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 Alexei Kojevnikov on 2000 June 29,
Niels Bohr Library & Archives, American Institute of Physics,
College Park, MD USA,
For multiple citations, "AIP" is the preferred abbreviation for the location.
Discusses his childhood and education in Illinois, undergrad and graduate work at Harvard; writing his thesis with Van Vleck; working at Bell Laboratoreis and the scientists there including William Shockley; the rise of interest in solid state physics in the early 1950s; research in superconductivity; the creation of theory groups at Bell Labs in 1956 and the relationship between theorists and experimenters in the lab; decisions on research topics at Bell; his year in Japan with Kubo; security restrictions at Bell and military research; collaborations with John Galt; experiments leading to localization of electrons in 1956-57; development of superconductivity theory; his visit to the Soviet Union in 1958; collaboration with Morel in 1961 on superconductivity; and research philosophy and approach to problems. Others prominently mentioned are: N. Bogolyubov; George Feher, V. Ginzburg, Gorkov, Charles Kittel, Lev Landau, David Pines, Harry Suhl, Gregory Wannier.
June 29th, the year 2000, and this is the fourth session of the oral history interview with Philip Anderson at Princeton University. I think we stopped last time around the early sixties, and but we didn’t talk about your first book Concepts in Solids which was based on lectures, and I am just wondering if you could tell about the context, how that book originated.
Well, it was very simple. I simply took my lecture notes and they were I think transcribed and slightly edited by a group of the undergraduates — a group of the students, I think they were graduate students actually who attended the lectures. Liu Sham was particularly helpful. He’s now at San Diego, quite a well-known band theorist. But there were a whole group of young people in Cambridge at that time many of whom I encountered in later life, like Morris Rice. There were students of Ziman and students of Volcker Heine who was there but whom I didn’t get really close to until quite a bit later. But a group of these students helped me and it entered my notes but they’re still very crude. There’s no index, there’s no — it’s not particularly organized. I hadn’t actually ever taught before, except in Japan, and there they just took my notes and wrote them, and copied them off directly. But in this case the notes were in fact cleaned up and edited and they were quite happy with the outcome.
It took five years, didn’t it, between the lectures with Livit [correct name?; spelling?] and the —?
No, the Livit lectures, it says 1963. I’m amazed that it – I don’t think it was 1963.
Oh, when you gave the course?
No, I gave the course in ‘61 of course, ‘60-‘61. I think it must have been five years. I think the date in my bibliography is wrong. We can look and see when [inaudible phrase].
That’s okay. That can be checked easily. That’s no big problem.
Because I promised the Benjamin people that I would — Oh, it was copyright ‘63, yes. But there was an editor there who kind of delayed and delayed but eventually it came out and I was quite happy with it. But it was just the lecture notes that I gave.
Was there any specific message you wanted to convey in this booklet say compared to the existing textbooks on solid state?
Several. One of them was that it didn’t need to be complicated. Basically if you just took, looked at a few of the conceptual ideas. Not all the concepts were developed at that time, but for instance the idea of elementary excitations looking at the low-energy excitations, which was very much the Russian point of view at that time — Landau, the Landau group’s point of view. But the idea of elementary excitations as a thing which was common to the whole field wasn’t in existence. I didn’t do band theory in great detail because that was done in the books, but what I did was to show that band theory could be done in a very extraordinarily simple way.
Did you yourself in your research work ever do the band theory?
Oh yes. There are things I’ve done that are at least closely related to band theory or one-electron theory. But no, it wasn’t my field, and so I kind of took a look at it from outside or above and would find interesting things about it or simple ways to look at it.
Did you feel that there was a split in the community between these several methodological approaches?
No, not yet. The band theorists were still many body theorists and the many body theorists were still band theorists. It was much more unified.
When did the split start developing?
The split started developing at least ten years later I think. There were formalists, but, you know, Abrikosov, Gorkof, Dzyaloshinsky didn’t exist. The many body theory was being created. And so it was still real physics, and people were not inventing models and doing any fancy stuff with them. There was a certain amount of formalism, but I think it was more, the field was more unified.
And when this split started to develop, was it just visualization [correct word?] or were there some, how you say, methodological issues involved — or simply because the community grew so much?
Oh, I think —
Or at least in what words would the disagreements be expressed better?
When I was first starting out in condensed matter physics, the idea was you could be a theoretical physicist and [inaudible phrase] the union card or the basic tools are group theory and second quantization. And if you knew those, that was all there was. And of course in concepts in solids I started out with second quantization. The older books didn’t even have that. And with the formal theory of Hartree-Fock equations as in a second quantized approach. Just about this time, well, in the fifties, the Green’s [spelling?] function many body theory approach had developed. There was Landau, there was Bowman Pines [correct name?], and so on, and some of the things I did. And so by the mid-sixties you began to need to know Green’s functions; you couldn’t work in superconductivity unless you knew Green’s functions but you would even do band theory with Green’s functions. But the renormalization group and many of the specialized techniques hadn’t yet been developed and we were just learning them. But now, well, I think when the renormalization group was taken over by the field theorists one began to have a kind of physicist who felt that he’d solved a problem if he found the exponents at the critical point. And that was I think — And another kind physicist as band theory became more and more efficient and the tricks became more and more esoteric. There was another kind of physicist who did nothing but band theory and didn’t really approve of many body theory and didn’t like many body theory.
Like who for example?
Well, they were followers of Slater [spelling?], and many of them were students of Slater. People like —
But Slater himself not included?
Slater — later Slater, yes I would include him. And of course there were always better band theorists, but Cohen[spelling?; Cowan?], Phillips[spelling?], but particularly Art Freeman[spelling?] was an example, Alex Zunger[spelling?]. The best of them of course did both things. Volcker Heine combined band theory with many body theory and developed the pseudo-potential method. That’s one of the major methods for doing band theory. Morrow[spelling?] Cohen went both ways. And then there were later people like a guy at — oh dear, the name is gone. He died. He was at Bell Labs. He was very, very good at combining the two.
And why do you think it wasn’t your field? Didn’t it have any appeal to you, or was it too formalistic at that time? What was —?
Band theory? Well, band theory, you know, what you’re doing is — as I say, I did some things in one-electron theory and it wasn’t that it wasn’t my field, but many body theory always fascinated me more because the problems were harder. And okay, that’s enough.
So should we then go to the general physics?
