Werner Heisenberg - Session X

Notice: We are in the process of migrating Oral History Interview metadata to this new version of our website.

During this migration, the following fields associated with interviews may be incomplete: Institutions, Additional Persons, and Subjects. Our Browse Subjects feature is also affected by this migration.

We encourage researchers to utilize the full-text search on this page to navigate our oral histories or to use our catalog to locate oral history interviews by keyword.

Please contact [email protected] with any feedback.

ORAL HISTORIES
Image not available
Interviewed by
Thomas S. Kuhn
Interview date
Location
Munich, Germany
Usage Information and Disclaimer
Disclaimer text

This transcript may not be quoted, reproduced or redistributed in whole or in part by any means except with the written permission of the American Institute of Physics.

This transcript is based on a tape-recorded interview deposited at the Center for History of Physics of the American Institute of Physics. The AIP's interviews have generally been transcribed from tape, edited by the interviewer for clarity, and then further edited by the interviewee. If this interview is important to you, you should consult earlier versions of the transcript or listen to the original tape. For many interviews, the AIP retains substantial files with further information about the interviewee and the interview itself. Please contact us for information about accessing these materials.

Please bear in mind that: 1) This material is a transcript of the spoken word rather than a literary product; 2) An interview must be read with the awareness that different people's memories about an event will often differ, and that memories can change with time for many reasons including subsequent experiences, interactions with others, and one's feelings about an event. Disclaimer: This transcript was scanned from a typescript, introducing occasional spelling errors. The original typescript is available.

Preferred citation

In footnotes or endnotes please cite AIP interviews like this:

Interview of Werner Heisenberg by Thomas S. Kuhn on 1963 February 28,
Niels Bohr Library & Archives, American Institute of Physics,
College Park, MD USA,
www.aip.org/history-programs/niels-bohr-library/oral-histories/4661-10

For multiple citations, "AIP" is the preferred abbreviation for the location.

 

Abstract

This interview was conducted as part of the Archives for the History of Quantum Physics project, which includes tapes and transcripts of oral history interviews conducted with ca. 100 atomic and quantum physicists. Subjects discuss their family backgrounds, how they became interested in physics, their educations, people who influenced them, their careers including social influences on the conditions of research, and the state of atomic, nuclear, and quantum physics during the period in which they worked. Discussions of scientific matters relate to work that was done between approximately 1900 and 1930, with an emphasis on the discovery and interpretations of quantum mechanics in the 1920s. Also prominently mentioned are: Guido Beck, Richard Becker, Patrick Maynard Stuart Blackett, Harald Bohr, Niels Henrik David Bohr, Max Born, Gregory Breit, Burrau, Constantin Caratheodory, Geoffrey Chew, Arthur Compton, Richard Courant, Charles Galton Darwin, Peter Josef William Debye, David Mathias Dennison, Paul Adrien Maurice Dirac, Dopel, Drude (Paul's son), Paul Drude, Paul Ehrenfest, Albert Einstein, Walter M. Elsasser, Enrico Fermi, Richard Feynman, John Stuart Foster, Ralph Fowler, James Franck, Walther Gerlach, Walter Gordon, Hans August Georg Grimm, Wilhelm Hanle, G. H. Hardy, Karl Ferdinand Herzfeld, David Hilbert, Helmut Honl, Heinz Hopf, Friedrich Hund, Ernst Pascual Jordan, Oskar Benjamin Klein, Walter Kossel, Hendrik Anthony Kramers, Adolph Kratzer, Ralph de Laer Kronig, Rudolf Walther Ladenburg, Alfred Lande, Wilhelm Lenz, Frederic Lindemann (Viscount Cherwell), Mrs. Maar, Majorana (father), Ettore Majorana, Fritz Noether, J. Robert Oppenheimer, Franca Pauli, Wolfgang Pauli, Robert Wichard Pohl, Arthur Pringsheim, Ramanujan, A. Rosenthal, Adalbert Wojciech Rubinowicz, Carl Runge, R. Sauer, Erwin Schrodiner, Selmeyer, Hermann Senftleben, John Clarke Slater, Arnold Sommerfeld, Johannes Stark, Otto Stern, Tllmien, B. L. van der Waerden, John Hasbrouck Van Vleck, Woldemar Voigt, John Von Neumann, A. Voss, Victor Frederick Weisskopf, H. Welker, Gregor Wentzel, Wilhelm Wien, Eugene Paul Wigner; Como Conference, Kapitsa Club, Kobenhavns Universitet, Solvay Congress (1927), Solvay Congress (1962), Universitat GGottingen, Universitat Leipzig, Universitat Munchen, and University of Chicago.

Transcript

Heisenberg:

... Nobody can hope that he can just easily change over from one tradition, to another one, and to a third one. Probably it wouldn't even be very agreeable to live if it were so easy to change over.

Kuhn:

Well, I deeply agree with that. ... But to put the tension as man's rational versus his irrational potentiality is to say that there are at least two realms. This I find very difficult. It seems to me one can say a great deal, and must say a great deal, about the nature of tradition, the essential role of tradition in science and other activities also, without trying to polarize the rational versus the irrational in quite this way.

Heisenberg:

Well, of course, it may be an over-sinplification to say there are just two poles. The pattern may be still more complicated. I would like to say that I sometimes see it in a same sense as one could see the relation between say, classical physics and chemistry or the relation between quantum theory and biology. I mean it in the following sense. When we speak about classical physics and chemistry — that is our dilemma of the twenties about which we spoke, so long as one started from the concepts of classical physics, one found almost everything fitted well together. At least whenever one could check, in some way it always came out alright. If, for instance, you studied the collisions between elementary particles, you would think, "Well, now that's something quite far away from classical physics." Still, when you checked the conservation laws, the laws of elastic collision, then everything fitted well. It was very difficult to see why this field of classical physics would not work everywhere. It is the same sense, I would say, at present when we discuss reality from the point of view of quantum theory. It is very difficult to see why this should not be sufficient to explain the organism. In fact, there are many biologists who would claim, "Well, that is everything, and you can calculate the human brain finally from some kind of theoretical equation. There is nothing else but just our physical chemistry." Still, I think everyone of us has the feeling that, for instance, something like consciousness does never come out from any kind of unified field equation or whatever else. Consciousness is something of a different plane. In the same way, I would say our rational approach to the world and to nature, to everything, is one approach which is perfectly consistent, so when you stay within it you never come out of it. In some way it's extremely difficult even to say that there is something which is not rational. Still, I feel that in some way the whole thing becomes more natural if one allows oneself to say that there are many things which cannot be copied by rational thinking. Just in the same way I feel that it's a simplification to say that beside physics and chemistry there is also biology. We are perfectly well entitled to use terms like life or human mind or consciousness and so on even if these terms do not occur in physics and in chemistry. Only at a much later stage we will probably be able to understand in which relation these two sets of terms are. Just as we have in the twenties understood how those qualitative concepts of chemistry relate to the concepts of pure physics. Finally, we have found a kind of unification so we can understand that both can be part of the system. Well, have I made myself understood?

Kuhn:

We must not go on with this. [Laughter] I think what you're saying is extremely clear and makes very good sense. If I understand what Pauli is saying this is not entirely a good expression of it. You would say that one can go this far this way, but then in order to get into other fields, we must add something, and addition out of which we may later make a greater synthesis. For one thing, I take it that what Pauli is insisting on is that although we have done our best to ignore the irrational elements, one need only to go back to Kepler, or just look at our present situation, to see that all of this time we have been dependent on elements which we have never admitted were there, but that have been there all the time.

Heisenberg:

Yes, that is exactly Pauli's position, yes.

Kuhn:

Now to say that is to say something rather different from the statement that we can go further by adding these additional ements.

Heisenberg:

Well, is it so different? I don't see why you think they are two different positions. In both cases, one says that there is something else besides that which we can cover by our rational thinking. Well, Pauli says that we actually have ignored the other part but we have lived from it. Well, in some way, you may say that any growth of science is just due to the fact that we've penetrated this rational thinking into other realms, but by doing so, we expand that which we may call rational. I've always liked a formulation as follows. We will never find an end to our attexpt to understand nature. The trouble is that only by doing so we learn again and again and afresh, what the word "understanding" means. The word "understanding" means in the sixteenth century something which is quite different from its meaning in the twentieth century and probably in the twenty-fifth century it will again mean something quite different from now.

Kuhn:

What you have just said I could subscribe to in every way, and would be delighted to have said it myself. I find it possible to talk about the nature of traditions by which these different ways of understanding come into being without trying to separate out what is the rational on the one hand from the "Urmythisch" of the western race on the other. I don't find that it contributes to my understanding of what we have just agreed to look for it in this reservoir of human perception that is somehow there by virtue of our being human. I certainly don't mean that because I feel some discomfort with this way I am then reduced to having to say, "We observe, and then we construct objective pictures of what we have observed." This is, of course, silly.

Heisenberg:

Yes, that's certainly silly. The process is more complicated, and it usually goes the other way. Of course, we have to observe something, else we could not understand. But then, we have sonwhere a picture which may or may not fit to this and only when we have somehow understood that it will fit, then we start with the rationalization. The rationalization comes in backward. But, of course, it is very definite criteria as to whether we have understood or not.

Kuhn:

I would love to go on with this conversation, and I hope when you come to Copenhagen then we may.

Heisenberg:

We might, yes.

Kuhn:

Most of the things you've said I thoroughly agree with and I'm delighted. There is one area in which I think we're in quite essential disagreement. It is the area that you talk about most directly in the piece in Dialectica. Yet, you have just said 11understanding meant something different in the sixteenth century from what it means in the twentieth. We've repeatedly found that we understood things differently, that we saw the world differently, that there were these really basically different points of view represented in the development of science that could not be bridged by saying that we have learned a little bit more, that we've become more precise. Rather, we've learned that we must do something different, and I would want to say that we have put aside what we were doing before to some extent in the process. As soon as one says that sort of thing, then I find it very hard to talk about this permanent content of the closed scientific theory, as you do in the Dialectica. So much of what you say are things that I think of myself as also wanting to emphasize in this book. But on this sort of point I would find that this undercuts that whole point of view for me. I can't reconcile these two things.

Heisenberg:

We'll have to discuss it more carefully. By that time I will have read your book more carefully; I have looked through it but have not studied it carefully and so we might have a good basis for a nice discussion when we are together in Copenhagen.

