Marvin Cohen

Zierler:

 Okay, it is April 7th, 2020. This is David Zierler, Oral Historian for the American Institute of Physics. It is my great pleasure to be here virtually with Professor Marvin Cohen. Professor Cohen, thank you so much for being with me today. 

Cohen:

Happy to be here. 

Zierler:

Can you tell us your title, and your institutional affiliation?

Cohen:

I'm University Professor of Physics at the University of California at Berkeley. In principle, being a University Professor, I’m attached to all the campuses of the University of California. I have travelled to and worked with people at the other campuses over the years, but I'm mainly localized and working here in Berkeley. 

Zierler:

Uh huh. Okay. So, let's start right at the beginning. Tell us about your early childhood in Montreal.

Cohen:

Well, I was born and grew up in Montreal, and the part that's relevant to my career, is that when I was a small child, I enjoyed playing with batteries, small light bulbs, buzzers, and other electronic items. I loved fixing and playing with mechanical objects like clocks and locks. I eventually became a theorist, but when I was young, I was an experimentalist. I think I actually know the day when my interest changed from experiment to theory. My grandfather had a farm not far from Montreal. In the summers, I would go there. One summer, I was playing catch with a friend. I was probably about 7 or 8 years old. I threw a ball in the air, and as it was coming down, a kid came by. He was of high school age, and he said that he learned in school how to determine the path of the ball while it was in flight. He claimed that he could tell you exactly the position of the ball as it moved through the air. At the time, I couldn't get over that. I don't think I ever got over that. The fact that you could use mathematics and really understand things from that point of view as opposed to making things and doing experiments was fascinating to me. I think from that day on I became a theorist. 

Zierler:

Were your parents from Montreal?

Cohen:

Yes, both parents were born in Montreal, but their parents were born in Eastern Europe. 

Zierler:

At what age did you start exhibiting academically, not just in your wonder at the natural world, but academically, when did you start excelling in math and science?

Cohen:

I think even as a small child, I was good with numbers and math in general, and I would build things. I was fairly sick when I was young. I enjoyed being home from school, even though I liked school. I would make a lot of model airplanes. During WW2, there were so many types of planes and ships to model, mostly using balsa wood. I also did a lot of reading at home. So, I was used to looking things up on how to build things like crystal radio sets. I read anything I could find in magazines about science, but I was mostly interested in the physical sciences and especially in physics, A precocious incident that comes to mind is that I remember that in the second grade I was trying to figure out why there weren't electrons in the nucleus of an atom. In other words, it all added up so well if people assumed there were electrons in the nucleus with the protons. 

Zierler:

In second grade you were thinking about this?

Cohen:

I couldn't understand why they had to bring in those neutrons when they just could make atoms and all matter out of protons and electrons. When I got a little older, I found out that you can't put electrons in there, because they won't go in. I think I understood why after a while. The bottom line is that I was very interested in physics at an early age, and I was persistent about trying to figure things out. I just turned 85, so let's say that for the past almost 80 years I probably thought about physics in some way almost every day of my life. Physics was always the prominent academic subject in my mind. But it was not always my major activity, especially in my teenage and college years. I should add that I was always interested in music, too. I won a singing contest when I was three years old. So, although I did not focus on physics at every stage of my life growing up, I always knew I wanted to be a physicist. 

Zierler:

Tell us about your education in Montreal. Did you go to private school or public school?

Cohen:

I went to a public school that was only a few blocks away from my house. Coincidently, a couple of days ago I noticed that you can use Google maps to look around cities. So, I looked at where my school was, and it's now an institute of some kind. It was a very good school, somewhat rigid, but I liked it. The curious thing about my early education is that when I moved to the United States at age 12, I was, according to the teachers here, far ahead in math and “things like that”, but since I came from a foreign country, they wanted me to “catch up socially”. They didn't seem to realize that Montreal wasn’t provincial; it was larger than San Francisco. I didn’t understand. I thought I was missing something, so I started by smiling at people in an effort to catch up socially. I think, this encounter had a big effect on me. I basically spent much of the next 10 years focusing on social things. I really got back to academics in graduate school. So, to answer your question, I guess I got a strong background in math, and it was a very good school.

Zierler:

Did you take physics classes in high school?

Cohen:

Yes, this was in San Francisco. It was terrible. 

Zierler:

When did you move to San Francisco?

Cohen:

My parents, my brother Gordon (born 1940), and I moved to San Francisco in 1947, and I went to junior high and high school in San Francisco. 

Zierler:

What caused your family to move to San Francisco?

Cohen:

My father had several brothers here, and they kept saying it's sunny and nice in California, it's too cold in Montreal, and it's America, and so forth. 

Zierler:

Did your parents become citizens?

Cohen:

Yes. 

Zierler:

When did you become a citizen?

Cohen:

When my parents became citizens, I was 18 years old. I remember, it was the time of the Korean War, and I recall that after they asked me the traditional question about what are the three parts of our government, they said, "Young man, are you willing to fight for this country?" And then the questioner said, “Speak up!” as I sat there realizing the gravity of the whole thing. Then I said, "Yes." At that time, as far as I know, you couldn't have joint citizenship, so, I gave up my Canadian citizenship and became an American. I was really happy to become an American citizen. As a child in Montreal, I actually built a model of the Golden Gate Bridge out of toothpicks. So, I was really looking forward to moving to San Francisco, and I knew I would like it here. Of course, I missed my family and friends in Canada, and it was a big change in my life, but I remember clearly that I was very happy to be here. I still am.

Zierler:

When you graduated high school, was Berkeley the be all and end all in terms of where you wanted to go to school?

