Gordon Baym

Thomas:

This is William Thomas with the American Institute of Physics. The date is January 24, 2024. And I’m speaking with Gordon Baym, who is going to be telling us essentially about his entire life. So, why don’t we start at the beginning with some of your family background and where you’re from?

Baym:

I came into this world in New York City, July 1, 1935. My parents were living in upper Manhattan in what is called Inwood. The area turned out to be a quite a nesting ground of young physicists. Shelly Glashow, Roy Glauber, and Leo Kadanoff all lived within a few blocks of each other and me. I went to PS 52 (Public School) at first, where Shelly was a few years ahead of me. My parents moved to Dobbs Ferry in 1943 and then to Brooklyn a year later, which led to my going to Brooklyn Technical High School. Tech, as it was called, was a wonderful place to be. I had a marvelous math teacher there, Isador Glaubiger, who gathered students who were interested in mathematics together and said, “Just learn the mathematics by yourself, just take the Regents tests (required tests in New York State to show that one was proficient in a subject) and we’ll learn interesting things.” Which we did for four years, including learning calculus, which at that time was not something that was normally taught at the high school level. He had a tremendous amount of influence over me at that point. Afterwards I went on to Cornell, which was relatively easy after Brooklyn Tech.

Thomas:

Can I stop you and ask you a little bit about your family? Were any of them involved in science or anything like that, or did you come to it on your own?

Baym:

My father was always interested in science. He wanted to go to Syracuse University, where he could study chemical engineering. And that being a land-grant school, it wouldn’t have cost anything. But he was told very explicitly, “Chemical engineering is not a field where a Jew could ever get a job.” So, he became a lawyer instead. But he certainly maintained an interest in things technical, and that had a strong influence on me. For example, he always helped me build radios when I was a little kid. During the Second World War, he got into more technology, first working in engineering drafting for the North America Phillips Company, which had just begun in Dobbs Ferry, and when we moved back to New York City, he continued working in engineering drawing. That technological aspect was very good, but I really had no model when young for what physics was at all. In fact, my dream at age 10 was to own a radio station. That was the closest I could imagine getting to technology.

Thomas:

You were more influenced by things like radio than you were by the atomic bomb.

Baym:

Well, I was 10 years old at the time of the bomb. No, that was a very memorable event. I actually remember the day it was dropped on Hiroshima very clearly.

Thomas:

You mentioned there were all these other people you knew later who grew up in a one-block radius. Were you with any of them at the time or in high school? Did you know people later on?

Baym:

Only later on. Roy Glauber was an exception, being the person who knew me the longest in the world because his parents and my parents were friends. He knew me from the time I was a tiny infant. Leo Kadanoff, I only met much later. We were contemporaries at Harvard and worked together in Copenhagen and Illinois. But I didn’t know him when I was little.

Thomas:

Okay, so now I’ll let you go on to Cornell.

Baym:

At Cornell, I began actually in electrical engineering, which I think I stayed in for maybe a day or so. Then, I switched to something called engineering physics, which I hadn’t known about. It would now be called applied physics. I found myself not wanting to take courses in foundry and other technical-type courses, but rather taking courses in art history and such. So, I switched into the arts and sciences program and became formally a physics major and math major at the same time.

I should have mentioned that my high school, Brooklyn Tech, was one of the three specialized science high schools in New York. There was also Bronx High School of Science, which is famous, of course, for producing so many great scientists, and Stuyvesant High School, which was very good but hadn’t produced as many well-known people. Brooklyn Tech was set up as a way of training engineers without their having to go to college. We had courses in machine shop, made our own drill bits, operated lathes, milling machines, everything under the sun. We learned to use Linotype machines, did metal casting, all kinds of very practical skills, which gave me a strong sense of being able to do things with my hands.

The atomic bomb, as you asked, did have an interesting influence on my education at Cornell, where I was from ‘52 to ‘56. Hans Bethe had come back after being director of the Theoretical Physics Division at Los Alamos during the War. He brought a vast number of people with him, experimentalists as well as theorists including Feynman and Philip Morrison. And they had a very strong influence on the way physics was taught at Cornell. I remember the senior honors program run by Philip Morrison, which focused on neutron diffusion, a topic of great programmatic interest at Los Alamos. But not obviously mainstream for a budding young physicist.

Thomas:

Not a very typical undergraduate topic, I don’t think.

Baym:

No, not at all. I actually started in my third year to take graduate courses there. At this point I met Shelly Glashow, and he had a large influence on me at that time. He and Steve Weinberg were a couple years ahead of me at Cornell. But Shelly kept giving me advice on what to do, which was very good. One time, he mentioned, “I’m going to Harvard to be a graduate student of Julian Schwinger.” I said, “That’d be good. I think I’ll do that, too.” So, I became a student of Schwinger. And then, at Harvard, I said, “What’s next?” He said, “Well, I’m going to Copenhagen to be a postdoc at the Niels Bohr Institute.” It was not called the Niels Bohr Institute then—Bohr was still alive—it was the University Institute for Theoretical Physics.” And I said, “That’s a good idea,” So, I did that as well. Then, in Copenhagen, I said, “Okay, what’s next?” He said, “I’m going to Berkeley.” I also went to Berkeley for a year, also. Actually, Shelly was at that time on the faculty at Berkeley and had wanted to keep me at Berkeley when I was there. But in retrospect, they had me give what would now be called a job talk, where I gave a colloquium on things I was doing in transport theory. It was not a talk designed to make myself appealing to experimentalists and such.

I had an extremely good education at Cornell. One of the wonderful people who was l there was Ed Salpeter. I often wonder whether I should’ve stayed there and become a graduate student at Cornell, and gotten involved in astrophysics at the very beginning. Salpeter was wonderful at the intersection of astrophysics and many-particle problems, at. But I didn’t, I went to Harvard instead.

Thomas:

Was there any question about where you thought you were going to go? You went to Cornell. I don’t think you might have mentioned any other places you might’ve been looking at. Was it always Cornell?

Baym:

That’s a very good question. Cornell was my first choice, and City College of New York was a backup, as was Union College. The only issue with Cornell was the tuition, which was a walloping $300 a year or so. But I did get a New York State Regent Scholarship. And I also got some support from what is called the Leopold Schepp Foundation, a similar amount of money. I was able to get through school without any financial questions or problems because of these. Financing my education was never a question.

Thomas:

Then, you went on to Harvard. There, you were following Glashow’s advice.

Baym:

Yeah, except I actually began Harvard in math, where I started working with George Mackey.

Thomas:

I saw you had a master’s in that.

Baym:

Yeah. I was interested in the mathematical structure of quantum mechanics. He was also. But I was also interested in physics, and I sort of fell in the crack between the two fields. The mathematicians felt I should be in physics, and the physics department felt I should be in mathematics. I don’t remember quite what instigated it, but I did switch to becoming a formal physics student.

Thomas:

Had you thought that you might have continued on all the way through in mathematics?

Baym:

I could well have. I should mention, as an undergraduate, that I got involved in a project to write a book on group theory with a wonderful physicist-mathematician named Harold McIntosh. He was sort of a dropout graduate student at Cornell. There were a lot of these people around. And he was the one who taught people like Shelly Glashow and me group theory. Unfortunately, that book project was never finished, but it would’ve been good to have done. I still have the written chapters and notes on my bookshelf.

Thomas:

You’ve moved from electrical engineering to engineering physics, and now you’re on the theoretical side of the physics to the point that it almost could be mathematics.

Baym:

Exactly!

Thomas:

Was there a point where it really occurred to you that this theoretical angle was really your thing?

Baym:

Theoretical angle in what sense?

Thomas:

Well, in the sense that you really wanted to do sophisticated theory, to the point that you’re really looking almost at the theoretical structures rather than…

Baym:

I really had no idea how one went about doing physics. There was one other important influence I have to mention. My mother’s best friend from high school named Gertrude Percus had a son, Jerry Percus who was becoming a physicist. Jerry was 10 years older than me or so. But every time we’d visit the Percus’s in Manhattan, I would end up in Jerry’s room, learning about physics. Jerry became a student of Yukawa at Columbia, and ended up being a rather well-known many-body physicist. He was based first at Stevens and then at the Courant Institute at NYU. He unfortunately died just a few years ago. It was Jerry who pointed out to me the possibility of being a physics.

Thomas:

When we were discussing your early background, did we mention your parents’ names?

Baym:

My father was Louis, which, of course, was Americanized. He was born Lieb. My mother was Lillian, an Americanized Libby. Her last name was Hymowitz.

Thomas:

We can always skip around. Back to Harvard, I suppose, and moving to become Schwinger’s student.

Baym:

I didn’t think about any other options but to become Schwinger’s student. Fortunately, at the time he was at Harvard, he used to have one student he would talk to a lot, which I became. I remember there was a constant line of people sitting outside his office, like a dentist’s waiting room. In my last year at Harvard, ‘59-’60, I was the person he talked to a lot. Schwinger gave me tremendous confidence -- confidence in using mathematical formalism, to see the structure of physical ideas without having to really worry about how to go about how to describing it mathematically. That part was very good. But everything I did was really quite formal.

Thomas:

I know that legendarily, Schwinger was very averse to people using Feynman diagrams and that approach.

Baym:

Indeed.

Thomas:

Did he enforce that?

Baym:

Not really. But he thought so differently from Feynman. Schwinger thought mathematically. Feynman was completely visual. I’m also much more in the visual class. I remember that Lowell Brown, who was a contemporary of mine as a Schwinger student, was talking to Schwinger, and he wrote a symbol for a Green’s function on the board, capital D sub F. And Schwinger says to him, “Do you have to use that notation?” A big breakthrough for me actually occurred when Dick Arnowitt, who was a general relativist, who worked with Stan Deser and Charlie Misner, wrote a set of lecture notes on pi meson physics, which really was in terms of Feynman diagrams. That was my first exposure to Feynman diagrams. And so, one thing I worked on right away was how to translate that into Schwinger’s language of variational derivatives and such. Seeing the connections was something I got rather good at, and which stood me in good stead after I left Harvard.

I could go on with the Harvard anecdotes. David Mermin, who’s also a rather well-known physicist, and I were at one point called on the carpet in the Physics Office at Harvard and that told neither of us were likely to have a good career in physics. Harvard was a tremendously unsupportive place. I remember David was advised at one point to stop being so close to me scientifically. That was a completely wacky thing to say.

I mentioned earlier in talking to you informally that there was three major bits of physics that were discovered during the time I was at Harvard, ‘56 to ‘60, that were not discussed: Landau’s theory of Fermi liquids, the Hanbury-Brown-Twiss experiment, which was explained by Purcell at Harvard, and also the Bardeen-Cooper-Schrieffer theory of superconductivity were simply not in the curriculum. It was quite amazing in retrospect. In fairness, Lee and Yang did come early to tell physicists at Harvard about their discovery of parity violation.

Thomas:

I was struck by that.

Baym:

Yeah. Harvard theoretical physics was, in a way, very insular. We were not encouraged to read the literature in theory, but rather to derive it ourselves. And I have to mention, because it was important to what came later, my PhD final exam. I had at that time started doing the theory of electrons in metals as a many-body system using Schwinger-like techniques, Green’s function and such. Nothing terribly ground-breaking. But Schwinger wanted everybody to do some practical calculation, so I derived the electrical conductivity of metal. Very cleverly. I actually got a finite answer. I shouldn’t have, looking back at what I did. [laughs] But I was off by a factor of four-fifths. Now, fives are very hard to get in physics. At my PhD final, Paul Martin, who was an assistant professor at the time, called attention to that. Schwinger said, “It’s okay, it’s good enough.” But that four-fifths really was a key. There was something much deeper going on. The third person on my committee, just for the record, was Curry Street, an experimentalist. The experimentalist on my prelim exam committee was Horst Meyer, who became an eminent low-temperature physicist. We were good friends over the years.

Thomas:

When we talk about four-fifths, that’s against the measured value?

Baym:

No, against Boltzmann equation derivations. And fives are very hard to get. The fives come about because one has basically Legendre spherical harmonics squared. If you have second harmonics, you get fives. Let’s switch the scene now from Harvard to Copenhagen. I asked Schwinger to write a letter of recommendation to the Institute in Copenhagen for me. He said, “Finish your thesis, I’ll write a letter.” That all went through. I was accepted to Copenhagen without any question. I meet Leo Kadanoff at our graduation from Harvard. He was sitting immediately in front of me. I said, “Oh, what are you doing now?” He said, “I’m going to Copenhagen as a postdoc.” I said, “That’s funny. So am I.” [laughs] We never talked much before.

Thomas:

You hadn’t interacted at Harvard at all?

Baym:

Yeah, we never interacted very much. It was very strange. But in Copenhagen, we started talking. And Leo had raised questions about understanding violation of conservation laws. If you do simple approximations to the exact mathematics in many-body theory, expected properties like the conservation of particle number would be violated. And so, I began thinking about what was going on there. I worked my way through this, with one of these “aha’ moments that came to me one evening. It all fell together nicely. I explained it to Leo, and then we wrote a long paper on conservation laws and correlation functions, which was very influential.