Yeah. Physics, that’s an almost a forgotten episode. I guess I was feeling pretty, I was feeling clearly high at this time in my life. I was feeling that really I had accomplished a lot and it was a little bit of an ego trip. But there was a party, and I don’t remember who gave the party — maybe Bob Showman, maybe Peter Wolfe, but we all were drinking quite a bit and lamenting the problems with all the various publications — Physical Review Letters and Physical Review and all the other alternative publications and the refereeing system — you know, that’s almost everybody’s pet peeve at any given time is the refereeing system — and how much junk was being published, and we all felt that 90 percent of what was — and continue to feel — 90 percent of what’s published is not worth publishing. And yeah, I guess Bern Matthias was — I later realized he had malice [correct word?] of forethought. He said, “Well, what would you do if you wanted to produce the ideal physics journal?” And so we started talking among us and invented a physics journal and we said it would have two or three things: one is it would be much more like Nature or Science now is but wasn’t at that time. Physicists didn’t use Nature or Science. But one thing we would want is very quick publication, but even more than that very quick decisions, and that the decisions should be made by a responsible person; that there should be someone who accepts your paper or rejects it and you know who it is. And then finally we said, “Well, if we are going to ask for the best papers in the world to be published, instead of having to pay page charges like you do in Physical Review, you would pay the authors.” And so I worked for, well– Bern said, “That’s very interesting. Maybe you’ll hear from me.” Turned out that he had met Captain Robert Maxwell, the famous Captain Maxwell who eventually got very fat and fell off a boat and defrauded his employees of their pension funds and became eventually very, very not particularly respectable. But at this point he was just beginning to grow in prestige and financial, in money, and he was very much involved in exploiting the resources of his little press that had begun to be a rather small press, Pergamon Press, which built its reputation actually on bringing the Eastern European literature to the West and then branched out and created a number of new journals like Physics and Chemistry of Solids. I knew some of the things that Sam Goudsmit said privately enough and publicly about Captain Maxwell’s business practices, but after all Sam was a competitor of Maxwell’s and one took that with a grain of salt. Anyhow, Bern knew Maxwell and took this idea of mine, which was basically mine, to Maxwell and said, “Well, Phil and I will run this journal for you if you’ll finance it.” And so Maxwell said, “Fine” and the message came back. And then he flew Bern and me first class over to Oxford and we went and spent a night in Haddington[spelling?] Hill Hall with Captain Maxwell, but unfortunately it wasn’t such a great party because he was in bed with the flu and his wife had to do the honors at dinner. And we had a very pleasant time nonetheless, and he signed an agreement. The contract I made conditional on the fact that we could keep personal control of the journal and if it didn’t work, if at any time Mr. Maxwell didn’t live up to the terms of the agreement we could pull out and the journal would stop. So I spent a summer making a list of names. I wrote maybe 600 to 1,000 letters to all the best physicists.
Prospective authors, explaining —
And who was supposed to be responsible editor? Just one?
Just me. I would send things out to referees, but I would be responsible for picking the referees. It was a crazy thing to take on, but for about two years it worked and many of the original papers in quark theory, many of the original papers on the Kondo effect, Robert Abrikosov(?), Pater Zul’s[spelling?] paper. There are quite a few famous papers. The first paper that ever used squids for measuring magnetizations by John Leedman[spelling?]. It came out, and we had a great success. But then suddenly Mr. Maxwell lost interest and starting getting farther and farther behind.
In terms of —?
In terms of publication, in terms of time, and so people complained and said, “You know, you’re not living up to this promise that you would publish on time.”
So what was your ideal in time?
We were going to get — well, I did. I read the manuscripts and decided on them. Mostly I decided on them myself, but if it was a field I didn’t know I would find one or another friend and say — I would just call them up and say, “Do you know this work?” and usually they knew the work. So I could do it. It wasn’t that much work, but it got farther and farther behind, and the quality of manuscripts we were getting therefore got worse and worse, and finally I just closed it down. But Captain Maxwell lived up to his later reputation. He refused to close it down. He just kept the — He never — I wrote a letter saying, “The journal is now gone,” and he would never give me the subscription list so that I could communicate with the subscribers personally. He kept that absolutely secret. I should have known that that was — that was a commercial secret so that he kept the subscribers sending in money long after the journal had died. Many libraries subscribed to it. Of course I never knew how many copies he even really produced, how many copies were in it. All this was kept from me, and I think the subscription list probably got fairly short fairly quickly. He wasn’t making a big profit on it. He had hoped I think that he could turn it into one of his high prestige major journals by — But he was not willing to spend money on publicity. All the publicity was my private business. Essentially like most publishers, he was happy enough to start a new journal, but he wasn’t about to spend any money publicizing it or producing the papers fast, producing the journals fast.
Who decided how much to pay to the authors?
Well, it was a nominal sum, but people were very happy to get it. I did. But it wasn’t New Yorker page rates, but it was New York Times say page rates.
Did you publish papers there?
I think I published a couple. I don’t remember exactly which. You’ll have to look through them to find them. But in the end I killed it. And the last paper was the 1965 issue. It had Bigley’s[spelling?] squid measurements.
And what do you think was the good time between the submission of the paper and publication? What is ideal?
The ideal would have been three months, two or three months. We were as much as a year late by the time I quit.
And what by that time was the standard for let’s say Physical Review? Oh, the standard of Physical Review Letters was of course they could do it in three to six months, but you would always have to argue with the referee and it delayed as much as a year and sometimes went more than that. They had a backlog. The nominal — I don’t think the issues came out of their nominal dates at that time. It went through a period when everything was slowing down.
But it was also a time when the European journals also started, the European Physics Letters?
Yeah, Physics Letters and so on, but at this time they were very low quality and so a lot of junk was published. Some very good papers were published of course, but a lot of junk was published.
And was Nuovo Cimento still an important journal at that time?
Yeah, Nuovo Cimento was important. But you wouldn’t have refereeing problems in Nuovo Cimento, but that was the problem. The stuff that appeared was just meaningless. We wanted to be, insisted on high quality, and we could maintain it, but for only about three years.
So the next topic is different paper “More is Different” and I’m just wondering if you could — Was it the first —? This is the —
Oh, [inaudible name] is different, yeah. [inaudible phrase]
Wasn’t it the first public statement on your part?
Yeah. That was the very first popular article I wrote.
How did this develop, or what was the context in which it appeared?
Well I’ll tell you the story. Actually it’s all — By this time La Jolla [correct name or word?] had started up. La Jolla, you know, this was the period when California started all its new campuses. The University of California multiplicated [correct word?] into — had been just UCLA. And Berkeley of course for science. Science was the high quality one. But they were starting up on campuses all over California and the third campus and one that had great ambitions to become a very high quality institution — particularly in the sciences — was La Jolla. At the same time they had Santa Cruz for humanities, starting Santa Cruz for humanities; Santa Barbara was being started, but that was very much a country club atmosphere. Davis was for agriculturalists. So they were starting a lot of new campuses, a lot of places were starting new campuses and new universities. But California was of course, had the big reputation and the wonderful climate and so lots of people wanted to go to California and start these new departments. And the best in them was La Jolla. And they had, well they had several advantages. One was that they already had the Scripps Institute there, which had very imminent oceanographers attached to it and geologists. Then they brought in a number of — I guess a lot of people got tired of living in Chicago, so they got a lot of Chicago faculty: the Mayers and Harold Urey, and quickly they had a couple, a few Nobel Prize winners and they were really active, interesting Nobel Prize winners. And then Walter Kohn, I guess he got tired of living at Pittsburgh at Carnegie-Mellon, so he took over the chairmanship of the department. And he had been coming to Bell Labs all these years and of course he knew who all the good people were, and so he reached in and picked out five or six of our best young people who were interested in moving. And so he got Harry Suhl [spelling?], he got George Feher. Bern Matthias was his big coup. And so all of a sudden there was a very active, interesting physics department there in La Jolla, but the big campus on the hill hadn’t been built yet. They were still working on the Scripps Institute. Living in the little town of La Jolla, very picturesque and a seaside town with a lot of fish restaurants and very fancy, jet set people around, so it was kind of a beautiful life and people were very eager to go. And I don’t know whether it was Walter or Bern — Oh, another Chicago member who went there was Zacharius, who was a close friend of Matthias. I don’t know which of these people it was, but they wanted me to come as a regents lecturer, which is a one-month appointment, very prestigious, and what you are asked to do is give one public lecture and then you can do what you like in terms of technical lectures. I think I did a couple of technical lectures as well and I spent a nice month. It was interesting. Munkee[spelling?] was there, Munkee and Christiane Corrolli[spelling?] was visiting Munkee, so there was a very strong superconductivity group. Keith Brueckner was to go there and he hadn’t I think gone yet, and his long time association with the JASON group hadn’t yet really been formed. So it was one of the strong new departments that had picked up condense matter physics because condense matter physics wasn’t being picked up by the prestigious old physics departments like Yale and Columbia and Princeton. So they did very well. Anyhow, I went to give the regents lecture and at the same time I had built up quite a head of resentment against particle physics because —
Were there any specific events?