Cohen:

Yes. As a high school student, I wanted to go to Berkeley. At the time, the difference between Berkeley and Stanford academically was large. Berkeley was really very good. The state was supporting UC very well. The tuition was $75 a year! It was a wonderful period for public education. However, one of the most amazing stories about educational institutions in California is really Stanford, because it was, in my mind, far below Berkeley at that time, and the rise in stature of Stanford over the years is extraordinary. The administrations that they’ve had were so good, and the planning was excellent. I believe that is why Stanford became such a high-ranking institution now. Berkeley was where I wanted to go. I didn't think much about it. It wasn't hard to get into Berkeley. I spent a lot of my time in high school playing music, particularly jazz. So, I was running around sneaking into the jazz clubs as an underaged person, and playing with groups, but I didn’t do all the rest like drugs and alcohol which often go with this. I still had an interest in physics, but it certainly wasn't connected with school at all. I was just reading about physics outside of school, because in the physics classes, we would just plot the temperature of the air each day, and study things like that. It was a turnoff in many ways, but I knew there was much more to physics than what we learned in high school. But the math classes in high school were really quite good. I was at George Washington High School, which was and still is a very good public school. I should add that in addition to playing jazz on the alto saxophone, I played classical music on the clarinet in the band and orchestra. I still play the clarinet (almost every day), but only classical music.

Zierler:

When you got to Berkeley, did you declare a major in physics right away, or that took some time?

Cohen:

No, right away. 

Zierler:

Because you knew that's what you wanted to do?

Cohen:

Yes, I always knew that's what I wanted to do.

Zierler:

Did you think that you were on the academic track from the beginning, or were you considering a career in industry at some point?

Cohen:

Always academic track.

Zierler:

So, the dream was essentially to become what you became, which was a professor of physics?

Cohen:

Right. I am really fortunate. After being a professor all these years, I've come across so many talented students who were so bright that they could do practically anything, but they didn't know what they wanted to do. That's tough, so I was blessed because from the beginning I wanted to be a physicist. But, when I was an undergraduate in Berkeley, I majored more in fraternity life than I did in academics. I missed classes, I played in bands, I did all kinds of things connected with my fraternity like pranks, dances, and parties.

Zierler:

How were your grades as an undergraduate? Did they suffer as a result?

Cohen:

My grades were not great. 

Zierler:

So, did you ever have any professors who told you needed to wisen up if you wanted to pursue a career in physics?

Cohen:

Yes, to some extent. Well, first of all, the nice thing about Berkeley, for me, was that it was big, and you were on your own in those days. I was managing my own life, and although I was not a very serious student by my standards, when I talked to some professors, they said, "You're doing fine." In general, I did fine, but I knew that my advisors were not thrilled with my performance. I was not motivated to excel grade-wise.

Zierler:

Berkeley in the mid '50s - who were some of the luminaries who were on the physics faculty at that point?

Cohen:

Luis Alvarez. I knew Luis then, and Emilio Segrè, both were in particle physics. Also, there were excellent faculty in a field that I didn't know much about at that time, which was the field that I eventually went into, condensed matter physics, called solid state physics in those years. For example, there was Charles Kittel who I didn’t know and my advisor Art Kip. Art was a solid state experimentalist, a very nice person, and he would be willing to sign things like dropping or changing courses for me, and he would give me good advice. He was always encouraging. Overall, Berkeley had a wonderful faculty at that time. 

Zierler:

At the time, what seemed to be -- I don't know if "trendy" is the right word, but what were the most exciting subfields in physics when you were an undergraduate?

Cohen:

It was particle physics. 

Zierler:

Why so?

Cohen:

First of all, there was this big laboratory, the Lawrence Berkeley Laboratory, up on the hill. They had huge machines, and they were discovering new particles, and there was a general feeling that particle physics was the most fundamental field. I remember Emilio Segrè saying things like that. I got to know Emilio very well when I was on the faculty, and we became friends. Early on he asked me to redo a calculation that Fermi and Teller did using modern methods. It was not an interesting calculation for me, but I so enjoyed talking to Emilio and hearing his stories about Fermi, who was my hero, I agreed to do it, and I did do it. In contrast, when I was an undergraduate, I was scared to death of him. When I first joined the faculty, calling him Emilio was very hard for me. Anyway, Emilio said that after he and Fermi and their group had done so much work on nuclear physics (for which Fermi received the Nobel Prize), it was obvious to them that they should move to particle physics, because it was more fundamental, and that was the fashionable field. This “Greek approach” of everything being built up from fundamental units is still a major part of the thinking of many people in physics. Nowadays, with thoughts related to emergence, and the ideas that have come from cosmology and condensed matter physics, etc., this hierarchy that was assumed at the time when I was a student is less prevalent. However, I should emphasize that when I was a student and later when I was a young faculty member, there was an implied hierarchy in physics. 

Zierler:

As an undergraduate, were you exposed equally to the experimental and the theoretical side of physics?

Cohen:

I just went to classes. I had courses with labs, but in the labs, many of us “dry-labbed”. We often didn't do the whole experiment, which we should have done. But I liked the labs, and actually liked experimental physics, but I liked theoretical physics much more. 

Zierler:

Why do you think that's so?

Cohen:

I liked the math and the fact that you could explain and predict things. I had a mathematical youth, so to speak. I think many theoretical physicists are enamored with math when they're young. The fact that a triangle has exactly 180 degrees in its three angles. The fact that geometry and other areas of mathematics were so precise and exact made that field quite beautiful in many ways. And I could use mathematics to tell where the ball was in the air just with an equation. Nowadays, I sometimes say, "A picture is worth a thousand words, but an equation is worth a thousand pictures." The fact that we can understand nature by just thinking deeply about it is a fantastic concept for me. 

Zierler:

Now, I've often heard this said from people that naturally gravitated towards theoretical physics that they had a comfort with math. That begs the question, why not just go into mathematics? Why take mathematics into physics?