Kuhn:

It would be a great pleasure and privilege for me if you have the time and would be interested in going on with this then.

Heisenberg:

That would be nice, yes.

Kuhn:

On the one hand, with regard to evolutionary views about the development of science, you and I say very much the same thing. I cannot reconcile that with the way of talking about the logic and content of an old theory which somehow wakes it simply a hard piece of modern knowledge so long as it's properly bounded. Now, if I may, I would like to pick up this morning and go ahead from a little bit more about the paper on "Schwankungserscheinungen", which we mentioned before and which takes up from the resonance paper. There are really two or three sorts of things I'd like to ask you about that. For me, this is a paper which I find extra difficult to understand. If it had been written three months later, then I can get all the results you get. But I'm not clear that I understand how you're getting them. You do indicate that this is an important paper for you because you were trying to show the opposition that one cannot now preserve a continuum point of view and particularly one can't do it by the expectation that you can get all these extra results by utilizing the Schrodinger or the wave approach. This is just not going to work. Now, I want to ask you first, "To whom is this addressed? Who are the enemy?" Surely, it is Schrodinger. Surely, it is also this experience that you have had with Wien. But how widespread is that point of view? What is happening to physicists at this point? Who are the physicists?

Heisenberg:

Well, I would say that it is still this view that the wave picture is an essential part of nature. In the wave picture, of course, everything is continuous and I think most physicists found it extremely difficult to think about quantum jumps as something which "jumps" — as a real discontinuity. Therefore, I don't know that I would have thought of any special enemy in that respect, I only saw that many people would still refuse to take the jumps as something quite physical. Well, perhaps you may say the real enemy is Schrodinger, and Wien. But I would say many physicists besides these two would still have felt, 'Well, this discontinuous point of view is, of course, very interesting and certainly contains a feature of nature, but there are the interference patterns and there are the radio waves so after all, there's not the slightest doubt that there is much continuous found in nature." Take, for instance, the radio waves. Whether the wave length is short or long can't make too much difference. Nobody would like to say that the radio waves are not quantized, but when you come to wave lengths, say below one millimeter, then all of a sudden you get the light quanta. That's absurd. You can't say it. Therefore, it was a steady point of concern to almost everybody how these two things could be combined. After all, instead of having two atoms being in resonance, I could say that here I have a radio emitter and here I have a radio receiver, and again the two things would come in. But what happens? After all, there comes the radio wave which is continuous and it is absorbed. But what in the world does happen there? Therefore, we found ourselves just tortured by this kind of dilemma. We saw in some way that both points of view are correct. But what is this "someway?" What is it? Here, I could see now that after all, there is a frequency, that is there is something, namely the matrix element, which moves with the periodic function. In spite of this, when I calculate from matrix mechanics the fluctuations, I do get the fluctuations of a discontinuous change. That seemed to me very illuminating. I saw now that these two things at least, in the mathematical scheme, are actually combined. The mathematical scheme does a thing which none of us could understand before, namely to combine these two apparently contradictory views. For me it was important because I felt that now I have learned that the mathematical scheme can do things which nobody expected it could. So this, I would say, was for me the content of the paper. May I ask when was it written?

Kuhn:

At the beginning of November, '26.

Heisenberg:

I see, yes. So it was written after my helium paper.

Kuhn:

After the helium paper and after the still earlier, but not very much earlier June paper on resonance. And, in fact, you refer to the resonance paper imnediately and say you were simply going to use those results.

Heisenberg:

Well, of course, this paper also goes back to the discussions with Jordan and the last part of the Drei Manner Arbeit which was written by Jordan. Furthermore, it goes back to the old paper of Einstein of the year 1909, where he showed that Planck's law forced some term on the expression for the fluctuations of a radiation field which is not present in the old Rayleigh-Jeans law, which means that we have fluctuations due to the discontinuities of the light quanta. Well, I wanted always to find this feature in the mathematical scheme and I was very glad to see that the mathematical scheme actually did contain both; it contains the continuous matrix elements and the discontinuous fluctuations.

Kuhn:

In the Drei Manner Arbeit in that last part which Jordan apparently did, there's also a reference to discussions in that summer with Ehrenfest. There is a paper of Ehrenfest which comes out about this time in which he points out some really pretty devastating mistakes in the Einstein paper which could be used to throw the entire result out and then he proceeds to show that it's still reasonable to hold the result. Do you remember anything? Were you in on those discussions with Ehrenfest?

Heisenberg:

Well, I'm afraid I don't remember that special discussion, no, no. No, but Ehrenfest would be a man to be interested in that kind of problem. That is quite true, yes. I might say that in connection with this paper I think that I have gotten a letter from Pauli once which was shortly after Born's paper on collisions. Born had stated in his paper that if meant the probability that the particle was there. I was only partly happy about it, as I said, because I felt this must come out of the theory, if it's correct. But then I had the letter from Pauli who said, 'Well, this Born now claims that | ψ(r) |2 is the probability for position r; but he should also remember that if I make a Fourier transformation and have ψ(p) then | ψ(p) |2 is, of course, the probability to find the electron in the momentum p." So Pauli already tried to use this probability concept in a little more general way and to say that it's not merely bound to the position, it is something which is true for any kind of transformation. Therefore, again I was glad to see that this view of Pauli came out by this fluctuation business. There the transformation is also a rather corrplicated one, but there one could see the square of the transformation matrix was the probability.

Kuhn:

I think I'm probably reading the paper badly. This was the point that surprised me when you said it the other day. [Zs. f. Phys. 40 p.505. If S is the matrix transforming the unperturbed eigen values into the perurbed, then, "Befindet sich das gestorte System in Zustand α, so ist | Sαβ |2 die Wahrscheinlichkeit dafur, dass (bei Stossprozessen, bei plotzlichem Aufhoren der Storung, usw.) das System in Zustand β betroffen wird."] I say you produce this result, but I can't see why the paper is an argument for it. I think this is just me, and I'm therefore hesitant even to ask the question. But it suddenly enters in this paragraph right at this point. It seems to me here that there are results around, including most particularly Born's and some of the transformation theory results already from Dirac, although I'm not quite sure how these —.

Heisenberg:

Well, it was a special case of Dirac. Dirac made the general transformation theory later on but this is a special case where this works, yes.

Kuhn:

There's nothing about probability interpretation in theresonance paper. It is this remark here that S alpha beta squared is the probability, if the perturbation is suddenly turned on, of finding the system that was in the state alpha in the state beta. As I read the paper, this just doesn't seem to me to relate to what's going on.

Heisenberg:

I believe that the argument is simple. The argument is the following. I may make a measurement on the atom which registers some function of its energy; not the energy itself, but some function, f, of the energy — any function. Then I can, of course, simply from the old quantum mechanics, ask, "What is the expectation value of this function of the energy?" Now this expectation value can, if the discontinuous view is correct, only be of the form 'probability of the upper energy times the function, f, evaluated at the upper energy + probability of the lower energy times the function, f, evaluated at the lower energy.' That is the point. So if this is true for any function of the energy then it means that we have the discontinuous change. I can apply it for the first power, the second power, the nth power of the energy. Since it is true for any function of the energy, then, of course, it settles exactly, first of all, that it is a discontinuous change; there can never be any value in between. Also, it says that the factor with which the function, f, evaluated at the upper energy appears must be the probability of the upper energy. The factor with which the lower energy appears must the probability of the lower. Now one could see from the general calculation that these factors are just the squares of the elements of this S. So the essential point, the surprising point, was that when you have any function of the energy and you would picture this as a [continuous] curve, you could, of course, never find [the time average of J the function [expressed entirely in terms]of the extreme values of the energy; and if you actually find the [time average of the] function [expressed in terms] of the extreme values of the energy, that must mean that there are only discontinuous changes.

Kuhn:

O.K.; no, you're quite right. I have been missing the relation of the probability to the diagram. [p. 504]

Heisenberg:

Oh yes, yes. The fact that this comes out from quantum mechanics is simply a proof that quantum mechanics states both that all changes are discontinuous and there are only the two values, and it also states that the probability of finding the upper value is the factor of the upper and so on.

Kuhn:

I'm impressed that having gone through this for quantum mechanics from the matrix mechanics point of view, you never here at all deal with the question whether one might not trace some sort of substructure if you had gone through all this in Schrodinger terms instead. ... It would leave open the question, for the people you particularly want to convince, as to whether with the Schrodinger formalism one did not have a device for getting still extra information. Is that at all a good hunch? It should have been represented in discussion about this at the time.

Heisenberg:

Well, I definitely wanted to keep always on the quantum mechanical side and not make any concession to the Schrodinger side which was not already contained in quantum mechanics. Perhaps this was just, psychologically, because I came from quantum mechanics. But at the same time I felt that whenever the people on the Schrodinger side would add something to it then I expected that it would probably be wrong. Well, again my impression was this. If the Schrodinger people do add some new information which is not contained in quantum mechanics, then this new information must be necessarily wrong. That was my idea. Of course, this conviction came from the fact that I thought that now we have a mathematical scheme which is consistent, it can be either wrong or right, but if it's right, then anything which is added to it must be wrong because it is closed in itself. Again, this idea of closed systems which has always remained very strongly in my mind. Also, I didn't want to say much about the Schrodinger picture because I was not quite sure myself whether or not I could understand every detail of it. For instance, at that time the complete transformability from Schrodinger's to our system —. Well, I think people talked about it and the Schrodinger paper had come out, but still it was not so clear yet. Only when the Dirac theory on the transformations had appeared was it quite clear how the things are related. Well, at that time, it was still a bit difficult, and therefore I felt, "Well, I do not want to say anything about the Schrodinger picture. I want to say what comes out of quantum mechanics. I will work with the hypothesis that quantum mechanics is consistent and is correct and therefore you can just not add anything to it."

Kuhn:

Do you remember reactions to this paper at the time?

Heisenberg:

Practically none. I think nobody has read it — I don't know. [Iaughter] Well, at least at that time so many people were engaged with Schrodinger theory and were enthusiastic about the new possibility there that they didn't worry much about whether quantum mechanics was in agreement or was not in agreement with this thing. Therefore, also, I think that most people thought that Born's paper about the probabilities is a very important new information about the interpretation of Schrodinger theory. I think only when Dirac's paper had come out about the transformations did some people look into this paper [of mine] again and find that it fitted in quite well with this general picture of Dirac. But I remember that also Born simply didn't know [my] paper at all. He had completely forgotten that it ever had appeared.

