From the Princeton Meeting to Hiroshima

Abstract

“Dr. von Neumann and myself felt it desirable to hold a joint meeting of all those interested in what we now call cybernetics, and this meeting took place at Princeton in the late winter of [1944–1945]”. (Wiener, N., Cybernetics). ... After Hiroshima and Nagasaki, a physicist writes to Einstein: “Here, at the Massachusetts Institute for Technology, [Norbert] Wiener stands aghast—as though a man in a confused dream—and wonders what we must do” (Einstein 1963, 342).

1 The Preliminary Phase of the Princeton Meeting

Concluding the passage in Cybernetics which tells us about the comings and goings for consultations with Aiken , von Neumann, and Goldstine , Wiener asserts:

At this stage of the proceedings, Dr. von Neumann and myself felt it desirable to hold a joint meeting of all those interested in what we now call cybernetics, and this meeting took place at Princeton in the late winter of 1943–1944 [sic, true date: 1944–1945] (61c [48f1], 15).

The early idea from which the Princeton meeting sprang was a conversation between Wiener and Aiken. On 17 October 1944, Wiener wrote to von Neumann :

I saw Aiken the other day and I am much impressed by the coincidence of his point of view with mine on the future of computing machines. We are thinking of trying to get some meeting of the American Society of the Advancement of Science after the war is over for the discussion of the whole complex problems relating to computing machines, communication engineering, prediction theory, and control engineering. I would like to get your point of view on that.¹

Two days later Wiener also acquainted Arturo Rosenblueth with his discussion with Aiken and the hypothesis of the conference at the American Society of the Advancement of Science, to deal with “the complex of subjects in which you and I are interested”. He added:

If we do that we shall try to make it a matter of invited addresses of perhaps a half an hour each by a mathematician, physiologist, electrical engineer, statistician, etc. Certainly, this wants to be done in such a way that you can participate in it.²

If we compare the two letters, there appears a duality in the way of understanding the new field that Wiener called “cybernetics”. In the first letter it is intended as a generalized form of communication engineering, centered on information, according to the ideas of the Yellow peril (1 February 1942). It was a theory that now firmly included both analog and digital computing machines. We could perhaps speak of “cybernetics on a large scale”, based on a rigorous and highly theoretical approach. In the second letter, another kind of cybernetics appears, in which physiology plays an important role, in line with the model represented by Behavior, Purpose and Teleology. Here we could speak of “cybernetics on a small scale”, based on a meticulous experimental approach. Wiener probably hoped that the two approaches could be brought together. For the moment he preferred not to clarify this duality.

However, as we read in Cybernetics, both von Neumann and Wiener “felt it desirable to hold a joint meeting of all those interested in what we now call cybernetics, and this meeting took place at Princeton” (61c [48f1], 15). Indeed, von Neumann asked to increase the number of guests, but required a very low profile.

On 4 December 1944 a letter of invitation was sent for a meeting to be held at Princeton University. It was actually written by Wiener, but also in the name of von Neumann and Aiken . Its recipients were Walter H. Pitts , Warren S. McCulloch, Rafael Lorente de Nó, Samuel S. Wilks, Ernest H. Vestine, W. Edwards Deming, and Leland E. Cunningham. A few days later Herman Goldstine was added.³ Wiener’s letter explained that:

Owing to the war, it is not yet the time to call together a completely open meeting on the matter, because so many researches developed in the war effort are concerned, but it seems highly desirable to summon together a small group of those interested to discuss questions of common interest and make plans for the future development of this field of effort, which as yet is not even named.⁴

The meeting concerned what I have called “cybernetics on a large scale”. As Wiener writes in the letter:

Communication engineering, the engineering of computing machines, the engineering of control devices, the mathematics of time series in statistics, and the communication and control aspects of the nervous system, […] have developed to a degree of intimacy that makes a get-together meeting between people interested in them highly desirable.⁵

It was stressed that this was a “field of effort, which as yet is not even named”.⁶ Cybernetics would return to this issue, speaking of an “essential unity of the set of problems centering about communication, control, and statistical mechanics, whether in the machine or in living tissue”, but hampered by “the lack of unity of the literature concerning these problems, and by the absence of any common terminology, or even of a single name for the field” (61c [48f1], 11; italics added), a problem, he adds, that had then been resolved by entitling the book “Cybernetics” (61c [48f1], 11).

