Reception and Rejection
 222 Downloads
Abstract
Pauli presented Heisenberg’s and his new approach to nonlinear spinor theory for the first time to a small crowd of physicists in Milan, on his way to board his ocean liner in Genova, on 18 January 1958. But the first presentation to a large audience was given on February 1 at Columbia University.
Pauli presented Heisenberg’s and his new approach to nonlinear spinor theory for the first time to a small crowd of physicists in Milan, on his way to board his ocean liner in Genova, on 18 January 1958.^{1} But the first presentation to a large audience was given on February 1 at Columbia University. This debacle is the stuff of physicist’s legends and various accounts of it are collected in von Meyenn (2005, pp. 871–872). I have also been able to track down some archival material (DBP) concerning this talk in the handwritten notes of Dieter Brill (now emeritus professor at University of Maryland). These largely confirm the established elements of the Pauli debacle: The sudden wavering (“Nowahsoah”), the growing unease in the audience (Brill notes that Lehmann, who was sitting behind him, remarked: I didn’t understand that at all—Das hab ich gar nicht verstanden), Pauli’s questioning of central elements of the proposal (“Heisenberg proposed degenerate vacuum. I’m not certain that this is necessary.”) while lashing out at the at those who found their approach mathematically lacking (“I tried to read some of the rigorous papers. It’s easy to be rigorous if one doesn’t discuss the problem.”). The resistance Pauli encountered in New York initiated a (very gradual) change of heart, which we shall discuss in Sect. 4.2, where we will analyze in detail which aspects ultimately led to his strong rejection of nonlinear spinor theory. But first we will discuss the rejection that Pauli encountered and more generally the reception of Heisenberg’s ideas by the physics community in the spring of 1958.
4.1 General Reception
In his autobiography, Heisenberg recalled that he “did not like the idea of this encounter between Wolfgang in his present mood of great exaltation and the sober American pragmatists” (Heisenberg and Pomerans (translator) 1971, p. 234). And this notion that what Pauli encountered was primarily a rejection of their highly speculative and philosophical edifice by the new pragmatic American spirit of physics survives to this day. However, in his immediate report on the talk in a letter to Heisenberg written on the very same day, Pauli recounted that the local Americans had shown a wait and see attitude, and that the main opposition had come from a group of mathematical physicists from the Institute for Advanced Study (IAS) in Princeton, all of them European, two of them German: Freeman Dyson and the former Heisenberg associates Harry Lehmann and Wolfhart Zimmermann.
As we have seen, LSZ had long opposed Heisenberg’s theory, had indeed trained themselves in opposing Heisenberg’s castles in the air. Heisenberg had written a letter to Zimmermann, on 6 January, giving the basic outlines of his work with Pauli. On the basis of this letter, a meeting was convened at the IAS on 21 January 1958 at 4 pm in preparation for Pauli’s talk 11 days later.^{2} During the meeting Frank Yang had asked whether the HeisenbergPauli theory contained electrodynamics, to which Lehmann had replied that Heisenberg indeed thought it did, citing Heisenberg’s derivation of the fine structure constant together with Ascoli. This appears to have turned off Dyson, who remarked that his opinion on this was “unquotable”. Indeed, Heisenberg’s derivation of the fine structure constant appears to have reminded Dyson of home, that is Cambridge, where the great astronomer Arthur Eddington had turned to constructing speculative theories, relating cosmology with the structure of elementary particles, and trying to explain why the fine structure constant was the inverse of an integer. After his talk, Pauli attributed Dyson’s rejection to his “seeing and fearing ‘Eddingtonianism’ everywhere” (Letter to Heisenberg of 1 February 1958), a view that Dyson later confirmed. So, one objection of these young mathematical physicists was clearly to Heisenberg’s shady approximation methods and the undue trust he placed in the numerical predictions derived from them.
This notion that one couldn’t just tweak one aspect of relativistic quantum field theory (the positivedefiniteness of the metric) and hope to leave all of the other desirable features intact appears to have quite pleased Bohr, who famously chimed in, declaring after Pauli’s talk that Heisenberg’s setup simply wasn’t radical enough. However, it needs to be emphasized that all Zimmermann was voicing was an “opinion” and that there was no way to formally prove the impossibility of an indefinite metric; this was merely an “empirical statement”. Just as Heisenberg could not show that his theory worked, LSZ were definitely not able to show that it didn’t. It certainly didn’t fit into the axiomatic framework they were constructing, but it was on the other hand so close to that framework that it did not seem to provide a promising starting point for constructing a consistent novel QFT.All these discussions have only confirmed my opinion that it is highly probable that there is no field theory with an indefinite metric that also fulfils Lorentz invariance, unitarity of the SMatrix and macrocausality. [...] [It] appears to be a general empirical statement: Too weak modifications of the principles Lorentz invariance, microcausality, and positive definite metric make the theory worse rather than better.^{4}
Another member of the Columbia audience, Yale physicist Gregory Breit, felt compelled to write to Heisenberg, inquiring about possible experimental implications of Heisenberg’s theory:I have also received a letter from three young physicists from the Institute (Princeton), that stronghold of experts. They courageously confess to be interested in the PauliHeisenberg theory and ask me for preprints. All this indicates that, in spite of the experts’ violent objections, you have attained your principal objective which was to free the unprejudiced physicists from the current dogmas.^{5}
Even more enthusiastic was the support that Heisenberg received from the Soviet Union, which arrived in the form of a letter from Lev Landau, sent just ten days after Pauli’s Columbia talk. There is no indication that Landau had heard of Pauli’s talk; he had simply finally gotten around to reading Heisenberg’s earlier papers and was thrilled by Heisenberg’s approach of regularizing the commutation relations:I heard Pauli speak at Columbia at the beginning of February regarding your joint work [...] I am, of course, much impressed by your general philosophy and the new encouragement regarding the theory. I have been wondering as to whether there is any hope of seeing by means of it what modifications in quantum electrodynamics you might expect [...] Is it still to early to attempt such a detailed application?^{6}
Indeed in the late 1950s, there was quite some interest in Heisenberg’s nonlinear spinor theory in the Soviet Union: Already on 19 October 1957, Heisenberg had sent some reprints of his papers to Dmitrii Ivanenko of Moscow University, apparently in response to a request that is no longer extant. And from Dmitrii Blokhintsev of the Joint Institute for Nuclear Research Heisenberg received another request for preprints on 10 February 1958, stating that the “theoreticians in Dubna are very much interested in them.” Heisenberg’s folder with USSR correspondence (in which all of these letters are to be found) even contains a reply (dated 10 July 1958) to a (not extant) inquiry concerning the nonlinear spinor theory by J. I. Granovsky and A. Yanof of the Institute for Nuclear Physics in AlmaAta, Kazakhstan.We have recently thoroughly studied your papers. They are written in a rather difficult fashion, and up to now I had not brought up the courage for them. The brilliant idea of modifying the Green Function of the primary particles in the manner you suggest made a truly shocking impression. I think it can hardly be doubted that, whatever further difficulties might remain, your idea is the true, unexpected solution of the problem. The photon and the fine structure constant are particularly nice. I find the comparison with experiment absolutely satisfactory, given that it is still only a model.