But what I realized looking back at my thesis was that I was not respecting the conservation laws correctly in the Green’s function formalism, and that was what was leading me to get this factor of four-fifths. That I got a finite number was, as I said, an accidental miracle. This problem set the path of my research for the next several years. Leo and I then began worrying about the general theory of many-body systems in a broader context. We were asked to give lectures in Ivar Waller’s theory group in Sweden at the Uppsala and also in Krakow and Warsaw in Poland and in Copenhagen. And we wrote these up as our book, Quantum Statistical Mechanics. Leo said to me very cleverly, “I think we should alternate our names on our papers. Our first paper was Baym and Kadanoff. I think the next one should be Kadanoff and Baym, back and forth.”

In fact we wrote only two things. The book became Kadanoff-Baym. But one of the major achievements in our book, which Leo gets a lot of credit for, was what has become known as the Kadanoff-Baym equations, a generalization of the Boltzmann equation where particles don’t have well-defined energies as they do in Boltzmann’s normal equations, but where the particle energies can have a spectral width as well. The formalism becomes much more complicate, too complicated at the time to carry out useful calculations. There were not modern computers then that could deal with these problems. It was only roughly at the turn of this century that people started using computers to solve the Kadanoff-Baym equations. And now, they are actually quite widely used from chemistry and theories of electronic devices to leptogenesis in the early universe -- as a way to deal with problems where there are not well-defined quasiparticles. It took a long time for them to get into use.

Thomas:

Did you have a sense at the time that it ought to have staying power, that this was something that was of fundamental importance?

Baym:

I think so, yes. The key was how to do a general transport theory in a many-body system. Of course, one limitation is that our starting point was fairly close to equilibrium. Nowadays, people study systems that are wildly out of equilibrium. We didn’t have the machinery at that point, or indeed the motivation, to study such problems.

Thomas:

The other thing I have to ask about is, this is a whole book, and you’re about two years out from your PhD. That’s, as far as I know, a fairly unusual thing among physicists.

Baym:

I got my PhD June of ‘60, and Leo Kadanoff went to Illinois in October of ‘61. We actually wrote the book in September-October of ‘61. We had no sense of it being unusual—we just wrote. But looking back on it, we could’ve done a much better job on the book now. [laughs] But it was very useful for the community, and remains so.

Thomas:

And did it kind of get you a reputation right away?

Baym:

Oh, yes, and I think that it led to my being hired in Illinois. I should mention that Leo remarked in Copenhagen that there were three good places to be doing many-body theory at the time. There was the Landau Institute in Moscow, Bell Telephone Laboratories, and Urbana, Illinois. Urbana, of course, because John Bardeen doing superconductivity and such. And Leo stopped being a postdoc and went to Illinois early in our second year in Copenhagen.

Thomas:

And by this time, BCS was known to you.

Baym:

BCS was ‘56, yes. Actually, I learned BCS in Copenhagen by studying Migdal’s papers on BCS pairing in nuclear physics, which is a rather indirect route into the problem. As I mentioned, Leo went to Urbana in October ‘61, and then a month later, I get a nice formal letter, standard business envelope all typewritten. I never got such letters. My mother would write handwritten things. I looked at the envelope and said, “Oh my God, it’s a job offer, I’m sure. and I’m going to have to take it.” [laughs] It is hard to believe now that there was ever a time where jobs just happened so easily. I did accept the job right away, but as I mentioned I spent a year at Berkeley, sort of as a one-year buffer zone before settling down in the Midwest. Every place I’d lived, New York City, Boston, and Copenhagen, were major seaports. Berkeley also. Urbana, Central Illinois was not renowned for its seaports. [laughs]

Thomas:

Before we move on from that, I want to ask you about the environment there in Copenhagen. This is around the last year or so of Niels Bohr’s life. Was he still a central figure there?

Baym:

Very good question. I came to Copenhagen in ‘60. Bohr died after I left in November ‘62. He would come to seminars and comment. But he was the grand old man of the Institute. He would have wonderful parties at his house. He was given the Carlsberg Mansion at the Carlsberg Beer Factory. It’s known in Danish as the Æresbolig, the House of Honor. We did not have any substantive interaction. But of course, I cherish the group photos with Bohr.

People who were important to me in Copenhagen were Leo Kadanoff, with whom I was working, as well as Vinay Ambegaokar, also a postdoc, who went to Cornell and was also doing superconductivity. The only person who really guided us anywhere was Gerry Brown, who, after having political problems, being too far to the left, left the U.S. for Copenhagen. Gerry was interested in nuclear physics, and he encouraged me at the time to start thinking about how to apply all the machinery that we had developed, the conserving approximation game and such, to nuclear physics as well. Aage Bohr (Niels Bohr’s son) and Ben Mottelson were around, and very, very active in nuclear theory. They were writing their big multivolume book on nuclei. Some of this rubbed off tangentially, but I was never seriously involved in what they were doing then. My real mentor was Gerry. He and I remained close friends and collaborators for many, many years, which became important to the later part of the story.

Thomas:

Okay, then it was Berkeley for one year. We discussed that a little bit, the job talk to you mentioned earlier.

Baym:

Yeah, to be absolutely honest, I didn’t do very much physics in Berkeley. I worried a little bit about Green’s functions. I had an interesting conversation with Andy Sessler at the time. But I don’t remember doing anything very serious. Berkeley was just seductive. The idea you can go to the beach in February. I was roped into teaching classical mechanics and statistical mechanics at eight in the morning, and afterwards, I would just have a relaxed morning shooting billiards, drinking coffee, and just talking with other physicists there, such as Danny Goldberger. But then, finally in September ‘63, I came to Illinois.

Thomas:

Let’s try and set the stage for Illinois. Bardeen is the key figure, David Pines as well.

Baym:

David, yes. Illinois was a tremendously welcoming environment. It was quite amazing. Very different from Berkeley, where I was officially a postdoc under Ken Watson, with whom I didn’t interact with at all, except at parties. He was off, setting up in General Atomics or whatever, I don’t remember the details. I continued doing fairly formal things initially at Illinois;. David Pines, at that time, was finishing his book with Philippe Nozières on the many-body problem, and he was very encouraging. If I explained some nice idea to him, he would say “Wonderful. Do you mind if I use it in my book?” [laughs] But for a young person, that was very good motivation, in fact. The big breakthrough at Illinois really came, I would say, around 1965, when dilute solutions of helium-3 in superfluid helium-4 were discovered.

To set the background, we were all thinking general things about superfluids, the main superfluid being helium-4 at the time. Helium-3 was not known to be a superfluid. It was discovered at Ohio State by Dave Edwards and collaborators that one could, in fact, dissolve up to 6% helium-3 in helium-4. Nothing else dissolves in helium, basically, except its other isotope. But that was a wonderful playground for me. John Wheatley was doing critical experiments on the first floor here, so Bardeen, David Pines, and I began worrying about the theory of the solutions. We did elementary theory. Bardeen was always very, very practical. For me, that was a real eye-opener of how to do practical physics.

He didn’t do theory because it was fun. He had very specific goals in mind. We really became quite expert on the theory of helium-3, helium-4. I continued in the game after the other two stopped doing things much on the solutions. Bardeen’s interest, by the way, was in whether helium-3 could itself become a BCS superconductor, a paired superconductor. And then, we would have a marvelous system, which is the helium-4 is a superfluid, with this Fermi superfluid as well mixed in. But unfortunately, we deduced that the critical temperature of this was way down in the sub-millikelvin regime. Not accessible to experiment at the time at all.

Thomas:

That came later, though, didn’t it? The superfluidity of helium-3?

Baym:

That’s right. Helium-3 superfluidity by itself did come later. The superfluidity of nucleons and nuclei was known. I mentioned that I learned about superfluidity from Migdal’s papers on nuclei. The question he was concerned with was the pairing of nuclei giving rise to reduced moments of inertia, which one would see by measuring rotational spectra of nuclei and such. The one lesson I got out of this is, it’s all the same physics. Whatever scale one is really working on, it’s still the same physics.

Thomas:

If we can jump ahead a little bit, as you move into things like neutron stars, did that continue to be the case? That you’re essentially…

Baym:

Oh, very much. That’s the direction I would continue on.

Thomas:

To stay on helium-3, helium-4 for a moment, as you’re developing this work over time, would it be in a continuous dialog with experiment? Or would it develop along theoretical directions?

Baym:

I was very in contact with experiment. I recently came upon letters I exchanged with John Wheatley. He was really one of the great low-temperature experimentalists. I was invited about this time (1965) to come to Japan by Ryogo Kubo. Kubo was one of the great statistical physicists ever, and a professor at Tokyo University. But I couldn’t go right away. The trip was put off for several years. I did go in the spring of 1968. Unfortunately, the spring of ‘68 was a time of student unrest, and Kubo spent all this time dealing with the unrest—first, with the medical students went on strike at Tokyo University. He became the main person who was dealing with them. But we did spend a lot of time together away from the campus, on excursions in Tokyo and further.

We actually ended up talking about physics less than I would’ve liked. I should mention I had met Kubo in the summer of 1964, which I spent at the Ecole Normale in Paris. He asked me what I was doing, I explained it to him, and he made some marvelous remark, “Oh, gaussians.” [laughs] I had no idea what he meant. I’m not sure if I do even now. But he was a very deep, wonderful man. In Japan, I did continue working on helium solutions, and that’s when I was corresponding with John Wheatley about it. I also had a rather long subway ride from where I was living to the University, and I began reading in the metro about astrophysics in the book by Iosef Shklovsky, called Intelligent Life in the Universe, which was translated by Carl Sagan. (As an aside. I met Shklovsky on a National Academy exchange trip to the Soviet Union in 1973, where he took me to lunch at the Soviet Academy. In the middle of which the famous and by then discredited agronomist Lysenko showed up for lunch. As I put on my glasses, he commented, “now all the optical systems come out.” He also commented that as a principle, one did not throw people out of the Soviet Academy!)

I became very interested in astrophysics. When I came back to Urbana in the fall of ‘68, it turned out pulsars were just discovered by Jocelyn Bell, in late ‘67. By this time Tommy Gold, and independently Franco Pacini showed that pulsars were neutron stars. A group of us got interested in neutron stars. It was lots of fun. The group was David Pines, Chris Pethick, and Goeff Ravenhall, who was a nuclear theorist. We just began teaching ourselves what was going on there, and ended up writing a couple of very good papers.

In one, we calculate the electrical conductivity of matter in neutron stars. We knew how to do transport theory. At that time, we were good solid-state physicists. That problem was interesting because we got a very high electrical conductivity. So, high that it became clear that one could not change the magnetic flux over large distances in neutron stars. Flux diffusion is determined by the electrical conductivity. The greater the conductivity, the harder it is to move flux. And so, it became clear to us you couldn’t dissipate flux in neutron stars and pulsars within the age of the universe. Neutron stars were born with whatever magnetic field they inherited from their predecessor, and you couldn’t get changes over 10-kilometer scales, 10 kilometers being the expected radius of neutron stars. The other paper was on states of superfluidity in neutron stars. There were people who had done it earlier including Vitaly Ginzburg But we wrote on possible superfluid states of the free neutrons in the crust of neutron stars, and of proton as well as neutron superfluidity in the liquid interior. That work is still pretty good. It’s lasted.

Thomas:

I guess that feeds into a question that I had. Was there any literature on this already? You mentioned a little bit. Of course, there’s some work on black holes. I don’t know if from a gravitational angle.

Baym:

The black hole, of course, was due to Oppenheimer in ‘39. But no, that was not on our radar screen at all. We were interested in the properties of neutron stars, but thought nothing about the structure of neutron stars. We didn’t know anything about that at the time. I have to record one hilarious thing that happened at the time of our study group. The head of astronomy, an ancient British cosmologist named George McVittie wrote a letter to the head of physics at the time, Gerry Almy, griping about this study group. I remember a quote from his note was, “A group of physicists is unilaterally usurping a branch of astronomy.” [laughs] Which was kind of nuts. [laughs] He didn’t last very long. Shortly after Icko Iben became the head of astronomy, and then we began to have very good relations with astronomy. I should mention, many schools have joint physics-astronomy departments. But the reason we didn’t in Illinois is that the Physics Department in Illinois is part of the College of Engineering. Which is a great advantage because our salary scales were always set compared to engineers, not compared to English professors and such. But astronomy was always part of Arts and Sciences, so administratively, it was very hard to put the two together.

Then, in the spring of ‘70, I went to Copenhagen on a sabbatical. That’s really when my getting into neutron stars began very seriously, which was not my immediate intention in going back to Copenhagen. David Pines also was there, and we had asked Chris Pethick if he could come. He had gone back to Oxford, but did come to Copenhagen. In the long run, Chris stayed in Copenhagen, but that’s a whole different story. Remind me at some point to come back to the helium-3, helium-4 question because that plays a role much later.

Thomas:

At Copenhagen, you were at NORDITA this time. I know that they were essentially co-located. I mention this because I actually read a short piece that you did for a volume that they had, so I think I know where you’re going, but I won’t stop you from saying it.