Well, there were several specific events. I had myself only just been elected to the National Academy of Sciences and I was experiencing the things which blocked people from being elected to the National Academy. And the physics section was almost all particle physicists and no condensed or very few condensed matter physicists.
Did you try to some —?
I tried to get some people in. Well, I did succeed in getting Conyers Herring, but it was ridiculous that Conyers Herring wasn’t already in it, because he was arguably the mostly highly regarded and most scholarly of all condensed matter physicists at that time. And then I was trying to get other friends in and finding that condensed matter physics was considered a little bit beneath people’s dignity. It’s a story which is not new. The National Academy is very slow at — And then they were still smart — The theory of the leisure classes, which is what Jeremy Bernstein [spelling?] called it, that all the lovely little summer schools and study groups and so on that people had all over the world in Aspen and Carchevs[spelling?] in Italy and Verona and so on, all these belong to the particle physicists and sometimes condensed matter physicists were allowed along on [inaudible word]. So I developed a head of steam against particle physics.
I’m just trying to think. You mentioned here in the paper some official in the field of material science urging the participants at the meeting dedicated to fundamental problems in condensed matter physics to accept that there were few or no such problems. I’m just wondering who that official was.
Oh. I see. [phone interruption; tape turned off, then back on...] ...well be Walter Kohn. That might have been a meeting in Seattle. I’m not sure. And this was quoted — of course the article was Vicky Weissopf’s article that I reference there. But I think it was Walter Kohn.
And among the particle physicists were there any particular persons or names who you felt responsible for that attitude or who would be most vocal in this?
Well, I didn’t know them as well then as I know them now. And in fact — well, I don’t know whether I knew at that time the long, long story of T. D. Lee at Columbia, but it certainly is true that he had over the years kept condensed matter physics out of the Columbia department. Again, it was much later that I heard a lecture by Sam [Samuel C. C.] Ting that was very much this way. You were hear remarks. When Weinberg[spelling?] said, “Well, you know, nucleon physics is really not interesting anymore. What’s really interesting is [inaudible word] reactions.” He’d gone already up to the scale where he was condemning even the physics that they had given up a few years before. Although Weinberg actually again is — well, he still has the attitude, and he was quoted by many people at that time and he certainly had that attitude. It came up, it was much more, much more evident and it became much more visible during the debate over the SSC [spelling?], which was much, much later.
I was planning to ask you about that, but I was just wondering whether we’ll do it now or later.
That’s going to be much later. But so I got to know these people only quite a bit later, and so I didn’t — I was just responding to the Weisskopf article. And for instance I didn’t go to Aspen. I didn’t know the people in Aspen — yet. I didn’t — My first year in Aspen was 1974, and Aspen was reserved essentially to particle physicists at that time. Later on I got to know them and some of my best friends are particle physicists but -
Could it be that part of this resentment came from the fact that the particle physicists appropriated broken symmetry?
No, no. That came much later. That was later. No, this resentment was straightforwardly the argument of the extensive versus intensive argument of Vicky Weisskopf, and I thought Vicky Weisskopf was [inaudible word]. He worked on line broadening, and I based my thesis to some extent on work of his. I thought he was kind of a real generalized physicist and here he was repeating this thing which was the attitude that you knew was — although not expressed directly — was there. So it was really a response to the Weisskopf article too.
And who was the audience at the lecture when you gave it?
It was just in the physics department and anyone else who wanted to come.
Ah. So basically those who would agree with you already.
No. Some of them would, some wouldn’t — because there were particle physicists there and there was Urey and I don’t remember who the particle — oh, Pitioni[spelling?] was there, [inaudible name]. But you know, you would meet them often enough and hear what they were saying. But you didn’t really — it was just, it wasn’t really a hidden idea; it’s just that I hadn’t heard it expressed as publicly at that time. But I knew it was there. And it was caused by resentment against particle physics, but of course the message is completely different. The message is a real intellectual point — the point about broken symmetry, the point about emergent properties which I didn’t know the word “emergence” at that time, otherwise I would have used it. In some ways I’m glad I didn’t use it, because what I mean by emergence is somewhat different from what a biologist means by emergence. Biologists are very much, when they’re talking about emergence they will talk about the — what is it? — the San Marcos Arches Syndrome [spelling and punctuation?]. Were the arches built in order to hold it up or were they built just as decoration and then incidentally they held up the dome. That kind of thing. They have this problem of acceptation and it becomes very complicated in their minds, whereas I had something very simple, I was talking, in mind. I was realizing that thermodynamics is something which happens in the n = infinity limit of atomic physics and but thermodynamics has a whole bunch of new concepts that are not implicit in atomic physics — and so on and so on and so on. But the message of course is in the record. So I gave a lecture and some of the, many of the people were totally uncomprehending. I thought all of them were until thirty years later I have a letter from Christiane Corrolli in which she said, “Oh, I really enjoyed that lecture and have always remembered it.” She was the only original hearer I ever found who actually understood what I was talking about and who was interested in it.
What did the other solid-state physicists say?
They didn’t really say anything. They just went and said, “Nice lecture.” That’s all. And nobody carried much away. And then it was just — I guess it was in the course of time that I, I kept the lecture, and that I decided that it needed to be edited and cleaned up and that I would submit it to Science, and I guess it was further brushes with the particle physics community that mostly caused that.
Do you remember any particulars?
No, no, I don’t.
And was there any problems with publishing it in Science?
No. It just went in. They accepted it, and it was no problem.
And what was the reaction?