I want to congratulate you from my heart on your brilliant successes. The old guard does not surrender!^{7}
And after Pauli’s talk, on 7 March, Weisskopf wrote to Pauli from CERN in Geneva:Pauli sees new ways to introduce representations of which we only have a smell. He[isenberg] is so attached to his earlier work that he tries to interpret it in that. Don’t apply Pauli’s enthusiasm to H’s letters.
And Pauli was well aware that it was now on him to decide the fate of the nonlinear spinor theory, especially within the physics community. In fact, Heisenberg’s grand claims had already traveled beyond the confines of academia and had reached the general public: a Göttingen colloquium talk held by Heisenberg on 24 February had received extensive news coverage, during which Heisenberg’s theory got the moniker Weltformel attached to it.^{9} Pauli reacted by circulating a now famous cartoon “to declare his independence (among physicists)”.^{10} The cartoon shows an empty square with the caption: “Comment on Heisenberg’s Radioadvertisment: This is to show the world that I can paint like Titian—Only technical details are missing.” In a letter to Rosbaud, Pauli further elaborated:In the meantime, Heisenberg has been here. [...] His talk and the following discussion were after all rather impressive. [...] When the group theoretical part was discussed, a number of difficulties were addressed, which Heisenberg did not answer well. [...] I only want to add that the independence of integer and halfinteger iso and regular spins was not explained by Heisenberg at all. He always tried to invoke several vacua, but that is a scam! If one introduces several vacua to get different particles, one is doing just as poorly as by introducing different fields. Where is the unified theory? One is back at the old number game with many fields. But probably there is something we (including Heisenberg!) don’t understand. Enlighten us.^{8}
While Pauli had sobered up, he had not yet fully made up his mind about the nonlinear spinor theory. It would take another month, during which he remained in the USA, now traveling to the West Coast, to Berkeley, for him to fully disavow himself from Heisenberg’s theory, thereby also sealing the fate of that theory for the physics community at large. We will study Pauli’s gradual disenchantment and its manifold reasons in the following section.My desire for “glory” is well satisfied; I just want a field of occupation that interests me (scientifically). For this purpose, H.’s optimism was useful, while certain groups of experts were boring me. In contrast, H.’s need for fame is insatiable. What does he want to compensate? Of course, he has inferiority complexes (as you emphasize in your letter). The hypothesis of the American physicists is that his trauma results from the fact that during the war he did not realize that one could produce plutonium. Incidentally, his theory of superconductivity also failed. Here [in the USA] I am told everywhere: “Nobody would believe H., if you were not in. [...]” I thus provide him with credibility.—Whether rightfully so, we shall see. [...] Up until now, I do not see that the program is impossible, as far as elementary particles + electromagnetism is concerned. The further development of the theoretical approaches will show whether it really works or if essential ideas are missing.^{11}
4.2 Pauli’s Turnaround
Pauli demanded that the fundamental equations of the theory, determining the energy eigenvalues of the Hamiltonian and thus the masses of the observed particles, should at least be defined and written down without making use of the TammDancoff approximation; and that there should be, at least in principle, a method for deriving, rather than postulating, the anticommutators. Their entire theory, Pauli feared, had not resolved any of the central issues of Heisenberg’s nonlinear spinor program (1 February):For you it is probably psychologically important that the new work is a continuation of your previous work. For me, on the other hand, that is a much less important aspect. I just want to have some mathematically well defined procedure.^{12}
Pauli had over the course of the preceding weeks, often remarked that he had “no intuition for those terrible TammDancoff approximations” (Letter from Pauli to Markus Fierz of 11 December 1957), and Heisenberg’s calculations were indeed not guided by strong intuitions, but rather by brute force, laborintensive calculations–it is no wonder that the paper that Heisenberg ended up publishing (after Pauli’s death) had five authors: Such calculations could only be performed with the postdoc manpower and the computational capacities that a large institute such as Heisenberg’s could provide. But ultimately the TammDancoff problem was not a dealbreaker. In his next letter to Heisenberg, of 10 February, Pauli merely demanded that the TammDancoff calculations be published in a separate paper, which he would not be cosigning:The mathematical foundations are not clarified, and one is still calculating with the old bad methods (are those even methods?).^{13}
Pauli was thus backing down on his strong demands about eliminating TammDancoff and was now merely asking to quarantine it. This was also because he saw that noone, especially not the sharpest critics of Heisenberg’s program, had anything better to offer with regard to the question of calculating energy eigenvalues in a nonperturbative theory. He made this point most strongly in a letter to Källén of 24 February 1958:[I]t is my wish that the TammDancoff calculations (to which I have anyhow not contributed) be sent out (and published) without my name. There I want to remain a mere observer! You need to separate them off!^{14}
and similarly in a letter to his assistant Charles Enz of 26 February:But, whatever the model may be (be it a spinor model or something else) one needs methods to treat an eigenvalue problem that are more decent than the ones we have now. [...] In my opinion also the Feldverein (LSZ) has utterly failed here. [...] Heisenberg is using the TammDancoff method for lack of something better.^{15}
Pauli thus, for a brief period, returned to the project. After all, all of the factors that had originally attracted him to Heisenberg’s proposal were still in place; it was just that he now had a more sober outlook on its (wellestablished) shortcomings. Indeed, Pauli felt that his new, more skeptical, approach might actually be beneficial to the research, writing to Heisenberg on 14 February: “Your optimism and my critical attitude is maybe a rather good combination.” But over the next weeks, Pauli began to have more substantial misgivings concerning the overall coherency of the theory, as he realized that there was a fundamental divide between Heisenberg and him concerning the role of the degenerate vacuum.The experts have the habit of assuring one that they themselves are not able to do it and that the others are not “rigorous”. If the Feldverein had produced something decent, Heisenberg would never have ended up using the TammDancoff methods (in which I personally do not believe).^{16}