Baym:

Really crucial to my subsequent career was meeting and working with Hans Bethe in Copenhagen. David Pines went off to Armenia for a couple of weeks, and they were two critical weeks. Because right after lunch one day, I was going through the mailboxes. Visitors’ mailboxes in Copenhagen were not assigned to individuals; there was an A box, a B box, and so on according to last name. I was going through the B box after lunch, and I discovered there a postcard addressed to Hans Bethe. What do you do? [laughs] He’s famous. You read it. I read it. It was an acknowledgement from Astronomy & Astrophysics, the journal, of a paper on neutron stars that he, Gerhard Boerner, and Katsuhiko Sato had just written on neutron stars.

And they sent it to the journal of Astronomy & Astrophysics. There were two messages in this. One, Bethe was working on neutron stars, and since he was not in Copenhagen but getting mail in Copenhagen, it meant he was going to come to Copenhagen. Chris somehow managed to get a preprint of the paper. I have no idea how he did it. The issue Bethe was concerned with was that neutrons have a crust like in the Earth, and a liquid interior. How do you go from the crust to the interior? How does the transition take place? The problem there was treatments were based on completely different physical models. You base the crust on nuclei you knew about, whereas the interior you would describe what are called nuclear matter calculations of liquids, neutrons, and protons, stuff that Bethe actually pioneered in ‘54 when he spent a year with Goldstone in Cambridge, England.

Chris and I began looking at this model. And being good solid-state physicists, we start making improvements to his description of the crust. So much so that the successful join of the theory he had gotten fell apart completely. It became self-inconsistent. Whereupon Bethe arrived in Copenhagen, and we told him right away there’s a serious problem. And he said, in his good German accent, “Ah, now, we must solve ze problem.” And that was an amazing, because neither of us knew anything about nuclear physics, we would spend the semester sitting at the feet of the great master of nuclear physics. And we started working together.

Bethe was phenomenal with a slide rule. We always joked that the thing that limited his accuracy with a slide rule was the Lorentz contraction. [laughs] If he went too fast, the slices would get too narrow. He was absolutely amazing. The only way we could keep up with his calculating was to learn to use a computer. It turned out, that in the basement of NORDITA, which was a separate institution from the Niels Bohr Institute [As an aside. the first time I was in Copenhagen in the early 60s, I was a postdoc at what became the Niels Bohr Institute, which is now basically the physics department of the University of Copenhagen. NORDITA lived in a separate building. In ‘70 when I was there, I was a visiting professor at NORDITA. In terms of one’s daily life, it was not critical. But administratively, it made a difference.] was a first-generation transistorized computer which Chris and I learned to program. We programmed in ALGOL. I immediately took to ALGOL because one course I loved as an undergraduate at Cornell was given by J. Barkley Rosser in mathematics on the Russell-Whitehead symbolic logic. I just found it tremendously appealing. And ALGOL is based on that language, in a sense. It was completely natural. So, we programmed neutron stars, and we got so good at it that Bethe left the calculating to us.

But this computer was wonderful. You describe neutron stars by the Tolman-Oppenheimer-Volkoff equation. You had to have an equation of state. Pressure versus density. You have to put that in as an input. To compute a neutron star using the Tolman-Oppenheimer-Volkoff equation, you choose a starting central density, and integrate outwards until the pressure fell to zero. The computer would keep putting layers on the neutron star, and you could actually hear the machine doing so—clunk, clunk—as it went through the neutron stars.

Thomas:

The noise is what it was making at the transition boundary?

Baym:

The transitions were making noise, yeah. [laughs] iPhones don’t do that. We got to the stage where we’d come out with pages of numbers and show them to Bethe. And Bethe would say, “How can we understand this?” And the was absolutely incredible; he would stare at a page of numbers for half an hour, just looking at a simple table. And then, he’d say, “Ah.” He’d write down a very elementary first approximation and get sort of the lay of the land. Then, next, he said, “Why is my first approximation not good enough? How can we improve it?” He would have a second layer of approximation. And after a few go-rounds, he really understood everything. It was absolutely remarkable. It’s an approach to physics that has suited me very well over the years.

The other amusing thing with Bethe is that Chris and I understood thermodynamics fundamentally. I remember that Bethe showed up one day with a calculation, which we didn’t agree with. A 10-page thing. In two lines, we showed him how there was a thermodynamic problem with what he was doing. His reaction was, “Ach, I am in a den of thermodynamicists.” [laughs] But it was wonderful. I actually stayed on in Copenhagen through October that year, calculating the equation of state of neutron star matter and the models of neutron stars it produced. We began looking at how individual nuclei would be contributing, and that really was the beginning of a serious career in nuclear physics. I’d never really worked in conventional nuclear physics studying energy levels of round nuclei. I was always more interested in properties of matter under extreme conditions, very high density and such.

Thomas:

And this would be something that would be studied within that field rather than, say, as an astrophysics problem, per se?

Baym:

One encountered such cold high-density matter, density above nuclear-matter density, really only in neutron star. There’s a cousin of it, of course, which is, in the early universe, if you trace backwards, the universe gets denser and denser. As we understood, basically from the mid-70s on, prior to a microsecond after the Big Bang, the basic degrees of freedom of quarks. But quarks in 1970, were not well established. We didn’t worry about any other arenas where high-density matter would be important.

Thomas:

It’s really the state of matter that you’re interested in rather than the object of the neutron star itself.

Baym:

Well, both. The reason for learning the state of the matter was to be able to make models of neutron stars, which meant just predicting their properties for a given equation of state.

Thomas:

I did notice in your list of papers, there are at least one or two that deal with pulsars that don’t behave regularly. The Vela Pulsar.

Baym:

Oh, yeah, that was a whole other game. Mal Ruderman at Columbia had come up with this very interesting idea, back in ‘68- ‘69, of starquakes. The idea was when neutron stars are very young, they’re oblate because they are spinning rapidly. As they cool down, the crust solidifies. But then, as they rotate more and more slowly, they want to be more spherical, and this puts stress on the crust. The crust should then crack, leading to what’s called a starquake. What was very interesting is, shortly after the discovery of the Vela Pulsar, which rotates 11 times per second, it was discovered one day to have sped up by sort of one part in a million, a tiny, tiny effect.

And what was wonderful about that, within the star quake picture, if there was a crack, the surface only had to move one centimeter—its radius was a million centimeters, 10 kilometers—and that would decrease the moment of inertia by one part in 10 to the 6, and by conservation of angular momentum, it’d have to rotate one part in 10 to the 6 faster. It was beautiful. Completely wrong, but beautiful. David Pines and I had worked out in Copenhagen in ‘70 the energetics of starquakes. The trouble was, Vela did it again, and there wasn’t enough energy stored to have quakes every couple of years for the 10,000 years that Vela was going already.

It was energetically impossible. So, we turned to other pictures, which depend on the interior superfluidity, there are—what’s fun here in the story is how knowledge of superfluids becomes influential in astrophysics phenomena. We knew that if you rotate a container of superfluid helium-4, it forms vortices. And indeed, even if you have a neutron superfluid, which is a BCS-paired superfluid, it too would form vortices. Ordinary nuclei are too small to have vortices, but in a large bulk chunk, you get vortices. And the vortices in superfluid carry the angular momentum.

The picture that got developed was that vortices could get stuck on nuclei in the same way of a snack swallowing a grapefruit. [laughs] But a vortice could have a lower energy if it was attached to a nucleus or possibly in-between nuclei. But in either case, the vortices become pinned to the nuclear lattice. As the nuclear lattice loses energy to the outside, the vortices are carrying too much angular momentum. The equilibrium vortex positions would be further out, but they can’t move because they’re attached to the nuclei somehow. Eventually, the stresses build up, the system breaks, a bunch of the vortices move out a bit. And that was the vortex picture of speed-ups. Whether, in the long run, it’s right or not, I don’t know. But it was a very good insight into possible effects of superfluidity within neutron stars. It could be that the crusts crack, and that plays a role as well. I don’t think the final story is in on this at all.

Thomas:

Is that why you wanted me to remember helium for later on?

Baym:

No, the helium-3, helium-4 for later on was in the neutron spallation source at Oak Ridge, which until a couple of weeks ago was mounting an experiment to look for the electric dipole moment of the neutron. It’s a wonderful experiment that would take me too much time to explain in detail, but one is measuring the interactions of helium-3 atoms in superfluid helium-4 with neutrons, as a way of measuring dipole moments. Our knowledge of helium-3, helium-4 became very, very useful in working out the plumbing of this whole experiment.

Interestingly is that at the time we worked on helium-3, helium-4 mixtures back in the 60s, the aim was to look for new states of superfluidity. In ‘72, when the superfluidity of helium-3 itself was discovered, everybody dropped the mixtures like a cold potato. But they came back. What became fun in the neutron dipole moment experiment was that one was looking at extremely dilute solutions of helium-3, one part in 10 to the10 helium-3, which was a completely different parameter range than was ever studied. That would have been a lovely use of the dilute solutions, were it not for DOE, in its wisdom, decided to abandon the experiment. What its fate will be, whether it will be done elsewhere or not, nobody knows.

Thomas:

This has just recently happened?

Baym:

Right now, yeah.

Thomas:

I know that the Spallation Neutron Source is shut down at the moment, but it’s just off the program altogether?

Baym:

Just to look for the neutron dipole moments, which is an extremely important experiment because it’s a way of using very low temperatures physics to study beyond the Standard Model physics. Physics up in the TeV range would give rise to the time reversal violation that you need to have a dipole moment. The more accurately you can put limits on the dipole moment, the more accurately you can eliminate certain beyond the standard model theories in high energy physics.

Thomas:

We were in the 70s with neutron stars. I don’t know if we want to transition to things like QCD at this point.

Baym:

QCD’s coming. In the early 70s, there was a wonderful proposal by Arkady Migdal, the older Migdal, of what is called pion condensation. Naively, you would think, “If you have neutrons, the neutrons can produce pi minuses, changing into a proton, and the pi minus being bosons would just form a Bose-Einstein condensate.” I worked on this problem as a field theory problem. It’s much more complicated than what’s happening in the simple picture I just mentioned. Rather, the neutron liquid develops a soft-mode instability, leading to a very beautiful state of matter which unfortunately has never been realized. But we worried about that as a possible state of neutrons in neutron matter. It would show up by giving greater rates of cooling in neutron stars. There are hints at how rapidly neutron stars cool, but not very accurate measurements at all.

To measure cooling you need telescopes that can do spectroscopy very accurately, and have very good spatial resolution as well as resolution of energy spectra emission. If you look at a neutron star, you can’t tell if the radiation is coming from the surface of the neutron, if it’s coming from a larger area, or if it’s coming from a hot spot.

At the time, the mid-70’s we were searching for other states of dense matter as well. At this time, quarks were well-known (even though Gell-Mann in his papers in the 60s writes of the possibility that experiment could disprove the existence of physical quarks. It a remarkable question why Gell-Mann did not believe that quarks would be real. [laughs]).

But studies of electron scattering on nucleons really showed that quarks had to be real things. So, the ideas of asymptotic freedom, QCD of Gross, Politzer, and Wilcek really came to the fore—around ‘73 or so. And then, not long after that, I had a postdoc named Siu Chin who is at Texas A&M. He and I asked the question of whether a neutron star can be a giant MIT bag. It was a catchy title, but the question is, could neutron stars have large regions of quark matter in them? We concluding that the answer was “no,” thereby setting the field back enormously. [laughs]

I had mentioned I was expert in thermodynamics. One thing I knew about thermodynamics is that if you have two competing phases, and you want to know which one wins, you compare their free energies, and you make a Maxwell construction between the two in some variables. We were making a basic assumption that at high densities, you compared quark matter with nuclear matter—a picture that now makes no sense whatsoever. At very high density, you can’t have nuclear matter no more than you could study the statistical mechanics of eggs, learning all about egg, egg scattering, then trying to predict what the properties of this matter are at 10 times closely packed egg density, which would be a yellow liquid instead. [laughs]

We were intrinsically making the assumption that there had to be a high-density nuclear phase, and asked, “Could the quark phase have lower energy?” Otherwise, we should not consider it. But lower energy meant that it was softer. That raised the question, how could a neutron star have a large mass of quarks within it? Where do you get the pressure? It was a brilliant paper, but it set the field back. At the time, we were really playing with different possibilities of high density matter.

Thomas:

That was what I was going to ask. Before, with helium, you’d been working in close dialog with experiment. In neutron stars, there are some observations like we were talking about of pulsars.

Baym:

Yeah, very few observations. The basic source of observation was pulsars, their timing, and the variations of the timing. I remember there was one instance where John Middleditch had reported a pulsar with something like a 1,700 Hertz rotation frequency. We all really got concerned, “Can a neutron star rotate 1,700 times a second? Or would it come flying apart, form weird Jacobi structures and such?” And nobody could make sense of it. Finally, it was discovered that the 1,700 Hertz was caused by some television camera, CCD, in this apparatus. It was wrong.

But at that period, we were really fishing around, trying to understand possible states of matter. Then, an interesting development came. In the mid-70s, T. D. Lee, who was at Columbia then, got interested in anomalous states of matter. What could matter look like at high density? These are things we sort of understand now in terms of chiral symmetry breaking giving rise to masses of neutrons, and at high density the chiral symmetry could go away, leaving a more favorable state of neutron matter. There was a crucial and wonderful meeting in 1975 on dense matter in Bear Mountain, NY. (Bear Mountain is up the Hudson River on the west side, and a place I knew from my childhood, actually.)