Cohen:

Well, some of the greatest theoretical physicists weren't the greatest mathematicians. Not even Einstein. There were many people with greater math skills in Einstein’s time who were thinking about the same problems he was. Some were working on relativity, but he came up with the solutions. The other thing is that there's this wonderful aspect of physics that Mother Nature rules in the end. In math, you can make up scenarios that are self-consistent, and that's fine, you can stop there. But, in physics, in the end, you have to agree with Mother Nature. Feynman made a relevant off-color remark about the difference, which he told me, and he probably told other people. I don't know if this is an appropriate place to say it, but he said, "Mathematics is mental masturbation, and physics is real sex." There's just something very real about physics. I have great respect for people who work in applied and pure mathematics. I love the field, and I love the concepts, and I love to think about and do mathematics. There's a joy in just thinking about mathematical things. For example, I was trying to explain to my grandchildren about the spread of the Coronavirus just using counting and what it means for things to double -- exponential growth. I used the old example that if you start folding and refolding a piece of paper, after you've done it, if you could, 50 times, you'd get a thickness equal to the distance to the sun. This is only an example of the power of counting, and there is so much more to mathematics. The concepts that come from mathematics are wonderful. 

Zierler:

Now, did you have a senior thesis requirement at Berkeley?

Cohen:

No. 

Zierler:

What was the decision making to leave Berkeley and go on to Chicago?

Cohen:

I didn't get into Berkeley for graduate school. 

Zierler:

Really? Was that uncommon for undergraduates to be accepted, or was it your grades that did it?

Cohen:

My grades were marginal. I think that's why. Also, students were encouraged to go elsewhere for graduate study, so they could get a broader view of the field. I'm not sure they had to do that, because the Department was so big that there were many different people students could work with and get a different view. So, I went to Chicago, and that's the first time I really worked hard at academics. Speaking of work, I always had an outside job up to that point. I was sweeping floors and delivering newspapers at age 13, I worked in music stores mostly selling records, worked in an insurance company filing papers, switching trains, and washing pots and dishes in college.

Zierler:

That's when you got serious?

Cohen:

Yes, and I really did work hard, and I enjoyed it. When I got there, I got all A's, and I also had to teach myself the undergraduate physics that I had missed out on. I flourished most when I got into research. In classes, it was a question of doing what I was supposed to do, and that's what I wasn't doing well earlier, but when I got to Chicago I realized the importance of this. It finally became clear to me. 

Zierler:

What were some of the differences of the departments coming from Berkeley to Chicago? What was the Chicago way of doing physics that may have been different from Berkeley?

Cohen:

Well, Berkeley was big, and basically, I was an undergraduate, so I had a different view. I was taking classes. There was a professor, there was a class, you take notes, you try to figure out what's going on, you take the test, and that was it. But at Chicago I was a graduate student, and at the time, I really cared about discussing and thinking about physics, so I sought out people to talk to. Then, when I started research, the atmosphere was fabulous, because there was a very intimate and active relationship between the research students. I have never been shy about asking elementary questions. As a result, when I was a beginner in the group, I kept asking “dumb questions” about what everyone was doing. As a result, I matured very quickly and learned a great deal about condensed matter physics from my fellow research students. I was also fortunate in having great mentors. I've been fortunate my whole life having great mentors. 

Zierler:

Who were your major mentors at Chicago?

Cohen:

My thesis advisor was Jim Phillips. Also, Morrel Cohen looked after me. There was a young faculty member, Leo Falicov, who also helped me. I eventually lured Leo from Chicago to Berkeley. After Chicago, just to list the mentors, I went to Bell Labs, and I was basically a postdoc with Phil Anderson, who unfortunately just passed away. To a lesser extent, I interacted with Conyers Herring, who was an incredible theorist, and I talked to Quin Luttinger who was at Columbia but consulted at Bell Labs for a few days a month. When I came to Berkeley, there was Charlie Kittel who brought me here, and I learned a great deal from him. I'm busy these days writing an obituary for Charlie. So many of these people -- Charlie, and Conyers, and Quin, and Phil -- are gone, and it's a new era. Anyway, I always had really great mentors. 

Zierler:

What was the process of deciding on a dissertation topic at Chicago?

Cohen:

Now, that was strange in several ways. As we discussed, I only knew about particle physics. I thought about cosmology, because Chandrasekhar was there. I got to meet him and I was very impressed. But the dilemma for me was that I expected that if I did something worthwhile in cosmology, even my great grandchildren won't see it verified experimentally. In my view, the experimental part of the field seemed to be moving extremely slowly. I had no feeling for how cosmology would change once better instrumentation was developed. But at that point, I was led to believe that it was really a field of the very far future, beyond my lifetime, perhaps. So, I set out to work in particle physics. I took a course on quantum mechanics from Richard Dalitz, and it was marvelous. I found out later, he had used the notes from a course Fermi had given at the University of Chicago. So, Fermi had this wonderful course that Dalitz gave, and as a result, I wanted to work for Dalitz. I went to see Dalitz, and he gave me a problem in particle physics, and I went off to do it, but around the same time, I went to see my advisor, who was at the time, Morrel Cohen. The Department assigned students to advisors. These were members of the faculty separate from your research advisor, just a “regular advisor”.

I walked in to see Morrel who assumed I was coming to discuss my choosing of a research advisor. I had just taken my first solid state physics course. I did very well in the course, and Morrel taught the class. Since he thought that I was coming in to get a research advisor in solid state theory, he informed me that he was full at the time, and he sent me down the hall to see this new faculty member, Jim Phillips. He said to go talk to Phillips. So, being an obedient student, I walked down the hall, I walked into Jim’s office and said Morrel sent me. Jim assumed I wanted to be his student, took out a book, Fred Seitz's “Modern Theory of Solids”, and said, "Go home and read these chapters and come back to see me in a week." I worked a little on the Dalitz problem that week, but I spent more time reading Seitz. After a week, I went back to see Jim, and he said, "Very good. Now do this." After about three weeks, I realized Jim was running my life better than I was, and I was happy with the physics and with Jim. Jim was extremely nice to me, and the reading material and the research were really interesting. 

Zierler:

What research was Jim doing at this time, himself?

Cohen:

He was working on the electronic structure of solids -- band structures, so, he gave me a problem, which, if you say it this way, it was the temperature dependence of an electronic band gap, it doesn't sound that interesting. But, if you really get into it, it had to do with finding the diagrams for the interactions between electrons, phonons, and photons. Curiously, it wasn't so different in spirit and style from what I was thinking about for Richard Dalitz's problem. So, I liked it, I did it, and I published it in 1962--almost 60 years ago. Since Jim was going off to Bell Labs for the summer, he turned me over to one of his postdocs to work on the electronic band structure of silicon carbide. There was no air conditioning in the building where I worked, and they “chained me” to one of those "ka-chunk" calculators, and I did the band structure calculation almost completely by myself. I really got interested in the details, so I worked hard on that problem. 