Kuhn:

Was there discussion in Copenhagen that lies behind this paper?

Heisenberg:

Well, I certainly think that I have discussed it with Bohr and perhaps also with Kramers and they did both agree so there was no difficulty. Nobody disagreed with the paper. It was only that people didn't find it especially interesting because after all, the Born interpretation of Schrodinger's wave mechanics was probably the correct one. Of course, Bohr definitely preferred Born's interpretation against Schrodinger's. There was no doubt about that. So people were not too much interested in this kind of approach. I think this paper was of much more satisfaction to myself than to any other physicist, which sometimes happens to papers.

Kuhn:

This whole concern with fluctuation problems was all terribly important back at the beginning of the century — right on through 1911 and somewhat after that. Ehrenfest stays terribly interested in them all the time, and so do some of his students.

Heisenberg:

Well, may I perhaps say also that in lectures, when I teach the students, I usually say that if you want to distinguish between continuous and discontinuous things you have to look for the fluctuations. That is the main point. You can always from the fluctuations definitely decide whether there is something discontinuous going on or something continuous going on. Therefore, from the very beginning I found this such a central point because it was again the problem, 'Are there quantum jumps or are there no quantum jumps?" I think everybody must agree that this is an important problem. It's actually a central problem. Then my answer would always have been that if you want to settle that question you look into the fluctuations.If the fluctuations are big, there are quantum jumps, if they are small there are no quantum jumps. That's quite simple. Therefore, I was so much interested in getting the fluctuations out. I wanted to say, "If my quantum mechanics does give the big fluctuations then it means that it does contain the quantum jumps and then any different interpretation, like for instance Schrodinger's, must necessarily be wrong." Therefore, I felt it was such a central point. Nowadays everybody is so convinced of the discontinuous things that nobody worries about the fluctuations and everybody thinks that obviously these fluctuations are big and so on. But at that time it was not so simple.

Kuhn:

But what I mean to ask you is that in your own education or in your own worrying about quantum mechanics before 1926, to what extent had you worried about this sort of issue before? My impression is that after the war the sort of thing you say to your students I would expect Ehrenfest to be saying to his students, but practically nowhere else.

Heisenberg:

That is quite true. Sommerfeld probably never would discuss it in his lecture. I don't think I have ever learned it from Sommerfeld. I do not know where I have first learned it, but I do remember that somebody told me about Einstein's paper on the light quanta. I think it was in Gottingen. It may be at the time when this discussion took place between Elsasser and anck and Born and myself. Anyway, in Gottingen, I think also that Born did discuss that kind of thing in his lectures. Born was interested. Born had mentioned the paper of Einstein and I then went to the library to look into the paper of Einstein of 1909. I remember that it made a big impression on me because I felt at once that there again this Einstein has just hit the central point of it. That is the center of the problem. I had always a very strong feeling for whether a point was very essential or not and this point impressed me as something which was essential. I could see that the whole question of quantum theory is concentrated in the question of the fluctuations. "Are there, or are there not big fluctuations?" If there are big fluctuations, then nobody can ever get rid of the quantum jumps. That was clear. Well, this has stuck in my mind and so this paper was partly also a result of these earlier expe'iences with Einstein's paper.

Kuhn:

It isn't clear to you at what point you were introduced to that paper?

Heisenberg:

Well, I should say around '24 or so. Around '24 probably. In the year '24 we discussed so much in Gottingen. All of these difficulties between waves and particins. I told you that I had some preference for the Duane idea of getting the waves out of the particles and not the other way around. Again, this was due to the Einstein papers, which had always impressed me. Then there was one such strange thing I always liked very much. There were some people who wanted to forget all wave theory; and against those people I always held that in the early Einstein paper, in 1909, he worked out the fluctuations in heat radiation and he found exactly two terms and the one term was actually the wave term and the other term was the light quantum term. I thought that this was a wonderful way of explaining how nature was — that nature was actually taking both views at the same time. Of course, that was a complete contradiction, but still I found it so impressive that in a simple formula nature would say, "Here you have the particles, and here you have the waves and no doubt about it, you have both."

Kuhn:

On this question of the Einstein. There is your paper in Die Naturwissenschaften in November, 1926, that is sort of a survey, and there you suggest that Einstein has indicated that the electron may have only that degree of reality that the light quantum had. From time to time people have suggested that Einstein has said not only that light is like particles, but he's exactly reversed this and has made the statement that, in effect, we're going to have electron diffraction. But it's not easy to find that in the literature.

Heisenberg:

Well, I would say that was probably in this paper of Einstein on the Bose statistics...

Kuhn:

There is a story told that when Einstein saw Elsasser's paper which is supposed to have been sent to him by Berliner to referee, Einstein said, "I didn't mean to be taken that literally."

Heisenberg:

Well, he may have been afraid for his own courage, you know. [Laughter]

Kuhn:

In the '25 papers you can surely find this and indeed that's where Elsasser finds it. I wondered whether there was anything earlier than that. That's the reasonable answer to that question. At the very end of the last session, in some part talking after the machine was off, I remarked, "Gee, this whole thing is begun in 1925, insofar as constructive new solutions are concerned, and before the end of 1927, it seems to really be all over."

Heisenberg:

Yes, in some way, yes.

Kuhn:

Now, I would like to get you to talk about the sense in which it was all over. If I may make that more explicit — what really happened to the profession? Is it that everybody was suddenly convinced? Is it that you suddenly picked up a whole lot of new people? There is that very sharp break-over, but I'm not clear what it was that established this as the end. I'm sure that it isn't that people like Wien were suddenly happy and came around to quantum mechanics. It isn't that Schrodinger was finally totally convinced either.

Heisenberg:

No, nor was Laue nor was Einstein, yes.

Kuhn:

And yet somehow the door is closed. Now what is that?

Heisenberg:

Well, I would say that by that time people recognized that there was a group of people in Copenhagen who claimed that they can answer every question and apparently nobody is ever able to disprove them. I would say that a change had taken which now I can only express in terms of lawsuits. That is "The burden of proof is reversed." The burden of proof suddenly went to the Wien people and so on because the rumor spread that there was a group in Copenhagen which can answer every question about experiments and, so far, all experiments come out correctly and they claim that they have no contradictions — they can actually explain everything without contradictions. So if you want to do anything against this view, you have to disprove them. And the rumor also spread that nobody has so far been able to disprove them, not even Einstein, who doesn't believe it. One knew that Einstein didn't like the new quantum theory, but at the same time one also knew that Einstein had not been able to disprove these people at the long conference in Brussels, nor has Schrodinger, nor has anybody else. That made a complete change of view among the younger generation because after all, the younger generation wanted to take part in the game. Now there were a few people from Copenhagen who told the younger generation, "You do this and you do that and you calculate this problem and so on and you can be pretty safe that nothing can go wrong with your calculation because, after all, every way of disproving this set of equations has been tried and nobody could disprove it. So we assume now that this is correct and you can go ahead." And then these young people, of course, found that they actually were quite successful and it was simply a new field in which you could work successfully. Well, look at the papers of Wigner, of Neumann, of many other people. Also it made a nice subject about which you could write textbooks. There appeared the textbook, I think, of Kemble, on quantum mechanics. A number of these things appeared. Somehow people saw that this thing cannot be disproved. It meant that actually the burden of proof had been reversed. That will always take place at some instant, and as soon as it has been reversed then this side which has not the burden of proof has a great advantage because then they need not worry, they can just go ahead. The other ones have to do something if they want an opposite opinion. Well, perhaps one should look into the earlier history of physics and other sciences for the same instant when the burden of proof changed. I think also with respect to relativity there must have been a definite time at which people felt, "Now those who object to relativity have the burden of proof, not the people who believe in relativity."

Kuhn:

I've never heard the point put that way and I think it's a terribly good way of isolating that time which is not often as precise as it is with quantum mechanics. You have convinced me of the importance of the Brussels meeting with respect to this problem, the sort of public knowledge in the profession of what had happened to Einstein. I suppose that if what Ehrenfest said was generally known it made an enormous impression.

Heisenberg:

That probably did, yes. And he definitely said that, "Einstein, ich schame mich." He was ashamed that Einstein could not in his mind make that turn which he asked others to make in the case of relativity. He could not see that something had really happened. And that he could not see in spite of the fact that he could not disprove it. He could see his arguments went wrong everytirne he tried.

Kuhn:

To what extent do you think that the profession as it existed then and as it rapidly expanded from that point on, really as a whole did accept the Copenhagen view, which is already to make the view somewhat more concrete, and to what extent do you think they just decided not to worry about it?

Heisenberg:

Ja, that is an interesting question. Well, at least I should say that what counts always is the activity of the younger generation. The activity of the younger generation was drawn into this field. The older generation had not so much activity anyway aid also they didn't want to learn so many new things. It was too difficult to learn them so they were now put into a very bad position. But the younger generation wanted to do something and they wanted to do something new and, of course, they could not do something new in the old fields of physics, at least not so easily. Here was a new field; there you could do new calculations and certainly many new things would come up in time. That was the decisive point, that the younger generation just decided, 'Well, now that's a thing where we can take part." So it very soon spread from these centers — from Copenhagen, from Gottingen, Ehrenfest in Holland. Then quite quickly also it came to the United States. I remember when I came to the United States in 1929, I found already quite a number of groups extremely interested in this kind of problem. In Chicago I gave a long series of lectures on quantum mechanics and I had many young people listening to it and being interested. But they knew already beforehand; it was not entirely new. That probably came to the United States by the young Americans who had been in Bohr's institute. There has always been a group — for instance, Mulliken and Keinble and Dennison and Slater.

Kuhn:

Kemble, I think, was not at Bohr's institute.

Heisenberg:

Had he not been? Oh, do I now get mixed up between Kemble and — who wrote this book on quantum mechanics? This rather early and good book.

Kuhn:

Kemble wrote that. He did have some time in Europe, but I think it was a little later than this and I think he was in Munich.