What was about that meeting, was that it brought together lots of experimentalists in nuclear and high-energy physics. This was essentially the beginning of the idea of trying to study high-density matter experimentally through nuclear collisions. These ideas continued to percolate around, certainly getting better defined around 1980. There were a fair number of meetings then, asking whether one could really begin to study high-density matter through very high energy of collisions of nuclei.

Around this time, in ‘82, I wat appointed to two physics advisory panels simultaneously. One was the NSF Physics Advisory Panel of the NSF Physics Division. One of the interesting outputs of that was approving the LIGO project. That was another story. But relevant to high energy collisions here, I was put on the Nuclear Science Advisory Committee, NSAC, being knowledgeable about high-density matter. Every five years, the Nuclear Science Advisory Committee would come out with a long-range plan.

Thomas:

The first was in 1979, right?

Baym:

The first was in ‘78-’79, led by Herman Feshbach. The question for our Long Range Plan, led by John Schiffer, was what should be the next big machine for nuclear physics? There were two proposals on the table. One was a sort of vast meson factory at LAMPF, the Los Alamos Meson Physics Facility. The other was possibility of colliding heavy ions to study high-density matter. What happened was remarkable. There was a meeting that was taking place in Aurora, New York, Wells College, which is just north of Ithaca on the east side of Lake Cayuga. We met there in the summer of ‘83. I was in charge of the subcommittee on high-density matter.

We arrived on the weekend. Monday, Arthur Kerman, who was a nuclear theorist at MIT, also one of our high-density people on the NSAC committee, came with news that the high-energy community had just had a meeting in, I believe, Woods Hole, where they decided to abandon the Isabelle project at Brookhaven in favor of building the SSC. (Actually, the name of Isabelle was not Intersecting Storage Array or anything, but it was the name of the sailboat of John Blewett, who was principle accelerator designer of the Isabelle project.) That was very relevant news because Wednesday morning, I was going to talk about possibilities of studying high-density matter, and my slides were just aimed at, “Eventually, we can look forward to building something.”

And suddenly, we all had tremendous opportunity dumped on our laps. Brookhaven had all the civil engineering in place. No ring was built, but the entire tunnel for it was there, and all the concrete was poured. And here was this tremendous opportunity to build a heavy ion collider in the tunnel to look for high-density matter. In my talk—and I take full credit for posing the idea because the words came out of my mouth, but other people were involved also, people like Larry McLerran. We had a lot of talks earlier with “BJ” Bjorken at Fermilab. It was really in the air. But it was an opportunity that was too good to miss. It was decided I would propose it, and it got this fantastic reception. “Wow, let’s do it.” That was the start of RHIC, the Brookhaven Relativistic Heavy Ion Collider.

Thomas:

There was really no proposal prior to that point?

Baym:

Well, Berkeley was proposing VENUS to do heavy ion collisions, but it was a rather awkward machine; Berkeley is in the hills, so one had the beam line dipping down in hill. It was actually a wacky design. But that was sort of the picture one was looking at at the time. (Some wit said that VENUS stood for Very Expensive Nearly Useless Synchrotron.)

It was just very good fortune for nuclear physics that the high-energy physicists were shutting down Isabelle. We were there to take advantage of that. And DOE loved it. Brookhaven loved it because it was something they could do with what they had already.

Thomas:

The termination of Isabelle was a huge blow to them, so to have this come in to them right away as an alternative use for that site…

Baym:

It softened the blow completely. From that day on, I became an accelerator physicist. I took part in all kinds of meetings at Brookhaven to ask the question of what the design parameters of the machine should be, whether we should be able to study different nuclei, different atomic number nuclei, and whether to design it do proton nucleus collisions—or as it turned out actually, deuterium-nucleus collisions. These collisions were easier since the deuteron would have nearly the same rigidity, charge per mass, as heavier nuclei. We were working through all these possibilities.

All the studies of what the basic machine should look like gelled with the formation of the RHIC Policy Committee, of which I was a member. In September ‘83, soon after our July deliberation, Brookhaven hosted a meeting on ultra-relativistic nucleus-nucleus collisions, known as “Quark Matter 3.” There were earlier meetings, certainly starting as early as 1980. If you look at the numbers of quark matter meetings, they don’t converge to zero, they converge to two or something. I gave the summary talk at the quark matter meeting in ‘83, about the machine. I remember I laid out a timeline for the machine of six years to beams. Hans Gutbrod, one of the wonderful experimentalists at CERN doing the WA80 fixed target heavy ion experiments at the time . looked at me in the middle of this talk and said, “Does it have to be that long?” This was six years. Turns out, it was 17 years in the end of trying to get the machine going.

Thomas:

We’ve wandered into my own interest area of science policy and DOE decision-making, that sort of thing. The first question that comes to mind for that is the relationship between the nuclear science and the high-energy physics community. Of course, they each have their separate PAP. There’s HEPAP, and then there’s NSAC, which you were on. Is that a division that’s just within DOE? Was there a division between communities in real life?

Baym:

There was certainly an intellectual division. It still is a very strong intellectual division. I remember I gave talks at HEPAP explaining what the machine was. Sam Ting pointing out that I mentioned producing10,000 pions in one collison, asked me, “Are you going to measure all those pions?” [laughs] It was intellectually a completely different project. Though, interestingly, in the LHC, the big experiments, ATLAS, CMS have been doing heavy ions as well, and there’s one dedicated heavy ion facility, ALICE. ALICE is a wonderful acronym. A Large Ion Collider Experiment.

My interest in heavy ions was discovering the collective states of matter, and answering the questions of what are the degrees of freedom of matter at the highest energy densities. What, when you put a lot of particles together, are its properties? The high-energy physicists don’t think about that. High-energy physicists want to know, “Tell me the basic Lagrangian. Once I know that, there’s nothing more to think about.” And part of this is that since the QCD Lagrangian is well established, QCD is no longer considered a respectable field of high-energy physics. It’s a problem that’s entirely for nuclear physics now.

Thomas:

The question originally occurred to me, we had Bob Jaffe on a panel with us here at AIP once. And I mentioned to him, “So you’re in nuclear physics?” and he said, “No, no, I’m in particle physics.” I thought, “Maybe these distinctions aren’t quite as strong as I thought they were.”

Baym:

They’re very strong. Nuclear physics, by and large, has become sort of a bastard area of physics. All the major departments, Harvard, Princeton, Stanford, Caltech, with the exception of MIT, the big strongholds of nuclear physics, don’t do it anymore. We still do in Illinois. But again, the interest is not so much around nuclei but other applications, e.g., the neutron electric dipole experiment, and neutrino physics. As an aside, the same problem comes up again in AMO -- atomic, molecular, and optical physics -- where the cold atom game becomes a new window into the many-body problem. And the atomic physicists and the laser spectroscopists didn’t really like cold atoms. They say it’s not part of what they do. There is the same dichotomy there as well.

Thomas:

The intellectual differences aside, is there a sense of competition for funding? Certainly, in the 1980s with building accelerators and colliders, the SSC was dominating the scene after that point. And they do have separate pots of money at DOE, but nevertheless.

Baym:

Yeah. DOE realized nuclear physics is important. Just things like isotopes for medicine. Nuclear tracking of illicit arms, nuclear explosives and such, which you could track by muons and such. It’s of tremendous importance. But the problem is, how do you keep this field going if you don’t support nuclear physics? And so, the field is fairly well supported for that. Even now, the only major accelerator on DOE’s table is the Electron Ion Collider, which will require solving a great many serious accelerator challenges. Things like cooling beams, getting high luminosities and such. These are problems of general interest. For the DOE, the EIC will keep accelerator physics going.

Thomas:

It’s the only collider facility left in the United States, really, right?

Baym:

The only collider facility, yeah.

Thomas:

Going back to the 1980s, on the nuclear science side, the big project at that time was CEBAF, so that had to finish up. Were you in line behind that?

Baym:

Well, that’s a good question. CEBAF was basically contemporary with RHIC. CEBAF was an interesting issue that was also decided by NSAC, the Nuclear Science Advisory Committee, at the time I was on it. There was competition between basically Argonne and starting a new lab in Virginia. And what won the day was support from 26 southern senators, their political power got the machine to the South (to the extent that Virginia was the South). But that developed well right away. But there was a long question of whether RHIC should be built. I remember Bob Hunter was, at some point, in the Reagan administration...

Thomas:

The director of the Office of Energy Research at the time, right?

Baym:

Yeah, thanks. I remember going to Washington to see him and trying to convince him that building RHIC was good, serious physics. And I did convince him. Eventually, this machine was built, but it just took forever. From its conception at Aurora, New York in ‘83 to first beams in 2000 was 17 years.

Thomas:

Reassembling the timeline, I understand that RHIC was involved—you know what the Trivelpiece plan was? Al Trivelpiece put together some sort of grand bargain, where there was a set of facilities that would go ahead together or in sequence. I think RHIC was one of those. The Advanced Neutron Source, which eventually became the Spallation Neutron Source, was one of those. Advanced Photon Source, I think. That, I think, was a moment when DOE decided that that was definitely a project they were going to do. I wouldn’t bet that that’s right, but I think that’s right. Of course, there are still subsequent layers of approval and congress appropriating funding.

Baym:

I don’t know the exact year when the first conceptual design for RHIC was approved (the code word is CD). But from ‘83 on, people were optimistic about building RHIC. There were very good staff of people at Brookhaven.

Thomas:

Who were your partners, the leaders of that project?

Baym:

There was, of course, the director, Nick Samios. He was a high-energy experimentalist, famous for discovering the three strange quark Omega minus baryon, but he was very much in favor of the machine there. Dave Hendry was one of the people in DOE who was a very strong backer of the machine. In general, the DOE Nuclear Science Office backed it very strongly.

Thomas:

RHIC is known for its experiments on the quark-gluon plasma and all that sort of thing. But that was under discussion, like, even in the 1970s. Is that right?

Baym:

If I look back at the long-range plan of 1983 which suggested building RHIC, certainly, the quark-gluon plasma in there. I drew the first serious phase diagram of what matter would look like at high-density versus temperature. That’s in the long-range plan at that point. Free quarks were definitely in the air. We talked about how RHIC would help solve problems of neutron stars; more exaggeration than I was aware of at the time. [laughs] The problem was that collisions at RHIC make hot matter. They don’t make cold matter. You don’t really ever get in the regime where you get particles overlapping, forming a Fermi liquid, and where Fermi statistics begins to play a role. There has been, over the years, a disjoin between what we’ve learned about quark-gluon plasma at RHIC and quark matter inside neutron stars. That’s the world, you can’t do much about that. [laughs] It’s a very hard problem.

Thomas:

What were your own theoretical interests, then, in this area as we move into the 1980s? You’ve had continuing interest in the neutron star problem.

Baym:

I began worrying about the physics of heavy ion collisions themselves. There was a first round of interesting experiments, which came out around ‘86 or so, from the fixed-target experiments at the SPS at CERN. You just tried to understand what the degrees of freedom were, how much energy was being produced, and such. I was certainly occupied in that game for a long time. I can’t think of anything tremendously profound that came out of it. Really despite the lack of convincing evidence, we always described dense matter in terms of quarks degrees of freedom.

Thomas:

In general, did you find—we’re now looking at the long term, from the 1960s through the 1970s. Did you find that fundamental advances in physics—of course, we’ve talked about QCD, but things like electro-weak theory, did that play into what you were doing at all?

Baym:

Actually, no. There were people asking what interesting things could you make in a heavy ion collision. Could you make Z’s, and W’s, and such. But it was not an important issue as far as I remember.

Thomas:

You have a couple of papers on cold fusion that I’d like to ask you about. I know it was on the list of things that you put down to talk about as well.

Baym:

That was a very funny business. Cold fusion was announced in ‘89. Of course, the first thing everybody worried about was, could it be real? Could it actually be something going on we don’t understand? And I was talking with Tony Leggett here, and Tony pointed out something very interesting. The claim of the cold fusion folks was that if you put deuterium into palladium, the deuterium likes to go together, fuse together, to form alpha particles, you get release of energy. Tony asked a very simple question, “Does deuterium like to dissolve in palladium?” which would be a key factor in palladium helping to fuse deuterium.

Well, it turns out, nobody ever did an experiment asking how much it likes to dissolve, but there were experiments in related materials that showed no, it was not very favorable. Tony invented an incredibly clever bound on what the cold fusion rate could be, something like a factor of 10 to the -23 smaller than the proponents of cold fusion, Pons and Fleischmann, were claiming. I played with it and managed to improve the estimate by another factor of 10 to the 23, so it’d be 10^-46. So, we wrote a short Physical Review Letter. the only Phys. Rev. letter that was published on cold fusion.