Zierler:

What were the details? What was it that was compelling to you about this project?

Cohen:

Well, first of all, I was learning a lot by setting up and doing a computer intensive calculation, and I was learning new physics. I was in a group with three other students, and I was the youngest. As I mentioned before, because I really never had a fear of appearing stupid, even if I knew I was asking an elementary question, I would keep asking questions. They would make faces or wink at each other, but pretty soon, I had learned a lot, just by asking questions, and really understood their projects at a deep level, and I matured intellectually. It turns out that Phillips, at the time, was trying to figure out optical spectra of semiconductors using electronic band structure. He had a student who worked on a computer program with a professor from Italy, Franco Bassani, a wonderful guy, who was visiting the Argonne Laboratory. I was involved with the student because I was interested in the problem. One day, the group working on the problem returned from a meeting at Argonne with a calculated spectrum computed on the Argonne computers. So here was Phillips, and Bassani and others. Everybody was sitting around in this room looking depressed. They were yelling, blaming, and arguing about why the calculation didn't work. I was off in a corner, like a young kid in a room of adults, not involved in the chaos. There was a computer output sheet of paper with all the numerical results. What they were trying to do was to obtain the optical dielectric constant as a function of frequency. This should in turn yield the reflectivity of the semiconductor as a function of frequency.

While they were arguing, I took the paper with the computer data to my corner and started working with the numbers. I looked up and added the values of the spectrum output at three frequency points, took an average, added up three more points, took an average, and out came the spectrum. So, there's the analogy with that ball in the air feeling. In other words, by doing this quantum mechanical calculation of how an electron in a solid goes from one state to another, you can tell people what reflectivity they'll get back from a piece of a semiconductor at different frequencies. I showed everyone what I did, and there was a long silence and finally a sigh of relief after they realized what had happened. After I put the whole thing together, I realized that this was an important area to work in, and then I sort of took over the problem, and began working on the optical spectra of semiconductors. Even now, I still get “turned on” when seeing details in optical spectra.

Zierler:

Is this the beginning of your own self-directed research?

Cohen:

Yes, it was for a while. However, an unusual thing happened. Jim Phillips left for Cambridge University on a sabbatical. I was on my own, and I told him that I would continue on problems in this area starting with work on hexagonal semiconductors. But I didn't. I decided that I wanted to learn about superconductivity. I learned a great deal reading on my own, and I decided that if you take semiconductors dope them enough, they'll become superconducting. I went to Morrel Cohen, and he thought it was a good idea, so I started working on this problem. 

Zierler:

What compelled you to get involved with superconductivity initially?

Cohen:

First of all, another student was working on superconductivity, and so in my usual fashion, I started asking him questions. I started learning about his project, making suggestions and really getting involved helping him, and he taught me a lot. Then I fell in love with the BCS theory, and I worked through it by myself. Once I understood it well, I refined my ideas about how semiconductors could be superconducting. As I said, Morrel was encouraging. At least, he didn't kick me out of his office. There was a semiconductor experimentalist on the faculty, Helmut Fritzsche, and he said he would try to explore my ideas.

Then I heard that Jim was coming back from his sabbatical, and I was scared to death, because I did not do what he left me to do. So, Morrel Cohen, and Hellmut Fritzsche, said, "Stay home for a couple of days, and we'll talk to Jim." And they did, and Jim was wonderful. He said, "Oh, that's great. Why don't you turn it into a thesis?” I wrote the thesis in a week, because my wife Merrill (she passed away in 1994) was pregnant, and we wanted to get to Bell Labs before the baby was born. So basically, I wrote it up and left for Bell Labs, That's the short story. The longer story is that earlier, once I got the ideas out, I started asking experimentalists to work on it. Then I went on job interviews, and I remember going to Westinghouse. I met Ted Holstein, and I was really impressed. He was at the University of Pittsburgh, but he would come periodically to Westinghouse as a consultant. He asked me dozens of excellent questions about superconducting semiconductors. And John Hulm at Westinghouse said he would try an experiment. So, that was great, and I decided that I wanted to go to Westinghouse, also partly because there were a lot of hills, and it reminded me of San Francisco. After visiting Westinghouse, I went to Bell Labs to interview, and I spent most of my time with Phil Anderson. Phil had a problem already picked out that he wanted me to do. I thought the problem was very interesting.

Zierler:

What was the problem?

Cohen:

The problem ended up being done by Bill McMillan, and it had a big impact. It was the inversion of the tunneling data in superconductivity using Eliashberg Theory. It turned out to be a very important problem. It would have been a natural for me, and I was interested. 

Zierler:

Why would it have been a natural for you?

Cohen:

Because I could do the computer part. I had already done things like that. I knew about superconductivity, I understood everything Phil said. We can talk about that later, about what it's like communicating with Phil, because I ended up interacting a great deal with him. So, it was just that my abilities matched, and my interests matched. But when I went back to Chicago, I said to Jim Phillips, Morrel Cohen, and Leo Falicov, who made up the theoretical faculty, "I'm going to Westinghouse." And they said, "No, you're not. You're going to Bell." I said, "But it was so nice at Westinghouse, and Holstein is great." And they said, "Yeah, Holstein is marvelous, but you're going to Bell." And then they kept asking me, "Was Phil Anderson happy?" I said, "Yes, he asked a lot of questions during my talk." They also asked if I felt comfortable at Bell. I said yes; I felt no pressure at all. Absolutely no pressure. Everybody talked about Bell as being a pressure cooker, however even when I was there as a postdoc, I never felt it. 

Zierler:

Pressure for what? What kind of pressure?

Cohen:

Well, pressure related to people being critical and competitive. In theoretical seminars, you're standing up there, and it's like The Bronx, or what I imagine The Bronx was like -- where did you grow up?