Heisenberg:

Was he in Munich? Well, there was quite a group in Munich with Sommerfeld. There was Houston, there was Eckart; it may be that Kemble was here, yes. That's possible. Of course, there was Laporte who later went to the States. Then I had, a number of Americans in my institute. There was Nordsick and Uhling. Then there was quite a lot of them. Feenberg and all the others. Yes, this sudden change was perhaps a bit unusual for the development of physics or science. Usually these changes are more gradual but in this case the Solvay meeting was really a turning point. Probably the fact that it came so suddenly was due to the other fact that the trouble has been too concentrated in many, many years. This dilemma between contradiction and paradox had accumulated, so to say, and had created an enormous tension among a rather large group of people. People were just in despair that nature could behave in such a strange way. One couldn't get rid of the contradictions. After this tension has grown so much then the moment when one could actually relax and could say, "There is a group now who says they know it all." That made an enormous change and a very sudden change.

Kuhn:

How big had that group been, that is, the group that was really in agony over these problems; I don't mean the group that was bothered because they could not make the helium come out. I mean the problems of principle — the paradox.

Heisenberg:

Well, of course, I don't know so many universities but I think that the knowledge of these paradoxes had been rather widespread. Say, for instance, Compton. Compton was interested in the paradoxes from his Compton effect. And there must have been a wide group of people who had been working on problems like the Compton effect. Then there were many physicists working on X-rays in many places over the world. Well, these people, of course, used, for instanc, the Sommerfeld formula and knew about the paradoxes; they knew the light quantum view and the wave view of the whole thing. Of course, the degree of worry may have been very different for different people. But the fact that there was something very paradoxical and almost contradictory was very widely known, I should say.

Kuhn:

This remains one of the most difficult things for me to get some concrete episodes that will help to pin it down. That problem, after all, had existed ever since 1903. Sure it's become clearer and clearer, but it isn't terribly long from the time Einstein enunciates the photon theory until it becomes quite clear that nobody's going to get around the equation. There has been even before that the whole needle radiation as a problem. I am not impressed by the extent to which in the early days more than a few people are really badly bothered by it, as distinguished from knowing it exists. You pointed this out when we spoke about your own education in Munich and Sommerfeld's attitude on the problem. It's perfectly clear that he knows it's there. I was much interested because it was the first such piece of information when you said Wentzel had really been bothered about it and had tried to work on it. But that's rare.

Heisenberg:

That is comparatively rare, yes. Well, I would say that many people have been bothered but at the same time they found the whole situation so paradoxical that they simply said, "Well, we can't do anything about it." That is, they didn't know how to attack this problem. They did not know where to get hold in the problem, where they could get any grip into the problem, and therefore, they just pushed it aside to some extent. I mean, Wentzel did not. He was interested in this Duane picture. I think many people did occasionally discuss it and then they would end up by saying, "Well, this is altogether so extremely strange and funny. We don't know what it's all about." Then they would push it aside simply because they couldn't do anything with it. Obviously, it was extremely difficult to do more than Einstein and these people had already done.

Kuhn:

I may be wrong and I say this in order that you'll tell me that I'm wrong, but my sense of the situation is that Einstein is deeply bothered about that. Schrodinger and de Broglie are deeply bothered about that in the sense that that funny route of the people who will take that problem so seriously that they'll try to do physics with it. But almost nowhere else was anybody trying to do physics with that anymore. The people in Copenhagen knew the problem was there, but even that doesn't seem to have been the really torturing problem in Copenhagen.

Heisenberg:

Not for some time. Later on, it was. It was a question of time. One might perhaps say that, logically, the history of physics could have come out in the following way. After Einstein's paper on the light quantum, somebody could have said, "Well, if such a thing is true of the light quantum, it might also be true about the electron." And he might even have found it by experiments and then from there on one would have had a direct route into the real quantum theory. But after all, it's asking too much that a human mind should have sufficient imagination to find such an extreme solution of paradoxes out of only a few data. Actually, only when one is forced by an enormous amount of data to go into that direction, only then one goes in the direction. So actually, it was necessary to have these many thousands of spectra lines of the atoms. By this accumulated empirical knowledge one was finally forced to go into a direction which was certainly there and which could have been found already from the beginning but which was too difficult. It was simply too difficult. Without an enormous amount of knowledge, you just can't do the thing.

Kuhn:

No. I thoroughly agree with that but what I do mean to suggest is that to some extent it is not only that one got all the knowledge, and then took the steps, but in fact one got all of this knowledge and because one had been dependent upon this knowledge, one took a different set of steps. I want to make it clear that this is a highly preliminary judgement in an area where I can't get information. I challenge you at this point. Because when you come back now and say there was, all this time, this awful paradox, "Is it a particle or is it a wave," I want to say, "Yes, it has been there, but exactly this group seems to have been paying very little attention to it." You had come another route.

Heisenberg:

Yes, we definitely had come another route.

Kuhn:

Yet, somehow or another, having gotten there, this problem suddenly becomes a great big problem again. It's how that hapens. In the first place, am I right in thinking that it had been brushed aside. I think it has very little importance that I can see in the origin of matrix mechanics.

Heisenberg:

Well, not to begin with, but in the end it had. I would describe the whole thing in the following way. When the physicists had met the existence of the quantum first in Planck's paper and then especially in Einstein's paper on the light quanta, then people were able to feel that this was something important and very new. But they were not able to do the other step which would have been absolutely necessary to come further and that is, to throw away the old physics, that is, to throw away all the classical concepts and replace them with new ones. This step — to throw away the old things — that could not be done at that time. Then what could the physicists do? They would, of course, try to use the old concepts and to add if possible, or to use if possible, these new ideas in places where they found them necessary. Of course, then the new concepts were only a kind of spice in the whole thing. It was not the real new situation. Bohr, for instance, in his model of the atom, just used this idea of the discontinuities to explain one essential fact, namely, the stability of the atom. The stability of the atom was a central point and Bohr saw that this could only be explained by adding this entirely new feature. But still, even at that time, people could not come away from classical mechanics. That was asking too much. Therefore, Bohr tried to reconcile classical mechanics with the new feature and that he did by the integral pdq business and so on. He realized at the same time that this was really not satisfactory. Somehow it didn't work out as he wanted it. But still one could not do otherwise. Again, it took twelve years until one really dared to go away and push all the old concepts aside and say, "Well, we have to give up classical mechanics entirely and replace it with something new." You must remember that not only the tradition of two hundred years of physics, but also ail the experience on the planetary motions and everything else just proved that classical mechanics is right.

Kuhn:

There was great agony, in particular over the whole question of how can we go on talking about orbits when all we've measured is stationary states. That sort of problem was deeply implicated in what was going on. It's really this one special problem about the photon that I mean to suggest has become terribly central again in '26 perhaps, but not to have been central in the Copenhagen group.

Heisenberg:

Well, there is one point which perhaps belongs to this problem. That is, that Bohr always tried to say in his early papers that the mechanics of the atom is a problem different from the theory of radiation. That is, as soon as the forces of radiation come into play, then you may expect some deviations from the old scheme. So, although he certainly was wrong in it, what Bohr had tried to do is this. In order to avoid a complete contradiction he said, "Well, mechanics is all right for the atom; the electrons move around the nucleus according to mechanical laws. On the other hand, we know that in the theory of radiation, there is something entirely new happening, namely, the quantum of action, and the light quantum. Therefore, he says that as soon as the forces of radiation come into play, that is, interaction between the mechanical electron and the electro-magnetic radiation, something new happens. So we can divide physics in two parts. One part is mechanics and Newtonian mechanics is perfectly all right even for the electrons. The other part is the theory of radiation. In the theory of radiation, Maxwell theory is wrong, or at least is partly wrong; it is correct for the interference and so on, but it is wrong for the energy distribution. The energy is distributed according to the light quantum picture somehow. This 'somehow' has not been understood. Bohr says clearly, "I have not understood it, I just te it as a result from what I get from Einstein's view of the light quanta." Bohr tried to avoid the complete contradiction. Actually Bohr did not succeed in avoiding it. For instance, in this discussion which I had with Bohr in 1922, it became clear in our discussion that you could not make the cut at this point; you could not make the separation between radiation on the one hand, and mechanics on the other hand.But that was only as late as that. That's just the point. For some time, people relaxed on this possibility that all the trouble came from the theory of radiation, that when you do mechanics everything is all right. So it took quite an effort by the physicists to see that this artificial cut which Bohr had tried to make between the radiation and mechanics was just inconsistent and couldn't be used. So that at least I think historically explains well why in the year 1922 again the photon versus the wave becomes more important. One could simply say it was because this cut between mechanics and the photon could not be upheld. This cut didn't exist.

Kuhn:

You would think then in your early days at Copenhagen, when you were there before the 1925 paper, that by this time again the photon was a central problem?

Heisenberg:

Well, the photon came more and more into the discussion. ohr still tried for some time to keep that view that you can make a cut between mechanics and electrodynamics and only gradually he came away from it. That is, by our discussions, and by discussions not only with me but with Kramers, with everybody else, also in the Bohr-Kramers-Slater paper, in some way he came closer and closer to the view that after all, there must be the same kind of trouble. It's not only trouble of electrodynamics, it is a trouble of mechanics as well.

Kuhn:

Just that '22 transition I find very helpful.

Heisenberg:

Yes, yes. For me it is almost located in our talk at the Heinberg in Gottingen, because I think that Bohr, perhaps not for the first time, just at this moment understood the separation which he always tried between electrodynamics and mechanics just doesn't work for this dispersion. This orbital frequency of the electron one can't get rid of; you cannot, just by saying electrodynamics doesn't work, explain that the atom has the wrong frequency in itself. If it has this orbital frequency then it must show up somewhere and you cannot simply say that it will never show up and what shows up is the wrong frequency. That made, apparently, some impression on Bohr to see that this cut couldn't be upheld. From there on the discussions came into this state, I would say, of despair. One had to do something very radical. From '22 on, I would say that the tension grew very rapidly and the despair always became bigger and bigger and people would say, "Well, how hopeless is this situation?" Therefore, also, there was such a sudden break in 1927 when there was a group saying, "Well, now we have gotten rid of all the inconsistencies. We can just answer every question you ask." So this explains why all of a sudden the general feeling changed about the problems.

Kuhn:

From '27, or thereabouts, one gets the feeling that every bright young man goes into this field. What I'm not clear about is whether there are suddenly more bright young men going into physics. Because you get a new generation in three years, and names that are still the big names suddenly become students in this period. To what extent do you suppose it is that one suddenly gets a lot more people going into the field, to what extent is it that they all go into the same area When you were a student yourself, to what extent were there bright young physicists who didn't go into anything having to do with quantum mechanics? There are two sorts of things that could have happened. One of them is that there are about the same number of people every year that there were before, but they now all go into this field. The other one is that there is suddenly quite a few more people.