This had the consequence that Tony and I became the official referees for Phys. Rev. Lettes on cold fusion, and we had arguments, I remember, with Willis Lamb, for example, the Nobel Prize winner for Lamb shift. We could not convince him and his colleague Bob Parmenter that their ideas completely violated our bounds on how fast you could get reactions here. And eventually, it became, I think, a religion, in a way. I was at a meeting at the EPR, the Electric Power Research Institute within the NSF, on cold fusion in October ’89, where I presented our ideas. I won’t tell you about everybody else, but Fleischmann of Fleischmann and Pons was fascinated. And he took me out for a lovely beer afterwards. He really felt that because we had shown that it could not be explained by ordinary physics, that they’d discovered something really fundamental. So, we all joked he probably went off and bought a first-class ticket to Stockholm the next day. [laughs] He really felt there was something going on.

Thomas:

Schwinger was involved in that, too, wasn’t he?

Baym:

Yeah, I don’t understand that, actually. Also, at the Electric Power Research Institute meeting, Teller was there. And Teller was a riot. (We’d interacted with him earlier in the 80s. I haven’t mentioned Los Alamos history project. That’s a whole other game.) Teller had the attitude, “Suppose this is real. How could we explain it? We could do it if we invented a new particle.” He called the new particle the meshuggahtron. [laughs] Meshuggah being a Yiddish word meaning crazy. [laughs] Very funny. During that meeting the great earthquake happened in Oakland, doing severe damage to the Bay Bridge and such. Teller was able, however, to call his family on military lines to check up.

Thomas:

Certainly, from an outsider’s perspective, something like cold fusion looks like it’s completely implausible. You’ve dealt throughout your career with things like very, very cold systems, things like Bose-Einstein condensates, really exotic states of matter. Did it look any different to you? Was there any avenue of plausibility for anything like this coming from that perspective of these exotic quantum phenomena?

Baym:

No. I certainly felt from the very beginning that it was crazy. It doesn’t mean that it has to be wrong. But all the standard tests did not hold up.

Thomas:

You mentioned the Los Alamos history project. I know this is a period where you were involved with the solid-state history project as well. Chronologically, it fits. Thematically, it’s sort of off to the side. I’m wondering if you’d like to discuss that here, or should we make it an addendum? I don’t know.

Baym:

Yeah, as I explained to you earlier, in the mid-70s, I started working on the history of solid-state physics with Lillian Hoddeson, studying the early history of the quantum theory of solids, starting with Drude through Sommerfeld and such in that period. And then, my next interaction with history was in the 50th anniversary of discovery of the neutron. The neutron was discovered in February ‘32 by Chadwick. In ‘82, there was a conference in Cambridge, England, where I gave a talk on the history of neutron stars. Now, I go back to the mid-70s, ‘73. There was a Solvay meeting in Brussels on high-density matter in neutron stars, and I was there, as was Leon Rosenfeld. And we were discussing the origin of neutron stars. And he recalled the night when a letter from Chadwick came to Bohr announcing the discovery of the neutron. That he, Bohr, and Landau spent the evening worrying about what this new particle could be good for.

It was that evening, in February of ‘32, that Landau proposed unheimliche sterne, strange stars. Chris Pethick went back to Copenhagen from Urbana, and I told him, “Check out what you can find in the archives there about this.” Subsequently, I was working on the history of solid-state physics, trying to track down what Landau was doing. Landau had spent two years, ‘30 to ‘32, on a fellowship in western Europe. Nobody had ever written down what his trajectory was, but like in the crime series we see on television with people putting pasties on a wall and trying to put the picture together, I figured out that in February ‘32 Landau was not in Copenhagen. Chris went back, checked it all out. He discovered that in February ‘32, neither was Bohr or Rosenfeld. They weren’t in Copenhagen either. They were off at a meeting in Kharkiv. This story of Landau was completely wrong. We eventually tracked down the whole story with the Russian physicist, Dima Yakovlev, and Polish physicist Pavel Haensel, and published it in Physics Uspekhi, getting the history straight. What happened is, I repeated this whole myth at the historical meeting in ‘82, and it became the origin story of neutron stars. I was completely wrong. For me, it was the quintessential example of not trusting oral history. [laughs] You can trust everything I say, of course. [laughs]

Thomas:

No need to check up on anything at all. I’m sure all the dates are absolutely correct.

Baym:

Yeah. The eventual story was, of course, Landau had done the work a year earlier. Rosenfeld remembered all the details, but didn’t have the right structure in it. I was very involved in the history of solid-state physics, which came out in the book, Out of the Crystal Maze, which I contributed extensively to as well. primarily chapters on the develop of many-body theories. Then, somehow, around ‘83-’84, Lillian and I were in Los Alamos, and Mike Simmons, who was deputy director of the Theory Division there then, came upon locked file cabinets, which in Los Alamos jargon are called safes, containing all the wartime papers. Here was the making of a wonderful history project. The issue was to get it approved.

Harold Agnew, who was the director of Los Alamos at the time, was dead set against the project. But he was eventually replaced by Don Kerr. His wife was an archivist, and they felt very favorable towards the project. In ‘85- ‘86, the project really got underway. Lillian and I spent the year in Los Alamos putting this all together, which was quite amazing because we had access to all the papers. But the principals, like Feynman, Bruno Rossi, and others, weren’t allowed to see their papers. We had to negotiate that. Bethe was okay because he maintained his clearance That was quite an interesting experience, putting that all history together in the book, Critical Assembly. That’s another story.

Thomas:

I have one general question about the relationship between scientists and historians. My sense is that in that same period of the 1980s was sort of a high watermark of, if not cooperation then at least, to a certain degree—there were people on both ends who were interested in it and working side by side. You had things like that project and also the Fermilab symposia on particle physics. And of course, the American Institute of Physics was very involved in both. Spencer Weart, my predecessor, was involved with the Crystal Maze book.

Baym:

Yes. Physics in that time was the science of choice, in a way. There was tremendous government support. Established physicists were getting older and felt some responsibility for what they’d done being recorded. I think things came together very nicely. There was money at the time.

Another funny little aspect of my interest in history dates back actually to 1960, when I was in Copenhagen. Have you seen the play Copenhagen?

Thomas:

Yes.

Baym:

It turns out that the author Michael Frayne of the play about this meeting between Heisenberg and Bohr, didn’t know what went on. And he imagines what could’ve gone on, playing on Heisenberg and the uncertainty principle. It was a gimmick. I knew exactly what happened, but I couldn’t do anything with the information.

Bohr, in 1958, reacted very strongly to the publication by Robert Jungk of this book called Brighter Than a Thousand Suns, which Bohr read as an apology of Heisenberg during the Second World War. It’s a revisionist history of Heisenberg in the Second World War. Bohr was really quite furious, and he wrote letters to Heisenberg. But Bohr didn’t actually do the physical writing. He dictated letters to his wife Margrethe, to his son Aage Bohr, and to Aage Petersen, his scientific secretary (or amenuensis). Aage and I became good friends, and we continued seeing each other through the ’70s. when he got early Alzheimer’s and died, unfortunately. It was very tragic. But Aage Petersen, to whom some of these letters were dictated, told me their contents back in 1960, including the one that said that Heisenberg telling Bohr at these meetings at the Æresbolig, the Carlsberg Mansion, where Bohr lived, that Heisenberg had been doing nothing for the past few years but thinking about how to make an atomic bomb.

I knew that since 1960, but didn’t know what to do with it. The papers of Niels Bohr were embargoed until 50 years after his death, which would have been December 2012. When Frayne’s play came out we put pressure on the Bohr family to release the papers early, and they were released then.

Thomas:

And then, of course, you served on our advisory committee a little.

Baym:

Even as chair! I guess we’re caught up through the late 80s.

Thomas:

I think that sounds about right.

Baym:

Let’s stop at this stage and continue very soon again. Next week and the first four days of the week after.

Thomas:

Why don’t we shoot for continuing on next week? Sometimes it’s nice to have a chance to digest what we’ve said.

Baym:

Okay, thanks so much!

[End session 1]

[Begin session 2]

Thomas:

The date is January 31, 2024, and this is session number two of an interview between Will Thomas, the American Institute of Physics, and Gordon Baym. Gordon, thanks once again for joining us. I think we might launch in — we were just discussing a few moments ago — with cold atoms. What would you like to tell us about that?

Baym:

There’s a very simple connection between my work in cold atoms and my work in high-density matter. It started with the institute in Trento, Italy, in ‘93 at the International Centre for Theoretical Nuclear Physics. Actually, the first meeting they held was not actually in Trento itself, but 10-15 kilometers north of Trento in a charming little town called Levico Terme. The meeting was on Bose-Einstein condensation. The Trento Centre is called ECT Star, the European Centre for Theoretical Nuclear Physics and Related. Cold atoms quickly became part of the related subject. The meeting wanted to summarize what had gone on in Bose-Einstein condensation. This was before the big splash.

Thomas:

By big splash, you mean the 1995 experiment.

Baym:

Right, 1995. This was ‘93. I was asked if I could organize sessions on Bose-Einstein condensation in nuclear physics. As you recall, I had worked on the problem of pion condensation for many, many years. There was also a suggestion of kaon condensation, condensation of K mesons, rather similar in nature to pion condensation. And then, there were questions of chiral symmetry breaking in nature and so on, which were all good topics related to Bose-Einstein condensation. So, I organized the session on this. It was a very interesting meeting because, as I mentioned, it was before the discovery of the atomic Bose-Einstein condensates. I remember a very young kid from MIT at the meeting who paid his own way to be there named Eric Cornell. who, two years later, together with Carl Wieman and Wolfgang Ketterle, did his groundbreaking Nobel Prize winning experiments.

Interestingly, this meeting led to my being invited to give the opening talk at a meeting two years later in France at Mt. Ste. Odile. This was the meeting on cold atoms and Bose-Einstein condensation, where Cornell, Wieman, and Ketterle announced their first results. So, I was actually in on it from the very beginning here. That summer, Chris Pethick -- my old colleague, who I’d been working with on neutron stars and superfluids over the years -- and I sat down and tried to understand the physical structure of the condensates using basically nuclear physics type arguments about energy versus size, and constructing explicit wave functions for the particles in the condensates.

This paper, remarkably, became a "best seller." It was, for a while, the most widely quoted theory paper in cold atoms published in Physical Review Letters. That really got both Chris and I into the field. He subsequently did a lot of great things and wrote a wonderful book on cold atoms. Over the years, I had quite a number of students working in cold atoms and exploring various topics, for example, the physics of spin-orbit coupling of atoms, artificially induced by lasers. Very clever techniques. That led to quite a few interesting, and still a few unsolved, problems.

Thomas:

Can I ask a very naive question? When we talk about something like pion condensation and kaon condensation, these are, of course, very short-lived particles. What should I be imagining when I think of something like that?

Baym:

That is hardly a naive question, it’s actually very deep. The naive way to think of it is, if you have a gas of neutrons, the neutrons can turn into proton plus a pi minus, and as long as this process is energetically possible, which means that the neutron’s chemical potential is high enough to supply enough energy in the process, then you would make all these pi mesons, and the pi mesons, being bosons, would likely form a condensate. But it actually turns out that’s a gross oversimplification. What is really going on at a deep level is that there is really a soft mode instability of the neutron gas in the spin-isospin channel.

It’s a collective mode of the system which goes to zero frequency whereupon the neutron fluid makes a transition to a new state. What’s really happening microscopically at a quark level is that the down quarks rotate slightly into up quarks. This sounds like it violates charge conservation, which it does. And it compensates it by creating a negative pi minus field. Neutral objects become positive objects, compensated by the formation of a charge field, which the pion field. It’s technically a very beautiful problem. And kaon condensation is rather similar, involving rotation of down quarks into strange quarks as well.

Thomas:

As you make the transition from a world where Bose-Einstein condensates are something theorized to something where they’re the object of experimentation, how radically does that change what you’re doing? Do a lot of ideas get rejected or confirmed? Or is it something in between?

Baym:

That’s a lovely question. In fact, we knew that superfluid helium-4 was a Bose-Einstein condensate. That had basically been known since 1938, when the superfluidity of helium-4 was discovered. And theoretically, it’s identical to what’s going on in the atomic Bose condensates. But the beautiful thing about the atomic Bose condensates is that one can vary tremendous the physical parameters. You can vary the density, do beautiful tricks with magnetic fields, you can vary the interaction strength between particles, change it from attractive to repulsive, and so on. You can study vast numbers of the different kinds of atoms. Initially, sodium and rubidium were two of the bosons. But the field then expanded into condensed fermions. What was a limited parameter space, from the point of view of nuclear physicists or condensed-matter physicists, where you had to deal with materials you were given, becomes a playground where you can dial in nearly any kind of system you want.

The cold atom field absolutely exploded into heaps of wonderful new physics, but as well, the ability to control atoms at the individual level, which is absolutely important. In a way, Theoretical problems in cold atoms are very similar to what one deals with in high-density matter as well. Even though they’re a different system with different names, the fundamental ideas are really the same. In fact, one of the big things that Chris and I did in our paper was introduce the idea that these systems had a relation to condensed-matter physics as well. Many people, including Jason Ho at Ohio State, have strongly developed this connection between condensed matter and cold atoms. That’s been a very important part of the field.

Thomas:

Is it to the point where if you make progress in one of these areas, it actually translates or inspires progress in the other field? Or are they not quite that closely connected?