Zierler:

My parents are from Brooklyn, so it's close enough. 

Cohen:

Okay. So, in my mind it was as I imagined -- you know, guys arguing in loud voices on the streets about such and such. I've been around aggressive people my whole life, and I don't have a problem with that. 

Zierler:

So, as far as your mentors were concerned, why was it no contest between Westinghouse and Bell?

Cohen:

Because Bell was superior as a scientific institution.

Zierler:

Did you look at it as an academic appointment that was hosted in an alternative setting, or was it an industry appointment?

Cohen:

I looked at it as a steppingstone. In other words, I was told that people went to Bell, and worked there for a few years, and then they took academic jobs. That was the image I had. But it was a two-year postdoc with the possibility of a permanent job at Bell afterward. 

Zierler:

Were you working on the same project the whole time, or were you sort of all over the place?

Cohen:

I was all over the place. What happened was, I got there, and I talked to Phil, and this postdoc was there, Bill McMillan. Bill and I hit it off, and much of my education at Bell in theoretical physics came from coffee with Phil and Bill, just sitting around discussing physics. Also, Quin Luttinger from Columbia, as I mentioned before, he would come in and visit every couple of weeks. So, I used to monopolize his time, and I learned a lot from him. What happened was, I told Phil that I really wanted to do more work on superconducting semiconductors, because, in fact, experimentalists had proved that it was possible. They got positive results, and I had new ideas. To be honest, although I knew that the problem that Phil had picked for me was important, I didn't think it was that difficult, and I didn't realize at the time that it was extremely important. I once asked Brian Pippard, "What did you think when your student, Josephson, came in and told you about the Josephson Effect?" Pippard said, "You know, when something great happens, at the time there's nobody next to you telling you it's great." So, Phil gave the problem to Bill McMillan, who just did an incredibly good job with it. I went on to do some work in superconductivity in semiconductors and electronic structure of semiconductors, and came up with the idea that an oxide could be a superconductor.

This is a fun story, if you have time for it. I decide strontium titanate could be an unusual and interesting superconductor. All you have to do is dope it by having oxygen vacancies, and look for superconductivity at low temperatures. So, I mentioned it to Phil, and he was sort of encouraging. Bernd Matthias was around at the time. I believe he had a joint appointment between Bell and UC San Diego. Bernd was the superconductivity materials and experimental expert. He's kept saying, "You're crazy. You've got your Green's functions, but I know superconducting materials!" So, I figured there was no way I was going to get any experiments done at Bell. But I heard that there was a group at the National Bureau of Standards that made strontium titanate, and I knew that they did low temperature research at NBS. Because of my early studies of particle physics, I knew that Ernest Ambler was there, and he was a great low temperature physicist. So, I went to the NBS, and I physically took a sample from the semiconductor group that made strontium titanate, gave it to Ambler, and Ambler’s group measured it, and it went superconducting. So, that was the first superconducting oxide. I came back to Bell, and Matthias was really mad. First of all, let me say that Bernd and I ended up being good friends, but at the time, he was vocally opposed to what I was doing. At the beginning he was not enamored with superconducting semiconductors, because it didn't fit into his view of superconductivity.

Zierler:

Why not? Can you explain that? What was the disparity there?

Cohen:

Bernd focused on trends in materials and used properties of atoms and their positions in the periodic table to explain the trends. He liked transition metals, and he had these plots of superconductivity as a function of valence which worked for him. He was an expert in materials, and the idea of just doping a semiconductor to get a superconductor didn’t appeal to him. But as it turned out, he was a very good friend of John Hulm at Westinghouse. John was another reason I was interested in going to Westinghouse. Hulm got together with Robert Hein at the Naval Research Lab, who did the low temperature work, and they found the first superconducting semiconductor, germanium telluride. So, for Matthias, since Hume was his friend, and he trusted him, it was okay but only after bringing up many very clever objections about the experiment which we were able to argue away. But Bernd insisted that strontium titanate can't be superconducting. Finally, I asked, "Why not?" He said, “Because it’s blue, and there are no blue superconductors!” It turns out that strontium titanate was so lightly doped that you could see through it and it did have a bluish tinge. Also, Bernd and probably everyone else at that time didn’t think an oxide could be superconducting. Anyway, it turns out that as a result of the skepticism, Bell Labs didn't want to let the paper be sent out for publication from the Labs. Although I wrote the paper with the National Bureau of Standards people, I was listed on the paper as being at Bell Labs. Since Bell Labs had an internal reviewing system, which was fabulous, and helped to keep the quality of their publications at the highest levels, I had to submit the paper for their review, and they said that they didn't want to let it out because Matthias said it had to be wrong.

So, I went to see Al Clogston, who was the boss of my boss, or higher. Al was a wonderful guy. I complained, and Al said, "Look, you're a young guy. Matthias has incredible intuition. Why don't you listen to him, and let's hold back on this paper?" I said, "It's right! The experiment is absolutely right, and I want to submit it." Clogston's was sitting there clearly uncomfortable, but very nice. Then I said, "If you don't let me send this paper out, I'm going to quit." Then Clogston relaxed and started laughing. He said, "You're a postdoc!" In other words, “This is the great Bell Labs! Who cares if you quit?” But he let the paper out, and it was published in Physical Review Letters. It had some trouble getting accepted, because I think Matthias was one of the referees, but it was the first superconducting oxide and the lowest carrier density superconductor. I continued with more work in that area, and I also started working on optical spectra again. And then, I got invitations to give talks at La Jolla, Stanford, and Berkeley. I'd only been at Bell less than six months, or something like that, and I had no idea that these were recruiting talks. I went first to La Jolla, and it was really nice. Then I went to Stanford, and I met Felix Bloch. Felix said, "I'm really looking forward to your talk. It sounds great." Then we started talking about what if I come to Stanford.

Zierler:

What did you think you were doing going out and giving these talks, if not for jobs?

