horace Barlow, Donald MacKay, Albert Uttley, with Alan Turing joining after the first meeting. Its historians record a rule that members who became professors had to leave, to keep rank out of the room.1016
The point of the room: small, informal, cross-trained, and allergic to hierarchy.
Beer and Project Cybersyn
Fernando Flores brought the British consultant Stafford Beer to Chile in 1971. The plan: telex machines in factories (most often cited as 500, though the count is disputed) sending daily numbers to Santiago, and an operations room with seven swivel chairs where managers could see the economy and act on it. During the October 1972 strike, the network helped keep the city supplied with about 200 trucks. After the 1973 coup it was abandoned and the room destroyed.1517
Beer's line from this world: "the purpose of a system is what it does."18
"It's called cybernetics and it was memory holed for being too dangerous over the course of the 60s and 70s."
Verdict: undercutsThe name faded. The science did not, and danger is not why.
Where "dangerous" is real
The Soviet state called it a "reactionary pseudoscience" in 1954, then reversed itself within a year and made it a national program.11
Wiener publicly refused military work in "A Scientist Rebels" (Atlantic, January 1947) and took no government money after the war.19
Cybersyn ended in a coup, and its room was destroyed.15
Why it actually went quiet
It won and got split up. Each field took the loop and gave it a local name (see the table below).
Rebranding. McCarthy said one reason for coining "artificial intelligence" was "to escape association with 'cybernetics.'"13
Money. In the early 1970s US military research funding narrowed to militarily useful results, and von Foerster's Biological Computer Laboratory closed in 1974.1420
"I wished to avoid having either to accept Norbert (not Robert) Wiener as a guru or having to argue with him."John McCarthy, review of Bloomfield, 2000
So the club name is a joke with a true center. Nobody buried cybernetics. It dissolved into every spoke of the ring until no department owned the middle, and the people who might have defended the middle lost their grants. A field killed by success and budget lines is a better secret than one killed by danger, because it can happen again to anything.
Where the middle went
The heirs, by spoke
Every name on the right is the same loop, wearing a department's badge.
Spoke on the ring
What it took
Under what name
Engineering
Feedback, stability, governors
Control theory; state-space control from Kalman (1960)21
Bateson's double-bind theory (1956), family therapy23
Management
Requisite variety, the viable system
Operations research, Beer's Viable System Model1215
Earth and ecology
Stock-and-flow feedback
System dynamics, which fed The Limits to Growth (1972)24
Everyone
The prefix
"Cyber," which outlived the field it named
What the club does with this
House rules
Start from the phenomenon, not the department
The Macy rule. The lab's ten things, one phenomenon run (PG-026) put tree rings, a Lorenz attractor and microtubule seams in one notebook and asked whether one structure explains them.
Grade your own lore
PG-026 failed its own test: blind decoys put the ten at the 25th percentile, nothing special. The club keeps that result on the shelf. Cybernetics got oversold once; the antidote is a run that can come back "undercuts."
Re-check from the center
Ashby's law, applied to checking: a checker has to vary as much as the errors it catches. So the lab runs two checkers written apart that must agree. The second wave solver matches the first to 4×10−7, and the OEIS term was recounted from scratch.
Judge a system by what it does
Beer's rule. A claim counts when something exists to show for it: a term live on the OEIS, a sheet packer that matches 134 shipped sheets, a rerun ENIAC Monte Carlo. Von Neumann sat at the Macy table; his Monte Carlo plan, coded by Klára von Neumann, ran on ENIAC in 1948, and the lab rebuilt its 1947 flow diagram as PG-025.
No professors
The Ratio rule. Rank doesn't settle anything here. Anyone can bring a claim, and the claim gets the same attack.
On the name
Two names, one club
FormalTechnosophists
For bios, papers, and anywhere a stranger might look. It says "wisdom about technique" and reads as serious.
HouseDangerous Secret Science Club
For the people who get the joke. The danger is the punchline; the secret is that the middle of the ring is still empty and nobody's job.
Reading list
Sources
Bollen J. et al. (2009). Clickstream Data Yields High-Resolution Maps of Science. PLoS ONE 4(3): e4803. doi:10.1371/journal.pone.0004803
Börner K. (2010). Atlas of Science: Visualizing What We Know. MIT Press, p. 13 (the image in the meme).
Wiener N. (1948). Cybernetics: Or Control and Communication in the Animal and the Machine. Hermann & Cie / MIT Press. overview
Maxwell J. C. (1868). On Governors. Proceedings of the Royal Society of London 16: 270–283. doi:10.1098/rspl.1867.0055
Rosenblueth A., Wiener N., Bigelow J. (1943). Behavior, Purpose and Teleology. Philosophy of Science 10(1): 18–24.
McCulloch W., Pitts W. (1943). A Logical Calculus of the Ideas Immanent in Nervous Activity. Bulletin of Mathematical Biophysics 5: 115–133.
Macy conferences: titles, dates, chair and participants. Wikipedia; transactions ed. von Foerster, Mead, Teuber.
Homeostat: built 1948, Barnwood House; Time, 24 Jan 1949; Ashby, Design for a Brain (1952). Wikipedia
W. Grey Walter's tortoises, 1948–49, Machina speculatrix. Wikipedia
Ratio Club, 1949–1958, founded by John Bates. Wikipedia
Cybernetics in the Soviet Union: 1954 Short Philosophical Dictionary; Sobolev, Lyapunov, Kitov, "The Main Features of Cybernetics," Voprosy Filosofii, 1955. Wikipedia; see also B. Peters, How Not to Network a Nation (MIT Press, 2016).
Ashby W. R. (1956). An Introduction to Cybernetics. Chapman & Hall. Law of requisite variety. overview
McCarthy J. (2000). Review of Bloomfield, The Question of Artificial Intelligence. Stanford
Biological Computer Laboratory, University of Illinois, 1958–1974. Wikipedia