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.
Please contact [email protected] with any feedback.
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.
In footnotes or endnotes please cite AIP interviews like this:
Interview of Philip W. Anderson by P. Chandra, P. Coleman, and S. Sondhi on 2002 March 8,
Niels Bohr Library & Archives, American Institute of Physics,
College Park, MD USA,
For multiple citations, "AIP" is the preferred abbreviation for the location.
Anderson discusses the theory of superfluid Helium-3; recalls germination of the idea and eventual publication of "More is Different"; reviews work on topological defects; discusses motivation for resonation valence bond work with Fazekas; talks about interaction with Lee and Rice on charge density waves; recalls foray into astrophysics with Pines and Alpar and theory of pulsars glitches.
This is the Evening News with Phil Anderson on his career in physics. The interviewers as usual are P. Chandra, P. Coleman, and S. Sondhi. Today is March 8, 2002. Our rough plan for this segment of the interviews is to discuss Anderson's work on Helium 3, his article More is Different, and his work on charge density waves and spin density waves, on pulsars, on the foundings of the RVB idea, and time permitting to move on to spin glasses and thereafter to localization. So I'll turn the floor over to Piers Coleman.
Phil, we’ve in previous interviews talked with you about your early contributions to the theory of anisotropic pairing in Helium 3. Now we are going to turn to the situation in the early 70’s, when super-fluidity was actually observed for the first time in Helium 3. So my question to you is how did you react to this discovery and what thoughts did it bring to your mind.
First where did I hear of this discovery? Somebody very early on said, “Why are there two phases in Helium 3 solid?” which is what the original letter claimed. But then I first heard in detail about the experiments at an awful Gordon conference. I don’t even remember what the Gordon conference was about, but it was a terrible one. It was held in Beaver Dam, Wisconsin, and there was nothing to do but play pinball. There were no sports facilities at that particular place, except there were tennis courts which were occupied by a tennis camp so none of us could play tennis. So we were all confined just to sit on the lawn. John Wilkins told us about the latest results from Cornell. At that point I said to myself, and I’m sure I said it loudly to all the others, there can't be two phases in the liquid unless it is the anisotropic superfluid because that’s in a sense the key result of our work way back in 1960 that any anisotropic superfluid had the same phase transition as every other superfluid with the same L value, but they then differed in fourth order in the order parameter. It did appear that the phase transition between the A phase and the B phase was absolutely continuous. At this point, I was quite excited. I somehow didn’t want to write a paper about just that. I became, as I learned more about it, fairly convinced that one of them was the Balian-Werthamer phase and one of them was an anisotropic phase. I guess what John told us about was the fascinating NMR results that Doug Osheroff was getting. I met Doug later at the International Low-Temperature Meeting in Colorado, I think it was at Boulder. I chatted with him and he told me that he was coming along to Bell Labs, and indeed rather remarkably he essentially carried the entire technology from Cornell to Bell Labs and had a new laboratory working within six months, which is an incredible achievement, but Doug is an incredible person. In the meantime I was puzzling about two questions. One was what were the two phases? We were by now fairly sure they were superfluid phases. The other was this intricate puzzle of the strange way that the NMR worked. That is in the A Phase there was a big shift in the NMR much bigger than the number that you would assign any of the possible dipolar interactions. Given Tc you know the coherence length, the size of the pairs, therefore you know the effective dipolar interaction between the spin and the orbit. So we knew that magnitude. And the shift was much bigger than that, and I thought to myself immediately about the famous antiferromagnetic shift in the antiferromagnetic resonance, which gives you a shift which is proportional to the square root of the anisotropy field instead of proportional to the field itself. In the early days of antiferromagnetism we assumed that the antiferromagnetic resonance was unobservable because it would be shifted away into the infrared. Paul Richards made his reputation by discovering methods of doing the infrared sufficiently sensitively so you could find it. So we kind of knew this was something like the antiferromagnetic shift. On the other hand there was the B Phase where there was no shift at all but an enormous drop in intensity, and the question was to find a sensible way to explain all of this. Chandra and I worked on that a lot during the fall, and then I went back over to Cambridge and continued to puzzle about it. Meanwhile Bill Brinkman had come over to Denmark for a sabbatical.
Phil, this is in fall of ‘73?