Cohen:

I thought I was giving research talks. People go to meetings, universities, and other labs to report on new research, and although I was young, Bell agreed to send me. The invitations indicated that they wanted to hear about superconducting semiconductors. So, Felix continues, and he says, "What you have done is marvelous. Of course, it's not based on BCS theory." I said, "No, it's a wonderful demonstration of the power of BCS theory." Felix being Felix -- as I said we became good friends later -- he looks at his watch and says, "Oh, I'm not going to be able to attend your seminar. I have a meeting that I forgot about." The point is, he just felt that BCS theory hadn't been proven. He was a big skeptic. At the time, I had no idea about what was going on. After I talked at Stanford, I came to Berkeley, to give the Department colloquium. Beforehand, I was warned, “Watch out, you'll get a hard question from Steve Weinberg." Steve did ask a polite and easy to answer question. So, it all worked out, and Charlie Kittel started talking to me about coming to Berkeley. And that's when I really realized what was going on, especially when Charlie started talking about parking problems I would have on campus. When I went back to Bell, I went directly to see Phil, and I said, "Phil, I'm getting an offer from Berkeley." Phil knew about everything that was going on. He had written a letter for me, and it turns out that Morrel had written a letter for me. They asked Morrel about my undergraduate record, and about what's going on in my case. The way Morrel tells it, he said, "Well, draw a line from where he was then to where he is now, and you'll get the trajectory," which was very nice of him to do. 

Zierler:

So, that served you well. 

Cohen:

Yes, and Phil evidently wrote a very good letter. John Hopfield was at Berkeley, and he was leaving Berkeley for Princeton. I was getting his job, and he told me that he had put in a good word for me too. Anyway, I thought Phil was going to be furious, but instead he sat down and said, "Who ever heard of somebody leaving Bell after one year, when you have a two-year postdoc?" I said, "I'm really sorry. I want to go back." He went on for a couple of hours trying to convince me not to go and gave me a bunch of reasons that I didn't fully understand at that time. He said the image of an industrial scientist is different than a professor, and is that why I want to leave? I said, "No, my family's out there." And, he went on and on, and finally, when I said, "Phil, you are using logic, but logic won't work here. I want to do it because my family's there, and I’ve always wanted to be at Berkeley." So, I kept working for a few more months at Bell, and then I went to Berkeley. I never regretted it, but Bell was wonderful, particularly for someone like me because I would walk down the hall and there would be all these experimentalists around. If you wanted an experiment done in the infrared, you'd talk to Paul Richards. If you wanted something in the UV, you'd talk to somebody else. I just absorbed the atmosphere.

Let me describe my first day when I came to Bell -- and then I'll stop on Bell -- I was moving into my office, and these guys were putting metal furniture in, so I said, "Oh, please, don't give me a rug." I had been in Phil Anderson's office, and when I walked there, I'd build up static electricity, and then I’d touch the metal file cabinet and get a spark and a shock. The movers started laughing their head off thinking that this new guy thinks he's getting a rug. A rug was more than one level up. Another level up was wooden furniture. On the first day, the theorists talked to me about what it was going to be like to be in the theory group at Bell. They said, "Well, we play chess before lunch, and we play Go after lunch." "We never come in when it really snows, but the most important thing is to be at 4 o'clock tea." And it was true. I was completely on my own. They didn't tell me to do anything. I could do anything I wanted, but at 4 o'clock, when everybody got together, there were people, who would travel to Europe and all over the world, telling about the latest research that was going on. After all, we didn’t have the internet then. And we would discuss physics. It was a very exciting, nurturing environment, and I thought it was just great.

Zierler:

Now, I want to ask you. When you said the moment when -- you didn't realize the importance of what you had hit on from what Phil Anderson had given you. Did you understand that Phil appreciated the importance of the experiment?

Cohen:

Yes. People kept warning me that Phil was hard to understand. However, because I spent time with him, talking over coffee, and other times, when he talked, I could understand him. For many people, he was opaque. Not because he wanted to be, but he could be opaque. I would sit there, and he'd say something, and someone would ask me, "What the hell is he talking about?" If you'd say, "F=ma," Phil might say, "I like to think about it in a different way." Then he'd tell you about his different way, and I got used to it and usually learned something. It was like when you're a kid and you can speak a couple of different languages. So, Phil always knew what was going on. He wasn't always articulate about it, but he understood what was happening, and he understood what I was doing. He understood the importance of it. Now, it's all done and accepted, but at the time my work was speculative. 

Zierler:

Did you ever resolve your disagreement with Felix Bloch when you met with him at Stanford?

Cohen:

Yes, we got to be friends in later years, and I would invite him to Berkeley to give talks. BCS ended up being so successful, but I don't know whether Felix ever came around about BCS. He had worried about the gauge invariance of the BCS theory. Later people worked these things out. Felix was very formal. He reminded me a lot of Gregor Wentzel, who I knew and really liked, so I was comfortable with Felix despite the uncomfortable time when we first met. Wentzel was on the faculty at Chicago when I was a graduate student. He was a co-inventor of the famous WKB method in quantum mechanics. I thought about working with him on a thesis project, and he was interested in me because he had given what he called a “major problem” in a class I took, and essentially, I was the only one who solved it. I didn't know this at the time, but the problem he gave us was a research problem that Fermi had solved, and it was in the literature. There was another student, who handed in the correct answer, but Wentzel knew that that guy got it from the literature, and he knew I didn't. Wentzel was like Felix, very formal and an excellent teacher. If I had worked for Wentzel, I would have worked on helium, and I wasn’t interested in helium at that time. But I think that solving that problem and getting Wentzel’s approval may have been my first “big recognition” in physics in my view. A funny story related to Felix was -- I don't know how long you want me to go on with stories -- I invited him to Berkeley a few times. One time, when I invited him, he said to me, "I’ll come to visit with you, but I'm not going to give another talk. I just gave a talk two years ago." I said, "Felix, they just want to see you. Please give the talk."

Zierler:

Because he's a super star? Why do they want to see him?