Baym:

Yes, one area we worked on extensively was vortices in cold-atom systems, which is actually very beautiful. Prior, in helium, one could make a few vortices and image them. This was very, very hard because helium has an index refraction of almost one. It’s very hard to see what’s going on in helium optically. But in cold atoms, wonderful techniques were developed. We really made a lot of advances in the properties of vortices of cold atoms. But this led, interestingly, to studies I was involved in with other people of what happens to vortices inside neutron stars. And in particular, what happens in a neutron star as the matter turns from hadrons, neutrons, protons into quarks in the interior. How would vortices be connected? That turns out to be a very beautiful problem.

But again, what’s nice about it, this is an area one would not normally have thought about, except for the cold-atom impetus. I should mention, one other aspect that I was involved in, just a wonderful problem, closely connecting cold atoms and nuclear condensed-matter physics. In 1998, I was at a big nuclear physics meeting in Paris that was had where I began talking to Dominique Vauitherin. He pointed out a very interesting question. Ideal non-interacting bosons undergo Bose-Einstein condensation, as originally written down by Bose and Einstein very early, at a temperature which is proportional, basically, to the density of atoms to two-thirds power.

Then, you ask, suppose you allow a little bit of interaction between these particles. If they are, let’s say, tiny, little hard spheres pounding into each other, does the transition temperature go up, or does it go down? And with what power of the density does this happen? There’s a dimensionless parameter, which is the density times the radius of the hard sphere cubed. How does the Bose-Einstein transition temperature depend on this parameter. Quite interestingly, this had been a problem going on since the mid-50s. Lee and Yang, in their classic papers on interacting bosons, worried about this problem. It turned out, they had something between the wrong answer and an unsure answer. But they missed it completely. And there were many guesses over the years, about how the transition temperature changed.

Then, my colleague, David Ceperley in Urbana, who does numerical computation of many-body systems, showed computationally that the transition temperature, when you added a repulsion between the bosons, would go up slightly with the first power of the density empirically. And so, Dominique mentioned this problem to me at the Paris meeting, and he mentioned also that Frank Franck Laloë at the École Normale, who was basically an atomic, more condensed-matter-oriented physicist, would be very interested in this. I spent the spring of ‘99 in Paris, nominally at Saclay working with Jean-Paul Blaizot, but also at the École Normale, shuffling between these two groups of people, Blaizot at Saclay, myself, and Vautherin at University of Paris Orsay, and Franck Laloë at the École Normale. We all came from quite different directions, but this problem brought out a strong connection between atomic, nuclear, and condensed-matter problems.

Eventually, we solved the problem. That was very nice. We managed to show that the transition temperature does go up somewhat, linearly within that density, and it has been confirmed in a lot of later work. This problem turned out to be a particularly hard critical problem—critical in the sense of phase transitions—in the sense that since it was linear in the radius of the hard sphere or the scattering length. And you might think that since it the interaction strength is proportional to the radius of the hard sphere that you could just use elementary perturbation theory. But when you work in perturbation theory, the first-order term, what you expect it to be, linear in A, the scattering length, is identically zero. And then the next order term diverges. The next next order diverges even more. It was a problem where every order of perturbation theory diverged more and more. It was a very beautiful and difficult problem, and a connection between the different fields.

Thomas:

The next thing on my agenda is to ask a similar set of questions about heavy ion collisions and RHIC starting up. RHIC starts up in 2000, after 17 years. I saw your paper, RHIC: From dreams to beams in two decades. I also read Bob Crease’s article on the origins of RHIC. Have you ever seen that?

Baym:

Yeah, he was writing to me at the time. I was supposed to give the opening talk at the 2001 meeting on ultra-relativistic heavy ion collisions at Brookhaven. I’d just come back from Taiwan and the next day, broke my wrist doing Tae Kwon Do. I was completely unable to travel. [laughs] So, I never actually gave the talk, though I wrote it up as the paper, Dreams to beams.

Thomas:

It’s good to have the paper. The one question I want to have about what we discussed last time is, when you were proposing RHIC at Lake Cayuga in 1983…

Baym:

Wells College, yes.

Thomas:

…had the non-Brookhaven people been aware at that time that there had been discussions at Brookhaven about possibly doing a heavy ion collider, possibly before or after Isabelle, or in a worst-case scenario, if it was terminated, as it ultimately was?

Baym:

There were certainly people at Brookhaven, Tom Ludlum, for example, who were quite aware of the possibility of doing heavy ions. But of course, until the Wells College meeting, nobody really thought doing something at Brookhaven would be a possibility. Because they were devoted to Isabelle. The main contender was Berkeley, with their project, VENUS.

Thomas:

RHIC does ultimately start up in 2000. But before that, there’d been things like fixed-target experiments at CERN. What were we learning from the colliding beams that we weren’t learning from fixed-target experiments?

Baym:

Good question. Colliding beams enable you to get to much higher energy densities. Effectively, when you have a fixed-target experiment—you know the experiment with the balls that hit each other. One balls comes and hits the other, that one carries all the energy away. That’s the quintessential fixed-target experiment. You may have a very high-energy beam coming in, but most of the energy is in recoil of the particles that are hit. In a collider experiment, with the center of mass fixed in the lab, the beam energy goes into exciting the internal degrees of freedom of the matter.

The particles deposit essentially all their energy in the collision. You’re able to get to much higher energy densities that way. In fact, one thing that did emerge is that one was making quark matter. There’s no evidence of making quark matter in the fixed target heavy ion experiments. They were very valuable, extremely valuable, for people to learn how to do heavy ion experiments, learning what questions to ask, etc. As Kozi Nakai, a wonderful Japanese physicist, put it, it’s not enough to make quark matter, you have to understand its properties as well. That has certainly been a prime focus of Brookhaven and experiments of the LHC as well.

Thomas:

What are some of the key things that we’ve been learning? RHIC is now coming up on the end of about a quarter century worth of runs. And we’ll get to the Electron-Ion Collider in a minute. But what have we been learning over this time through our access to quark matter via experiments?

Baym:

Let me first discuss what we haven’t learned. My primary interest in heavy ions was the possibility of making very high density cold matter. And you can’t do that very easily because at high densities, it becomes very hot as well. You really are studying hot matter in the collisions. One of the marvelous things that was learned was that quark matter produced at RHIC looks almost like a perfect fluid in the sense that it has extremely low viscosity. There is a very interesting bound that emerges at string theory level in various models, that the ratio of the viscosity to the entropy density was bounded below -- the viscosity could not arbitrarily low. The bounds is basically 1 over 4 pi. It turned out that the viscosity of the quark matter that was being produced at RHIC came very close to satisfying this bound, within a factor of two. Interestingly enough, in cold atoms, there’s a wonderful connection. One could deduce the same ratio, and it also very close to this bound. First of all, this result is a lovely demonstration of the universality of problems of physics, be they at micro-Kelvin level or multiple MeV temperatures. It’s all very much the same physics going on there. A lot of the RHIC studies have been sort of bread-and-butter topics. How many different kinds of particles would collisions make and so on.

Thomas:

You’ve also been very closely involved with what’s going to be the next step, when they install the Electron-Ion Collider. That was a process that began almost as soon as RHIC started operating. I think they started mentioning it in the long-range plan in 2007, was it? The NSAC?

Baym:

That’s correct. Interestingly, I took part in the physics decadal survey of the National Academies from 2010 that dealt with nuclear physics. Our committee very carefully studied the NSAC long-range plan and really felt there was not a very good argument or physics basis why one should be building an electron-ion collider at that stage. I think what happened since then is, the electron-ion proponents went back to the drawing board and really worried much more about making a good case for it. Then, in 2016, the Department of Energy came to the National Research Council of the National Academies. wanting to build a big accelerator and maintain forefront accelerator physics in the United States. They asked the question, "Is the science of an electron-ion collider good enough to justify building such a collider?" A committee was formed, and I was asked to chair it, which I did. It was a very interesting committee. A lot of people just felt quite agnostic about the question. There was a contingent, mostly from JLab, Jefferson Laboratory, the electron accelerator, which was very much in favor of building an EIC. Over the course of a year and a half, we heard lots of arguments and eventually came to the conclusion that yes, the science was justified.

We phrased the scientific issues in terms of three challenges. One question was, how do we understand the mass of the proton? The famous Higgs mechanism gives rise to the mass of the elementary particles. If you look at the elementary particles within the nucleon, there are three basic quarks, plus lots quark-antiquark pair. If you add up the masses of the three quarks, you would conclude that a normal person would weigh something on the order of a kilogram. In terms of the mass of the nucleon, it’s about a factor of 100 bigger than the mass of its three quarks. What’s interesting about the nucleon, if you go back and look at other systems like in atoms, you discover that the mass of the atom is a little bit less than the mass of the nucleus plus the mass of the electrons, That’s the binding energy. Similarly the masses of nuclei are less than the masses of the nucleons they contain. That’s also binding energy.

One the other hand in the nucleon itself, you have the quarks, and discover that the mass of the neutron and proton is 100 times bigger. It’s a very unusual system. What is going on there is, of course, that lots of components in the nucleon are relativistic. The quark-antiquark pairs, lots of gluons and such. The question is, what and where are these ingredients? For example, are the gluons concentrated to the center, or towards the outside? One just didn’t know these questions at all, how the distribution depends on the energy scale which is studied—how fine a microscope are you peering at things with? There are a lot of really very good and hard questions here.

A second interesting question is the spin of the proton. Again, if you naively add up the spins of the three quarks inside, you don’t really get the right answer. That problem is really basically still open. You don’t know what the gluons are contributing, you don’t know how much orbital angular momentum the particles, the quarks and the gluons, are contributing to the spin of the proton. We all take the spin for granted every time we lay down in an MRI machine to do nuclear resonance on our body. We take for granted that our protons spin around, but why they have the spin value they do, we don’t understand. That’s another thing that the EIC will study.

A third very interesting question, which is driven in part by the connection with many-body physics, is that, as was discovered at HERA in Germany, if you look at very short timescales, you see lots and lots of very low-energy gluons present in the system. So many that they form an interesting many-body system. They’re interacting together. And the properties of the system are completely unknown here. That was another of the “cartoon” ideas, motivations for future experiments.

Thomas:

You had two long-range plans, the 2007 one and the 2015 one. And it was only really with this National Academies study that—I’m sure all these ideas were in circulation, but this is the real articulation that’ll get you to CD0, in the DOE lingo, for this.

Baym:

That was amazing. Our report was issued the summer of ‘17, and CD0 went in almost immediately. The Department of Energy was anxious to do the project. And it’s been quite on track ever since. I’ve never been on an Academy study, or any study, where the government responded so rapidly.

Thomas:

I was actually following this in my previous policy job, and I had the same impression, that they did move very quickly on it at that point. They hadn’t done the site selection at that point. There was a proposal from Jefferson Lab and a proposal from Brookhaven. And it seems to me that the Brookhaven one was a lot more straightforward. Maybe they were looking for something to do after RHIC at that point.

Baym:

That played a bit of a role. JLab, Jefferson Laboratory, had just gotten an upgrade as well. But a big driver was that I think it is technically far easier to put an electron tunnel inside the RHIC structure, the civil engineering that was already present at RHIC, than it would’ve been to construct a proton machine at JLab, where people had no experience whatsoever in hadron physics. Building a tunnel is—I was going to say an easy job. It’s not easy at all, but there are a lot of technical issues to be solved if one wants very high-luminosity beams, which means the beams have to be really focused. Which means that you have to do cooling of the beams to ensure straight motion as much as possible. There are many challenges there that are really still under development. As an accelerator project, the EIC is a wonderful machine for solving a broad range of accelerator problems. I can understand the DOE’s deep interest in doing this.

Thomas:

The other interesting aspect—I have no idea if you have any insight into this—is that if you look at the previous long-range plan, not the one that just came out, the 2015 one, there was a notion that they would do the neutrino-less double beta decay experiment before going on to EIC, and that has not happened. Again, I don’t know if you have any insight.

Baym:

In neutrino-less beta decay, the nuclei are not cooperative. The decay rates are very low. [laughs] In fact, one doesn’t know the rates at all, presenting enormous uncertainties in the rates at which nuclei will undergo double beta decay.

Thomas:

The last transition that I want to look at, and this is based on another paper that you’ve written, The Golden Era of Neutron Stars. We’ve entered a new phase of observation. Of course, we’ve been able to observe them for many decades now, especially with pulsars, but things are changing. Could you walk us through that?

Baym:

Where to begin?

Thomas:

Well, for one thing, would you say it’s been a real step function, or is it something that’s just been accumulating over time, our capabilities?

Baym:

I think one of the big step functions has been in observation. In the last 15 years, we’ve discovered heavy pulsars, neutron stars at masses over two solar masses. And so, the question is, how do you support two solar masses? The standard picture was always nucleons would be the source of the pressure. And at higher densities, that picture begins to break down. What was wrong with pure nucleons is, first of all, the sound velocity in all nuclear calculation becomes greater than the speed of light. While nobody’s ever proved deep down that the sound velocity, which is basically a thermodynamic derivative, has to be less than the speed of light, it makes good sense that you can’t send messages faster than the speed of light, in accord with special relativity. Well, at high enough density, the nucleon picture breaks down here. One can keep decorating the nucleons in various ways, assuming for example that the effective mass of the nucleons becomes small. However, what is the basic physics behind such a conjecture is up in the air.