This is in the fall of ‘72. All this is happening in the fall of ‘72. In a rather short time after the discovery, essentially when I knew the facts about the resonance, I knew that the B Phase had to be the Balian-Werthamer because it didn’t shift, and the Balian-Werthamer is essentially isotropic, and I suspected that the A Phase was an orbitally ferromagnetic one. Chandra came over to Cambridge and stayed for a brief period during which he was extraordinarily uncomfortable because we put him in the local inn and it turns out that anti-Indian prejudice is alive and well in Grantchester in spite of its being so close to Cambridge; and he was very uncomfortable there and he became quit antsy, and I think that had something to do with what happened between us later. Anyhow, he tried to persuade me that we had developed some kind of theory which could explain the shifts, and so we wrote out a letter and popped it off to Nature. And then fairly soon, against Chandra’s explicit opposition, I phoned Nature and said you mustn’t publish that letter I think it’s wrong, but Nature said we have already got it on the cover, you can't retract it and so it got published. So that was the first run-in I had with the magazine Nature. I have never had a good, close relationship with the editor of Nature except personally. Lois Garwin is the daughter of one of our oldest friends and she was the editor for a long time, but I think therefore, perversely, she always picked the wrong referees for my papers. Anyhow, so I tried to withdraw that one. And I did develop a theory of the shift, which I don’t remember where I published it or if it’s even on this list. I think I talked about it at the Nobel Symposium but probably didn’t really — well, the Varma paper came out. I think I put my new theory of the shift in a footnote or something like that. But it was not the correct theory which was discovered by Leggett essentially simultaneously. It was close enough to correct that it did explain the two shifts and did make it clear that the A Phase was what we now call the ABM Phase, was the orbitally ferromagnetic one. I was dealing with it as a thermodynamic problem. I was assuming it was a hydrodynamic oscillation like ordinary sound rather than a quantum oscillation like zero sound, so I was doing a theory which would have been perfectly all right if we were at frequencies I guess low compared to the appropriate relaxation times. And some of the experiments were in the region, so in some sense it was more or less correct. It was in a region where zero sound and first sound would have had the same frequency, and this is where you would put it. So my theory was in some sense correct, but it certainly was completely finessed by Tony’s. Then I had this idea, which was a letter published in Phys Rev A8 that may have been a short communication. Anyhow, we published it very quickly. Bill Brinkman came over from Denmark, and he had been working on some things in Helium 3, and I had had this crude idea that because we thought the appropriate model for the interactions was the spin fluctuation model that the interactions would depend on the susceptibility and realized that was enormously changed by the presence of the order parameter, namely the measured susceptibility was down by a factor of three in the B Phase, and we knew that in the A Phase apparently the orbits strictly ordered along an external magnetic field, so that the constant susceptibility with a big shift was only along one direction, and we knew in the other two directions it must be considerably smaller. So we knew the susceptibility had changed, and we realized that if the interaction was due to the spin fluctuations that the change in the susceptibility would change the free energy of the ordered phases and wouldn’t change Tc because Tc only came from the linear terms, but this would provide a large non-linear term. So the simple assumption for the non-linear terms that they were just thermodynamic and just essentially those of the original theory, we would also have a large non-linear term that came from the susceptibilities. Bill worked out a number of details and he was already an expert on the spin fluctuation theory, and so between the two of us again during that semester at Cambridge we composed what is the paper that appeared in Phys. Rev. That was the work I talked about at the Nobel Symposium at Goteborg and Bill did too. We were both very happy to be invited to this Nobel Symposium where the Helium 3 heroes were going to talk about their work and Tony Leggett was going to talk about his work. Actually John Wilkins, I think it was, phoned me up during that time and said we’ve got a Nobel Symposium, will you come and talk, and you know, I was as usual kind of stiff-necked about various things and said well I won’t just be one of a myriad of speakers and be allowed ten minutes, and he said it’s not that kind of thing at all and this should be just fine. So I suggested that Bill be invited too, and he allowed as how that was a good idea. After I had gotten back to Bell Labs that spring Bill and I worked out more of the details of this feedback mechanism for the free energy. He’d been in touch with the group at Cornell, one way he’d been in touch was that they had worked out this fourth order free energy, also at the suggestion of our pre-print, and parameterized it, and Bill was the referee on their paper and he suggested that they had one of the parameters missing, and they thanked the anonymous referee. So although I think the parameterized fourth order free energy was first published by the Cornell group, Ambegaokar and Mermin, the correct parameterization came from Bill and the idea that there was a fourth order term from the feedback, came essentially from me. They didn’t have any discussion of the actual predicted values. So Bill and I went to Goteborg, and it was a very exciting meeting because this wasn’t long after Wilson, Kadonoff, and Fisher. The three of them came and talked about their work on the renormalization group, and the Helium 3 work was discussed and I gave a talk and Bill gave a talk and Tony gave a talk. In the middle of Tony’s talk the light dawned and I jumped up afterwards and said, no, this is antiferromagnetism, it’s just the same equations, but he said, no it’s not antiferromagnetism, it’s the non-linear sigma model, and that was the first instance in which I had ever heard of the non-linear sigma model and I did discover that the non-linear sigma model was my own model in antiferromagnetism. The equations are identical to the equations for NMR or EPR in antiferromagnetism.