Wiener raised the issue of the name for the field again in a letter dated 28 December 1944, proposing to discuss it during the meeting. For the moment he suggested establishing a permanent society that could be called the “Teleological Society”, while “Teleology” or “Teleologia”, could be the title of the associated scientific journal.⁷

2 The Princeton Meeting

The meeting was held on 6 and 7 January 1945. Aiken was unable to participate, but does not seem to have withdrawn because he was assigned to one of the four working groups formed at the meeting. On 24 January 1945 Wiener wrote to Rosenblueth :

[It] was a great success […]. The first day von Neumann spoke on computing machines and I spoke on communication engineering. The second day Lorente de Nó and McCulloch joined forces for a very convincing presentation of the present status of the problem of the organization of the brain. In the end we were all convinced that the subject embracing both the engineering and neurology aspects is essentially one, and we should go ahead with plans to embody these ideas in a permanent program of research […]. While we are going to meet again in the spring, we have not organized in a formal permanent society.⁸

We can interpret that Wiener must have provided a unified basis for discussion when presenting his idea of a generalized communication theory, with a focus on the ideas of the Yellow peril, and their extension to computers, in short what I have called “cybernetics on a large scale”. McCulloch must have spoken about the properties of the neural net of A logical calculus, explaining their similarity with networks formed by relays in the digital computers. It remains to explain the presence of Lorente de Nό. An eminent neuro-anatomist and pupil of Ramon y Cajal , he had studied in vivo the reverberating circuits in the brain, demonstrating their importance in the vestibular nystagmus (involuntary rhythmic oscillation of the eyes) and in other neurological situations (Lorente de Nó 1938). We have already mentioned the relation between the supposed memory function of closed circuits of neurons, and the mercury delay line of ENIAC and EDVAC.⁹ Therefore we can suppose that Lorente de Nό spoke about this theme.

In this period von Neumann had taken the entire field of “computing machines” firmly in his hands. We know nothing directly of his talk. We can form a draft idea of what he said, considering the content of seminars and informal conversations he gave around that time in Los Alamos, as related by Nelson and Metropolis . He regularly use to discuss the latest developments in computing, including the Harvard Mark I, BTL relay computers, and the ENIAC, and had formed a personal view of the whole subject (Metropolis and Nelson 1982, 351). In particular, he had become to see:

[…] the technical links between the separate independent developments. He also described his ideas on the computer of the future, outlining his single-address architecture, later implemented in the IAS computer, in the IBM 701, and in other computers. […] In addition to information about digital-computer development, von Neumann told us about conversations he had had with McCulloch and Pitts, who were investigating brain functioning using Boolean algebra. Their ideas stimulated his thinking about advanced, digital concepts (Metropolis and Nelson 1982, 352 and 354).

Wiener was very satisfied with the meeting, which he considered “a great success”.¹⁰ “This meeting”, he would state triumphantly in his autobiography, “I may consider the birthplace of the new science of cybernetics, or the theory of communication and control in the machine and in the living organism” (64g [56g], 269).

Regarding the “Teleological Society”, its establishment was postponed until after the war. “The reasons are”, Wiener explained to Rosenblueth , “that owing to the control of different and by no means unified government departments over parts of our program, it is best not to stir up a fuss until the military situation makes matters of classification less important”.¹¹ Furthermore, the choice of name for the new field and the public communication of research in it had also been postponed. The conference was held in a strictly confidential manner at the behest of von Neumann . “However”, Wiener assured Rosenblueth, “we definitely do have the intention of organizing a society and a journal after the war, and founding at Tech or elsewhere in the country a center of research in our new field”.¹²