Cohen:

He was a super star. The Bloch theorem, Bloch wavefunctions, his Nobel Prize in NMR ---yes, he was a super star. Although Felix was one of the fathers of the field of solid state physics, many Stanford faculty felt he was a hindrance for the development of the field at Stanford in his early days there. They referred to a luncheon club of Felix Bloch, Leonard Schiff, and others who frowned on hiring faculty doing solid state physics in the Physics Department. For a long time, solid state people were hired into the Applied Physics Department. It wasn't until later that Felix et al loosened up. So, back to my story. Felix agrees to talk, and when he opens his talk, he says, "Marvin has invited me here again, so I apologize, that I will be saying some things that I said before. But my friend Heinrich Casimir told me, "Never underestimate the value of hearing something twice.” Everyone laughed, and I thought it was a great line. So, when high temperature superconductivity in the oxides hit, I went to Europe to a big conference to give a talk reviewing the status of the field, but I had just given the same talk someplace else. I didn't have time to change it very much, and I saw some of the same people in the audience who were at my earlier talk. So, I got up and said, "I'm giving you a review of what's going on in high temperature superconductivity. I apologize if people just heard me talk, but Felix Bloch told me that Heinrich Casimir told him, "Never underestimate the value of hearing something twice.”.” At the end of my talk, this guy walks up to me and he says, "I'm Casimir. I never said that." 

Zierler:

That's great.

Cohen:

And then Casimir starts questioning me about superconductivity, and he was a brilliant guy. It turns out that you could see where he had left off in this field years ago. They made him head of the Phillips Lab to protect it from the Germans. He went into other fields, but his superconductivity knowledge up to that point was probably right at the cutting edge. So, he asked me questions that they had at that time, many really penetrating questions. It was an incredible exchange. Once I got back to Berkeley, I read everything I could read about Casimir because he was so impressive. So, back to Felix, the bottom line is that Felix and I became good friends, and when we discussed physics, I never mentioned BCS. 

Zierler:

Now, when you got to Berkeley, did you ever have a concern that some of the faculty members who knew you as an undergraduate might have trouble accepting you as one of their own? Did that ever cross your mind?

Cohen:

It was interesting. First of all, there's the emotional thing, seeing the people there I'd taken courses from. As I told you, I became a very good friend of Emilio Segrè's, but, of course, I never told him that I dropped his class on the last day of the course. But Eyvind Wichmann, who knew about my past as an unexceptional undergrad, came up to me and said, "Were we terribly wrong denying you admission to grad school?" I answered, "No, you were absolutely right. What you thought at the time was correct." He said, "Okay." and that was it. So, my slacking off as an undergraduate didn't play a role in anything that went on after I joined the faculty, except that I probably had a little more compassion for students who were late developers. 

Zierler:

And then the following year you became a senior scientist at LBL. Was that the plan the whole time? Was that part of the attraction, so that you could work at Berkeley Laboratory?

Cohen:

No. Basically, it was a question of trying to get LBL to support condensed matter physics. 

Zierler:

That was your interest? You wanted to get LBL to support condensed matter physics?

Cohen:

Yes. There was a group of experimentalists in the Department who needed support, space, and equipment. As a theorist, I didn't need that much money. I just needed computer time. But for the experimentalists, by getting a Department of Energy program going, I could get LBL to help out, and that's what happened, and it turned into a big deal. It's a major avenue of support from the Department of Energy for condensed matter physics in our Department now, and the funding comes through LBL. I had an office at LBL, but I rarely used it. LBL, or more correctly LBNL, or the “Berkeley Lab” is a wonderful institution with extraordinary scientists in a wide variety of fields.

Zierler:

Why did LBL not have an interest in condensed matter physics before you got involved? Was it a bureaucratic thing? Was it a funding thing? Technological limitations?

Cohen:

No, historical. It was tradition. They worked on particle physics, big machines, big physics was the Lawrence style. Then, they start expanding their efforts in different fields like chemistry, biology, energy, and now they're in everything. 

Zierler:

What was your pitch to LBL to embrace condensed matter physics?

Cohen:

I didn't need the pitch. They went after me. I think Luis Alvarez liked me. As soon as I got here, he had me give a talk at a locally famous seminar series that he held at his house. People said that I should anticipate that it would be a grueling experience, but it was very nice. So, he may have put in a word for me. I don't know what went on behind the scenes. I was very happy about how the LBL connection worked out, but I preferred to house my group on campus, and that's where I basically was during my entire Berkeley career. At the time, there were particle theorists here that I could talk to without having to go to the hill to see them. Steve Weinberg and Shelly Glashow were here, and I'd have lunch with them and other particle theorists. I really enjoyed that. Regarding the level of particle physics familiar to me, relative to the what was happening in the field, I knew much more then than I do now, because the field has changed, and I haven't kept up on the level that I did in those days. 

Zierler:

Now, when you received tenure only two years later, that's a relatively quick process, right? Two years?

Cohen:

Yes.

Zierler:

How did you interpret that? What was the message behind honoring you with tenure that quickly in your career?

Cohen:

Well, I was publishing a lot which helped. The fastest way to tenure was, and still is, getting outside offers, but I never used outside job offers to boost my level here. I was getting outside offers, but I refused to use them. Kittel was disappointed that I didn't, because outside offers make the tenure promotion a lot easier. I remember that at that time I'd gotten an unofficial offer from USC to double my salary, and to help me to finance an expensive house in Pacific Palisades --I would have been rich today. Charlie said, "You have to take that offer to the dean." I said, "I can't. I just won't do it." I usually followed Charlie’s suggestions. This was a rare refusal.

Zierler:

You mean, you refused to leverage the offer?

Cohen:

Refused. I never did that, through my whole career. I don't know whether the faculty and administration knew about my outside offers. I suspect they had some information because when Chancellor Tien called me and told me that he wanted me to be a University Professor, he also said that he did not want me to take offers I had just received. I said that I wasn’t leaving anyway. When people would call to recruit me, I always said something like "I plan to stay here; I plan to die in Berkeley." I was where I wanted to be. I also did not put any pressure on anyone to help me be promoted, given tenure quickly, or anything like that. However, I guess that there were people taking care of me, at least it seemed that way. 