Was there any Russian presence at this meeting?
Yes there was. Dzyaloshinski was there, Gorkov is in the picture, so he was there; but definitely Dzyaloshinski, he was there and gave a talk. I gave the summary talk for that meeting, which was a very interesting experience. They had a banquet, which was a very bibulous banquet and a lot of fun, and in the end they brought in Swedish folk dancers. That isn’t necessarily a lot of fun. Swedish folk dancing is among the least advanced, but it wasn’t as bad as the folk dancing when we went to Graftavallen. Northern Swedish folk dancing is something else again. But anyhow they did bring in the folk dancers and so in my talk I said they even brought in the dancing girls. Then after the banquet I couldn’t sleep because of course the sun didn’t set and I had an east window with no curtain, and so I sat there overnight and composed the summary talk in which I first mentioned the idea of unstable fixed points determining possible forms of the eventual stable fixed point among other things. But it was great fun. — Well, Seb Doniach was there and he talked about the Doniach-Sunjic , another x-ray anomaly. There was a talk by Walter Kohn on the LDA, which was at a very primitive state then, but he predicted some of the things that were actually going to happen. There was a nonsense talk by a Belgian guy on the surface energy which ascribed it all to essentially the zero point energy of fluctuations in the electric field, which probably was nonsense, but everyone had to be polite to him because he was the only Belgian there. Leon Cooper talked about neural nets; it was one of the very early mentions of nets. So it was a very exciting meeting and I had a great deal of fun summarizing it. I went from there to Belgium to a meeting run by de Vreese where I talked about what my student John Inkson had been doing where he essentially was doing what we would eventually realize were corrections to the LDA, the corrections which were responsible for the error in the gaps in the LDA. Later on it was worked over and improved — well not improved on but redone — by Steve Louie and others and I believe that John really had it right. We were discussing surface energy and the field gradient in surface layers, surfaces between semiconductors and insulators, and of course therefore we had to understand in a many body sense what the self-energies of particles in the insulators were and we discovered the term that came from the long range Coulomb effect which can't possibly be summarized in a local self-energy, so long as it’s insulating you can't use a local self-energy, and therefore you have the wrong gap. So we identified that shift in the gap. I think that my notes for that talk are the only really published discussion of John Inkson’s work in full. He published a couple of papers, and there was his thesis. But John had wife problems. His wife was adamant against his going overseas on trips and so he never really published his work as strongly as he should have, and I don’t think it’s known to this day that he basically solved the problem of the gap in semiconductors, although he has published papers which do explicitly that. My only contribution to that, I don’t think any of the papers were really published in my name, I hope not my only contribution to that was that I kept saying there has to be some correction, and I brought in some, well I don’t know, half right, half wrong, data from Jim Phillips who had been observing that the Fermi level always naturally stuck somewhere near the middle of the gap, no matter what metal you put on the surface. Basically the old problem that had stymied Shockley in the old days from making a field effect transistor was the fact that the Fermi level got stuck in the middle of the gap because essentially the long range term was not in the metal and was in the insulator, and so as far as quasi-particles were concerned, the insulator appeared to have a much smaller gap. Although I didn’t put it in those terms; I didn’t understand it as completely then as I do now. But there was no question John Inkson correctly solved the LDA problem.
Now we were going to ask you a little bit about More is Different, although I gather a lot of the history is recorded elsewhere. Obviously at this time you were looking at the great diversity of collective phenomena that you could see occurring in nature, and you formulated it in this now famous paper “More is Different”. Can you tell us more about that?