Possible funding sources for the project were also discussed: the J.S. Guggenheim Memorial Foundation, through one of its leaders, Henry A. Moe and the Rockefeller Foundation through Warren Weaver . In addition, Wiener had “heard from von Neumann mysterious words concerning some 30 megabucks which is likely to be available for scientific research. von Neumann is quite confident that he can siphon some of it off”. In general, Wiener was enthusiastic and confident about the future of the project, whereas “McCulloch and von Neumann are very slick organizers”.¹³

Von Neumann prepared a “Memorandum Summary of the meeting”, of which a copy was sent to each participant with the date 12 January 1945.¹⁴ He focused on the division of several tasks and the future agenda, referring to the letter of invitation for the more general view regarding the new field.¹⁵ The working groups were the following: “1” on “Filtering and prediction problems”, comprising Pitts and Wiener; “2” on “Application of fast, mechanized computing methods to statistical problems”, with statisticians Deming, Vestine , Wilks , and if possible, von Neumann himself; “3” on “Application to differential equations (astronomy, ballistics, hydrodynamics, etc.)”. It seems to me that here von Neumann took for implicit, without repeating it again, the specification “Application of fast, mechanized computing methods to”. It is no accident that the components of this group were experts in computer design, such as Aiken of the Harvard Mark I, Leland E. Cunningham , astronomer in charge of the BRL automatic computing section (Cf. von Neumann 1944) , Goldstine , and von Neumann. Finally, there was a group “4”, concerning the “Connected aspects of neurology”, comprising Lorente de Nό, McCulloch , and Pitts.¹⁶

Each group had the task of drawing up “preliminary draft memoranda”, within 3 months, i.e., by the end of March or the beginning of April 1945. After 6 months, so by the end of June or the beginning of July, there would be a new general meeting, similar to the January one, where “a joint final memorandum of the research program will be undertaken and a small group put in charge of writing it”.¹⁷

3 The Work of the Four Groups

What results did the four groups produce? The usual thesis among historians of cybernetics and computer science is that the project ran aground. Considering the secrecy in which the groups worked, the fact that there is a shortage of documents (not an absolute absence, as a matter of fact) does not seem to authorize such a drastic conclusion. Bigelow, asked by Aspray about the fate of the teleological society, “attributes this failure in part to a clash of personalities among the organizers”.¹⁸

The fact is irrefutable, but the clashes—we have to use the plural because there were at least two—which actually provoked a cataclysm within the new field of cybernetics occurred after August 1945, i.e., after the nuclear bombing of Japan. As we shall see later, there was a clash, or at least a strong cooling of their relations, between Wiener and von Neumann in the Fall of 1945, after Hiroshima and Nagasaki. A second clash, with Wiener on one side and McCulloch and Pitts on the other, happened between December 1951 and January 1952 (Cf. Conway and Siegelman 2005, 213–234). However, from the amount of letters in our possession, no sign of a clash emerges at least before August 1945, the date of the nuclear bombing of Japan. Indeed, the 7 months between the conference in Princeton and the end of the war were marked by a genuine honeymoon between Wiener and von Neumann, two men forming a sort of hub around which the activities of the others were spun enthusiastically.

Nothing suggests that the program proposed by von Neumann in the Summary was not applied in the seven first months of 1945. After reading it, Wiener quickly replied to von Neumann, stating:

I found one thing missing in your assignment of topics: namely, there was no single place where the problem of transition from the computing machine to the control machine was discussed. I think this is one of the most important aspects of our project and as it is closely related to the prediction and filtering problems, I have assumed that it goes to our subcommittee for a report.¹⁹

On 1 February 1945 von Neumann answered Wiener, saying he was “most interested to see the memorandum by you and Pitts as soon as it is ready”.²⁰ He also claimed to have “thought a good deal more about transitions and intermediate forms between counting and continuous processes, and I have some ideas on the subject I should like to discuss with you”.²¹

Moreover, von Neumann informed him: “after some discussions with Cunningham and Goldstine I think it would be desirable to add Eckert of the University of Pennsylvania and Chandrasekhar to our group, and possibly also G. Stibitz ”.²² On 23 January 1945, von Neumann wrote to Deming expressing his wish to include Harold Hotelling (1895–1973) in the group of statisticians.²³ It therefore appears clear that the groups were working.