You can make models that sort of work. But another problem is that you approach nuclear matter by solving the Schrödinger equation for a batch of nucleons, with two body forces between the nucleons, and you discover is that it doesn’t work all that well. You get approximately the right binding energy for symmetric nuclear matter, but you find the density is about double what it should be. If you try to calculate light nuclei, helium-3, helium-4, etc. you discover that they are also under bound. What this tells you is, there must be some three-body forces. There’s a big industry of people implementing three-body forces to make this picture look good, and that becomes a little bit phenomenological at best. But as you go to higher density, what about four-body forces, five-body forces?

And then, you have to ask the question, does it really make sense to be describing high-density matter in terms of forces between nucleons at all? You have to worry about retardation effects and so on. And then, you ask the question, does it really make sense to be using nucleons themselves at all? One of the sources of everything we know about nuclear matter calculations—well, input for nuclear matter calculations, are nucleon-nucleon scattering experiments. Suppose that you knew everything there was to know about niobium-niobium atomic scattering. Would you predict that niobium metal was a superconductor? The answer is no. And if you did sodium-sodium scattering, would you predict that sodium metal is not a superconductor? No. Once you go to higher densities, new degrees of freedom open up. And you have to take this into account.

Let’s go back to my involvement with high-density matter. I’d always been interested in the possibility of quark matter. I had a wonderful postdoc, Toru Kojo, who came to Urbana around 2015. We started thinking about the possibility of quark matter again. Tetsuo Hatsuda, with whom I have been collaborating since the very early 2000’s on neutron stars and cold atoms over the years, first at the University of Tokyo, and then at RIKEN, where he is now, was having interesting ideas in this area. Well, my Japanese colleagues and I began to develop the picture of nucleons going continuously into quark matter with increasing density, taking the point of view that we know, at low density, what’s going on because it’s nuclear matter. In fact, we don’t know what’s going on fully, but that’s another story. [laughs] At high density, we know that at high density one has quark matter, which we understand well at extremely high density, well beyond that in neutron stars. We understand the low and high density ends. and one can just interpolate in between. it turns out that the interpolations are very highly constrained, both by thermodynamic stability, and by the requirement that you can explain heavy neutron stars, above two solar masses. We don’t understand the interpolation region. How the degrees of freedom go from nucleonic to quark is a major unsolved problem And in fact, we don’t understand quark matter well enough.

The primary way of describing quark matter, of solving problems of QCD, which is a highly nonlinear theory, is through lattice gauge theory. But lattice gauge theory breaks down at finite density. There’s this funny thing called the fermion sign problem, which is too complicated to talk about now which prevents calculating at finite density from working. While one can understand the properties of the hot vacuum very well by lattice gauge theory, but one cannot do so at high densities. That’s been a big problem. One is forced, in a way, to use semi-phenomenological type theories. We’ve been using the Nambu—Jona-Lasinio model applied to quarks at high-density quarks, a phenomenological model in which the gluons are effectively integrated out model. This works very well for an enormous variety of QCD problems. That was our picture. We developed that, and produced our best equation of state, which we called QHC 21, ‘21 being the year, QHC being “quark hadron crossover.” Actually, the crossover from hadrons to the quarks. It seems to be working very nicely.

There were two interesting breakthroughs that came here. My earlier paper with Chin way back in the ‘70s, had as I mentioned, the field back because took the rules of the game to be that you had to look at quarks that had lower energy density than hadrons, on the (false) assumption that there there are states of hadrons at very high densities. But there are no states of hadrons. No more than there are states of atoms at metallic density. And no more than, as I mentioned, if you pack eggs at 10 times the normal density, you can’t describe egg matter in terms of these little oval white things that are called eggs at all. One of the good things about our equation of state approach is we allowed all possible quark-matter equations of state in the game. They were not required to be soft. And we were able to explain several high-mass neutron stars.

A very interesting feature of this, is that to explain the high masses we also had to include BCS pairing between quarks. And the role that BCS plays is another beautiful example of the influence of condensed-matter. Pairing means that the particles like to form pairs; they are attracted to each other. And what we found was that the strength of the pairing correlations was stronger than you would predict on the basis of perturbation theory in very high-density matter. This is completely natural because as you decompress quark matter down towards the density inside a normal nucleus, of course, the quarks want to be bound together, to form triplets. The fact that strong two-body correlations want to come in was a very satisfying result. We have a handle on ensuring that quark matter is consistent with high-mass neutron stars.

Then in 2017 came the famous multi-messenger event, seen by LIGO. This was two neutron stars undergoing a merger, and was seen immediately in gamma rays and then all across the entire electromagnetic spectrum. That event was one of the most beautiful discoveries ever, I think. It gave some insight into the possible equations of state, but not very accurately. The basic game was, as the neutron stars come close to each other before merging, they get distorted, and the distortion depends on what’s going on inside them. And their shape is also reflected in how fast they rotate about each other. LIGO deduced some information about the tidal deformability of the stars, and guesses as to what the equation of state had to be, both of which were And that quite consistent with what we were doing. That was very good.

But the really great discovery was NICER. NICER, the Neutron star Interior Composition ExploreR is a small washing machine-sized experiment that was put on the International Space Station in 2016.

Thomas:

NASA acronyms are the best.

Baym:

That’s how you get NICER. And NICER’s been doing almost nanosecond-time resolution of events. It is able to measure radii and masses of neutron stars. So far, only two neutron stars have been fully analyzed and published. One was a neutron star of about 1.4 solar masses and a radius of around 12 to12.5 kilometers. What’s interesting is, this is a little bit bigger than the radii predicted by nuclear matter, which is something on the order of 11 kilometers. But the differences are sort of in the noise. Then, NICER analyzed another pulsar, the heaviest neutron star known at the time, a mass of about 2.08 solar masses, and found that when all the dust settled its the radius was probably about the same as the 1.4 solar mass star. It was quite striking. You ask the question, what happens as you put more mass on a star? It gets heavier with more gravity and should get smaller. The same way as when you lie down on your pillow, if your head is very heavy, you’ll sink far down. The pillow gets much smaller. But how could it be possible that stars are keep roughly the same radius as they get more massive? It means that new degrees of freedom must be coming in, and they become more and more repulsive. And these have to be the quarks beginning to play a role here.

Thomas:

Obviously, NICER is a very specialized instrument. Who is pushing that particular project through?

Baym:

I think the original ideas—one of the people involved was Fred Lamb in Urbana, whom you may know through the APS ballistic missile study. And Cole Miller, a former postdoc in Urbana and now at the University of Maryland has been one of the major data analysts. There’s also been a group run by Anna Watts in Amsterdam. NASA likes this project. It’s been immensely simple and has been yielding a wealth of data. And there will be other neutron stars analyzed, giving more mass-radius information in the future.

Thomas:

Can you explain briefly how it works? Is it just focused on the neutron star all the time? How does it determine the radius of something so small at that distance?

Baym:

I don’t remember exactly how long they observe each star, but the picture is the following. You know how light is bent by the sun, and other gravitational sources. Light is bent in gravitational fields. Because neutron stars have enormous gravity, because of the light bending, you can see a little bit behind the star. If I look at you, I can see your ears. I can’t see the back of your head. But if you were a neutron star, I’d be able to see a little bit of the back of your head as well because of the light bending. The idea is to follow bright spots on the neutron star. The source of the spots would be involved with the pulsar emission process, with very high-energy-density electrons being ripped off the surface by gigantic electric fields of 10 to the 18th volts difference between the north pole and the equator, for example. The particles that are rippedreturn to the surface eventually. And so, beams would fall back onto the neutron star surface, making hot spots. You look at a hot spot in the back, and follow it to see for how much of the 360-degree rotation you actually see the hot spot. Naively, since that’s a pure number, it can only tell you the ratio of the mass and the radius. You have to go beyond lowest order General Relativity, looking at properties of rotating neutron stars, where the rotation speed also enters in order to be able to deduce a mass and radius. Now, that’s the easy part. The hard part is, these spots—is it one spot, or two, or 10 spots? Are they oval-shaped? What do they look like? In the end, you really have to make models of what the possible spots could look like. It turns out, the results seem fairly insensitive to the details of the models.

A few round spots seems to work rather nicely. There are certainly levels of uncertainty in all this. But I think it’s the best insight we have into the equation of state of high densities, and it’s much better than RHIC or LHC can offer because there, you’re stuck with high temperature simply because so much kinetic energy is dumped into the collision. You can’t get down to zero temperature. In a sense, our advertising that RHIC would solve problems with neutron stars was a good selling point, but a little bit of an exaggeration.

Thomas:

If we can return to LIGO for a second, I think you’ve mentioned that you were on the NSF panel that approved the project. Is that correct?

Baym:

Yes, that was in 1983. October ‘83, Kip Thorne, Ron Drever, and Rai Weiss came to the NSF on G Street in Washington to describe this project. Everybody said, "Wow. The probability of anything happening is zero. But the benefit of anything happening is infinite, so we have to support it." And that became the first large-scale NSF project. It was originally $50 million. Little did we imagine what a wonderful, wonderful idea it was, and what we would learn about the Universe from detecting gravitational waves/

Thomas:

It took a lot of resolve to push that through. On the kilonova, just going back to that for a second, how much of the information that you’re getting from that comes from the actual gravitational wave signal itself, and how much comes from the observations that were triggered by the signal?

Baym:

Oh, gosh. Not much in the gravitational wave signal. The problem with the gravitational wave signal is that at present resolution, the detectors are sensitive up to around 500 Hertz, whereas it’s probably double that where the neutron stars begin to merge. So, you don’t really see in the gravitational wave signal what is happening in the merger itself. There, the information is electromagnetic.

LIGO comes back on the air within a couple of months. Virgo unfortunately has been shut down for technical reasons, but it will be up soon again. And KAGRA, the Japanese experiment, will also be functioning shortly. One expects, within the next year or so, to start to see many more binary neutron star mergers. If all goes well, you could see maybe up to 100 binary neutron star mergers per year, which would be an incredible wealth of information.

Unfortunately, it’s not good enough simply to detect a binary neutron star merger. There was another one a few years ago, but it was seen with only the LIGO Livingston detector; the Hanford detector was down. The one detector could not localize the event in the sky. And without sky localization, you cannot tell the electromagnetic observers, the telescopes, where to look. The one we had in 2017 had good localization, and was also seen in gamma rays by the Fermi LAT and also by Integral, two gamma ray telescopes. That one was very well localized in the constellation Hydra in the southern hemisphere.

Thomas:

And a lot of observatories found it very quickly, so it was a very nice and lucky observation.

Baym:

Yeah. But unfortunately, Hydra went behind the sun. The Earth moved to where the sun was in between, so electromagnetic observations just stopped after a short while.

I would like to just mention more recent stuff that’s kind of interesting that I got involved in.

Thomas:

Are we going to be talking about primordial neutrinos?

Baym:

Yeah.

Thomas:

The other messenger, yes.

Baym:

Primordial neutrinos from the Big Bang turned out to be a very interesting problem. I have a wonderful colleague, Jen-Chieh Peng. Who just before the pandemic, in February 2020, said to me, "There are all these relic neutrinos from the Big Bang coming. What happens to them en route?" The question was that neutrinos are always made left-handed; their spin points opposite to their direction of motion. But in General Relativity the orientation of the spin and momentum be changed by gravitational fields. This was sort of presented in a not-very-clear set of papers early on.

Interestingly, if a photon is bent in gravity, a right circular polarized photon remains a right circularly polarized. On the other hand, if you just throw a ball up on the Earth, and it’s spinning one way, it goes up into the air and comes back spinning the same way, but its momentum is reversed, changing the relative orientation of the spin and its the momentum. A right-handed ball comes back left-handed, simply because its momentum’s reverses. Neutrinos, having a small mass, lie in between these two results. Once the pandemic came. I sat at my dining room table, basically, for a year and a half working on this problems, finding in great detail the amount of bending of the spin and momentum that a neutrino coming from the Big Bang would experience en route. One of the important inputs here was through the Planck Observatory, which measured density fluctuations in the early universe, the fluctuation in the microwave background. Jen-Chieh and I were able to use that data to predict how much the neutrinos would come out with a chance of being right-handed instead of left-handed. Unfortunately, it’s low, and it’s not something we’re going to measure very shortly.

But neither are we actually going to see the neutrinos from the Big Bang themselves for a while. The most promising method of doing this is through inverse beta decay in which neutrino is captured by tritium, which turns into a helium-3, emitting an electron. You measure the electrons. But you have to have very good electron energy resolution to distinguish it from the background of tritiums ordinary beta decay, which also emits electrons. You get slightly higher-energy electrons from capturing the neutrino than you would from the beta decay. That’s due to the fact that in the capture process, the neutrino mass makes the electron more energetic. But one is nowhere near the capability of resolving the energy electrons well enough to see the difference. That’s one problem. The second problem is that nobody has ever even seen the inverse tritium beta decay, this process where a neutrino comes into a tritium and makes helium-3. That process hasn’t been observed. One of the things I’ve been worried about is ways to see processes like this as well. Very strangely, over the course of the last few years, I’ve unexpectedly become a neutrino physicist. [laughs]

Thomas:

I was going to say, it seems like a bit of a departure for you. Neutrino physics is quite different, at least in my view, from condensed matter.