Well the history is very explicit; I think I must have talked about it in the previous interview. The paper was really written in 1967 during the period when I had decided to go to Cambridge and Joyce had flown off to England to close the deal on the house that we thought we were going to buy, and I flew off to La Jolla to be Regents Lecturer and then eventually Joyce came and joined me in La Jolla after a rather notable night in which Bernd Matthias and I and the Zachariasons and I guess Joan Matthias sat around playing Hearts. I think I told you about this. In La Jolla I was working away on my Regents lecture and also thinking about various other things. On the same floor were Maki and Christiane Caroli who was having a sabbatical and there’s a paper of Caroli and Maki which Ong just photo-statted for me which is the first paper on the Nernst effect in superconductors, and they were working on that, although I could never really get very excited about Maki’s methods; they are orthogonal to mine, and beside nobody could understand what he said. But the amusing thing is of course he has a very fine contralto voice and he was always singing and I could hear this singing going up and down the halls. At first I thought it was Christiane because she’s got a very fine voice also and occasionally they sang together, but not often. So he would work away and sing. And I only lately realized what they were doing in physics. But then Joyce told me that my talk was rather incoherent and I said it can’t have been incoherent, a few people understood it, and she said I don’t think many, and so I rewrote it and finally thought well I must publish this because I have rewritten it so carefully, and sent it off to Science and I was very surprised when they published it. Of course the scheme is that “More is Different”, that large complex systems can do things of which the atoms they are composed of can't do it all. I had just begun to understand the true conceptual depth of the idea of broken symmetry. So I used the idea of broken symmetry as the prime example. I kept talking about broken symmetry in those days to people, anyone who would listen. For instance when I got invited to give an auditorium talk at Bell after I got the Nobel Prize I based it on broken symmetry, but I think that fell rather like a lead balloon. But I still felt it was very important. I really only got it properly formulated when I went off in 1980 to a meeting in Paris in honor of Pierre Curie, and the best discussion of broken symmetry is in that. But I had the essential concept of here is this thing that happens, and that the symmetry has to be restored in some way. The final piece was put in the puzzle by the realization that the rigidity could be broken only by defects in the ordered state, and that happened around 1975.
I was just about to suggest that it was the set of papers from then, one of these with Gerard Toulouse on textures, and then one an article by Richard Palmer again on textures and then one on Boojums with Dan Stein and Rob Pisarski. Exactly how did you get in the thick of this? Was it connected to the more mathematical developments of the defect theory at that point, topological ideas?
Yes. Well, I was at Bell Labs in the summer of ‘75 and I think Gerard Toulouse came by for a visit, and he had been thinking about these topological things and I said, “topology, schmothology.” It’s not very interesting. And he said, “Yes it is interesting,” and he tied four strings to the arms of a chair and he showed how he could rotate the arms of the chair and twist the strings and then you could undo it. And I said, “That’s a good trick!” and I began to have much more respect for topology. So there is no question the people who really invented the idea of using the mathematical theory of topology in discussing ordered phases were independently and simultaneously the Toulouse-Kleman paper and the Vladimir-Mineev paper that both came out in 1975. There was another thing that was very exciting that happened at this Goteborg meeting which was incredibly rich was that Pierre de Gennes gave a talk about possible defects in Helium 3. I don’t think he had them right, but anyhow he said what you need to do is to find an appropriate free energy and an appropriate Landau-Ginsburg free energy, and he put up some suggestions. He had an analogy of free energy that he had written down for liquid crystal phases. One of the sayings that I quote in my summary talk is de Gennes, who said, “This is superconducting, and it’s a liquid crystal. Obviously it was made for me,” and he talked about the geometry of liquid crystals and, what might happen in defects in Helium 3. Bill Brinkman and Doug Osheroff at the same time were having ideas about the possible textures that Helium 3 would have in various kinds of vessels. They realized that the A Phase for instance would always have a radial texture, and we knew at least enough topology to know that spherical containers should contain at least one defect and so on. There were some topological terms that we were using, but we weren’t using them as completely as we should have. Although Brinkman did invent the equivalent of a hedgehog in Helium 3; we never figured out how to find the hedgehog for the B Phase which has a vector order parameter. He and Doug showed that some of the NMR results were quite definitely caused by the texture that Helium 3 had to assume in the vessel that you were doing the experiments in. The only thing I contributed to topology was the idea that there was a connection between topology and dissipation. If you had line or point defects they could be pinned, but if you could destroy the order parameter with a sufficiently extended texture, that couldn’t be pinned. So Helium 3 A for instance couldn’t really be as perfect a superfluid because there would always be a texture in it that could be moved by external currents which were driven, and that some of the defects were textured defects. In other words, you have a classification of defects into the first homotopy group, the second homotopy group, and the third homotopy group is the group of textures in three dimensions. So we did write a letter saying that Helium 3 is going to be unique, as