I do not know of letters between Wiener and von Neumann from 1 February to late March. On 24 March 1945 Wiener wrote to von Neumann: “our meeting last Monday was a grand success, and hope we shall have many more of them”.²⁴

It seems to me that these words could not refer to a face to face meeting. It is easier to imagine a conference, in which Wiener would also have taken part in a active way. Through the perpetual calendar we discover that the previous Monday was 19 March 1945. On that date a conference was held at the IAS in Princeton on “Shock-waves and Supersonic Flow”, where von Neumann had presented a paper on “Refraction, Intersection and Reflection of Shock Waves”.²⁵ Wiener also tells of having spoken with Oswald Veblen , professor at the IAS in Princeton, and one of its main fathers, along with ballistic research in the United States. Veblen had also been a real mentor for von Neumann in the United States. On 26 March 1945 von Neumann also sent Veblen a memorandum entitled “Use of variational methods in hydrodynamics”, in which he discussed nonlinear PDEs, especially of hyperbolic or mixed type, that is, those PDEs that one typically meets in the problems discussed in the paper he gave at the conference.

In the memorandum to Veblen, von Neumann concluded that he would make a systematic attack on hydrodynamics, both by introducing new analytical perspectives, and by using the new high speed computing machines. He began to question the inner nature of the numerical difficulties posed by hydrodynamic problems.²⁶ Metropolis and Nelson call von Neumann’s new approach “experimental mathematics” (Metropolis and Nelson 1982, 348). It implied that computational work should not just be a shortcut to get solutions to mathematical problems for which analytical solutions were not available. On the contrary, von Neumann thought that, using results obtained numerically, one could find new analytical methods. At about the time of the memorandum to Veblen, von Neumann began to think, as he had then written to Commodore Strauss in October, that:

An electronic machine of the most advanced conceivable type should be constructed, not for use on specific applied mathematical or physical or engineering problems, but with the purpose of experimentation with the machine itself in order to develop new approximation and computing methods, and generally to acquire the mathematical and logical forms of thinking which are necessary for the really efficient operation of such a device, with the methods it will have brought into existence.²⁷

Von Neumann believed that this line of research would best be conducted within an academic environment, rather than in military or industrial laboratories “which have definite, and necessarily relatively narrowly defined, applied problems to which they must devote all or most of the time of their equipment”.²⁸

Wiener immediately welcomed von Neumann’s ideas in an encouraging way. On 24 March 1945 he wrote to von Neumann: “Veblen told me about your post-war plans for hydrodynamics. I think your balance between pure and applied maths is the right one and our little control scheme fits perfectly into the picture”.²⁹

Wiener considered that their common plan for the Teleological Society would be perfect for this project. However, he thought that the IAS was not the right place to carry out von Neumann’s ideas, since they would require laboratories, and noted that “labs don’t grow in ivory towers”.³⁰

Wiener informed also von Neumann that he had spoken about him with George R. Harrison , dean of the MIT Faculty of Science, in view of an appointment of von Neumann as director of the Department of Mathematics, to replace Henry B. Phillips who was about to retire.³¹ Wiener also proposed to von Neumann that MIT should be the headquarters of the Teleological Society. However, he considered it a disadvantage that, on the physiological side, they would have to support the Harvard Medical School, who in his opinion, now without either Cannon or Rosenblueth , had become a primarily clinical institution. Another option would be to convince MIT to establish a small physiological studies program.³²

On 21 April 1945 von Neumann told Wiener he had been at MIT to discuss his employment (Cf. Heims 1984 [1980], 189). He had been fascinated most of all by the meeting with Richard Taylor . Von Neumann wrote:

I was in Cambridge and saw R. Taylor and the Tech mechanico-electronic analyzer, we spent two days together. It was very interesting, particularly considering what Taylor might do in the future. We should by all means have a talk ‘à trois.’ I think there is more to learn from Taylor than from Aiken—and not on analyzers only or mainly.³³

The letter confirms once again the special intimacy that built up in this period between von Neumann and Wiener, around the themes of cybernetic and the idea of a future joint collaboration at MIT.