Again very little reaction. I didn’t hear much from anyone for a long time. The first — actually the very first time I ever heard from it independently of my own using it not as a text but to give out to students and things like that — the very first time I realized anyone had read it was in early 1977. I had a call, a completely blind phone call from a man named Eugene Yates who was — I don’t know whether he’s a neurophysiologist or a bioengineering specialist. He’s kind of a bioengineering or biology generalist who I guess he started in neurophysiology. And he called me up and said, “Are you the Anderson who wrote ‘More is Different’?” I said, “Oh yes, I am.” He said, “I’m glad to find you are still alive.” He didn’t know who I was. I don’t think he even knew — I think this was shortly before I received the Nobel Prize. It was 1977. And so after I received the Nobel Prize in a couple of weeks he called me up and said, “I didn’t know who you really were, that kind of importance.” And I said, “Yes I am.” But I went to the meeting in early ‘78 and there was a big neurophysiology meeting in Keystone, Colorado. It was great fun and people I was on the stage with were all crazy as far as I can tell, they were half crazy. Arnie Lindell [spelling?] who later got a MacArthur [correct word?], but he was all interested in psychedelic drugs and the physiology and psychology of psychedelic drugs. Ibie[spelling?] Iberall was an old Marxist who was interested in his own version of a physics theory of life. Well, it’s interesting. He’s kind of an old systems theorist, a student of McCulluck [spelling?], Oren [spelling?] McCulluck. Iberall, I-b-e-r-a-l-l. And Yates himself who was sane enough but had some far out connections. So that was the very first point at which I became interested in the complexity of general physics, where I got socialized into that group of people. But that was much later of course, but this was the very first time that I realized that it had become kind of a word-of-mouth classic — without my knowing about it at all. I never hear from it until this time in ‘77 when suddenly there began to be people coalescing around this point of view. An interesting experience. Well, I was in La Jolla. Joyce, my wife, didn’t arrive to La Jolla until later because she was in Cambridge trying to buy a house for us. I had about a year before in ’66 — In one of his many visits to Bell Labs, Nevill [F.] Mott dropped in and said he had managed to get a new professorship established in Cambridge and he wanted to use it and these new professorships were meant to be used for reversing the “Brain Drain.” The government was concerned about the Brain Drain and they wanted to get the Englishmen back to England. And there had been a very considerable Brain Drain; he was right. And he said did I have any suggestions, did I know of anyone who was interested — oh and then of course it’s a matter of courtesy when you are asking that, you say, “You wouldn’t want it, would you?” And I said, “Well no, not really,” but he pressed it a little bit, and it turned out he would be [inaudible word]. I said, “I couldn’t possibly. I couldn’t even afford to come full time, but if you can make some kind of part time arrangement, maybe I’d be interested.” And so he made this. He did fight through this agreement that he would create a temporary and a visiting professorship, backing it up with this unoccupied chair. In other words, I was seated a few feet above this chair in a different professorship that existed only because the underlying professorship was there. See, a professorship in Cambridge at that time was a very considerable job; it was essentially I guess in prestige, not in pay. In prestige equivalent to a named professorship in the U.S. It’s a very senior job. The original professors were all department chairmen — the Cavendish professor, etc. But they had just begun having professors, but these professors were still supposed to be more or less institute professors, group leaders. And so I became ex post facto the group leader of the theoretical group within the Cavendish Laboratory. And so I was in principal Brian [D.] Josephson’s boss, Volcker Heine’s boss, and anyone else that was likely to be around.
Did you have to teach with this arrangement?
A professor is not required to teach. He is required to give I think a total of eight lectures, which is, that means eight academic. If he gives eight seminars — a seminar or anything. He’s not part of the regular teaching. But in Cavendish at that time it was they had just developed the custom that everyone did supervise a few undergraduates, so I had the experience of supervising undergraduates. Anyhow, Mott worked this all out and I flew first class to Cambridge and saw the Vice Chancellor in Masters Lodge of Queen’s College and flew back, and after seeing a real estate agent and asking him to look out for a house for us. And he found us a house and so Joyce in principal was going to the auction where this house was to be sold, and I was going off to La Jolla. But she didn’t buy it, because it had dry rot. It was quite unsatisfactory. But she actually found another house and we were all set to buy that one. She came home and — well, never mind my housing problems. We didn’t end up with a house until a year and a half later. So, I had this job in Cambridge. Bell Labs kept my salary and I still earned seven months salary at Bell Labs because I was only on duty at Cambridge six months and also I was on one-sixth time during that time, so I retained seven-twelfths of my salary, and then the probable error of that salary was when I was paid by Cambridge. I almost — well, I took quite a financial beating, but it was not as severe as it would otherwise have been. I certainly couldn’t have afforded to go there full time. Cambridge salaries changed in the course of time, but my salary there never matched anything like the salary at Bell Labs.
How about the relationship between particle and solid-state physicists in Britain? Was it similar to American?
Well, it was institutionalized, and you realize that given this professorship scheme, the fact that a professor becomes – there is a professorship which is the Cavendish professor. He is the head of the Cavendish Laboratory. It is a big institution. It has several, 120-200 undergraduate students, it has large core of graduate students. It’s twice the size of a typical American department, twice or three times the size of a typical American department. And when it gets a new professor, he is perfectly free to change the entire character of the department. Mott had just become professor when I was there the first time. He followed Bragg who followed Rutherford. Under Rutherford we know what it was; it was strictly nuclear physics — except that the Mond Laboratory was established under a separate grant, separate professorship, a low temperature laboratory, by Pyotr [L.] Kapitza. The Mond Laboratory — but it was a part of the, it was a low-temperature part of Cavendish Laboratory. Then Bragg took over and he changed if almost completely. He retained a certain amount of particle physics, but he let essentially Oxford take over the nuclear particle physics. He was interested in X-ray diffraction of course and he was driving in the direction of the great discoveries in molecular biology that they did indeed make. He also allowed a, sponsored a big initiative in radio astronomy which came right after the war. They invented radio astronomy there, I mean the real radio astronomy, and so Bragg simply abandoned particle physics and started these other two things. When Mott came in, he was happy to have radio astronomy and kept it on. He would have been happy to have the biophysics stay on. Mott was very enthusiastic about that work, but it was getting too big for him. It needed a new building, and he was trying to get a new building built as fast as possible, but biophysics became too big and too unwieldy and they moved out of the Cavendish — to everyone’s great unhappiness, but there was no way we could have kept them. And Mott of course switched a lot of the laboratory to condensed matter physics. And there was a small nuclear physics/particle physics competence, but it was just left over. So there was a lot of bitterness against Mott and the condensed matter physics and Robert Frisch for instance almost never spoke to me. At the same time of course there was a totally separate physics department, the division of applied math and theoretical physics. This is the great Cambridge Physics Department that had Dirac in it and that actually Dirac has the Lucasian professorship, which is Newton’s professorship. [Inaudible word] always had Pokinghorn [spelling?], students were, I think Green was a student there. But Eden [correct name?] Pokinghorn — well, many very competent, very famous particle physicists, particle theorists were attached to DANTP. It also had a lot of applied science. It was an interesting mixture of pure mathematics, very applied science like turbulence and fluid dynamics and geophysics for instance and very high brow theoretical physics, dispersion theory and all that kind of thing. And Hawking [spelling?] of course, general relativity. Hawking is now the Lucasian Professor at [inaudible word] there. Oh, and I think the astronomer — continuous creation.
Hoyle was in the DANTP. So it was a crazy mixture, and well we cooperated with them. We taught courses with them. We taught a course in Part III Mathematics, which was their specialty. Their catchall degree course. You could study anything in Part III Mathematics in DANTP, but you didn’t study experimental physics, and there’s always been an excessively wide split between experimental and theoretical physics in the Cavendish. But I did teach a course and I continued to teach a course that fit into Part III Math and the physics graduates, many of the physics graduate students would come to theoretical physics. I tried to do Green’s functions. This was my period when I learned Green’s functions in order to teach them, [inaudible word] AGD and Kadanoff and Martin, and Kadanoff and Baine [spelling?] and so on, in order to teach them [inaudible word].
What did you use to learn the Green —?
I used AGD — well, actually I used Kadanoff and Baine, which is not a well-known book but I think it’s a good one. And —
And did your collaboration with Josephson continue?
I didn’t collaborate with him, but he worked with a man who was hired to be my assistant, John Leckner. And Josephson and Leckner wrote some very interesting papers on ions in helium III, and he was working critical phenomena at that time — some of the early exponent relationships [inaudible word] Josephson at that period. Josephson was in principle my responsibility and Heine was a very close friend, became a very close friend. We worked very closely.
Yeah. Heine, H-e-i-n-e [pronounced hy’-nuh] for your transcriber. And his first name was Volcker, V-o-l-c-k-e-r. He was a German exile from New Zealand. His family was exiled because of Quakerism, not because of being Jewish. And got his degree in New Zealand and came to Cambridge and never left — and he’s still a very close friend.