Heisenberg:

Now I come to a very general problem of the history of culture in general. I'm personally convinced that the number of interested people is probably about the same all the time, in any century, and so on. But may I make a very general statement? I would say that the very well-gifted people in the fifteenth century went into painting or they went into sculpture. In the eighteenth century, undoubtedly they went into music and they became Mozarts and Beethovens. Well, I would say that I see the whole development rather differently from the idea that it is very well-gifted people who do the thing. I would say that there is a historical process going on. This historical process puts questions and the really well-gifted people see, "There is now a point at which I am asked to do something. There's a problem which I possibly could, with my gift, attack myself and could do something with it." I, for instance, do not believe that we have a smaller number of people very well-gifted for music than we had in the time of Mozart and Beethoven. It is only that at that time the historical processes had just given an enormous possibility, had given the possibility to do that kind of thing which they did in Vienna — Haydn and so on. Therefore, every young man who really felt that he could do something in music would, of course, try to go to Vienna in that time. In the same way, in the fifteenth century, probably they would have gone to Holland and also Italy. Therefore, I would say the well-gifted people realized that something was going on in physics. Actually, that was my situation when I came from school. Well, I had learned that I could do mathematics comparatively easily and I had seen some books on physics. When I had read a few books on this modern development, when I had seen the book of Weyl on relativity, when I had seen the popular lecture of Sommerfeld on atoms, then I felt that this is now a field where I could try to do something. So I was caught by this possibility of doing something and I'm convinced that for these excellent people like Beethoven and so on it was the same thing. They felt that now there was something that they could take part in and do something. Therefore, it looked as if the number of gifted people all of a sudden increased but that's probably not so. The number has always been more or less the same, but the historical processes had led to a critical point in which one could do something and so the process collected the gifted people. I mean it's a bit extreme view, but I think it contains a large part of the truth.

Kuhn:

As a basis for operation, I would be very glad to accept this view which I think must contain a very large part of the truth. But, for example, you entered the university in 1920. The number of people who come out of the university between 1920 and l926 and worked in this new field is still comparatively small. I would guess that if one counted the number of people in the generation that enters in about '24 and '25 and comes out between then and '30, one would find that there is several times as many people. I'm not sure about that. I don't mean that there are necessarily more people, but I get the impression that if one looks for people who are the senior figures today, and the senior figures for the last twenty years, there are a disproportionate number of them exactly in that age group which perhaps includes yours, but I'm not sure it doesn't start up two or three years later.

Heisenberg:

Yes, but you know it may be that in the years between 1920 and '23, still many people did not go into this new field because they still felt everything seems to be so unclear and difficult and so, "Better not go into that mess now."

Kuhn:

That is what I'm asking you. Do you remember people like that?

Heisenberg:

Well, I certainly know that there have been very clever students who actually later went into industry or did some other job. So it was not obvious that a young man was attracted by that kind of physics when it was still in this vague disagreeable state of contradiction and paradoxes. But a few minds were attracted by this, like Pauli and myself. Well, it' s almost whether you like that kind of very unclear state or whether you rather prefer a state where it is somewhat clearer. Apparently the number of physicists who liked the state where it is somewhat clearer is larger. That I would agree to. Still, yes, very briefly after that there was a general impression that one could do something in this field, and probably it will get clear in a short time, and therefore many people wanted to do that sort of thing. To make a comparison in the present time, there are many, many really good experimental physicists in the last ten years who just want very urgently to work with the big accelerators. Simply because, first of all, they are probably very clever people and also because with these big new machines you really can do something new about the elementary particles and can find very important data. That's interesting stuff and therefore we go into it.

Kuhn:

This, I think very probably, is just right. I'd love it if one could make it more concrete. There must have been a number of other people who were good physicists, of another sort perhaps, but who simply looked at this and backed away. That stops happening a little later. Can you remember some people like that? There is nothing harder to find out than about the people who didn't go into the field. One's quite convinced that that happened.

Heisenberg:

Oh, it definitely happened. Well, let me tell one case. That is perhaps a good case. That is a young man who told me about the gamma ray microscope. That was (Bochard) Drude. Well, I had a visit from him just a few days ago. What happened to him was this. He was a student in physics. He discussed with me all kinds of quantum physics even as late as '25 or so. Well, one must say that he came from a family where there was some trouble among themselves. Anyway, his father had committed suicide. His father, Paul Drude, was a very well-gifted physicist, who had written the theory of dispersion. He had, all of a sudden, committed suicide. The son lived with the mother and three sisters and apparently he had difficulty agreeing with the mother. In any case, all of a sudden — I was on leave from Gottingen at that time, otherwise I would have tried to change his mind — he went out of Gottingen and out of the house of his mother and was simply lost for some days. He didn't like the whole thing anymore. He just hiked by foot somewhere. It was not the problem of a girl, it was just the question that he hated Gottingen and he didn't want to go into physics anymore. He just hated all these old professors talking nonsense and so on. So he went away. When he was found again and his mother asked, "Well, what's the matter with you?", he said, "Well, I just refuse to do science anymore. I want to do a practical job. I hate all this atmosphere of Gottingen, of these very sophisticated people. It's just awful." He went into the "Kaufmannslehre," you know. He wanted just to make some money. Later on, he went into the Siemens factory. Actually, at present he has quite a high position in the Siemens factory, not as a physicist, but rather as a kind of organizer. Well, this, of course, may also be due, say to some rather unusual conditions of the mind. Still, the man has never become crazy. He's a perfectly normal man, and very successful in his business, and a very nice man too. So I still like him and I'm glad when he sees me. But I think that's one of those cases where you see a man with certainly very high ability suddenly go out of physics. But that's perhaps a bit unusual and we should try for other cases where people went into industry or took jobs of a slightly different —.

Kuhn:

Or worked on some other part of physics.

Heisenberg:

Yes. The trouble is that those people you don't remember. They go out of your mind. What one could do is look through the list of doctoral students of Sommerfeld or of Born and then see what happened to these people. If I were to see a list of the doctoral students of Sommerfeld, I probably could pick out names and say, "Well, this was actually an excellent man — so he went into a different field."

Kuhn:

In the long run, that systematic name by name statistical investigation will need to be done. What one tends to lose that way is the individual story.

Heisenberg:

Yes. I just wonder —. No, the trouble is that I had to do only with those people who really were interested in quantum theory. The other people wouldn't interest me, so I would probably not discuss with them physics any longer. So they are forgotten again. Oh, I do not remember any names, for the simple reason that I had no connection with these people. You know, that is the trouble. There were certainly always other doctoral students in Sommerfeld's institute. Quite a number of them. And I'm sure that some of these people were very good, and they would get very high degrees, and so on. But I wouldn't remember the names. Well, some people, for instance, went into industry after having taken part a bit in this development of theoretical physics. For instance, Welker. Welker was a pupil of Sommerfeld who did something on the theory of metals but later on he went into industry and he was extremely successful in making transistors and that kind of thing.

Kuhn:

There was a terribly good physicist, in a slightly older generation than this, P. P. Ewald.

Heisenberg:

Oh yes, he was.

Kuhn:

He was always considered as one of Sommerfeld's very famous and very favorite students, who was clearly a person of very great ability, but who very early took up something that wasn't so problematic, made his career of it, and had almost nothing to do with the fundamental problems from there on. And in certain very strange ways the story of Debye fits that.

Heisenberg:

Yes, even Debye fits, yes. He went into different branches, and Debye was absolutely first-class, yes. Well, then there was Lenz and there was Kratzer. Lenz did take part for a very short time in the problems like that. He did this crossed electromagnetic fields problem, but later on never published anything and he always felt a bit unhappy about it. It was one of those cases where a man always felt that he couldn't quite fulfill what he wanted to do. He was professor in Hamburg and I think he was quite successful in teaching. He never took part very seriously in this development. We could to some extent also mention Richard Becker. He certainly was very good, but he never took part in the development of the principles. Later on he did very good work on solid state physics and so on, but not on these fundamental questions.

Kuhn:

Do you have the feeling that if these same people had started in '27 —.

Heisenberg:

Yes. For instance, if Richard Becker had started in '27, he undoubtedly would have taken part in all this development. Yes, yes. Well, it's difficult to pin down names and to say, "This history would have gone differently." I think, as a general attitude, it must be true that people are attracted by something which goes on. If there is an interest in problems put by the historical process, then the gifted people are just collected, are attracted, and will do it.

Kuhn:

Let me send you off on one other tack. One says, "In '27 we had everything." To a great extent this is remarkably nearly true. Fundamental applications come right on, one after the other, in this period. The one thing that doesn't behave quite that way is electrodynamics. I think in this area even still, even with Dirac's electron theory, we don't know what questions to ask. I hate to miss the chance to collect one thing like the story you told me the other day when I'm afraid the machine was not on about the bet over the Klein-Nishina formula. Well, there are other things of that sort. I would particularly ask you for early recollections on the Dirac electron theory. We had a very interesting letter from Delbruck who said that he really had almost nothing to tell us but one thing he thought would interest us was certain recollections that he had about early discussions — not as early as '28 — in Copenhagen at the time of the discovery of the positron. He said that Bohr was initially very sceptical about the existence of a particle at all but that he also insisted finally that maybe it exists but it's got nothing to do with this stuff of Dirac's, with Dirac's hole.

Heisenberg:

Oh, I see.

Kuhn:

I really would simply invite anything, hopefully starting early, starting '27, '28, and going on forward, if you can, on this range of issues, i.e., quantum electrodynamics, but also the field that is giving the most fruitful surprises, like the positron.

Heisenberg:

Well, I should perhaps say this. In the early days, say in '22 and so on, we discussed this question whether one could separate electrodynamics from mechanics and actually we decided that this could not be done. That was one of the central points. But then this thing was in a very strange way reversed. I remember that when I heard first about Dirac's paper on the electron. Of course, everybody was enthusiastic about this way of getting a relativistic equation of the electron and also these very strange arguments of Dirac's. He said that the wave equation must be linear in the operator d over dt, that is, the energy can only come in as the first power, and therefore he had to take square roots and therefore he had to double. The old problem of the doubling came in and thereby he got the electronic spin.

Kuhn:

Was everybody happy about that?