far as we know, in that the superfluid isn’t perfect, is never going to be perfect. So I guess that was my sole basic contribution to this sequence of papers. The other thing about Helium 3 I probably should mention is a little bit of history; it’s always been a chip that I have had on my shoulder. In the summer of 1974, I went to the Saint Andrews Summer School, which was a very pleasant school. I don’t think even Mike Cross came along, but I stopped by and saw Mike Cross after the summer school. For that summer school I wrote out a full joint review paper with Bill Brinkman in which we put all of our results. Incidentally, Bill had found a student, interestingly, from Cornell. I don’t know how Cornell felt about us, Bill absorbing one of their best students, but Joe Serene came and worked with Bill and they really powdered the problem of these fourth order terms in the free energy, plus a few other terms, plus a little bit more. My name was on it because the original idea had been mine, but it was Bill and Joe Serene who together worked out the consequences of the spin fluctuation theory and put that in a big paper; here it is, Spin Fluctuation Stabilization appeared in 1974. But that was strictly Bill and Joe. And Mike Cross contributed some further work on the free energy in which we felt we could really compute the A/B Phase difference, or at least its dependence on magnetic field, and we felt that was complete proof of the existence (but not dominance!) of the spin fluctuation terms. This paper then went to the editors, who were Bennemann and Ketterson, but they had to wait a long time for the experimentalists to write their piece, and so the book appeared in Physics of Liquid and Solid Helium. I have the history wrong. Bill and I presented this to Armitage and Farquhar, and they were good editors and they got the thing out in 1975, but the total press run was probably a few hundred just for the libraries, there was a book on that meeting. At the same time we submitted it to Bennemann and Ketterson, but they waited an extra three years and published it only in 1978. I didn’t want to put it in Bennemann and Ketterson. I submitted it actually to the Reviews of Modern Physics. It was accepted by the science editor who was Bert Halperin and then as a chief editor’s decision, David Pines said, we are having a full paper by Tony Leggett and we can't have two papers on The Theory of Helium 3 and so he rejected it which is the reason it appears only in the Physics of Liquid and Solid Helium. I have never really forgiven him for that. I think it was the wrong decision because the two papers actually deal with two totally different aspects. Tony’s is a marvelous paper about the delicate questions of symmetry and the very fine, very small energy terms that affect NMR and various other effects and about the longitudinal resonance and various things, and it’s absolutely perfect in covering that, but he says nothing, not a word about spin fluctuations, for instance, not a word about the microscopic theory, and so it was a mistake because they would not have been conflicting in any way. He was at Illinois and Pines had an obvious conflicting interest, and I am not sure he acted without bias. Perhaps, in view of the two Nobel prizes which eventually resulted from He3, I should comment further. It was a very competitive atmosphere at the time, both in experiment and in theory; but the competition remained friendly and constructive, totally unlike the atmosphere surrounding high Tc; everyone was willing to acknowledge progress due to others’ work, and in fact progress was extremely rapid because of the different methodologies applied by the different groups: resonance at Bell by Doug, mostly sound by the original group at Cornell, and SQUID magnetic measurements, and magnetic cooling, by Wheatley. That was experiment — in theory the Cornell group of Ambegaokar and Mermin and our group at Bell were more focused on microscopics, Leggett did beautiful work on symmetry and heuristics, and I remember only deGennes’ characteristically original view as having much relevance otherwise. I think all parties agreed that it was a very painful decision whether or not to include Wheatley in any prize, but in the end it was solved tragically by his death. He had after all originated the field of millidegree cryogenics and He3 as a Fermi Liquid, and measured Leggett’s predicted longitudinal resonance. But no one I talked to felt that theory should share the primary glory, including me, even in my most private thoughts. My talk at Goteborg, and the introduction to the Brinkman-Anderson review, were unconcealed advertisements that I, with Morel and Brinkman, had some rationale for coming next after the experimentalists. Although the specific application to He3owed a lot to Brueckner and came only weeks before Emery and Sessler, in the papers with Morel I had been the first to probe the nature of the superfluid state as genuinely an anisotropic superfluid though violating Landau’s criterion, and to propose the near-degeneracy of phases which was such a prominent feature; and Brinkman and I identified the microscopic mechanism which controlled the phases which actually appeared, and proposed the non-trivial ABM phase. On other matters, such as the fourth-order free energy, the A1 phase, and topology and textures, I think it was a wash between Cornell and Bell, and on the dynamics between deGennes and Bell. I think Tony’s great contribution to the microscopics was the inclusion of Landau parameters, and I continue, just as a matter of taste, to see microscopic understanding as more physical than heuristics and symmetry, no matter how delicate and beautiful. So if I hadn’t been taken out of the running by my 1977 prize, I would have been antsy about Leggett’s recent one, if only because it neglects his greater contribution.
Switching topics for a moment, I want to take you to a couple of papers which sort of stand out in this period, it’s returning to your old love anti-ferromagnetism. So these are the papers, Resonating Valence Bonds: A new kind of Insulator that you did by yourself, and then the longer paper on again, the Resonating Valence Bond state. What brought you to this?