It is not hard to see why Pauli had such high hopes for this conjecture: It would after all have tied together the symmetry structure of the model and the taming of the indefinite metric, thereby establishing an intimate connection between the consistency issues and the empirical adequacy of the theory. Heisenberg, however, never actually endorsed these ideas, and his initial response (on 2 January 1958) was hesitant at best:The following I have not yet calculated and state it rather as a conjecture, which still needs to be confirmed by further calculations. There will indeed be a decomposition into two distinct subspaces and one can probably arrange things so that [...] Q and N (baryon number) provide a “superselection rule” between the subspaces. [...] [I]n each subsystem the metric will be that of BleulerGupta [i.e., the metric of QED, which did involve negativenorm states, which could not, however, propagate to infinity and thus never induced negativeprobability transitions]^{18}
But Pauli was not yet deterred. To Källén he wrote on 10 January: “The conservation laws of Q and N ensure that [...] unitary S Matrices exist.” It was only when reevaluating the theory soberly after the shock of the New York talk that Pauli realized that Heisenberg had been pursuing a diametrically opposed path. Not only was Heisenberg not convinced that one could combine Hilbert Space structure and degenerate vacuum; he no longer believed even in the connection between the degenerate vacuum and the conservation laws. Indeed, already in a letter of 30 December 1957, which Pauli had essentially ignored, Heisenberg had distanced himself from his original idea of using the degenerate vacuum to restore the U(1) symmetry that was broken by the anticommutator:You are completely right that the doubling of the vacuum implies that the question of the Hilbert Space metric appears in a different light. However—as we already discussed on the phone—one must not connect this metric with Q and N [charge and baryon number] [...] Rather, the metric is only connected with the notion of probability.^{19}
But why then hold on to the idea of a degenerate vacuum at all? In the same letter in which Heisenberg disavowed himself from the idea of restoring symmetry through the degenerate vacuum, he had identified a new problem for the degenerate vacuum to resolve: the theory was lacking not just in symmetry structures, but also in representation content. This needs some unpacking. In quantum mechanics, one would determine the representation content of a theory (by which I mean the representations under which the states in the theory’s Hilbert Space transform under the relevant symmetry transformations), by finding the eigenbasis of some maximal set of commuting operators, one of them being the Hamiltonian. In perturbative QFT, on the other hand, where this was not an option, the representation content was instead determined (e.g., by LSZ) using the asymptotic condition: Asymptotically, any state approached a state composed of a number of free particles. Thus, a complete set of states (and their transformation properties) could be determined by taking all of the free field operators (or the corresponding creation operators) of the theory and acting with them on the vacuum state, to construct first one, then two, then general nparticle states.The quantity V in the commutation relations is after all at first something like, shall we say, an external electric field \(\mathbf {E}\) in the Hamiltonian of hydrogen. V disturbs the invariance with respect to the [PauliGürsey and chiral U(1) groups], just like \(\mathbf {E}\) disturbs the invariance under spatial rotations. Of course one can perform the corresponding transformation of \(\mathbf {E}\), i.e., also rotate \(\mathbf {E}\), then everything remains invariant. But this kind of invariance doesn’t guarantee any conservation laws [...]. In our case, one can of course also ‘rotate’ the V [...], i.e., perform the [...] transformation of the vacuum. But that also does not guarantee any conservation laws [...] So, for the time being, I don’t see the sources for the two strict conservation laws for Q and N at all.^{20}
Now even this construction method for the theory’s Hilbert Space was really not available to Heisenberg: In his theory all of the observed particles were actually strongly bound states, while the elementary \(\psi \) particles were not supposed to appear in asymptotic states at all. Still, Heisenberg expected to construct his Hilbert Space (and thus determine the representation content) in almost full analogy to weakcoupling QFT with an asymptotic condition: States were to be constructed by acting on the vacuum with (a power series in) \(\psi ^{\dagger }\) and \(\psi \), the only difference to QED being that one did not renormalize in order to make \(\psi ^{\dagger } \left 0\right\rangle \) into a oneparticle state. Rather, even the oneparticle states were obtained by acting on the vacuum with complicated polynomials in \(\psi \). Combining this view of the Hilbert Space with the theory Pauli and he had constructed, Heisenberg noticed that there was an issue: by identifying the PauliGürsey group (which acted on the spin degrees of freedom) with the isospin, one was inextricably linking spin with isospin, implying that there could be no particles with spin but without isospin or vice versa, in blatant contradiction to the empirically established spectrum of hadrons, which contained both a spin 1/2 particle without isospin (the \(\Lambda \) baryon), and scalar particles with isospin (such as the K mesons). What was even worse, once one had interpreted the state whose leading term was \(\psi ^{\dagger } \left 0\right\rangle \) as the proton (the corresponding antiparticle being the antineutron, due to the identification of PauliGürsey and isospin) it was fundamentally unclear which state to identify with the neutron. The degenerate vacuum offered a whole host of new possibilities here; in particular one could simply identify \(\psi ^{\dagger } \left \Omega _1\right\rangle \) with the proton and \(\psi ^{\dagger } \left \Omega _2\right\rangle \) with the neutron, thereby effectively identifying the vacuum as an isospin doublet. This further implied that a state could get its isospin either from the field operators or from the vacuum those operators acted on, thereby eliminating the overly restrictive spinisospin correlation.