Heisenberg:

Well, the point was that in some way everybody admired this ingenious way and this possibility of getting the problems of relativity out and especially that he could get the Sommerfeld formula. The fact that now we had the Sommerfeld formula but with the right number of levels so that it fitted with the X-ray spectrum was such an enormous success that everybody said that it must be right. At the same time, I do remember that I felt very uneasy about it. I saw, "Well, this Mr. Dirac has not only introduced one doubling — that of the spin, which we know and which Pauli had explained to us was necessary — but he had to introduce another doubling, that between positive and negative energies." I felt at once that now we are getting into trouble. I think not only I, but also many other people would feel that something has happened which is disagreeable. It was the first time that there could creep in the feeling that even with quantum mechanics not all problems are solved. Now all of a sudden we get into a new kind of problem, a new kind of difficulty, which is very typical just for quantum mechics and not for classical physics. In classical physics, you could separate positive from negative energies quite easily because every change was continuous and therefore you just were boumd to one scheme, say, positive energies. In quantum theory this could not be done, and if there are negative energies as solutions then why should you not have transitions? This was, of course, also recognized by Dirac at once and so for the first time one got again into disagreeable problems in quantum mechanics. Therefore, one had now to take up these old ideas of Bohr, namely, that radiation theory is something different from mechanics. There may be difficulties in the theory of radiation which are not present in mechanics. Now everybody hated by that time such a kind of argument. Still, these were the facts. You couldn't change them. So when Dirac theory of radiation came out — you remember he did publish a number of papers on dispersion and so on — then in some way we all felt, "Well, it's all right if he gets good results but it's not as nice as we hoped that quantum mechanics would be. Somehow it contains features which are a bit artificial." Actually, I wrote then a paper in which I tried to do it in a more sincle way than Dirac did. Actually it didn't contain anything new but again I felt that, in the theory of radiation, as soon as you bring in the Lorentz group something funny happens because you have the square root. One saw that the Lorentz group introduces a square root somewhere and that is very disagreeable. For some time Dirac, who was very worried about his negative energy solutions, tried the statement that these negative energy solutions were the protons. I remember discussions in Leipzig — at that time I was already in Leipzig. I had probably gotten a letter or so from Dirac saying that he now identified the negative energy solutions with protons. Somebody had to tell about this paper of Dirac in our seminar and then I tried to argue that this must be complete nonsense because one could see from the symmetry properties that these negative energy solutions must belong to the same mass. There is absolutely no way of getting a different mass for the two kinds of solutions just on account of trivial symmetry. So then we said finally, "Well, first of all, we cannot get the proton. We just don't believe Dirac in this. Also we don't know what he can do with this hole business." We didn't know what to say about it. We said, "Well, if that idea of the hole is correct, it just means that they are not protons, but there must be other positive particles." But then again, this idea of the holes was so odd so that everybody said, "Well, that's so terribly artificial." At least nobody was too happy about it. Then the positron was found, but that was rather much later in '32. I remember that Bohr did not believe in these things. We had brought one picture of a track — I think by Anderson or somebody — to our skiing hut here in the Bavarian mountains. I went with Bohr and Bloch and Euler and Weizsacker and one of the sons of Bohr to our skiing hut here. One of us had brought a picture of this track and we had long discussions in the hut whether ikis was actually a positron or was it not. We gradually succeeded in convincing Bohr that it might be a positron. He was not so clear then but it took some time until Bohr would be willing to agree that there was such a positive model of the electron because Bohr, also, felt that now we're getting into a part of physics which again contains difficulties, contradictions. It doesn't really work out as one likes it. So one had this uneasy feeling of' getting into a mess again. That kind of thing.

Kuhn:

Why should he have been sceptical about the existence of a particle of this sort?

Heisenberg:

Well, first of all, at that time, nobody liked new particles. Everybody felt, "Well, we are so happy that we now only have the protons and electrons and probably every atomic nucleus just contains protons and electrons. Why should there be any other elementary particles of this world besides these two?" You know, at that time, people felt that protons and electrons were something once and forever created by the Lord and there's nothing to be discussed about it. So the idea that the elementary particles are themselves again solutions of some fundamental law — that was very far away at that time. Any new particle was a complication to the picture, which nobody did like. So also with the neutrons. Actually, the year of '32 was a shock for many people because it all of a sudden shoved that there may be many more elementary particles than we had thought. That again means that our simple picture which we had believed for some time is out and we have to think of something quite new. I think there was always this game of simplification and complication. There was once the idea of Prout in 1800 that everything was composed of hydrogen. This didn't work out but then at least for sometime people thought, "Well, everything is composed of protons and electrons." Then there came the neutrons and it was very disagreeable. Then there came the positrons and very soon after that we got the mu mesons and all the rest of it. Whenever one felt that one had come to the highest degree of simplification then everybody was happy. As soon as one saw that this is not true it was very disagreeable and therefore Bohr was not inclined to believe it to begin with. Then soon afterwards, of course, he realized that this was the situation.

Kuhn:

What sort of special problems were raised with respect to Dirac's paper by the fact that he doesn't contribute the way anybody else has ever done physics.

Heisenberg:

Yes, ja. You mean what the reactions were of people?

Kuhn:

Well, one has the feeling that if you or Pauli or anyone of three or four other people came out with certain of these things, in the first place it will look very different — the totally new notations of Dirac. He invents his own mathematics, or at least to a physicist he seems to be making up brand new mathematics, with the result that he's giving totally condensed presentations of things which, if other people have seen at all, has taken them weeks of paper work to bring forward. This is, on the one hand, a sign of his great genius, but he also makes it perfectly possible for people to hold his physics at arm's length.

Heisenberg:

You mean to just forget about it?

Kuhn:

Yes. You know, to say, "Well, that may be good mathematics, but it isn't physics."

Heisenberg:

Yes, but I would say only a few people would think that. I think most people agreed that Dirac did very good physics. It was very much his own style, but still they realized that he had always something very substantial in his papers. Of course, the way that he got his results seemed to me almost absurd. I attended many discussions with Dirac. The way, for instance, to get the spinning electron by saying, "I must linearize an equation. Therefore, I have to introduce these operators sigma," and so on. That all seemed so absurd to me. Also, I felt that as soon as one introduced, for instance, these negative energies, then one gets into a mess; then one has to introduce something which one ceases to understand. When I discussed that with Dirac I remember that he was a bit angry about my criticism in that sense. He said, "Well, you start off at once connecting one problem with other problems and saying 'Well, if I get trouble here, I get trouble there and so on.'" "But," he said, "that's absurd, because one can only solve one problem at a time." And that is one of his "Glaubensbekenntnisse". He would say that anybody who tries to solve more than one problem at a time is just a fool because he can't do it anyway. Therefore, Dirac and I had always been rather opposite to each other in this question of how to attack problems. I would always look for the connections. I would say, "I can never solve one difficulty at a time. I can solve only 100 difficulties at the time, or I can never do anything. Everything is connected to everything, and the whole thing must give a consistent picture." But Dirac would say, "No, that is impossible because nobody is so clever as to solve 100 difficulties at a time; you have to concentrate on one point." He had concentrated on the problem of a relativistic wave equation of the electron and there it was. And he didn't mind at all whether anything goes wrong at other places. He just said, "This is my equation and that is right." Well, he was right. I felt from the very beginning a bit uneasy because in some way one could see that introducing the Lorentz group meant the real trouble. Well, the real trouble came in then later quite clearly in the papers of Weisskopf about the self-energy of the electron. Also in those papers which Pauli and I wrote on the quantization of fields we saw quite soon that after all it doesn't look too well. It is true that for the free light quanta everything could be made to fit, but as soon as interaction came in it didn't look right. I think it's so interesting to analyse this question, "What does it mean that the theory looks correct or does not look right." The Dirac theory of the electron was, on the one hand, absolutely convincing on the ground of the agreement with the Sommerfeld formula, X-ray spectra, and so on. On the other hand, I felt at that time that somehow it doesn't look right. It does contain things which are too odd like these negative energies, and now it gets dangerous. So ever since we have been worrying about this side of the problem. Even nowadays we have no complete answer to the problem. That's also quite interesting in that we have an analog to the situation 30 years ago. The fundamental feature of quantum theory has been known since 1900, but it took 20 years until one really came to see that it was a central problem which one could not avoid. So for 20 years the fundamental difficulty was known but only after 20 years things got concentrated so much that you really could try to solve it. Also here the fundamental problem of electrodynamics is now known — 30 years already — the paper of Weisskopf was around '32. But only now people really come to terms with it. Only now one sees that this is a fundamental structure of all the elementary particle business but even now it's not quite clear how the solution is. So I find that very strange how long a fundamental difficulty can stay in physics until the physicists really can come terms with it.

Kuhn:

You say when you and Pauli did the papers on field quantization, it felt wrong?

Heisenberg:

Yes. You know, it was not like in quantum mechanics. In quantum mechanics everything came out much simpler and much better than what I expected. Somehow when you touched it and you had a disagreeable difficulty at the end you saw, "Well, is it that simple?" Here, in electrodynamics, it didn't become simple. Well, you could do the thing but still, it never became that simple. For instance, you had to introduce this supplementary condition and you had to make some kind of limiting processes — first introducing an epsilon and at the end you put epsilon equal to zero. You know, that kind of stuff didn't look right.

Kuhn:

Did Pauli feel that way too? You saw very much the same way.

Heisenberg:

Oh yes, yes. Well, later on one learned then that so far as one had to do with particles without any interaction — the free particles — then everything was all right, but in the rather trivial sense that you just had representations of the Lorentz group and so on. Everybody had to agree that free particles is just no physics because physics is possible only on account of interaction. Therefore, the interaction is always the central problem. A free particle is just nothing because a free particle cannot interact. As soon as you bring in the interaction you get all the trouble. Well, that is probably the same situation over again in the old time before one had the theory of relativity. One had seen that the electron, by moving, changes its mass. Then one had to introduce terms to explain this change of mass. Then you have these very long and complicated formulas of Abraham for the mass, for the dependency of the mass on the velocity and so on. But as soon as you have relativity, then it became quite simple. It looked right. There you could see, "Well, that is the solution to this problem." To put it into more trivial terms, Pohl in Gottingen — you know the old Pohl, you probably have seen him — he loved that kind of general philosophy. Therefore he had put up in the lecture room in Gottingen a Latin sentence saying, "Simplex sigillum yen", that is, "The simple thing is a sign that it is true." But since, except for this, Pohl didn't line theoretical physics, but only loved experimental physics, the students, of course, translated this differently. Didn't you hear uhe translation of the students? Well, that was old-time experimental physics; "Simplex sigilium yen" was translated as "Siegellack ist einfach das Wahre." Now "Siegellack" means usually this wax which you use to make apparatus vacuum tight. [Laughter] Still, I mean there's no doubt that this kind of simplicity shows that you are on the right track. This simplicity was there in quantum mechanics but somehow was not there in quantum electrodynamics.