What I remember bringing it up was reading some experimental data, I think it’s referred to in the paper, Japanese measurements on some low spin antiferromagnetics, particularly something which was planar. So Patrick Fazekas had arrived from Hungary to post-doc with me. He came on his own funding: we had people come all the time on their own funding and show up at the door and say, Look here I am; what shall I do? Patrick was scruffy looking being from behind the iron curtain, but I got him an assignment to Jesus College and saw to it that Jesus College somehow housed him. I assigned him this problem of what about these Japanese data that seemed to indicate that this particular spin half antiferromagnet in two dimensions was not exhibiting antiferromagnetic ordering. The data are very questionable. It was polycrystalline samples and so on, and I think in the end it turned out the speculation that they were not spin half antiferromagnets was probably true, but by this time I also knew that CuO was very unhappy with being an antiferromagnet and entirely different from the four which I had begun studying many years before. Manganese, iron, cobalt, nickel oxides are all straightforward antiferromagnets, and Cu is not straightforward, which for reasons which we now understand and perhaps are quite different. There were all these experimental data that indicating that spin half systems don’t like to be antiferromagnetic, and that they particularly don’t like to be antiferromagnetic in two dimensions. I was perfectly aware that there was at least one spin half antiferromagnet that worked beautifully which was a copper chloride dihydrate that Gorter had found the first antiferromagnetic resonance on many years before. That was the case in which anti-ferromagnetic resonance was within range and it checked the square root of HAdata and the S magnitude I had predicted years ago. Anyhow, so I said to Patrick maybe these things aren’t antiferromagnetic, and there are these old papers of Kramers and Hulthen that proves that one dimension isn’t antiferromagnetic, so let’s have a look and see whether two dimensions is antiferromagnetic. I had been reading old papers of Pauling for several reasons; I was conscious of his old papers that claimed that the metal was just a resonating valence bond state, which I had actually given as one of my first Journal Club talks at Bell Labs way back in the ‘40s. With the supreme overconfidence of a 25-year-old, I condemned it out of hand. But then I looked at some point into Pauling and Rumer and their work talking about valence bonds, as aware of various facts: that they are over complete and nobody ever managed to sort them out into a really satisfactory theory, and that Pauling had developed a heuristic, which worked very well for chemistry, but Pauling’s heuristics are the kind of thing no one else should ever try to do because he knows his chemistry too well. It works but in mysterious ways. For instance, he explains the octahedral coordination of say nickel or manganese in the shell of ligands around manganese, and he says, Oh yes of course, there is a d orbital in the oxygen shell and a d valence bond he sticks in there which has no reality whatsoever, it’s just there is a bit of d in there because of perturbation theory. I was aware of all of these things, and so I suggested to Patrick that maybe we should test whether two-dimensional systems really were antiferromagnetic. Actually at that point I was a little shame-faced about my naiveté about Pauling and so since I had been invited to submit a paper to a Festschrift for Pauling in the Materials Research Bulletin of all things, I sent it off. But I didn’t write in the covering letter that it was for the Festschrift for Pauling so it didn’t appear in that, but in a later issue it would with apologies. It appeared in Materials Research Bulletin for that reason, because I had learned how hard valence bonds were to deal with from Pauling.
Now another set of papers in this period.
Patrick eventually became a very good friend and he still sends Christmas cards.
Is he still scruffy?
A little bit. Well, he was very communist in those days. Well, I think he’s from peasant stock or lower class stock and probably had gotten into the upper class universities because he was that, and so he was quite a defender of the system. I can remember his feeling that we gave him a rock from the beach and his great happiness of getting this rock because it was a gift he could accept, as a proper communist because it was not anything that was too valuable. He was so happy with this that he lugged it up this five hundred foot cliff from the beach. He and Ali Alpar came together on a train and spent a few days with us.
Let’s get back to Ali later. Let’s move on to Patrick Lee and Maurice Rice and your work, a set of papers on the conditions for giant conductivity in TCNQ.
I have never understood why they put me on that paper. I guess Patrick came in and talked it out with me before — Well, a couple of things: there was the awful TTF-TCNQ experience.
You were forced to memorize the full chemical expression for that?