And in a letter to his assistant Charles Enz, who had remained in Zurich, Pauli wrote on 22 March:The root of the problem appears to me to be this. The idea, which we already discussed during the dinner in Zurich, namely: a general bisection of the world [...]—that this more general idea, in close connection with the division of the Hilbert Space into I and II could hardly be developed: It was in connection with this general idea that I then situated also the degeneracy of the vacuum, which is now just appended as a trick to obtain the combination of halfinteger isospin (regular spin) with integer regular spin (isospin).^{21}
To make matters worse, Pauli soon realized that the degeneracy of the vacuum didn’t even work as a formal trick, because it could not properly reproduce twoparticle states. Pauli outlined his misgivings in a letter to Heisenberg’s new assistant Dürr of 26 March 1958. The problem in a nutshell is the following: If indeed \(\psi ^{\dagger } \left \Omega _1\right\rangle \) is the proton and \(\psi ^{\dagger } \left \Omega _2\right\rangle \) is the neutron, then \(\psi ^{\dagger } \psi ^{\dagger } \left \Omega _1\right\rangle \) should be two protons and \(\psi ^{\dagger } \psi ^{\dagger } \left \Omega _2\right\rangle \) should be two neutrons. But there is no state that can be identified with, say, deuterium, i.e., a state with one proton and one neutron.The entire spinor model of Heisenberg and myself appears to me to be losing its attractiveness as I get to know it better (the indefinite metric is not integrated as nicely as I had hoped), but we will probably have to publish something after all, so that nobody has to repeat the same considerations.^{22}

He no longer believed that Heisenberg’s proposal could be empirically adequate, because the degenerate vacuum was an insufficient tool for reproducing the known particle spectrum from a single underlying field. There was “a discrepancy between the mathematical possibilities [...] and the physical facts.”

He disagreed with Heisenberg’s (TammDancoff) methods for extracting quantitative predictions from the theory.

He no longer believed that Heisenberg’s treatment of the indefinite metric was mathematically consistent, and that Heisenberg had, after winning the Battle of Ascona, decided to rest on his laurels and to no longer think about the indefinite metric at all.
His invectives against LSZ and the QFT “experts” in general were actually far more severe and elaborate than those against Heisenberg. Consider, the following assessment of Symanzik, the only member of LSZ who was still in Göttingen, which Pauli gave in a letter to Bruno Touschek on April 14 (PSC IVIV):But let me first make some supplementary negative remarks about LSZ [...] I never thought of holding it against them that they do not join Heisenberg and his ideas. (To the contrary, I always very much liked this strength of character in LSZ, and it was even one of the reasons for me to so strongly recommend Lehmann for the professorship in Hamburg.) But what I do hold against them is that they do not have any initiative or path of their own^{24}
On 29 April, in a letter to Enz (PSC IVIV), he dismissed the entire axiomatic approach^{26} to QFT: “[A]xioms are at best useful in a finished theory (and even then their value is questionable); but never do they point to a good way towards further development.” An assessment that was supplemented by a couplet, which I feel unable to translate and will leave to the Germanspeaking reader to enjoy:[H]e appears to have reached a position where “rigor” becomes identical with intellectual nihilism! Well, since no (physically relevant) statements on quantized field theories can be proven or disproven, it becomes permissible to make arbitrary claims. It is therefore easy for him to live in “rigor” next to Heisenberg in the same institute!^{25}
Also, Pauli was still interested in central elements of the framework he had developed with Heisenberg, such as the indefinite metric, the identification of the PauliGürsey group with isospin, or the regularized commutators. He merely believed that Heisenberg’s radical monism had failed and that the only way to incorporate these features in a model was to introduce more than one field (and ditch the degenerate vacuum): “Farewell unified quantum field theory!”, as he wrote to Fierz.^{27}Süß wie Honig entströmt dem Mund der Experten der Humbug.
Fern vom Grund der Physik scheint er als bläulicher Dunst.