Kuhn:

Wail, now at first it must have looked as though it was going to come out?

Heisenberg:

Well, to begin with, of course, we felt that by quantization of the electromagentic field you must get the light quanta. That did work out and so far it was all right. Actually that was more or less already contained in the old paper of Einstein. May things did come out all right, but as I said, very soon one saw that the application of the Lorentz group is something dangerous. I think mostly one saw it from Dirac's paper on the electron. The negative energies definitely showed that something comes in now which doesn't fit so well with what we know. Then soon afterwards there came —.

Kuhn:

Well, you always had the option that there you didn't have to insist on linearizing.

Heisenberg:

Well, that one did not know. Of course, Dirac made a strong argument in favor of linearizing. his argument was roughly this. The standard quantization procedure which we had in quantum mechanics was always having an Hamiltonian operator, H, being qual to d/dt . If one went to the Klein-Gordon equation, then it meant that one went away from quantum mechanics. One did not have this simple formula anymore. Well, it introduced the two signs and everything else. On the other hand, Dirac couldn't avoid the two signs anyway. It was then quite an inortant point that Pauli and Weisskopf in their paper could show that even if one started with the Klein-Gordon equation, one got exactly the same troubles as Dirac did with his hole theory, so the two things were nearly equivalent. But that belongs already to the next step. That all belongs to this recognition that as soon as the Lorentz group comes into play then we get into troubles again.

Kuhn:

At what point was that pretty clearly established?

Heisenberg:

Well, I would say just like in old quantum mechanics, it's difficult to say one point where it was clearly established. It was quite clear when one had this paper of Weisskopf on the self-energy of the electron. It was less clear, but already seen in the negative energies of Dirac. Then Pauli and I felt that it was also present in these problems of having supplementary conditions in the electrodynamics. You know, the old Lorentz condition. One had to introduce some light quanta which were not there. Well, in a more modern fashion you would say that the adequate way of describing the electromagnetic field is really the indefinite metric of (Laurel and Gupta). They showed that when one introduced an indefinite metric and thereby introduced some ghost light quanta which don't exist, then formally everything works very smoothly. Later on you can see that you can throw out the ghost light quanta and everything comes fine. I should perhaps explain more clearly that when Dirac had published his theory pf radiation, dispersion and so on, he got results but his mathematical scheme was not Lorentz invariant. Even if one took Dirac's equation for the electron it was not Lorentz invariant for the following reason. He had to introduce in a very odd manner the Coulomb forces, What he did was this. He had introduced interactions with the light quanta, so the light quanta were separate entities coining from the electromagnetic field. But besides the light quanta he had to introduce just the normal Coulomb force which is the action at a distance. So this certainly did not look Lorentz invariant. The funny thing is that one can afterwards prove that it is actually the Lorentz invariant. But in order to prove it, the simplest way is that of (Laurel and Gupta) by introducing the indefinite metric. Also, Pauli's and my intention was, first, to prove that one can make a Lorentz invariant consistent scheme by which then Dirac's theory of radiation would come out. Therefore we made a great effort to get out these Coulomb forces as a part of the formal procedures. We did succeed to some extent but again it didn't look nice. You could do it, but we had always the impression that you have to force the issue a bit. Therefore, nobody was too happy about it still. As I say, it never looked so nice as we wished it to look. Still, in Leipzig we worked on this problem. Then I got interested in the theory of the positron. I did, together with Euler, some work on the scattering of light quanta by light quanta, and these more subtle problems of quantum electrodynamics. It was a very interesting field, but definitely a field in which everybody felt, "Well, it's all right, but still we are not at the end of the story. That is not in a state yet as we had in the old quantum mechanics."

Kuhn:

Tell me again about that bet with Pauli.

Heisenberg:

Well, the Klein-Nishina formula. I don't recall exactly, but I think it was this. Klein-Nishina had written a paper in Copenhagen, and they came out with emission of radiation which was too big by factor 137. That, of course, could not be true. One did not see any single way of getting rid of this factor 137. So I thought, "Well, that's again showing clearly that these things cannot be made all right. After all, the theory of radiation is something new, and we'll have many troubles from it." But apparently Pauli just thought, "Well, Klein-Nishina must have made a slip. If they do it properly it will come out." And he also had given some reasons for it. Finally we came to this bet. I said, "All right, you get a bottle of wine if finally Klein-Nishina will get the correct radiation formula. If not, you have to pay a bottle of wine to me." Then shortly afterwards I got the news from Copenhagen that actually the formula came out right and it was due to a rather ugly mistake. Well, it was a difficult calculation after all. If you got rid of the mistake, everything was fine. So I had to pay a bottle of wine to Pauli.

Kuhn:

This shows a significant difference in attitude between the two of you as to whether things really were in a mess.

Heisenberg:

Yes. Well, Pauli apparently at that time was pretty optimistic with respect to the theory of radiation and thought, "Well, if only we apply the rules of quantization to radiation then everything will come out." I apparently already hesitated at that time to believe that. I don't know why. Klein-Nishina must have been around '28 or '29. Anyway at that tine I had already from Dirac's theory on the dispersion the impression, "Well, it doesn't look nice, you know. It doesn't look as it should." Therefore, I would not be surprised if that would come out the wrong way. Pauli probably soon also discovered that there were troubles in the theory of radiation.

Kuhn:

Did you ever have contact with Walter Gordon? He's a person who starts out doing fundamental work and then just disappears.

Heisenberg:

Yes, yes. I remember. Well, I know how he looked. I do remember that I have seen him. You mean the man with the Klein-Gordon equation? Yes. Yes, he was in Hamburg, so far as I recall. Privatdozent in Hamburg. I don't know why he completely disappeared again. I remember that I had once or twice seen him and perhaps had a short discussion. There was no close contact. Another story, speaking about physicists who disappeared, was this. There was a very good physicist in any laboratory in Leipzig with the name Majorana. You know the so-called Majorana representation of the Dirac particle. He came as a young Italian physicist to Leipzig. He was a very brilliant man and at the same time a very nervous type of a man. He did excellent work. He was always extremely pessimistic about physics. I tried always to induce him to write papers and so he did finally write a very good paper. That was this Majorana paper. And then he went for a leave of absence to his home which was in Italy and nobody has ever since heard a word from him. There was one version that he had been killed by these Mafia people in Sicily because he had some connection with these Sicilian families. Another version was that he had committed suicide by jumping from a boat. He went from Naples to Palermo by boat and apparently nobody knew after that time where he had gone. There was a third version that he lived in a monastery as a monk in Sicily and just didn't tell anybody. It is not absolutely impossible that he is still alive in one of these monasteries. I don't know.

Kuhn:

I had known that there was mystery with his disappearance. All of the people who knew him speak of him in just the most glowing terms. He was pessimistic about physics at a time when almost everybody else was violently optimistic?

Heisenberg:

Yes, but I would say that he was perhaps not pessimistic about physics especially but rather about life in general. He was that kind of difficult fellow. Well, sometimes I thought perhaps he had had very difficult experiences in his life with other people, perhaps his girls or so. I don't know. Anyway, I couldn't make out why he, being such a young man, and such a brilliant young man, could always be so pessimistic. He was a very attractive fellow, so I liked him in our Leipzig group. I tried to see him frequently, and we had him at our ping pong. games. Then I would sit down with him and ask him about not only physics but his private things and so on. So I tried to keep in touch with him. He was a very attractive fellow but very nervous so that he would get in a state of some excitement if you talked to him. So he was a bit difficult. Then all of a sudden he disappeared. I was very sorry to hear of it. Then I, of course, asked Fermi and all the other Italians about it. You know that his father was quite a famous man and also played quite a big role in political life.

Kuhn:

I didn't know that.

Heisenberg:

Yes. Then, of course, we inquired of his family, but the family didn't know. There I think they believed finally that he had committed suicide, but nobody knows. I only once have heard the rumor that he might be in one of these lonely monasteries in Sicily. But I didn't know that Gordon had disappeared.

Kuhn:

I'm interested, for reasons I think you see, in this story of the bet. Because this is just the sort of thing that indicates a parting of the way — difference of outlook. And, of course, it doesn't begin to show in published papers. Do you remember other things of that sort? Often it is a bet that will make one remember.

Heisenberg:

No, I wouldn't say so much. Well, there certainly have been bets among physicists quite frequently but I don't remember such cases very clearly.

Kuhn:

Let me pick up one point which really is the point at which we were stopping yesterday. I had said how curious it was that there was this whole way of doing business that seems suddenly just to vanish after '27 — vanish to the extent that I think most people would say, "That wasn't physics." Would you talk about that a little bit?