No, no. Alan Heeger published this letter in which he had found two samples of TTF-TCNQ which had extraordinarily high conductivity which appeared at very, very high temperatures with a big peak in conductivity, and they said this is superconducting fluctuations. Various of us were outraged at these ridiculous claims. What we didn’t realize was that TTF-TCNQ was the first organic metal or more or less metal and would eventually become a gigantic field and would eventually earn Alan the Nobel Prize, of which more later. But Gordon Thomas was particularly outraged by the idea of publishing something like superconductivity at 30 degrees on the basis of two out of a large number of samples. And we were also aware of the trap of four terminal conductivity measurements which they had been using. The trap of four terminal conductivity measurements is that in any material with anisotropic conductivity, and particularly in the presence of magnetic fields, is that you may not happen to put your two terminals on one of the current paths the current would spread out and maybe it doesn’t get into that piece of the stuff and so it will look like a very high conductivity because you don’t get any voltage. You measure the current on the outer leads and the voltage on the inner leads, and anyone who knows anything about transport measurements knows that you must reverse the leads in some way and make sure that the current is flowing past your voltage lead, and we suspected that that had not happened. So Gordon got himself a lot of crystals of this stuff and measured them and showed that all the crystals that he had that were properly measured had no peaks. Then he made a lot of them write a joint letter together so that he wouldn’t be the only one who weathered the wrath of Garito and Heeger. Then we all went to a meeting on the top of the San Gabriel’s on Great Bear Lake. I arrived and was lying there sleeping peacefully, and the door opened and in came Maurice Rice and I had no idea that we were all sleeping two to a room instead of one to a room. It was all right; I mean I like Maurice Rice but it was just a shock. And this was the most god awful contentious meeting I have ever experienced before 1987. I want to say there have been plenty after 1987. And there was one session of comment on TTF-TCNQ which was chaired by Tony Garito, and there was supposed to be five minutes per speaker and it was always ten minutes for any speaker from Penn and two minutes from Bell or any other place. And there was this awful woman from Israel; she turned out later to be very nice, but she was an Israeli, a real Israeli, and she wasn’t taking Garito’s squashing her either. I got to meet her later and I remember realizing that it had only been his treatment that had made her seem to be such a tough cookie. So there were three or four different sides squabbling at each other, and I was thus prepared for high Tc. So we got back and said to ourselves what is Alan seeing, and maybe he is seeing sliding charge density waves, and then we said to ourselves, or rather Patrick, who knew the literature better than I did, long ago Larkin and Ovchinnikov said that the density waves would be pinned, and so let’s do a theory of the AC behavior of charged density waves. I had some long discussions with Patrick, but the paper was written entirely by Patrick and Maurice and my name was put on it. I submitted it as just a footnote to be on the other work at the Nobel Symposium too, but I didn’t really talk about that. But there was work on charge density waves at that symposium. By this time they were only seeing these so called superconducting fluctuations in high frequencies and maybe even in the infrared, and we said well that’s what he seeing. But that was Patrick and Maurice really almost entirely and my role was to publicize it by talking about it at the Symposium.
The third of our topics from the >70s is work on a neutron stars, and you mentioned Ali Alpar previously, with him and Tosatti and Palmer.
That started when David Pines was coming to whatever meeting he came to and we would talk about how he does solid state physics in neutrons stars. And Pines and Jacob Shaham wrote a long paper which he came to Bell Labs and gave talks about, and then I guess he gave talks at the next academy meeting about solid-state physics and neutron stars and pulsars ascribing all these interesting glitch phenomenon to what he called star quakes. He and Jacob initially said well there is the crab pulsar and we know it’s slowing down and it’s noisy, and that works out reasonably well with star quakes in the crust. He had made himself familiar with Mal Ruderman’s theories of structure of pulsars, and we all talked to Mal from time to time. This was a group based on the Aspen Center for Physics. So I set Richard Palmer to thinking about it, and then I thought well, they must be right. At least if there are glitches that’s it’s a proof that the star is solid, that it has broken symmetry, because otherwise you can't have a glitch, you can't have a sudden event happening in a purely gaseous sphere, so this is a nice proof of solidity. It all looked fine until they began to look at the other pulsars, particularly Vela, which is a fairly regular glitcher, and by now we know there must have been recorded several dozens of glitching pulsars. The Vela glitch was a steadier pulsar, something that didn’t have a lot of noise to its rotation. And it (the glitch)was an enormous thing. It was 10-6 of the total rotation energy. If you figure out how much that means, 10-3of the strain energy in the crust or in the whole star, and the crust is in a kind of like a tiny ice crust on a deep pond. Most of it was thought to be fluid or superfluid. So their idea was that somewhere deep in the star, this tiny crust didn’t even have anywhere near that amount of moment of inertia, so that it had to be something deeper in the star and it had to be really the core, so they called it a core quake and said, “Hey look, we found a core quake.” I became interested in the equation of state of nuclear matter, and assigned Richard Palmer to that problem and suggested that he look at some work on scaling of quantum solids that I forget who had done. So we started from the heavy rare gas solids and scaled to solid hydrogen and solid helium and we decided that the neutron star if it was a solid was very much farther on the quantum level, and so we worked for quite a while on quantum solids in possible high-pressure phases of nuclear matter. This was the Palmer/Tosatti paper in Nature [Physical Sciences]; I don’t know why I submitted it to Nature because I was still mad at Nature, but they published it, thank goodness. Later on Palmer wrote a very nice thesis on quantum solids and various possible things. At the same time Gordon Baym and the gang at Illinois was doing things, and we talked to them, but I guess never collaborated with them. The corresponding states approach was again Palmer, and Palmer was a very impressive student and we did very good work. But then I suddenly realized that what was causing these glitches was something entirely different, which was pulsar glitches as a hard superfluidity phenomenon, and Naoki Itoh was visiting me. He was a non-scruffy post-doc who I also put up at Jesus, but very Japanese, and well, not the quality of Patrick. He has found himself a niche in astrophysics and doing very well at a minor University. He talked me into at least one paper that I have ever since regretted this, a paper on nuclear physics here with Itoh, The Theory of Zero Plus States, and that turns out to be quite wrong. Itoh’s, well, I’ll take responsibility for it: I was over optimistic about how much I knew about nuclear physics. But the other was a nice paper for Itoh to get.
