Wiener, McCulloch, Pitts, and a message from Mexico City · an essay from the house

his is the story of how the people who started the science of brains as machines stopped speaking to each other, and of what it cost the youngest of them. It is told by friends, decades later, and the friends do not agree on dates. We give every version we found and invent nothing to join them. Brief quotations from modern writers are attributed and linked; everything else is our paraphrase.¹

I. the household

In 1941 Warren McCulloch, a psychiatrist and neurophysiologist in Chicago, met two young men through his colleague Gerhardt von Bonin: Jerome Lettvin, a medical student, and Walter Pitts, a runaway from Detroit who had never enrolled anywhere and was the best logician either of them had met. McCulloch and his wife Rook took both into their house (Lettvin, J. Y., 1999)(Gefter, A., 2015). Two years later McCulloch and Pitts published the paper that made neurons into logic (McCulloch & Pitts, 1943).

In 1943, visiting Lettvin in Boston, Pitts met Norbert Wiener. Wiener was already famous and was about to name a new science, cybernetics, which needed exactly what McCulloch and Pitts had. He praised Pitts extravagantly and took him on (Gefter, A., 2015). By 1951 or 1952 the whole group was in Cambridge: Jerome Wiesner, at Wiener's urging, had brought McCulloch, Pat Wall and Lettvin to MIT's Research Laboratory of Electronics, so that Pitts could work with Wiener. Lettvin gives the earlier year, Gefter the later (Lettvin, J. Y., 1999)(Gefter, A., 2015).

II. the message

Then Wiener ended it. From Mexico City, where he and his wife Margaret were visiting the physiologist Arturo Rosenblueth, he sent word that he would have nothing more to do with McCulloch or anyone in his group. The sources do not agree on what the message was or when it went:

• Lettvin: a letter to Wiesner, late in 1952, "viciously phrased" (Lettvin, MIT Encyclopedia of the Cognitive Sciences).
• Gefter: a telegram to Wiesner, 1952 (Gefter, A., 2015).
• Dyson, summarising Conway and Siegelman's biography of Wiener: a letter to the president of MIT, 1951 (Dyson, F., 2005)(Conway, F. & Siegelman, J., 2005).

Wiener never told Pitts why (Gefter, A., 2015). McCulloch, according to Dyson, never knew (Dyson, F., 2005).

III. the reason given

The reason that has come down to us has one source. About ten years later Rosenblueth told what had set it off. As Wiener's biographers Conway and Siegelman report it, Margaret Wiener had told her husband that men in McCulloch's group had seduced the Wieners' daughter Barbara (Dyson, F., 2005)(Brown, 2006).

Only after a decade did Rosenblueth reveal what had set off this explosion
— Lettvin, MIT Encyclopedia of the Cognitive Sciences

Margaret Wiener gave no account of her own to the biographers, and none has been published. We have nothing to add to what she is reported to have said, and we do not describe her.²

Those who weighed the story did not believe it. Lettvin called it "a deliberate and cynical manipulation designed to sever Wiener's connection with McCulloch and his group", and added:

The details are not edifying; Wiener was victimized as much as the group.
— Lettvin, MIT Encyclopedia of the Cognitive Sciences

This story had no basis in fact
— Dyson, New York Review of Books, 2005

The reviewer for Physics Today read Conway and Siegelman's book the same way: Lettvin "suggests that Margaret's accusation was a lie" (Brown, 2006).

IV. what it cost

For Wiener, the loss was a field he had named and then walked away from. For McCulloch, a patron and a friend, and no explanation. For Pitts it was worse. He had been Wiener's student in everything but paperwork. In 1954 Fortune listed him among the most promising young scientists in the country (one of ten, says Smalheiser; one of twenty, says Gefter). He stopped publishing on his own account. He destroyed his thesis:

At about the time of the break with Wiener, and arguably because of it, Pitts destroyed his thesis
— Smalheiser, 2000

Others put the burning later, after the frog work of 1959 had undercut the logic of 1943; the sources disagree on timing and on cause. In 1955 he would not sign the papers that would have given him an MIT doctorate (Smalheiser, 2000). He went on working with Lettvin on the frog (Lettvin et al., 1959), drank, and withdrew.