4 The “Dark Side” of Cybernetics

Von Neumann was a great scientist, and like Wiener, aimed to get general results. In his hands, although with the help of others, and in particular Wiener, the design of computers became a science. But he was engaged in a military project of extreme strategic relevance. In Los Alamos throughout 1944, theoretical and computational work had focused mainly on the plutonium bomb and the implosion problem. In early 1945 Oppenheimer scheduled the various stages leading to the July deadline for the completion of the uranium and plutonium bombs. By then, it was expected that the industrial production centers would have reached a sufficient amount of uranium for Little Boy. It was also believed that a sufficient amount of plutonium would be available for Fat Man, and for a real scale test. The latter was carried out on 16 July 1945, at the Alamogordo air base, New Mexico, at the so-called Trinity site, producing as expected a destructive power of 21 kilotons (1 kiloton is the equivalent explosive power of 1000 tons of TNT). This test alone guaranteed the effectiveness of the method and allowed physicists to evaluate the consequences of the explosion (Cf. Frank, Cochran and Norris 1996).

In the aforementioned letters to Wiener of 1 February 1945, von Neumann informed Wiener: “I am leaving on February 4 for Aberdeen, and on February 6 for the West, and expect to be back in the first days of March”.³⁴ He concluded: “I hope that we shall see each other very quickly thereafter. I am sure it would be very profitable for me at least if we could have another conversation”.³⁵ The “West” was of course Los Alamos; in fact, on 12 February 1945 it was from there that von Neumann wrote to Goldstine , offering a number of suggestions about the development of the computers at the Moore School.³⁶

Being a great scientist, just like Wiener, von Neumann would certainly also have been motivated by the discovery of new knowledge in the emerging field. However, in the course of 1944 and 1945 von Neumann had been engaged with the activities at Los Alamos and his main interests were the theoretical and computational needs of that laboratory, in charge of the most important military project in the United States. Therefore the Princeton meeting in January 1945, like the talks before it and the activities following it, must have been strictly designed, in his mind, to cater for those more immediate interests.

It is rather clear that the changes von Neumann had asked to introduce in the ENIAC and the new EDVAC project were connected to the computing needs at Los Alamos. It is in my opinion possible to hypothesize that the contribution of the working group formed at Princeton in January, especially Group 3, was related to those projects.

On the one hand, von Neumann needed analytical and numerical tools, and on the other, he needed to know how best to build the new electronic digital general purpose computer.

In this regard it is interesting to consider the co-optation of the astrophysicist Subrahmanyan Chandrasekhar into Group 3. Von Neumann had long been in close contact with him. They were both speakers at the 4th Annual Conference on Theoretical Physics on “The Problem of Stellar Energy”, held on 21–23 March 1938, focused on nuclear fusion in stars. Among the other speakers, there was also Teller , the father of the H-bomb.³⁷ In 1943 he had published with von Neumann’s blessing an important essay on “Stochastic Problems in Physics and Astronomy”.³⁸ In that same year, Chandrasekhar had become a consultant for the BRL, for which he dealt mainly with shock waves, his first encounter with hydrodynamics, in which he would later make his career.³⁹ It does not seem unreasonable to assume that von Neumann’s interest in involving him concerned the mathematical demands at Los Alamos which related to the problems of hydrodynamics. In particular, he would have been able to help in connection with nuclear fusion.

On 12 February 1944, from Los Alamos, von Neumann told Goldstine : “I am also working on the problem of formulating a two-dimensional, non-stationary hydrodynamical problem for the ENIAC”.⁴⁰ Metropolis and Nelson add that “in early 1945” von Neumann asked Frankel and Metropolis “to perform the very complex calculations involved in hydrogen bomb design” on the ENIAC (Metropolis and Nelson 1982, 353).