Before we get over to papers and work, should we taught about the summer schools like Verona and others or shall we go to papers now?
Okay. We should go back and fit in a couple of things. Mostly we covered superconductivity, superfluidity — or superconductivity in the myuse [correct word?], and my first incarnation in superconductivity last time, where it talked about the tunneling spectroscopy. And also the work with Kim [spelling?] on flux flow and flux creep. But then I gradually began to develop a general point of view and I was mostly occupied at that time with codifying it and organizing it in my head. And the result was a couple, several articles on coherent matter field phenomenon and superfluids. Then there was this little article with Bern Matthias, joint article about why or summarizing why we really believed in superconductivity, in the BCS theory and all that. Then finally I saw the analogies — [inaudible word] been making a lot of analogies, and one I became interested in was the analogy between superfluids and superconductivity, and I suggested this experiment to Paul Richards, the Richards-Anderson experiment, proposed this experiment as in an analog of the A. C. Josephson effect experiment of Shapiro [spelling?] and Damm [spelling?] and myself. So Paul Richards worked it up and in the first paper we used a very crude apparatus and we seemed to be seeing the effect.
Was it at Bell Labs?
At Bell Labs. This was — I think the world as a whole considers that this is one of the many times when I have been actually wrong, because what happened then was that Paul Richards refined the apparatus, the original apparatus, and had small — We were measuring the chemical potential difference by just measuring height difference in a couple of capillaries, and they were open capillaries and it seemed a very noisy measurement so Paul thought it would be much cleaner if we closed those capillaries and made it a nice clean experiment. And then he saw a beautiful step [spelling?] phenomenon and he published that under his own name a couple of times. But I was reluctant about that, because I was not sure that one could get proper measurements of pressure differences with the apparatus closed off in that way. Well it turned out that he later proved that — or someone else later proved — that the experiments were merely measuring acoustic resonances in these closed capillaries. But nobody ever proved whether or not we were measuring acoustic resonances in the open capillaries, because we couldn’t have had sharp resonances. And we might have had broad resonances and maybe nonlinear effects or zeroed in on them. Anyhow, I’m not sure whether the first letter [correct word?] was right or not. In any case the effect — and I didn’t make proper estimates of the magnitude of the effect, and it probably much later, later work by other people showed that it was probably much weaker than what we were likely to see. So it’s very likely that we never saw it. But that did stimulate me to write this coherent matter field phenomenon in superfluids paper which has some very interesting ideas about superfluid helium, about vortices in superfluid helium and in fact about motions of incompressible fluids as a whole, and it’s the only paper I ever wrote in hydrodynamics and it has in fact, as far as I know, a unique equation of the hydrodynamics of incompressible fluids. Where is it? [flipping through pages] No, no, this is not the article. “Considerations of Superfluid Helium” [correct words?] — that’s the article. And this new corollary in classical hydrodynamics which you can derive from [inaudible word or phrase] equation, which is the equivalent of the Josephson equation for a superfluid helium. Then I also wrote a review article for Gorders’ [correct name?] series “Progress in Low-Temperature Physics,” which is the coherent matter field phenomenon. So I was busy codifying, but there was not much science to that. Then the other thing that got done in that period when I was at Bell Labs and not, I hadn’t gone off to Cambridge again, was I worked a lot on informal theory of resonances in metals. You asked whether I’d ever done band theory. I haven’t really done band theory, but I have done one-electron theory. I studied the resonance phenomenon, and I even tried to do some [inaudible word] scattering theory with Bill McMillen [spelling?] that I’ve talked about in Verona. Again that contains an error. The error was a programming error by McMillen, so the actual numerical results we had – we had a little model of random scattering by resonances which we worked out, and we just got it wrong as a graduate student of Walter Kohn’s later proved. It was a matter of at some point there was one line of code that said that theta = sine theta [correct equation?; θ = sine θ], and it meant what it said. We were substituting theta for sine theta and it just didn’t work very well. So in the numerical calculations there is a wrong, some wrong computing. But in the theoretical calculations I did some very careful mathematical derivations of Friedel’s [spelling?] sum rules and things like that in a very general form. And that is an interesting and complicated — although in some senses trivial, because it’s not interactive; it’s a linear theory. But it has some very interesting consequences which I used a lot of later, and I’m all for it. Those three, those were two or three things that I did in this [inaudible word] period before I went off to Cambridge.
Did you change the topics or the direction of [inaudible word] once you moved to Cambridge?
Well, what I did was all of a sudden there were graduate students. I had never had graduate students, never had anyone that I could tell to do something and–
Approximately how many?
Well, what we did was I took on two a year — I think maybe three the first year and after that two a year. I kept a steady flow of two a year, and we got them out in three to four years. That would mean I had seven or eight when you give them time. And some of them were very good, and some of them were to say the least not. It was a very mixed bag, because — well, initially we were a little subgroup of the Cavendish Laboratory which was not the famous center of theoretical physics of the University of Cambridge. The famous center of theoretical physics was the DANTP. We had had a strong and a very good series of students and postdocs. And Ziman had been there, Ziman and Heine. They were a very powerful team. But Ziman had gone off and become the professor at Bristol and taking his students with him, so Heine was left with a group of students, some of whom were good and then a couple of students in lots [correct word?]. So we tended at first to get the ragtag and bobtail people from all over the world. We had a Singaporean and we had a Turk and we had an Israeli, a crazy Israeli, and we had a little woman from Singapore who talked very high just like this [imitating high and squeaky voice] and was terribly shy, and we had, Volcker [Heine] had a Ghanian and a Sri Lankan. He called him Apple Pie, but his name Apapillye [spelling?]. And it wasn’t until I’d been there about five or six years that we began to establish a reputation — or three or four years. Suddenly the top flight English, first from the University of Cambridge, began to come to us. And then we had a sequence of real crackerjack people, the best students I’ve ever had. But initially they were good students but they were a strange lot. And so I had to think of a lot of things for them to do, and I did, and some of them were good and some of them were not.
Did you think of some strategy where to develop or where to direct them? Or was it [inaudible phrase]?
No, no. I had a lot of backed up ideas, and you know there was no lack of ideas for them to work on; it was just what did that particular [inaudible word] want to work on.
I was just thinking about maybe some idea about where the field in general is going — or was it all different kinds of problems?
The most desirable theoretical consultant at the Bell Laboratories, which was by far the biggest generator of condensed matter physics problems or condensed matter physics in the world and also knew everything that was going on in condensed matter physics in the world. And so basically I could just listen to what came through my office. People who dropped into my office I would talk to —
So you basically profited in your work with graduate students from your work at Bell Labs.
Oh yeah. Of course, of course.
Was there also a flow in the other direction, in the opposite direction?
Oh yeah, yeah. Well, for instance I took many of the graduate students — in fact I think one a year was the average — one a year would come and work for the summer with me at Bell Labs. I did that with Davidson, yeah, Davidson, I did it with Armatage [spelling?], Armatage, I did it with Mike Cross [spelling?], I did it with Omaine [spelling?], Richard Palmer, I think Gideon Yuvall, and Gideon Yuvall a home in — well, we didn’t put him up. We found him an apartment. Joyce became very good at finding apartments for them or places to stay. Gideon Yuvall was the crazy Israeli, and we took our lives in our hands when we taught him to drive. But he drove — well, never mind. Never mind. He was wild. So, they would come. Mike Cross for instance came and worked with me on, first on the liquid helium ideas, but then he began to get interested in helium III and picked up a collaboration with Bill Brinkman [spelling?] at Bell Labs, and there were several papers out of that. The big paper on the Kondo effect is Anderson, Yuvall and Hayman [spelling?]. Don Hayman was from Bell Labs. It was a consequence of one of these summers. And so on. So Volcker came over sometimes and some of his band theory students came to the labs.