This blinkandyoumissit Nazi reference^{30} to LSZ is, however, to be contrasted with the crass wording Pauli used when talking about Heisenberg in a letter to Paul Rosbaud of 18 April:[T]he theoretical physicists of that eternally divided people [the Germans] are divided into two halves: one (Heisenberg) has betrayed mathematics, the other (LSZ and followers [Mitläufer]) has betrayed physics.^{29}
Even if Pauli believed that Heisenberg’s research program might still actually be more fruitful than that of axiomatic QFT, at least LSZ were not trying to coopt him. While motivated by clear intellectual differences, the forcefulness of Pauli’s break with Heisenberg was certainly driven by his desire to distance himself from Heisenberg, who was scientifically and politically compromised, in the perception of both the physics community and the general public. So, on the day after breaking with Heisenberg (8 April), he wrote a short memo, which he sent out to a large number of physicists, announcing that the publication was off and giving reasons, though merely emphasizing the first point above (the impossibility of reproducing the elementary particle spectrum with just one field), which was the central point in which Pauli went beyond the criticism he had received from the “experts” in New York.You have characterized Heisenberg well with “...trying to force things that can’t be forced.” “And if thou’rt unwilling, then force I’ll employ!” is also very German. Göring once called this “icecold realism.” Icecold it was indeed, but the Germans tend to have difficulties with the “realism.” Is it realism to want to conquer the entire world, without seeing how one should actually be able to do that as so small a country? Or a different formulation, which I recently wrote to Weisskopf—in connection with H.—: “German endeavors have a marked tendency to end in a Twilight of the Gods, with the Rhinegold sinking back into the river!”^{31}
Similarly, Pauli also tried to distance himself from Heisenberg in the eyes of the general public. The first press reports on Heisenberg’s Weltformel had been based on a Göttingen colloquium talk for an audience of experts, and Heisenberg had insisted that it had not been his intention to get the press involved. Pauli’s reaction to the hullabaloo, which also led some journalists to contact him in the US, was to give curt answers and dampen the enthusiasm (Rettig 2014, p. 170). Heisenberg’s reaction was different: he was not so much angered at the media interest per se, but rather by the fact that they were getting it wrong, publishing “dreadful nonsense” (haarsträubenden Unsinn).^{32} And he knew the perfect venue to set the record straight.
From the published version of the talk (Heisenberg 1958), it appears that Heisenberg indeed did not mention Pauli as a collaborator, only as a source. But he most certainly did not comply with von Laue’s request; indeed, the Planck centennial is now mainly remembered for Heisenberg’s presentation of his theory. In any case, Heisenberg’s attempts at placating Pauli were unsuccessful, as were Dürr’s attempts at repairing the defects that Pauli had pointed out: Dürr first tried to pitch an infinitely (rather than doubly) degenerate vacuum (letter to Pauli of 7 April 1958), but in the paper ultimately published by Heisenberg, Dürr, Mitter, Schlieder and Kazuo Yamazaki, a Japanese Alexander von Humboldt scholar, Pauli’s point was conceded (Dürr et al. 1959, p. 443):It will be easy for you to mention my contribution in a footnote of a paper authored solely by yourself. What is much more important to me is that my name is not written above a thing that I can no longer take responsibility for, and I want to urgently ask you to also take this into account in your lecture at the Planck celebration.^{36}
Pauli publicly announced his break with Heisenberg at the 8th Rochester Conference, the first one in the series to be held in Europe, in Geneva, from 30 June to 5 July 1958. This was the first time that Heisenberg and Pauli met after the breakup and Pauli’s return to Europe. Pauli, who had chaired the session in which Heisenberg presented, described the events in a letter to Fierz on 9 July:One will thus need at least one more continuous oneparameter group [...] in the mentioned sketch of Pauli and one of the authors it was attempted to get these groups through a doubling of the vectors in Hilbert Space, in particular the vacuum. [...] [T]his no longer appears to be possible to us.^{37}
When Pauli and Heisenberg met again shortly afterwards at the summer school in Varenna, 21 July to 9 August 1958, Heisenberg felt that Pauli had again softened toward their theory, an impression he even stated in his autobiography (Heisenberg 1969, p. 319). But Pauli was insistent that this was not the case.^{39} And it is safe to say that while Pauli’s break with Heisenberg may not have made an overly strong impression on the general public (with Pauli’s name hardly being mentioned in the press anyway), it certainly defined the physics community’s view of Heisenberg’s theory to this day, where the nonlinear spinor theory, if it is known at all, is merely an amusing, if lamentable anecdote. To quote one final letter, to Paul Rosbaud on 3 May 1958, where Pauli wrote:[T]he session I chaired was as satisfactory as could be expected, and I rather enjoyed it. My only concerns were to
(a) prevent that Heisenberg keeps on going around telling everybody that I agree with him (b) move the attention and the interest away from the spinor model to the more general problem of the indefinite metric [...] Both goals were fully achieved (for the future I now plan to no longer take part in the discussions after Heisenberg’s talks. [...])^{38}
This universal rejection of Heisenberg’s theory was given further emphasis by the story’s final tragic twist. On 21 November 1958, Pauli was in Hamburg to receive an honorary doctorate. Bert Schroer,^{41} then a young PhD student of Harry Lehmann, remembers how Pauli sank into a swivel chair (known as the Pauli armchair, as it had already been at Hamburg University back when Pauli was a professor there), visibly exhausted, exclaiming: “I am having a hard time digesting that Heisenberg.” (Der Heisenberg liegt mir schwer im Magen). Five weeks later, Pauli was dead of pancreatic cancer. Killing Pauli was the final sin of Heisenberg’s Weltformel.So it appears to me that the whole unpleasant journalism will fall back badly on H. For in his case it is not relevant what the average newspaper reader thinks, but rather what the physicists think. I have the impression, H. is entirely isolated even among the German physicists (not to speak of the rest of the world; basically all physicists know that I will not be publishing with him) (also Landau!).^{40}
Footnotes
 1.