There are also papers from around 1980, the spin vorticity and so on.
Oh yes. This eventually we did publish the paper, Pulsar Glitches and Restlessness as a Hard Superfluidity Phenomenon, and that one earned a prize. The Observer had a contest for the most opaque scientific title, and they found this one in Nature and it actually won that year. I got nothing for that prize, but I did win a prize. But in fact it’s quite correct. Then I guess I had been communicating with David Pines through this whole thing, and Jacob. Jacob was always eager to come up from Columbia and talk to me. I guess he started out as student or post-doc of Mal Ruderman. So we had done a lot of chatting, and David therefore said, “Why don’t you come out to Aspen?” and I said,” Aspen, that’s a highfaluting place for big shots in particle physics, isn’t it?” David said no, Elihu and I are out there, but in fact it turns out that he and Elihu were the only solid state physicists who had ever been there, but I learned later that they had been instructed by the NSF that they had better get some solid state physics on board, that they were too theoretical and too impractical and they were just a mountain climbing club for the particle physics community, for the particle physics elite actually, which was very true. Although actually one of the founders was Mickey Cohen who was also a condensed matter physicist, but he never had much to do with the physics. I guess there was a lot of nuclear physics. Bethe was an early founder. Anyhow, they had gotten to the stage of needing an NSF grant and NSF said you better get in some condensed matter physics so I accepted this invitation and stayed for a week and had a lovely time walking around in the neighborhood and playing tennis and driving around in David’s Jeep. We talked and I came out with this pulsar glitch idea. I was staying there in the astrophysics months, not as a solid state physicist. They wrote me a letter on Aspen stationary and said we hope you will become a trustee of the Aspen Center for Physics, and when I looked on the side of the letter where it had listed the trustees my name was already printed there, and only later did I understand what was going on. But that was the start of a very close, very good, association with Aspen that lasted from ‘74 through the present. I did become a trustee and eventually became an officer. Elihu was the president of the institute for a long period, somewhere around one and a half terms, five years or something like that. During most of that time he and I and to some extent David organized whatever was to be organized in Aspen, more or less on our own recognizance, in the way of condensed matter physics. In the initial years, year after year the gang of Shaham and Alpar and Pines and myself and Ruderman would gather at Aspen and think about pulsar glitches, at least part of the time, and a whole sequence of papers came out of that, which eventually I think ended up being right and probably even being proved out, although there were details in some of them that I am not happy with. In particular in one of them David insisted that we must make a prediction, otherwise no one would ever believe us, and so we made a prediction which turned out to be completely wrong, that the glitches in the Vela pulsar would be fairly regular at two year intervals, and then it immediately stopped having a two-year interval and started having a four-year interval and then it had a six-month interval and it turned out that they were totally chaotic of course! But we did in the end develop a theory which had actually one predictive fact, one predictive consequence. The rest was essentially constraints. I mean there was almost nothing that you could find that you could explain these glitches except this theory. But in the end we predicted that about one percent of all the slowing down of all old pulsars not pulsars with stuff around them that spun them up but isolated pulsars would glitch. In other words, there would be a glitch in each one every hundred years, and they have been watching them and sure enough it’s about one per hundred pulsars per year. So we got the right percentage and that’s about all you can predict about glitches. Vela is relatively new and that’s why it’s glitching so much faster. But each of these glitches is about the same size, 10-6, and you can do the numbers; as to the order of magnitude we were right on the button.
So it would be a good point to stop for today and we will finish the other topics next time.