And he was doing all the time mostly band theory?
Well, we divided the field very much. He was already – he had invented the pseudo-potential method and was refining the pseudo-potential method, so he was already in that business, although he was interested in many different phenomenon, but he became more and more into that, and therefore I tended to move over to the many body side and localization and so on. So it just worked out very well. It was very, very convenient. It was kind of — everybody wins all at once. The students got the Bell Labs experience as well as Cambridge; Bell Labs got me plus students, rather than having half of me they had more than me because they had me plus students and postdocs. Cambridge got a very effectively working condensed matter theory group where a lot of things were discovered in the course of eight years or so. One of them still may make a Nobel Prize; we don’t know. They keep putting Sam Edwards [spelling?] up for that. Well, let’s do —
Shall we go to the Infrared catastrophe?
Yeah. This is — the paper is there, 1967, which of course means it was submitted in ‘66. This was actually what I was thinking about the rest of the time in La Jolla, that month in La Jolla, so I can remember very well the period when I was thinking about this problem. And I just didn’t think adequately, but it did start off a whole field of physics as a matter of fact. It started with John Hopfield [spelling?]. Hopfield had been a postdoc with us and then gone to Berkeley and then finally went to Princeton and achieved tenure. He was a condensed matter theorist and therefore he was kind of low on the totem pole in Princeton, but he stuck it out and he had a lot of very good students. And one of his students, Gerry Mann [spelling?], became interested in the problem of what became known as the X-ray edge problem. It has a couple of versions, but the one that he noticed was in doing X-ray photoemission spectroscopy on metal — which means that you are — Let’s see, how do you do it? You insert — well, you certainly hit the metal with an X-ray. Yeah, it’s straightforward. You hit the metal with an X-ray photon and the X-ray photon kicks an electron from an inner shell to the neighborhood of the Fermi level. There is an empty band of electrons above the Fermi level. The electron falls back from that empty band and emits an X-ray proton, falls back into it. No, no. When you kick an electron out I guess, then the electron falls back from the band into the X-ray and that’s it: it falls back from the band into the X-ray and so you have a filled band. You fall back from the filled band into the empty state, empty core state and emit an X-ray photon. So you’re inserting an X-ray photon and you get out of it a photon. You are not measuring the photo emission; you are measuring the X-ray edge with X-ray. There are other experiments. You can emit the electron from the filled band directly to the vacuum. I don’t remember all of them, but the basic thing is that with this X-ray you have changed the occupancy of an inner shelf and so the potential that the electrons and metal sees — not the potential of the, not the bare potential of the potential of an unperturbed metal — but the potential of unperturbed metal with a missing electron in a core level. So there’s a different potential there. When you look at these spectra you expect to see just the electrons are filled up to the Fermi level, you expect to see the electrons from the Fermi level down coming back [inaudible phrase] and the corresponding photon [inaudible phrase]. You don’t. You see a singularity at the end [correct word?], at the edge of that spectrum at the Fermi level. The electrons which come from near the Fermi level, sometimes the electrons are particularly are particularly stormy [correct word?] but if you see [inaudible word] big rise or sometimes a big drop at the Fermi level. This phenomenon had been known for a while. Of course, you know, there is the usual experimental fog and but it was beginning to become clear. And Gerry Mann had been working for John Hopfield and developed some very complicated theory in which he said, “Look. There is something quite interesting going on here. The electronic state of the metal with a change in potential at one of the sites is different from the electronic state of the pure metal in some interesting way.” And he was able to, he had some kind perturbation theory which gave an enhancement of the intensity or a decreasing. Well, I think he saw an enhancement of the intensity right near the Fermi level, and by something, you know, doing second order diagrams. And then he said, “Look, the third order diagram is even worse,” and he kind of speculated that there is a similarity there. Walter Kohn, who was still visiting Bell Labs, said, “No, no, no, poo, poo, poo. I proved with majumdar [correct word?; spelling?] that nothing happens when you change the potential of an impurity.” He showed that changing the potential of an impurity or introducing an impurity potential into a metal, the potential just gets screened and then he claimed he proved that nothing further happened. What he did was to calculate the one-electron Green’s function under those circumstances, and he seemed to find that it was completely – well, at least that it had no singularity when you caused a balanced state by your [inaudible word] potential. Therefore [inaudible phrase]. And we all knew that there were many, many examples where you could get roliances [correct word?; spelling] using the Fermi surface as a source of singularities. There is a kind of an obvious logarithmic singularity that is likely to happen in perturbation theory. And everybody thought Gerry Mann had found one of these phony algorithmic similarities. And John Hopfield and I got to talking about this and I suddenly realized that, well, phony similarity or no phony similarity, there was this strange phenomenon that when you added an impurity potential of any kind the state of the electron, of the Fermi sea is orthogonal to the pre-particle state. In other words that a potential, any potential always causes enough displacement of the electrons that it is orthogonal. So this was the origin of the physics of what we came to call Fermi edge singularities. This was the very first paper. I found that although there is nothing that happens in the energy, if you look at the normalization constant or at the overlap between the starting state and the ending state, that that overlap has a logarithmic divergence in the number of, in the size of the sample. And that was the start of it. But I never was able to get the actual X-ray of singularity. I could say, “The edge itself, that’s going to be singular,” but [inaudible phrase] divergence. Because of course the thing which was [inaudible word] the electron was emitting with no energy, that’s the [inaudible phrase] itself. But it wasn’t until quite a bit later that I understood what was really going on. In the meantime Nossiere [spelling?] and Gagiray [spelling?] had written two very, very long papers in which they derived a singularity, and then Nossiere de Dominici’s [spelling?], spoke up in the back of the room and said, “I can do that on the proverbial document envelope” and [inaudible word] the real paper and [inaudible phrase] de Dominici’s was published and showed that there was indeed such a singularity. Well, this was a problem that actually called for me to have a student work on it, and so I had this very bright and crazy student Gideon Yuvall from Cambridge. He started out working on exactly how this problem was to be done. The second thing had been happening during these years, the condo effect had been discovered, and as I said a couple of —
Was it in ‘64?
Sixty-three, ‘64. And the acclimation of the Kondo effect had been discovered, but nobody understood what the ground state of the Kondo system was and so there was a series of papers by Abrikosov and by Suhl [spelling?] and by many other people trying to find what happened at absolute in the condo problem. At the same time I had been, of course had the Anderson Model, the magnetic impurity model, and I was interested in what many body effects might happen in the Anderson Model. There was work by Hayman with Schrieffer trying to do the Anderson Model by Albert Structonovich [spelling?] techniques. And both Schrieffer and I came to the conclusion that the Hartree-Fock solution of the Anderson Model wasn’t really the answer, and we both realized that some of this had something to do with the condo problem, but neither of us went very far with it. And so in the end I wrote up this paper “The Ground State of a Magnetic Impurity in a Metal” which had a very refined version of the infrared catastrophe in it. [Inaudible phrase] just the mathematics of the infrared catastrophe in a very neat, beautiful form. And this, well I think came out simultaneously with Nossiere de Dominici’s, who both realized that it is the face shift and not the side [correct word?] of the face shift; it’s the s-matrix [punctuation?], the scattering matrix and not the t-matrix [punctuation?], which is relevant for the X-ray edge problem. And this is one of the two interesting theorems that I was mulling over in my head at this time. The Friedel theorem which says the number of particles is equal to the face shift and this other theorem which says the orthogonality undergoes a logarithmic divergence and the orthogonality [inaudible word] is equal to the phase shift squared. And both of them obviously depend on face shift and not on trigonometric functions of face shifts. And so they are nonperturbative. They are intrinsically nonperturbative results, and one is quite interested by that when you get a truly nonperturbative result. Perturbation theory of course always gives you this scattering matrix [inaudible word], works from the scattering matrix [inaudible word] and does things in a sequence of the Feynman diagrams, but no sequence of Feynman diagrams will ever give you something that’s linear in the face shift or quadratic in the face shift itself — because the shattering potential, the scattering is the same for face shift = 0 and face shift = π. So this is outside Feynman diagram theory. So I was puzzling about — I actually took this paper and a bunch of other stuff to the Les Kouche summer school the first time I went there.