As recounted in a letter to Viktor Weisskopf of 16 February 1958.
 2.
Our knowledge of this preparatory meeting is based on handwritten notes taken during the meeting by Dieter Brill, who was then a graduate student at Princeton University (DBP). I want to thank him once more for making them available to me.
 3.
That Lehmann’s main objection to Heisenberg’s theory was the use of the indefinite metric was also independently confirmed to me in private communication by Bert Schroer, who started his PhD work with Lehmann in 1958.
 4.
Alle diese Diskussion haben mich in meiner Meinung bestärkt, dass es sehr wahrscheinlich keine Feldtheorie mit indefiniter Metrk gibt, die Lorentzinvarianz, Unitarität der SMatrix und Makrokausalität erfüllt. [...] [Es] scheint offenbar ein allgemeiner Erfahrungssatz zu sein: Zu schwache Änderungen der Prinzipien Lorentzinvarianz, Mikrokausalität und positiv definiter Metrik machen die Theorie eher schlimmer als besser. Letter from Zimmermann to Heisenberg, 16 June 1958, WZP, Heisenberg Correspondence.
 5.
Gürsey to Pauli, 6 February 1958, PSC IVIV.
 6.
Letter from Breit to Heisenberg, 7 March 1958, GBP, Heisenberg Correspondence.
 7.
Wir haben in letzter Zeit Ihre Arbeiten gründlich studiert. Sie sind ziemlich schwierig geschrieben, und ich habe früher nicht den Mut dazu aufbringen können. Die glänzende Idee, die Green’sche Funktion der primären Teilchen in der von Ihnen vorgeschlagenen Weise zu ändern, macht einen wahrhaft erschütternden Eindruck. Ich glaube, dass es kaum noch zu bezweifeln ist, welche Schwierigkeiten auch noch zu beseitigen wären, dass Ihre Idee die wirkliche, unerwartete Lösung des Problems darstellt. Besonders schön ist das Photon und die Feinstrukturkonstante. Ich finde den Vergleich mit dem Experiment vollkommen befriedigend, da es sich ja bis jetzt nur um ein Modell handelte.
Ich möchte Ihnen mit [sic] Ihren glänzenden Erfolgen herzlichst gratulieren. Die alte Garde kapituliert nicht! Letter from Landau to Heisenberg of 11 February 1958, WHP, Folder 1864.
 8.
Inzwischen war Heisenberg hier. [...] Sein Vortrag und die darauf folgende Diskussion waren doch recht eindrucksvoll. [...] Als der Gruppenteil diskutiert wurde, kamen allerlei Schwierigkeiten zur Sprache, die Heisenberg nicht gut beantwortete. [...] Ich möchte nur hinzufügen, dass die unabhängigen ganz und halbzahligen Iso und Normalspins von Heisenberg gar nicht erklärt wurden. Er redete sich immer auf verschiedene Vakua heraus, aber das ist doch ein Schwindel! Wenn man verschiedene Vakua einführt um verschiedene Teilchen zu kriegen, dann ist man doch genauso schlecht dran, wie man mit verschiedenen Feldern wäre. Wo bleibt dann die unified theory? Dann ist man doch wieder beim alten number game mit vielen Feldern. Aber wahrscheinlich verstehen wir da was nicht (inklusive Heisenberg!) Klären Sie uns auf.
 9.
For overviews of the press coverage of Heisenberg’s theory, see von Meyenn (2005, pp. 989–994) and Rettig (2014, Sects. 15.3.4 and 15.3.6.1).
 10.
Um meine Unabhängigkeit (unter Physikern) zu deklarieren, letter to Paul Rosbaud of 5 March 1958, PRP, PauliRosbaud Correspondence.
 11.
Mein Bedürfnis nach ‘glory’ ist ja gedeckt, ich will nur ein Betätigungsfeld haben, das mich (wissenschaftlich) interessiert. Hierzu war mir H.s Optimismus günstig, während gewisse ExpertenGruppen mich gelangweilt haben. Dagegen ist H.s RuhmBedürfnis \(\underline{uners\ddot{a}ttlich}\). Was will er damit kompensieren? Natürlich sind Minderwertigkeitskomplexe bei ihm vorhanden (wie Sie in Ihrem Brief betonen). Die Hypothese der amerikanischen Physiker ist, sein Trauma rühre daher, dass ihm während des Krieges \(\underline{nicht}\) eingefallen ist, dass man Plutonium herstellen könne. Übrigens ist ihm auch die Theorie der Supraleitung schief gegangen. Überall höre ich hier: “‘Nobody would believe H., if you were not in. [...]” Ich helfe ihm also zu \(\underline{Kreditf\ddot{a}higkeit}\) —Ob mit recht, wird sich zeigen. [...] Bis jetzt sehe ich keine Unmöglichkeit des Programmes, was Elementarteilchen + Elektromagnetismus betrifft. Ob es wirklich geht oder ob noch wesentliche Ideen fehlen, muss eben die Weiterentwicklung der theoretischen Ansätze zeigen. Letter from Pauli to Rosbaud of 5 March 1958, PRP, PauliRosbaud Correspondence.
 12.
Für dich ist es wohl psychologisch wichtig, dass die neue Arbeit eine Fortsetzung Deiner früheren Arbeiten ist. Für mich dagegen ist das sachlich viel weniger relevant. Ich will nur irgendein mathematisch wohldefiniertes Verfahren haben.