Baym:

1,000% departure. But it was a wonderful thing to be able to be doing over the pandemic. Just sitting at my dining room table, where I’m sitting right now, just scratching away. Another interesting thing we looked at was neutrinos, because they have a finite mass, have to have a finite magnetic moment. Just like the protons, with their spin. And the question is, as the neutrinos go past magnetic fields in the early universe and in galaxies, how much does that rotate the spin compared to the momentum? The momentum’s not changed at all going through a magnetic field because the neutrinos are neutral. If they were charged, then they would. But the spin can be rotated by the magnetic fields. And so, that also produced spin rotations that may have been comparable to the gravitational rotations. We just don’t know the neutrino magnetic moment. This was an absolutely wonderful, fun project. I think in my later years, I’ve become one of the world’s experts on things that are very, very hard to measure. [laughs]

Thomas:

It’s sometimes been the case with you. When you started with neutron stars, they were not very easy to measure. And still aren’t, but maybe a bit easier than they were.

Baym:

Yeah. It takes time.

Thomas:

Yeah. That’s really brought us up to the present. Unless you want to dwell on some other aspects of the science, I thought we might kind of cool down with some discussions about some of your collaborations, the University of Illinois department that you’re in, that sort of thing. You’ve mentioned Chris Pethick on a number of occasions. As I was looking at your CV, I note that over the decades, that’s been a continual collaboration for you. I’m wondering if you can just bring out your relationship with him a little bit.

Baym:

Yeah. Chris came to Urbana as a postdoc in the mid-60s, and we talked on and off. And then, things really got going when we started doing neutron stars together in 1970. Over the years, we’ve always been in close touch about everything. Eventually, we worried about superfluids, properties of helium-3, and such. Then, when cold atoms began, we started working furiously together. And we still are. A very interesting problem we’re still concerned with that we’re working on very recently is understanding the conservation of angular momentum in simple beta decay. It turns out, understanding angular momentum there eventually brings one down to basic measurement problems in quantum theory. Which is a place I never expected to be landing. We’ve just have a really wonderful working relationship over the years.

Thomas:

Is that a case where you’ll be talking about what you’re working on, and you say, "Why don’t we work on this together?" Or are there certain aspects the problems you would consult on?

Baym:

A little bit. He’ll say, "I’ve been thinking about this." "Oh, yeah, so have I." That was an interesting question. When gravitational waves were discovered in ‘15, we both started thinking independently of how they are damped going through matter, through the cosmos. I was in Copenhagen then. We just started talking and eventually wrote a lovely paper on the subject.

Thomas:

Another figure you mentioned in the first session that we did is Gerry Brown. We discussed him briefly, discussed how he wound up in Copenhagen. You mentioned that he was a mentor-type figure to you. I know that you’ve done at least a couple of papers over time.

Baym:

Yeah, Gerry was a very interesting physicist. He had very little critical sense, but he had wonderful ideas. He sort of operated in a way of coming up with outlandish ideas, so one was immediately devoted to proving him wrong. [laughs] That was certainly part of his role in high-density neutron star matter, where he was quite provocative. Then, we also worried about things like the Ericson-Ericson Lorentz-Lorenz effect (a funny title for a paper). There were two Ericsons, Thorlief and his wife Magda, one Lorentz and one Lorenz, the former the famous Lorentz of the transformation, and the other a teacher of Niels Bohr. In nuclear physics, Ericson and Ericson found an effect in pion nucleon interactions that was the analog of the Lorentz-Lorenz effect in dielectrics, based on the difference of the local electric field and the mean field. We related the effect to the Landau-Fermi liquid theory, showing how the original theories was in fact first attempts to derive correlations functions in many particle systems. This work was wa nice combination of ideas, which we worked out in considerable detail. And it had to do with the observable effects in pion-nucleon physics.

More generally I’ve really enjoyed, especially lately, talking to experimentalists. They really know what’s going on. [laughs] Theorists can be all over the place. My neutrino collaborator, Jen-Chieh Peng is one of the great neutrino experimentalists, who did crucial measurements at the Daya Bay reactor near Hong Kong, of the PMNS mixing matrix elements for neutrino mass states in terms of neutrino flavor state. An electron neutrino is a superposition of three basic mass eigenstates, which go by the dramatic names of one, two, and three; the muon neutrino is another combination, and measuring these combinations are has been very important, which Jen-Chieh contributed immensely to. Another person I’ve had very nice collaborations with is Barbara Jacak, who was a nuclear experimentalist in the Brookhaven AGS fixed-target project, and then at RHIC. And now, she’s at Berkeley. But what’s wonderful in working with experimentalists, is that they really understand the physics, the basic ideas of what’s going on, what can be measured, what are the important problems to look at.

I should mention, one other person I’ve been collaborating a lot recently is Doug Beck in Urbana, who’s a nuclear experimentalist. He came to me because of the helium-3, helium-4 problem that I worked on, because he and collaborators were designing the experiment at the Oak Ridge Spallation Source to search for the electric dipole moment of the neutron -- a very subtle problem. Deep down, the neutron can have an electric dipole moment only if time reversal invariance is violated, something beyond the Standard Model we don’t understand. This is a beautiful experiment, using polarized neutrons interacting with helium-3, sitting in a bath of helium-4. He wanted to understand the general plumbing of what happens to helium-3 in helium-4 as a function of time and so on, which opened a wonderful range of problems. He, Chris Pethick and I wrote a series of quite lovely papers on the properties of the dilute solutions of helium-3 and helium-4 that were applicable to this experiment here.

Thomas:

As I was looking through your CV, I noticed that many of your collaborators or coauthors have been international people, like Jean-Paul Blaizot, Markus Holzmann, Leonid Frankfurt, Henning Heiselberg. They keep appearing again and again. I’m curious, do you like working internationally? Is that incidental?

Baym:

Well, let’s go one at a time. I actually me Jean-Paul Blaizot in 1977 at a summer school at Les Houches, when he was still a graduate student. He’s an amazing mountain climber. He said, "I will teach you mountain climbing if you teach me physics." It turns out I’m a much better teacher of physics than he was at mountain climbing. [laughs] Only in the sense that my skills were not that great, but he really learned physics very well. [laughs] After he got his PhD, he came as a postdoc to Urbana. Jean-Paul was very involved in the collaboration I mentioned earlier to study the effect of transition temperature on Bose-Einstein condensates. One of the roles I played in this game was teaching a lot of people around the world condensed-matter physics and bringing it all together as unified problems from nuclear physics, to high-energy physics, to cold atoms, to real condensed-matter physics. Markus Holzmann was a graduate student at the Ecole Normale who got involved in this Bose-Einstein transition temperature calculation, and we’ve really remained good friends and done a lot of work together. He also spent a lot of time in Urbana. Henning Heiselberg, whom I knew on and off in Copenhagen, was also my postdoc in Urbana for a number of years.

Thomas:

That seems to be the common theme, that people have gone through Urbana as postdocs or graduate students.

Baym:

Yeah. Gerry Brown described Urbana very nicely as a "finishing school" for physicists, [laughs] in the sense of the old English aristocracy finishing schools, where they’d send women to learn how to set tables, be good hostesses, and such. In the sense that people came to Urbana—I think this is the important point—and they’d learn a much broader view of physics than simply sticking in one little narrow area.

Thomas:

I’ve seen reference to an “Urbana style” of physics, and I gather it’s become a bit of a tagline for the department recently. But I also got the sense that way back when you joined, they were trying to create something like that, Bardeen, and Pines, and those fellows.

Baym:

I think, in a way, it just happened, through the hires when I first came. When I came, I had an office suite, which I’m still in, with Bardeen, Pines, and Leo Kadanoff. The four of us were together. That was a rather nice nucleus. But our interests went off in different directions. Certainly, our adventure early on into neutron stars was the first broadening out that we were doing at the time. But another part of Urbana that was actually marvelous was that people of all fields talked to each other. We’d all meet at lunchtime, say, near the mailboxes. We would have lunch with people from all different discipline areas, theorists and experimentalists, and there was always the sense that everybody was trying to teach everybody else what was going on. That was quite wonderful, but it really doesn’t exist anymore. Things have gotten too fragmented. The pressures on people are too large, in a way. People are working through lunches, Zooming, whatever. [laughs]

Thomas:

One question that I’ve always been curious about, they call it the department head at the University of Illinois rather than the department chair. Is that just a nomenclature thing? Or is there some significance?

Baym:

No, it’s very, very important. Department chairs are generally selected from among the faculty. I can’t tell you what the choosing process is at other schools because I didn’t have a part in it. But they serve a fixed term, and then somebody else comes in. And department chairs operate by consensus. Department heads have absolute power. As long as you have a good department, it’s good. It relieves the faculty of a lot of burden, like endless faculty meetings to decide what to do. But our heads traditionally here do operate always by consent of the department anyhow.

Thomas:

Can you describe any regional connections? You’re not all that far from Chicago. I went to undergrad at Northwestern, so I know it’s a few hours.

Baym:

When were you there?

Thomas:

I finished in 2001. Before I got into history of science, I was a physics major.

Baym:

Okay. My older son and my younger daughter both went to Northwestern. One a little before you, one just after you.

Thomas:

I worked in Bill Halperin’s nuclear magnetic resonance lab.

Baym:

Oh, okay. Did you know the low temperature physicist John Ketterson?

Thomas:

No, I don’t think I had any classes or interactions.

Baym:

He was at Northwestern. He sort of roped Chris Pethick and me into writing two review articles, one on Fermi liquid theory, and one on the dilute solutions, which Chris and I eventually turned into our Fermi liquid book. Our connection with Chicago has not been that good. You can get from here to Chicago in three hours by car. Immensely boring drive. It’s a lot easier to get in the plane, change planes at O’Hare, and go somewhere else. [laughs] There are occasional regional meetings between Chicago condensed matter and Urbana condensed matter people. But it’s not a closely knit community by any means.

Thomas:

Something you just said reminded me, I wanted to ask you, you did textbook lectures in quantum mechanics. This is way back in the late 60s, so we’re jumping back.

Baym:

That’s a very good question. I came to Urbana in ‘63, and I was given the task right away of teaching quantum mechanics. I studied many different views of quantum mechanics, having been through Cornell and then Harvard. I debated, "Should I be following standard textbooks, or should I do it my own way?" I decided, basically the morning of my first lecture, I was going to do it my own way. And I started with giving a number of lectures about two-level systems, the photon. How much quantum mechanics can you extract from the photon? It turns out that’s what everybody does now, particularly the cold atom people. They know less about wave functions than they know about two-level states of atoms, in terms of states of atoms. Then, after a while, the students said, "We need notes. We need a textbook. We can’t find any of this stuff."

I started writing notes after each class. Immediately after each class, I would sit down and write some 10 pages of notes, and that was a very interesting exercise because I saw the image of the students looking at me. It’s typical when you write papers to be very clever and put in some throwaway remarks to show how bright you are, not really getting into the details of the arguments because you could assume the reader must be bright enough. But the students would all look at me, these virtual students, saying, "We don’t get it. You better explain that better." The notes were very carefully explained. I ended up after a year and a half with some 800 pages. The notes were mimeographed. You remember mimeograph machines?

Thomas:

As a historian, I run into mimeographs quite a bit.

Baym:

Mimeographs are blue; my notes were reproduced by what I think was called a Ditto Machine, where one typed the “ditto master” on a beige sheet of paper. I ended up with 800 pages of notes, and somebody suggested I publish them. It might’ve been David Pines. And a lot of people said, "No, no, no, we have enough quantum mechanics books." But I eventually did publish them, and it turned out it was very well-received, and it’s been quite successful over the years. I agreed 15 years ago to do a new edition of the book, but several things have happened. I’ve just kept going doing physics. I’d much rather keep working on things I’m working on than writing textbooks. Another thing is that, recently, quantum mechanics has run away. Between quantum information, quantum cryptography, quantum chaos, quantum computing, there are just so many new things that also need to be talked about. Being able put all that in a revision is a daunting task

Thomas:

It’s very interesting how those aspects of quantum mechanics have come to the forefront so much in recent years. They essentially weren’t taught, except as a conceptual thing, for a long time. Entanglement and what not.

Baym:

Precisely. Even the question I mentioned of the apparent non-conservation of angular momentum in beta decay gets one into the basic measurement problems of quantum mechanics, which are still not universally agreed upon. But it remains fun.

Thomas:

Gordon, I’m out of questions. Is there anything you would like to say?

Baym:

I can’t think of anything at the moment. If I come upon something…

Thomas:

Yeah, we can always insert it into the transcript. It actually happens more than you would think. [laughs] People are very eager to write more.

Baym:

Yeah, I’ve been on your side of it, too.

Thomas:

Yeah, I’ve noticed you have some in our collection where you’re the interviewer.

Baym:

Yeah. And then, also, working with Lillian Hoddeson. You’ve been great fun to talk to. I really appreciate it.

Thomas:

I really appreciate it. I’m glad that you got in contact with us to recommend this. I think this will be useful. Again, if we come up with more later, there’s no law against tacking on a third session even.

Baym:

I hope we get to meet before too long.