What year was that?
That was ‘68, the summer of ‘68, so I had been in Cambridge a year. And I talked all about magnetic impurities and about the Friedel theorem. I didn’t do the mistake in multiple scattering theory, fortunately for [inaudible word], but I did this ground state theory and I talked about the face shift squared theory and various other aspects. So it was about that time, well, Gideon was thinking about it. He had already made one contribution which was a very nice one. He said, “Look, you really did Nossiere de Dominici’s. All along you had it right, because obviously in a Fermi system there is a unique velocity and so if you have the dependence on the size of the system, you also have the dependence on time because the size of the system and time are related by the velocity.” So if you have a linear velocity a logarithm of system size turns into a logarithm of time and an exponential of a log of time gives you a power of time and that’s exactly what, that is essentially what Nossiere de Dominici says [correct word?]. So Gideon was already showing up as very bright. But then in the course of ‘68-‘69 Gideon and I had a sequence of brilliant ideas. The first was this equivalence to the classical one-dimensional cooling gas which was Gideon’s — well, Gideon got it by what he — he just loved the Muskhelishvili technique, that certain mathematical technique which I still find difficult. It involves the Weiner-Hoff [spelling?] equations and things like that, that I find very hard. But Gideon used the Muskhelishvili equation to show that the Kondo problem was equivalent to a time, a problem in which the spin is flipping back and forth as a function of time and the flips are connected by a certain potential, a certain power law potential. And if we could only solve the problem of these flip objects, these pluses and minus-plus, minus-plus, minus along the chain with the interactions within the chain we could solve the condo problem. And then he and I both realized that this is the same as the one-dimensional cooling [correct word?] gas, and we looked in the book for the one-dimensional cooling gas and found that that is the case which — well, I guess Dyson was just in the course of solving one-dimensional gases with power law potentials. But he said the case we had was still indeterminate; nobody had solved it. So we were stuck with that. And then I began to think, and I thought, and I thought, and I thought, and I reproduced essentially the thinking that Gelman [spelling?] and Goldberger [spelling?] had created some ten years before when they were trying to solve renormalizing electrodynamics, and so I reinvented the renormalization group for this problem – which was the first time the renormalization group had been used in a condensed matter problem. I did it some various ways, but in the end we essentially did it in space in time and under the first paper was this 1970 Physical Review B1 [punctuation?]. The second paper was “Scaling Theory Qualitatively Correct Solution and Seeing the Results” which was Yuvall and Hayman. And he had come over that summer. And Yuvall was very nervous about getting it published, because actually the Physical Review — I have never forgiven Bob Schrieffer for this, because the Anderson, Yuvall and Hayman paper would have had priority on the renormalization group if he had not blocked publication of it because he didn’t believe it. He heard me present it at a little meeting that we had in summer 1970 and decent enough to stand up and say “I’m sorry.” Well, I said during the talk, “Some idiot is blocking this.” He stood and admitted to being the idiot and he went through. But I don’t claim. You know, essentially my renormalization group theory was a one-parameter theory that was very like the original Gelman-Goldberger theory and was not the Fisher [spelling?] and Wilson and Kadanof renormalization group theory. It didn’t have — I mean, essentially I assumed universality; I assumed that certain things were irrelevant. I knew they were irrelevant, but I didn’t give an argument why they were irrelevant. So I used the concepts of universality and irrelevancy, but I just assumed them without giving them any real discussion. So those concepts — I think those are the important things — those concepts are what Wilson and Fisher added to the renormalization group that made it an important theoretical, truly important theoretical development. And I used the concepts but I didn’t know that I was — you know, I was speaking prose [correct word?] but I didn’t know I was speaking it. If you look back at that paper you will realize that I say, “It doesn’t matter whether you have sharp jump or a rounded jump, it’s not going to be relevant,” but I didn’t use the word irrelevance and I didn’t say, “I can prove that it’s irrelevant by showing that it gets smaller as [inaudible word] longer times.” I just — well, I could prove that but I didn’t say that I could prove that. And I had irrelevance, I had universality, and I had the renormalization group. Gideon and I had it — because Gideon did contribute much of this. But particularly this insight about the time and size being the same and the insight about the power law interaction, and I think I mostly contributed the insights on the renormalization. And that of course was that. Scaling theory and so on.
The second paper, right? The exact results [inaudible phrase]?
Yeah. Well, the first one was the one body theory, the second paper was [inaudible word] theory, and space [inaudible word] and scaling I summarized it at the ILT [spelling?] in Japan in 1970. And Gideon wrote it up also in 1970 for solid-state communications because he was so — it was his thesis after all, and this damn referee was in the way, so he wanted to make sure that he found out. And finally this poor man’s derivation of scaling laws, that’s the only thing that Wilson has ever acknowledged that I contributed, but in fact that paper did. That paper two was equivalent, except for numerical calculation and his numerical facility with a computer, numerical tricks. That’s absolutely equivalent to Wilson’s solution of the Kondo problem and has in fact some results that Wilson was unable to get. But Wilson has never been eager to acknowledge that contribution. Incidentally this is the beginning of the great explosion of equivalence making transformation back and forth from quantum problems to classical problems, from one problem to its inverse, and so on. And this was a problem which starts out as a quantum problem in quantum fluctuations in the spin model, the [inaudible word] equivalent to the classical one-dimensional gas. That in turn is equivalent to a certain Ising model, although these are exact transformations among these three models. It turns out also that it’s equivalent to the signal model, the quantum mechanical signal model, and in the course of doing it I believe we were the absolute first to write down the famous [inaudible two words] renormalization diagram for the signal model in one-plus-one dimensions — which has infinity applications to one problem or another. We also in the inverse square one [inaudible word] of the [inaudible phrase] paper we also showed that the statistical mechanics of this ferro magnet had this peculiar structure where there is this —
What paper is this?
The J. Phys. C. [correct name?] paper.
Oh, the poor man derivation?