 13.
Eine Klärung der mathematischen Grundlage erfolgt aber nicht, es wird nur mit den alten schlechten Methoden (sind das überhaupt Methoden?) weitergerechnet. Letter to Heisenberg of 1 February 1958, PSC IVIV.
 14.
[E]s ist mein Wunsch, dass die TammDancoffRechnungen (zu denen ich ja auch nichts beitrug) ohne meinen Namen verschickt werden (und erscheinen). Ich möchte bei diesen bloßer Zuschauer bleiben! Ihr sollt diesen abtrennen! PSC IVIV.
 15.
Aber, was immer das Modell sein mag (ob Spinormodell oder etwas anderes) man braucht anständigere Methoden, um ein Eigenwertproblem zu behandeln, als diejenigen, welche jetzt vorliegen. [...] Es ist meine Meinung, dass auch der Feldverein (LSZ) hier ganz versagt hat. [...] Heisenberg wendet da in Ermangelung eines Besseren die TammDancoffMethode an.
 16.
Die Experten pflegen einem da nur zu versichern, dass sie selbst das nicht können und dass die anderen nicht “streng” sind. Hätte da der Feldverein etwas Anständiges produziert, so wäre Heisenberg nie dazu gekommen, die TammDancoffMethode anzuwenden (an die ich persönlich \(\underline{nicht}\) glaube). PSC IVIV.
 17.
This evaluation, and the quote below, is to be found a long letter written to Heisenberg on Christmas over the course of three days, from 25 to 27 December 1957, PSC IVIV.
 18.
Das folgende habe ich noch nicht gerechnet und sage es mehr als \(\underline{Vermutung}\), die noch durch weitere Rechnungen bestätigt werden muss. Es wird in der Tat ein Zerfall in zwei getrennte Termsysteme herauskommen, und zwar kann man es wohl so einrichten, dass [...] Q und N (Baryonenzahl) für eine ’Superselectionrule’ zwischen den Termen sorgen. [...] in jedem Teilsystem die Metrik die BleulerGuptasche ist...
 19.
Du hast völlig recht damit, dass durch die Verdopplung des Vakuums die Frage nach der Metrik im Hilbertraum in einer neuen Weise gestellt wird. Aber—darüber sprachen wir ja schon am Telefon—man darf diese Metrik \(\underline{nicht}\) in Verbindung bringen mit Q und N [...] Vielmehr hat die Metrik allein mit dem Wahrscheinlichkeitsbegriff zu tun. Letter from Heisenberg to Pauli of 2 January 1958, PSC IVIV.
 20.
Die Größe V in den VertauschungsRelationen ist doch zunächst so etwas wie, sagen wir, ein äußeres elektrisches Feld \(\mathbf {E}\) in der Hamiltonfunktion des Wasserstoffs. V stört die Invarianz gegen die Gruppen (A) und (B) ähnlich wie \(\mathbf {E}\) die Invarianz gegen Raumdrehungen stört. Natürlich kann man eine entsprechende Transformation von \(\mathbf {E}\) vornehmen, d.h. \(\mathbf {E}\) mitdrehen, dann bleibt wieder alles invariant. Aber \(\underline{diese}\) Art der Invarianz garantiert ja keine Erhaltungssätze [...]. In unserem Fall kann man natürlich auch V [...] ‘mitdrehen’, d.h. die [...] Transformationen am Vakuum ausführen. Aber auch das garantiert keine Erhaltungssätze [...] Ich sehe also zunächst überhaupt noch nicht die Quellen für die beiden strengen Erhaltungssätze für Q und N.
 21.
Die Wurzel des Übels scheint mir \(\underline{die}\) zu sein, dass die Idee, die wir bereits bei jenem Abendessen in Zürich besprochen haben, nämlich: eine \(\underline{allgemeinere}\) Zweiteilung der Welt [...]—dass diese \(\underline{allgemeinere}\) Idee, in enger Verbindung mit der Teilung des Hilbertraumes in I und II, sich so wenig entwickeln liess: In Verbindung mit dieser allgemeinen Idee habe ich mir damals auch die Entartung des Vakuums gedacht, die jetzt nur als Trick angehängt ist, um das Zusammengehen von halbzahligem Isospin (gewöhnlichem Spin) mit ganzzahligem gewöhnlichen Spin (Isospin) herauszubekommen.
 22.
Das ganze Spinormodell von Heisenberg und mir scheint mir bei näherer Bekanntschaft stark an Reiz zu verlieren (die indefinite Metrik ist nicht so schön eingearbeitet, wie ich gehofft habe), aber wir werden wohl doch etwas publizieren müssen, damit nicht andere noch einmal dieselben Überlegungen wiederholen müssen.
 23.
Letter to Markus Fierz of 6 April, PSC IVIV.
 24.
Vorerst aber noch ergänzende negative Bemerkungen über LSZ [...] Nie habe ich daran gedacht, diesen vorzuwerfen, dass sie sich Heisenberg und seinen Ideen nicht anschliessen. (Im Gegenteil hat mir diese Charakterfestigkeit bei LSZ stets gut gefallen, und sie war sogar mit ein Anlass, dass ich Lehmann so sehr für die Professur in Hamburg empfohlen habe.) Dagegen \(\underline{werfe\,ich\,ihnen\, vor, \, dass \, sie \, keine \, eigene \, Initiative \, bzw. \, keinen \, eigenen \, Weg \, haben.}\) Letter to Fierz of 6 April 1958, PSC IVIV.
 25.