He destroyed all of his past work that he could find and became a ghost long before he died.
— Lettvin, MIT Encyclopedia of the Cognitive Sciences

He died on 14 May 1969, alone, in a boarding house in Cambridge, of bleeding from oesophageal varices. McCulloch died in September of that year. Wiener had died in 1964. Margaret Wiener lived until 1989 (Dyson, F., 2005).

V. what we can say

Very little for certain. The account of the cause rests on one informant speaking a decade later; the dates rest on recollection; the effect on Pitts is told by those who loved him. What is not in doubt is the shape: a young man whose standing depended on one older man's regard, and who did not survive its withdrawal as a working scientist. Science likes to think of itself as a system that does not depend on any one relationship. This one did.

VI. what was lost

Authored: the house's view. This section is ours. It adds no facts to those above; it says what we think they add up to. Where it says might, it is a guess. The first loss was a piece of work. In the early 1950s Pitts was writing a doctoral thesis about large nets of neurons whose connections are random and three-dimensional, and what such nets would do on average (Gefter, A., 2015). Wiener, who had made the mathematics of random processes his own, was the reader it was written for. Pitts burned it (Smalheiser, 2000). No draft is known to survive, and nobody can now say how far it had got. The second loss was a household. For most of a decade four men worked as something closer to a family than a department: McCulloch, who opened his house; Lettvin, who brought medicine; Pitts, who brought logic; Wiener, who brought the mathematics and the name. After the message there were two camps, and the one who had belonged to both had nowhere to stand. Conjecture. The third loss cannot be counted. Pitts might have been among the first to work out how a mass of randomly wired cells settles, holds a pattern, and fails. Others reached those questions over the next thirty years, by other routes; whether he would have got there sooner, or somewhere else, is a guess. We set out that guess, labelled, in the thesis that wasn't, and do not repeat it here. None of this needs a villain. A thesis is paper, a friendship is habit, and both were ended by a few lines sent from Mexico City. The science went on without him. What it would have been with him was never tried.

in the tour: the room called "Building 20" has a placard called "The letter" ···

notes

1. Every claim on this page is a row in a ledger of sources we checked; where they disagree we give all versions. The quotations are brief, from the works linked, and used for comment.
2. The published accounts name Barbara Wiener. We name her for that reason only, and say nothing more about her.

Sources

1. Lettvin, J. Y. (1999). "Pitts, Walter." In R. A. Wilson & F. C. Keil (eds.), The MIT Encyclopedia of the Cognitive Sciences. Cambridge, MA: MIT Press. ai.ato.ms/MITECS/Articles/lettvin1.html
2. Gefter, A. (2015). "The man who tried to redeem the world with logic." Nautilus, February 2015. nautil.us/the-man-who-tried-to-redeem-the-world-with-logic-235253/
3. McCulloch & Pitts (1943). A logical calculus of the ideas immanent in nervous activity. The Bulletin of Mathematical Biophysics. doi:10.1007/BF02478259
4. Dyson, F. (2005). "The tragic tale of a genius." Review of Conway & Siegelman. The New York Review of Books, July 2005. www.nybooks.com/articles/2005/07/14/the-tragic-tale-of-a-genius/
5. Conway, F. & Siegelman, J. (2005). Dark Hero of the Information Age: In Search of Norbert Wiener, the Father of Cybernetics. New York: Basic Books.
6. Brown (2006). Review of Conway & Siegelman, Dark Hero of the Information Age: In Search of Norbert Wiener, the Father of Cybernetics. Physics Today. doi:10.1063/1.2216967
7. Smalheiser (2000). Walter Pitts. Perspectives in Biology and Medicine. doi:10.1353/pbm.2000.0009
8. Lettvin et al. (1959). What the Frog's Eye Tells the Frog's Brain. Proceedings of the IRE. doi:10.1109/JRPROC.1959.287207

see also

	Walter Pitts
	Norbert Wiener
	Warren McCulloch
	the thesis that wasn't — what might have been, labelled as a guess
	timeline: cybernetics
	annotated: Pitts and the frog