It seems unlikely, however, that the purpose of this computation concerned the H-bomb as such. This bomb was definitely not a priority at that stage. Regardless of the real moral reasons for a bomb thousands of times more powerful than a fission bomb, the low priority placed on it by Oppenheimer is easily explained by the fact that a fusion bomb needs to be triggered by the explosion of a fission bomb, and this fact remained uncertain until the Alamogordo test.

On the other hand, Goldstine assures us that “this problem was of great importance, since it was to test out a dramatic new idea for Los Alamos and an equally dramatic one for the Moore School, the ENIAC”. (Goldstine 1980 [1973], p. 246). In an interview, Metropolis clarified: “We tried to run a set whose ensemble would enable us to make certain inferences about what the prospects were for the possibilities of thermonuclear ignition” (Metropolis 1987).

In my opinion it is more likely that the computation aimed to assess the consequences of the explosion of an atomic fission bomb, and in particular the verification of the hypothesis of the “ignition of the atmosphere”. Actually, since 1942, physicists had been considering the risk that the very high temperatures and pressures produced by the explosion of an atomic fission bomb could trigger a chain reaction of thermonuclear fusion in the light nuclei of the atmosphere. Only very accurate computations could solve the question. Frankel and Metropolis went to the Moore School in the Spring of 1945 (Metropolis 1987 and Goldstine 1980 [1973], 214–5), and various computations were carried out under the supervision of Teller up to the first months of 1946 (Goldstine 1980 [1973], 258). In August 1946, a report by Teller and others on “Ignition of the atmosphere with the nuclear bombs”⁴¹ was distributed at Los Alamos. No doubt the same computations could have been a prelude to the construction of the H-bomb itself (Fitzpatrick 1999, spec. 118), but it seems quite possible that, prior to the Trinity test, consisting in the explosion of a real plutonium nuclear bomb, the ENIAC and all the computational and intellectual resources available at that time, were being used to describe in detail the entire course of the nuclear explosion, from its triggering through gun or implosion methods up to its final effects on the atmosphere.

5 Wiener and Rosenblueth Study Cardiac Conduction

In the letter to von Neumann of 24 March 1945, Wiener also announced that on 31 March he would leave for Mexico City, going as invited speaker to the Congress of the Mexican Mathematical Society at Guadalajara, scheduled from 28 May to 2 June 1945. He would in fact stay from 9 April at Instituto Nacional de Cardiologia.⁴²

On 27 April he informed von Neumann that in Mexico City he had teamed up with Rosenblueth on the “mathematics of nerve conduction”.⁴³ In another letter, Wiener described this as “the study of the applications of the theory of networks of sensitive tissues to the flutter and fibrillation of the heart”.⁴⁴ The provisional results of the research were presented at the Guadalajara Congress and then in a definitive form in [46b].

It was in effect the first true experience of “cybernetics on a small scale”. The research was discussed by Wiener and Rosenblueth together with Walter B. Cannon himself, taking advantage of his visit to Rosenblueth. Cannon would die in October 1945. On the one hand the research can be understood as the continuation of an experimental study, begun by Rosenblueth and Cannon in the year 1941 (cf. Rosenblueth and Cannon 1942), based on studies of the response to stimulation of the central nervous tissue in some animals, observed by oscilloscope. Between 1942 and 1944 Rosenblueth also continued his experimental studies of rhythmic responses in the nervous tissue and the striated muscles.

In his research with Wiener, Rosenblueth took into account two pathological forms of heart rhythm, flutter and atrial fibrillation, diseases of the heart rhythm that are both characterized by a much higher frequency than the normal rhythm (the so-called “sinus rhythm”). Fibrillation is characterized by variable and substantially random periods. The main purpose of the research was not really to study the two diseases, but rather to develop methods that could be useful for studying the neural net, taking advantage of the fact that, in the heart muscle, stimuli spread in a similar manner to the way they do in the cerebral cortex, but without the complex situation created there by synapses and the on/off behavior of the neurons.