Heisenberg:

You mean what was the change? Well, I would put it simply this way. After '27 again one had solid ground from which you could work. You had a basis and you could be pretty certain if you did your mathematics well, or when you had understood your experiment properly and did the nathematics well, then you would come out with a reasonable result. So that was a new state which again resembled the very old states. For instance, in the year 1890, that was the situation if a man had a problem in theoretical physics. It was called mathematical physics at that time because it just meant that you had to take either Newton's or Maxwell's equation and apply these equations to the problem. If you do it well, then it will come out as the experiments say. But in the period before, that was not the situation. You had just no basis to work with and you knew that whatever you did you were already in a contradiction or at least in a paradox. You couldn't distinguish between paradox and contradiction. Therefore, I would say that this stage before 1927 was a much more interesting phase of physics than this later stage. I loved that situation in which one had to judge the weight of this argument against that argument, in which you could not simply work it out from some definite basis. In some way I always found this state of physics in which you can start from a solid basis a bit annoying. I thought, "Well, after all, you know your assumptions. It's only a question of doing good calculations and asking a sensible question. But that's almost the thing which a calculating machine can do and why should I do the work of a calculating machine?" In the years, say, between 1920 and 1927, there's no doubt that these problems could not be solved by calculating machines because you wouldn't know what question to put in them. Therefore I have made always a very strong distinction between those parts of physics where everything is closed, is settled, and other parts where everything is in a mess. People, of course, have definitely different preferences and other people would certainly like those states in which everything is clear. I must say this is also a problem which has worried me in a philosophical sense. Take, for instance, the theory of numbers. If you assume that there are any people living on the Mars or on some other planet and if you assume that at least they have something like a brain so that they have found the idea that one can count 1, 2, 3, then, if that should have happened to them, there's no doubt that for these people the theory of numbers must be exactly the same as ours. If only they ask our questions, they must undoubtedly get to the same answers. That is a very strange thing that by a few very simple axioms, namely, the axiom that you can count, you have already introduced this enormous wealth of connections. The same is true in Newton's physics. In Newton's physics, when you write down mass times acceleration is force, then you have not only been able to solve the motions of the moon around the earth, but also all the complications of planetary motion, all hydrodynamics. Just everything is contained in this very small equation. That is extremely interesting in some ways, and, of course, enormously encouraging in that sense that when you have started with such an equation you can still find many, many things which nobody had solved before and which come out of it. Still, I felt, "Well, it's doing the job of a calculating machine because there I have just to apply these laws in a rather complicated problem and if I do it well it will come out and I have absolutely no choice to make. It will just come out as a calculation runs." I must say that I have always been uneasy about this feature of such a closed system that you are forced to these results and there is nothing that can be done about it. In some way I loved that state of physics in which you really had choices, in which you could see that there's this possibility and that possibility. Of course, the end must be that closed scheme. And only when you have a closed scheme, only then you can say, "Now I have understood." Before you have this closed scheme you cannot avoid always talking in contradictions. Your words have no definite meaning and you just touch things but you don't get hold of the things. Still, that is such an interesting and very exciting state of physics. Then you have the choices and you must make the choices really not on rational grounds but rather on feelings, on impressions, on the hopes for connections between different things. Just in that stage, of course, such a sentence as that of Dirac that you can only solve one difficulty at a time is completely wrong. Because in the time between '20 and '27, you had to solve all the difficulties or no difficulties. Only the connections counted. The single thing was nothing.

Kuhn:

Yet, certainly in the period since '27, physics has again gotten at least once, perhaps several times, into a mess.

Heisenberg:

Yes, yes. There's no doubt. [Tape runs out]

Kuhn:

... Well, let me ask you simply then as a question. Would I be right in guessing that since '27 the thought experiment has not again come to play the sort of role that it played in quantum mechanics before '27?

Heisenberg:

I think that is correct. Still, I feel that in the time since it has played too small a role. This has to do with the other fact that that kind of physics which Bohr and Ehrenfest, especially, have liked is someway perhaps out of date, and at least it was not practiced by many people. I feel, for instance, that that is practiced too little. Well, I may mention one point from a later time at which I definitely tried to practice it but it took a long time until people could agree about it. That was this problem of multiple production of particles. When Weisskopf had written this paper on the self energy of the electron, I felt that this was now a central problem of a similar type as the early twenties. That is, this was a point where our concepts just go wrong and where something disagreeable comes into it, something new. So I tried to localize and concentrate, and say, "Well, what has gone wrong in Weisskopf's paper?" We had the whole theory of Dirac which more or less was satisfactory. Again, as I said, it didn't look right. It wasn't too good. Then we had this problem of Weisskopf and we couldn't get rid of these infinities. I started thinking,"What can I do? Can there be an experiment from which I can see how nature avoids the difficulties?" These difficulties are not only difficulties of our own stupidity with regards to mathematics, but apparently they are difficulties in nature. Then I came to the problem and said, "Well, alright. If two elementary particles — whatever they may be, electrons or mesons or whatever — collide at extremely high energy, then nature is forced to give an answer." I saw that, if one had a certain type of interaction such as one had in the Fermi theory of beta decay, then that must simply mean that very many particles can be created at one instant. Then I saw, "If nature does a thing as radical as that, that in one single act many particles are created, then it's quite possible that actually nature never is bound to produce particles with infinitely small size." I could put it this way: in quantum mechanics, it is, of course, essential that the electron is just a point. That you can see most clearly in the papers of Klein, Jordan and Wigner, for when you have the computation rule, you get a delta function in the computation rules. So that is very nice. But then only it turned out in Weisskopf's paper that the electron cannot be a point, because after all, its self-energy would be infinite. So you had to ask, "How can nature arrange such a thing so that on the one hand nature had the quantum rules and thereby some discontinuous structure of the delta functions; on the other hand, nature apparently does have a finite side of the elementary particles which then probably means that you never can measure with higher accuracy than l0-13 centimeters." Therefore, I felt, "Well, the crucial experiment is what happens if two elementary particles hit each other." Therefore, I suggested from that paper that probably there is a multiple production of particles. Now you probably know how it was. It was a long story. Heitler and Oppenheimer just at that time had found the cascade theory of the cosmic radiation so they could explain the showers in the cosmic radiation without this idea of multiple production and they were certainly right in that case. Still, I always believed in the multiple production, and it was only after 1950 or so that one really saw that multiple production exists. At that time already I felt that it was difficult that I could not convince people of this multiple production business. I remember in 1959 I was in Chicago for a meeting on cosmic radiation and I gave a paper on this multiple production business. I got some opposition from Oppenheimer and we had some discussions in the lecture room. In some way, I was disappointed that I could not convince people that this was now a very interesting and exciting point. This was a crucial experiment by which nature had to confess what nature does with respect to this business of the local interaction. But people loved mathematical schemes and so it was out of the picture again. Well, I just wanted to give this as an illustration that I, with some regret, realized that this kind of physics which we did in the old times has come out of the picture. These young people, extremely good physicists, don't like that kind of physics. They like the mathematical schemes and they like phenomenological descriptions of experiments, but they won't think in terms where one forces nature to confess, that is, to think out crucial experiments by which one can see what happens. Well, of course, a later stage of physics may come back to it. One never knows. But I was surprised that the style of physics has changed since. Probably the reason is that — and that is a thing which I feel in discussions with the younger generation — these younger people again have grown up in a state of physics where you had to work from a solid basis, namely that of quantum mechanics. I quite frequently feel when I discuss the modern problems of field theory with people like Lehmann and Szymanzic that these people believe much more than I do in quantum mechanics. They believe that physics must for all time be just on the same solid basis as the old quantum mechanics. Also a man like Wightman, for instance, an extremely good mathematician, but he believes too much in this solid ground. I just don't believe it. I would say, "Well, quantum mechanics was a very good scheme for some time but now we have a new situation and so we can just as well expect that something new happens." It took inc some time to realize that these young people like Wightman or Lehmann and Szymanzic had grown up in a period of physics like that period in which Planck has grown up. Planck also had great difficulties to get into this new state of physics between 1920 and 1927. Even afterwards he was not so happy about quantum theory. Apparently, as soon as you have started in your youth doing physics on this solid ground, then you never dare later on to come away from it. Einstein and Planck had these difficulties and again the younger generation have the same difficulties. Only our generation, since we have grown up into a complete mess, is in the happy position that we are quite willing to give up schemes if necessary.

Kuhn:

Well, I take it that what you're really saying here is two things. One of them is the willingness to give up schemes. The other thing — the thing I'd like to be sure that we are pinning down — is that there was a different way of doing physics, a sort of investigation.

Heisenberg:

Yes, a sort of investigation which I had ascribed to Faraday as originator. It's a kind, of way of arguing about physics. Faraday, Bohr, Ehrenfest — that is just this style — one doesn't bother too much about the mathematical scheme. That is a later trouble. One first tries to see how things are connected — what they really mean. I would say that is really quite contrary against that kind of thing which Dirac does because Dirac starts from extremely nice mathematical schemes and never starts from the connections. This kind of physics which Faraday and Bohr and Ehrenfest tried to do really starts from the connections. "How nature does act in this experiment and, if nature acts this way, must it not act in another experiment that way. How are these things connected? Can we understand the one from the other? What is the relation between Planck's law and the fluctuations in light quanta and the stability of the atom?" Such terms of principle like "the stability of the atom" and "the quantum condition" is a different style of physics because one tries to make it more difficult by forgetting about mathematical schemes and at the same time one comes to a kind of substance of the things which one is inclined to forget if one works in the mathematics alone. By forgetting the language of the mathematician, one gets to difficulties because, after all, it's difficult to describe physics without having a logical connection. Still, just by doing so one is forced to do certain things more carefully. One is forced to think very carefully about what will actually happen in this experiment. How does nature avoid this trouble?

Kuhn:

Isn't it also essential — and this I think is missing in Faraday — that this be focused around a trouble. That is not there in Faraday generally.

Heisenberg:

No, that's quite true. It is focused on the trouble, yes. Well, that was the historical situation in which we found ourselves at that time. So I think it is most fruitful when it is focused around the trouble because a trouble is something that is not in mathematics, but in physics. That is one point about which I always fight even nowadays with our present day physicists. For instance, last year in Brussels. I was so interested, when I read my own statement afterwards, to see that I always tried to convince the other physicists that, for instance, this problem of the finite size of the electron, the localization of causality, was a problem of physics and not of mathematics. I don't think that you can ever hope to solve it in mathematical schemes before you have understood the substance of the problem in terms of physics. Then, of course, you come to the funny concept of the "substance" of the problem. What is that? What, for instance in the old times, was the substance of the problem of the dualism between waves and particles? Well, later on, in the mathematical scheme, we would say that the substance was either the transformation property of this mathematical scheme or you could say it was going away from Aristotelian logic to quantum logic. So it was a very deep substance. By talking about it without mathematics, one at least got the enormously paradoxical side of the problem out of it. One saw how terrible it was, or to use the words of Bohr, how difficult it was even for nature to avoid contradictions. After all, even nature must think about the contradictions, so to say; it must try to avoid them. Well, later on, of course, when you have the mathematical scheme, then you can see how it is done. I think it's quite useful for some time not to try mathematical schemes, but to see how nature does avoid the contradictions. So in this more modern development, I found it so important tosee that nature does avoid the contradiction of the finite size of the localization by making multiple production of particles. Therefore, I always prefer to talk in these terms, but I do realize that when I now talk to the younger generation they very frequently feel that I'm talking in too vague terms and they say, "Well, your vague ideas are all right, but put it down in mathematics, then we will believe you." Then I have to say, "Well, that is my trouble; I first have to understand it from the side of physics."

Kuhn:

We've come close to this several times and I did not want to let it get away from us.

Heisenberg:

After all, if we have problems left, we might do some in Copenhagen.