No, no, it’s numerical results on the Kondo problem [inaudible word] the inverse [inaudible word] one-dimensional icing mode. And that is also the origin of Costowitz [spelling?] and Thalmus [spelling?], the Costowitz-Thalmus theory of two-dimensional condensation of vortices. This was the first rigorous theory of a defect-dominated phase transition. This is the one-dimensional phase transition. It’s a phase transition in inverse square Ising model, and the phase transition is dominated, is caused by the appearance of the spin flips. Either — or the spin flips in one region can renormalize to irrelevant, they disappear and the spin, the outer spin stays finite. At a higher temperature the spin flips proliferate and well in the end you have zero average and the phase [correct word?; face?] transition occurs by both a singularity and a jump. And this jump occurs at a — well, it was proved by David Cowlis [spelling?] who became interested in our work on this problem, and he proved that this jump had a certain universal numerical size, had to have that, and it was his knowledge of this problem that led him to be able to solve the Costowitz-Thalmus problem, the vortex problem, vortices and then thin film in a superfluid film. So this is the origin of Costowitz-Thalmus and uses the same techniques that are in this J. Phys. C. paper. So, it was the first renormalization group in condensed matter physics and it was the origin of the, it was the first theory of a defect-dominated phase transition. Well, maybe not. In a sense there are ways of looking at the two-dimensional icing model that would give you similar results. But it’s not much different. In 1970 this was the period when Josephson was breaking down mentally. He had a really bad period in there for a while. He was doing better and better physics, he was doing wonderful physics, but he, you know, the Josephson Effect became bigger and bigger and he was made an FRS and he was given all kinds of prizes. He was awarded the Medal by the Franklin Institute even though Pippard [spelling?] and I wrote and said we don’t think it’s good for him to receive all these medals so quickly and to have to try to do all of this traveling. Well, I think we were wrong to do so. But he did receive all these medals so quickly, and at the same time his mental health was deteriorating. And I don’t know exactly how bad it got. Pippard knows; I don’t. Pippard essentially was his original Ph.D. advisor, the person to whom he actually talked. But anyhow, when he finally was selected for the London Prize at the International Low Temperature Meeting, he was in a nursing home and it wasn’t possible for him to go to Tokyo. So I went and gave his lecture for him.
It was about how he discovered the Josephson Effect?
Discovered the Josephson Effect. And at the same time I had forgotten my [inaudible word] time in a totally different session I talked about my space time scale stuff. That was the first time I’d been back to Japan. [Inaudible word] was still almost itself. It still, it still stayed in the Japanese [inaudible word] and we visited the Geisha Quarter which was not a red light district but an entertainment district. I had a lot of other students of course, did some various other interesting things. One other student actually based his thesis on this work, John [correct name?] David Armatrage who later left physics and became a paint executive. He actually did a rather bang-up numerical job of the renormalization group, of the calculations that Yuvall and I just talked about. But he published that on his own. In fact maybe he only wrote a thesis about it, because by this time Wilson had done his numerical job, which he didn’t do until 1972. It was three years later or so. So we actually did the numerical work, but eventually Wilson got the credit. And he deserves the credit. His techniques were incredibly cute, and he understood some of the — there are many things to understand about this thing. It contained a wealth of interesting detail, we had this interesting idea of effect of the energy levels changing their order and [inaudible word] and alternation and so on but was really very cute. And is important, because it shows that in the end you renormalize from a system that has a spin to a system that doesn’t have a spin and that was [inaudible word] done. But in principle I think we discovered more about it than he did. Okay, spinning glass [correct word?]. This was the student Wai-chao Kok, K-o-k for your transcriber, W-a-i-hyphen-c-h-a-o. She was a very shy little girl from Singapore and I had always been puzzled about the problem of random spinning purities in any system. The copper manganese problem was the conical [correct word?] one, we’d have manganese spins, presumably described by the Anderson Model, and if they are concentrated enough they never condoize [correct word?], never get down to the condo model, so you get some kind of randomly arranged spin structure. I was conscious of that and conscious that I didn’t really understand how to deal with it, and a lot of things that — I mean, I had all kinds of thoughts about it. Suhl and I had once done a paper on the effect of these random impurities in superconductors. I don’t remember where Anderson and Suhl is. It was way, way back, ‘62 or ‘63. I don’t even know whether it got into the — [shuffling through papers] No. It doesn’t seem even to have gotten into it. Bibliography. Anyhow, it doesn’t matter. What we showed was that if you have a superconductor then you get a long range interaction and it may have some — Well, we didn’t quite understand the effect. We were supposing — well, yes, therefore that would enforce a certain scale on it and you might get a spin density wave. But I don’t think you can actually do. And I don’t know why I published it in the Materials Research Bulletin. I at this point took up the localization theory again. There was a discussion, there was an article about localization and now it didn’t happen by Browers [correct word?; spelling], and at this point I really believed in localization, I began to believe in localization, and I woke up again to localization and there are two papers about localization here from 1970. So I began to be interested in localization, and I thought, “Well, maybe localization theory will tell us what’s happening in this copper manganese problem” and did essentially a, I made the argument that since the problem is nonlinear if you have any localized item values they are going to gradually wipe themselves out. In the end you may arrive at an extended but randomly arranged arrangement of the spins. And so I suggested that maybe that’s why we actually could have a phase transition with a randomly — I didn’t say “phase transition,” unfortunately; I just said [inaudible word] with the copper manganese problem maybe you do get rigid random [inaudible word] structure. I did not suggest that there was any phase transition. But I was thinking about it and I had a little idea about that and I set Wai-chao Kok on that idea and gave a couple of talks about it. Here: “Comments on the Perimagnetic Cury [correct word?] Temperature and [inaudible phrase],” [inaudible publication name], 1971. That was the reason for playing with the copper manganese with this random spin problem was having tended, having to have something that was easy enough for this very shy woman lacking in self-confidence, a girl lacking in self-confidence, what she could do. One of the most excruciatingly painful dinner parties I think we ever had was one where we invited both Brian Josephson and Wai-chao Kok, and neither one of them could somehow speak above a whisper. And so Joyce and I talked to each other for an entire evening. Josephson was painfully shy, particularly in the presence of women before whatever happened to him when the nursing home happened, and she was painfully shy in the presence of anyone senior like myself and it was hopeless. What happened here was, of course you see in 1970 that is — I didn’t invent the transition and I didn’t such the transition, but I did say the words spin glasses, so that’s where spin glasses entered the universe is in this interesting idea that it’s an extended — it’s kind of like an extended state of the interaction Hamiltonian. I was suggesting it’s a random [inaudible phrase] of the random orientation [inaudible word] value of interaction Hamiltonian. I kept puzzling about that problem until four years later. This wasn’t a graduate student; this was Sam Edwards. I was on the committee which appointed Sam Edwards to come and be the successor to Robert Frisch and one of the named chairs in the Cavendish, but he was head of the — the English equivalent of NSF, he was University Grants Committee. Not the University Grants Committee — the SRC, Science Research Committee — Commission. And so he came up only on Saturdays, and he thought about physics on his way back and forth on the train. And so every Saturday we would meet in the department and over coffee we would discuss this serious problem of this random interaction, random exchange interaction between spins. And I had my ideas, this idea of the extended [inaudible word] value, and by this time there were experiments by Mydosh, M-y-d-o-s-h, and others. I shouldn’t forget the others because they get very unhappy, but I forget their names. There were several other, two or three others who eventually showed experimentally that there was a sharp transition in his materials. And since that I’ve got this model, this technique that I’ve been hiding away in my notebooks and trying to use it on gels and polymer gels and other problems for solving random Hamiltonians, random statistical mechanical problems. Brought it out and dusted it off and tried it on this problem and it worked, so we published the replica method together and the idea that there is a phase transition in a spin glass can occur. In this random Hamiltonian or magnetic spin system with random sine [correct word?] frustrated spinning directions. Also at the same time I invented the term “frustration.” I think I used it in that. And frustration has become a buzzword in the whole question of complex systems.