Wiener introduced a strong mathematical component into the research. In particular, the equations to explain fibrillation were based largely on the ideas of a paper on “Discrete Chaos” (43a), written with A. Wintner . In addition, Wiener states: “The statistical technique used in the study of heart-muscle nets has been extended to the treatment of neuronal nets by Mr. Walter Pitts ” (61c [48f1], 17). This would become the theme of Pitts’ doctoral thesis: from the collaboration between Pitts and Wiener, the project of a brain based on stochastic neuronal nets had sprung.

6 At the Height of Enthusiasm

Both von Neumann’s “experimental mathematics” and Wiener and Rosenblueth’s work on “cardiac conduction” can be considered in the framework of post-war plans. At that point, the end of the war seemed very close. On 28 April 1945, Mussolini had been killed, while on 30 April the Soviets had taken the Reichtag and Hitler had committed suicide. Between 17 July and 2 August, the last inter-Allied conference was held at Potsdam, where on 26 July the new President Truman , former vice-president of Roosevelt , who had become president on the latter’s death on 12 April 1945, had ordered the surrender of Japan.

At the beginning of July, Wiener went to talk to Henry Moe of Guggenheim Foundation about the plans for the new field, and also for a scholarship to allow Pitts to complete his Ph.D. at MIT, since his relationship with the Kellex Corporation was nearing its end.⁴⁵ Moe showed interest in the project and submitted the forms for Pitts’ scholarship.⁴⁶ In the same month, Wiener and Pitts went to visit McCulloch at the University of Illinois. McCulloch wrote to Rosenblueth that they “were full of your experiments and Wiener’s calculations on the theory of flutter and fibrillation and they showed me swell manuscripts under way”.⁴⁷

Wiener’s enthusiasm arose from the prospects opened up by this “field yet unnamed”, so full of promise. An important aspect for the success of the projects in this field was certainly von Neumann’s appointment at MIT, which was now considered as the most probable place for the “Teleological Society” or whatever it would be called. On 1 July 1945 Wiener wrote to Rosenblueth: “I have had several consultations with von Neumann […] and it really looks to me now as if the appointment and his acceptance were in the bag”.⁴⁸ On 11 July he added: “it is quite clear that if the appointment comes through, all of our ideas concerning an organized collaboration between physiological and mathematical subjects will follow as a matter of course”.⁴⁹ The same day he wrote to another person: “Johnny has been and gone and it looks as though he is in the bag. Everybody is delighted and we are going to go places”.⁵⁰

Wiener was ready, together with his collaborators, to begin to work with athletic spirit on the new field which had emerged during the war. On 22 July 1945 he wrote to ask Rosenblueth to consider coming back to Mexico. Meanwhile, their research manuscript had also been subjected to von Neumann’s criticism before being sent to Rosenblueth for publication.⁵¹

Further evidence of Wiener’s enthusiasm consists in a curious letter of 26 July, in which he replied at length to a young student who had asked for advice on what were the most promising fields of study for the future. Wiener spoke of his research with Rosenblueth , concerning “the region in which physiology and mathematics come together. In particular, both in the nervous system and in such muscular systems as the heart”.⁵² This was clearly cybernetics on the small scale, but bound up with cybernetics on the big scale. In fact he added:

Closely related to the problem of the analysis of organization in living tissue is a problem of the synthesis in organization in such devices as computing and control machines”,⁵³ requiring “a revision of statistical theory in which the procession of events in time is fully considered.⁵⁴

7 The Nuclear Bombing of Japan and Wiener’s Crisis

Wiener could not have dreamt how the political and military leadership of the United States had chosen to put the word “end” to the war: with the nuclear bombing of Hiroshima on 6 August 1945 and Nagasaki on 9 August. This event deeply marked the fate of cybernetics, and even our way of understanding it.