...er mir einen Standpunkt erreicht zu haben scheint, bei dem \({\underline{``Strenge{\text {''}}\,\,mit\,\,geistigem\,\,Nihilismus\,\,identisch\,\, wird!}}\) Nun ja, da keine (für die Physik relevanten) Behauptungen in quantisierten Feldtheorien bewiesen oder widerlegt werden können, ist es eben erlaubt, darüber beliebige Behauptungen aufzustellen. Leicht ist es ihm daher in “Strenge” neben Heisenberg im gleichen Institut zu leben.
 26.
“Opportunistic Axiomatics” as Stoelzner (2001) has called it.
 27.
Adieu, einheitliche Quantenfeldtheorie! Letter of 6 April, PSC IVIV.
 28.
It should be noted that, at least in private, LSZ were willing to admit to the dismal state of and the dismal prospects for axiomatic QFT. In an interview conducted by Dieter Hoffmann and Ingo Peschel [Max Planck Institute for the History of Science, Preprint 485], Peter Fulde, who was a student at the University of Hamburg in the late 1950s, remembers expressing his interest in doing a PhD thesis in field theory and being rebuffed by Lehmann: “Field Theory? They’re stuck. Axiomatic field theory is stuck. The best mathematicians are trying their hardest, but no progress is being made.” [Die besten Mathematiker beissen sich die Zähne aus, aber es geht nicht weiter.]
 29.
Letter to Fierz of 6 April; [...] die theoretische Physiker jenes stets gespaltenen Volkes in zwei Hälften zerfallen, von denen die eine (Heisenberg) die Mathematik, die andere (LSZ und Mitläufer) die Physik verraten haben.
 30.
In postwar denazification, Mitläufer was the term used for people who were not directly tied to war crimes, but still had sufficient ties to the Nazi regime not to be exonerated.
 31.
Heisenberg haben Sie gut charakterisiert mit dem “...versuchen, Sachen zu erzwingen, die man nicht erzwingen kann” “Und gehst [sic] du nicht willig so brauch’ ich Gewalt!” ist auch sehr deutsch. Göring nannte das einmal “eiskalten Realismus”. Eiskalt war es schon, aber mit dem “Realismus” pflegt es dann zu hapern bei den Deutschen. Ist es Realismus, die ganze Welt erobern zu wollen, ohne zu sehen, wie man das eigentlich als ein so kleines Land machen könnte? Oder eine andere Formulierung, die ich kürzlich—im Zusammenhang mit H.—an Weisskopf schrieb: “deutsche Unternehmungen haben eine ausgesprochene Tendenz mit einer Götterdämmerung zu enden, wobei dann das Rheingold wieder im Fluss versinkt!”.
 32.
Letter to Pauli of 5 March 1958, PSC IVIV.
 33.
On the historical background of this event, in particular in the context of German separation, see Hoffmann (1996). I would like to thank Dieter Hoffmann for making the official program of the Planck centennial available to me.
 34.
seine neue mathematisch formulierte Theorie zur Erklärung des modernen physikalischen Weltbildes am 25. April in Berlin ausführlich und allgemein verständlich erläutern. This article from the Welt is reproduced in Eckert (2000).
 35.
Letter from von Laue to Pauli of 18 March 1958, PSC IVIV.
 36.
Du kannst sehr leicht meinen Anteil in einer Arbeit von Dir allein in einer Fußnote anmerken. Viel wichtiger als dies ist mir, dass mein Name nicht über einer Sache steht, die ich nicht mehr verantworten kann, und ich möchte Dich dringend bitten, dies auch in Deinem Vortrag bei der Planckfeier zu berücksichtigen.
 37.
Man wird daher noch mindestens eine kontinuierliche einparametrige Gruppe [...] brauchen [...] In dem erwähnten Entwurf von Pauli und einem der Verfasser war versucht worden, diese Gruppen durch eine Verdopplung der Vektoren im HilbertRaum, insbesondere des Vakuums, zu gewinnen. Dies erscheint uns aber [...] nicht mehr möglich. The authors did, however, realize that the symmetry breaking scheme, which was still being upheld, in itself already implied a degeneracy of the vacuum, and the concept of breaking symmetries with a degenerate vacuum, inspired by Heisenberg, would go on to have a glorious career in particle physics (Borrelli 2015b).
 38.
war mir die von mir präsidierte Sitzung soweit \(\underline{befriedigend}\), als dies zu erwarten war, und ich fühlte mich \(\underline{recht \, wohl}\) dabei. Es handelte sich nur darum, (a) zu verhindern, dass Heisenberg weiter herumläuft und allen Leuten erzählt, ich sei mit ihm einverstanden, (b) die Aufmerksamkeit und das Interesse vom Spinormodell weg auf das allgemeinere Problem der indefiniten Metrik [...] zu lenken. Beide Zwecke wurden voll erreicht (in Zukunft beabsichtige ich nun nicht mehr, in Diskussionen nach Heisenbergs Vorträgen zu sprechen. [...]).
 39.
See the discussion of the Varenna meeting in von Meyenn (2005, pp. 1238–1239).
 40.
So scheint es mir, dass die ganze üble Publizistik schließlich schlimm auf H. zurückfallen wird. Denn in seinem Fall kommt es \(\underline{nicht}\) darauf an, was der DurchschnittsZeitungsleser meint, sondern was die Physiker meinen. Ich habe den Eindruck, H. ist auch in Deutschland unter den Physikern ganz isoliert (von der übrigen Welt ganz zu schweigen; dass ich nicht mit ihm publiziere, wissen praktisch \(\underline{alle}\) Physiker)(auch Landau!).
 41.
eMail from Bert Schroer to the author of 22 February 2018.