At first it did not seem that Wiener suffered that much over the still fragmentary news of the nuclear bombing of Japan. Even on 11 August 1945, he was referring to Rosenblueth with the usual refrain about von Neumann : “Johnny was down here the last 2 days. He is almost hooked”.⁵⁵ During the visit to MIT, von Neumann had also reassured him of the internal consistency of the manuscript on heart conduction.⁵⁶ Three days later, a letter from Harrison reached von Neumann, offering him a mathematics professorship at MIT, with a good salary of $ 15,000, and a commitment to appoint him as Director of the Department of Mathematics, after Phillips ’ retirement.⁵⁷

However, as the news about the atomic bombing of Japan became more precise, the broadly cheerful atmosphere due to the end of the war dissolved in Wiener’s soul. He began to build up a better understanding of the circumstances. His mood began to darken, and finally he fell into an acute crisis of conscience. In the letter of 11 August, he had confided to Rosenblueth that “in the present almost certain to come interval between wars (and I hope to goodness it will be a long one), I think we can do an enormous amount with our new schemes”.⁵⁸

He had a presentiment, common to other enlightened spirits of the time, that the end of the war would have been only the prelude to a World War III. He just hoped that it was a long interim, in order to cultivate the new projects in the best possible way. A further point followed, however, and in response, Rosenblueth tried to comfort him:

You sounded rather pessimistic in your last letter. That is wrong. The war could not possibly be going better […] Your work seems to be going along beautifully from what you tell me. Your family is doing handsomely. Our projects, although still in the realm of the ‘we shall see’ are alive and kicking (or maybe I should say, and wagging their caudal appendage). Your novel is still trying to crack its shell. You have friends and they don’t forget you—witness thereof, the pleasant time I’m having writing to you. What the Avernus can you crab about, anyhow? It’s really a great world and a great life, my dear Norbert, notwithstanding their occasional infirmities.⁵⁹

Before Wiener’s eyes, the doors of Avernus were opening wide. As in 1942, he was experiencing a rude awakening from the enthusiasm of his tireless research. But this time the vision of reality had the apocalyptic colors of a nuclear bombing that had erased the lives of about 200,000 people, mostly civilians, with only two bombs. And maybe a vague feeling of having been an unwitting accomplice. He felt in a deep crisis of conscience.

On 21 October 1945 the physicist Daniel Q. Posin of the MIT Radiation Laboratory, illustrating a widespread uneasiness with regard to nuclear weapons among scientists, wrote to Albert Einstein :

Here, at the Massachusetts Institute for Technology, [Norbert] Wiener stands aghast—as though a man in a confused dream—and wonders what we must do, and he protests at scientific meetings the “Massacre of Nagasaki” which makes it easier, for some, to contemplate other massacres.⁶⁰

On 18 October 1945, Wiener presented a resignation letter to the MIT rector Karl T. Compton , claiming that he wished “to leave scientific work completely and finally. I shall try to find some way of living on my farm in the country. I am not too sanguine of success, but I see no other course which accords with my conscience”.⁶¹

Some reasons for the crisis leaked out in a letter written on 16 October to his friend Santillana . This was a long letter in which Wiener made a list of the results he had been working on, which he had been so proud of until a few months before, and which had formed the basis for all his projects after the war. Wiener explained:

Ever since the atomic bomb fell I have been recovering from an acute attack of conscience as one of the scientists who has been doing war work and who has seen his war work a[s] part of a larger body which is being used in a way of which I do not approve and over which I have absolutely no control. I think the omens for a third world war are black and I have no intention of letting my services be used in such a conflict. I have seriously considered the possibility of giving up my scientific productive effort because I know no way to publish without letting my inventions go to the wrong hands.⁶²

Wiener had not directly participated in the Manhattan Project, but must have been aware that “his war work a[s] part of a larger body” had been used in a way that he did not approve, that is, of course, to destroy Hiroshima and Nagasaki. But his war work coincided with the “science still without a name”, namely, cybernetics. What frightened him more about this research was the question of the “wrong hands”, an expression he would use several times in future writings.

Considering all these facts, it is easy to understand the gloomy language used by Wiener to announce the birth of cybernetics:

Those of us who have contributed to the new science of cybernetics thus stand in a moral position which is, to say the least, not very comfortable. We have contributed to the initiation of a new science which, as I have said, embraces technical developments with great possibilities for good and for evil. We can only hand it over into the world that exists about us, and this is the world of Belsen and Hiroshima (61c [48f1], 28).