Zierler:

Cleary, you had such a tremendous amount of loyalty to Berkeley. I wonder how much of that was emotional, that you just loved the place, and how much of it was that it was simply the best place for you to do your work?

Cohen:

I think it was both, but I would say it's more the former. I didn't think so much about my working conditions or whether I was treated well. I think I can work almost anywhere, but I’ve liked working here. A major point is that the graduate students here have always been outstanding. In the end, I knew that there were many good institutions around, but I just always liked Berkeley and felt fortunate to be here. My late wife Merrill liked being here too since she was born in San Francisco. We raised our two children in Berkeley, Mark, born when I was at Bell Labs in 1963, and Susan, born when we were in Berkeley in 1965.

Zierler:

You spent so much of your career doing visiting professorships and fellowships really all over the world. Broadly speaking, what's the value of taking your work to other institutions? What do you get out of it, and what do those host institutions get out of it?

Cohen:

Well, I always wanted the latest data and opportunities to talk to experimentalists. So, if there were things going on elsewhere of interest to me, I would get involved. For example, I spent some time in France, and the theorists in France were very formal. They didn't use computers much, and they weren’t as coupled to experimentalists in my field as I was. So, I could go there, and I would get these beautiful experimental curves that I could interpret, and I had it all to myself because of the lack of local theoretical interest. So, I often traveled to get data and to interact with other experimentalists and theorists. Because of this, I made lifelong friendships with people from other countries. And then there were special times. For example, in May of '68, when the world went crazy, and there were planes flying over Berkeley dropping tear gas on everything, my friend in Paris called and said, "Why are you staying in that crazy place? Come to Paris for the summer. I have all these experimental results you could work on." So, I gathered up my wife and two kids, and we went to Paris, and believe me, the tear gas in Paris was much worse. I couldn't get anything done there, so I came back. That was the only time I was trying to get away from Berkeley because I was having difficulty working here, but soon there was chaos everywhere. 

Zierler:

How did you deal with all of the unrest at Berkeley, in terms of your role as a professor and a mentor to undergraduates? How did you respond to all of the unrest that was going on?

Cohen:

I was much more connected during that period to graduate students. I behaved very much like the rest of the members of the Physics Department. We wanted to keep the educational program going, and at the same time, we wanted to be supportive of the students. So, I became an advisor, and I thought I could try to help by talking to students. That was my major connection with undergraduates during that period. I quickly developed an appreciation for what psychiatrists must go through. The students came in, and they're depressed, and some were on drugs. I remember a student saying “My wife ran off with a ski instructor. What should I do?” All these kinds of things came up, and I wasn’t trained in how to help them. I was really sympathetic, so it was very hard to not get too involved. But then, after, I don't know, two weeks, or a month, I became sort of, "Okay, next?" I hardened, because you can't “let that stuff in” all the time if you're going to help. Eventually I decided that I wasn't very good at psychological advising, but I would still have my door open to students.

I would still get involved. I’d go down to Sproul Plaza where the action was. I would talk to reporters to try to see if I could get them to paint us in a better light. I remember going to Sproul Plaza, and these reporters were running around having a great time showing the chaos of Berkeley, and I said, "What are you doing here? Why are you giving us such a bad image?" And they said, "We get our stories here." I said, "Why don't you go down to Stanford?" One reporter said, "It's a long drive from San Francisco down the peninsula, and who knows if I'll get anything. I can come here and have a complete story with little effort by 1PM." I was very upset at the way we were pictured. Therefore, I got involved in that aspect, gave interviews, and tried to talk to students. But through it all, I was still focused on my work.

To give you a personal example, I attended the 40th anniversary of Mario Savio’s speech on a police car. Let me remind you that when I was an undergraduate at Berkeley, it was all about the dance Saturday night. We had to get the decorations up. We didn’t lock our doors. It was the '50s. Then I came back in '64, and the place was totally different. All of a sudden, there's chaos, and Mario Savio. I remember standing there and watching him on the police car. So now I am out there watching them reenacted it. I looked around and realized none of those kids in Sproul Plaza now were born at that time. I thought about whether everything had become different than in the past. And then I realized that even if it had, I wasn't different, because I was thinking the same thing at that moment as I was thinking 40 years before. That is, that I’ve really got to get back to my office, as I've got a hell of a lot of stuff to do. As you know, it was a crazy time in the 1960s. We kept our venetian blinds down, so if a bomb were thrown, the glass wouldn't come into our offices. Universities like Berkeley have great inertia, and they keep going. And fortunately, I had inertia too, so I kept doing physics. That didn’t change, so life went on for me without getting too sidetracked like some of my colleagues and many students did.

Zierler:

I wonder if you can talk -- you said when you got to Berkeley, and you got tenure so quickly, that you were "publishing like crazy." Right? Now, obviously, part of that is that you're young, and you're full of energy, but also, you feel a need to publish because there's a lot to write about. So, what explains from a discovery perspective, all the things that you were doing that really needed to be communicated to your colleagues? What were the things that you were involved in that required that intensity in writing?

Cohen:

First of all, there were a couple of psychological things. At Bell Labs, there was the feeling that our mission was to discover the secrets of nature before anybody else. There was also a feeling that you're in competition with the world, so you should publish your results. Some of this must have rubbed off on me. Also, the first paper that I worked on as a grad student at Chicago, where they chained me to a computer in a hot room, was turned down by a competitor who reviewed it for the Physical Review, and it never got published. And it was a correct paper. That feeling of having worked hard and then not getting it out there was not a good feeling. Also, I observed colleagues and fellow students who worked on things and then put them aside for later and never got around to publishing them. So, I developed a work approach where I worked as hard as I could on the research I was doing at the time, wrote it up, and got it out. This is not easy because when you finish a project, you are most interested in the next project not the one you just did. The danger is that if you wait, you will forget details.

I’ve published over 850 papers. My rate of publication since arriving here didn't change much over the years, so the publication rate I had when I was young was not because of my youth or motivated by a desire for early tenure. Now, it's slowing down somewhat because I used to publish a lot of my papers with graduate students based on problems I suggested and supervised, and now I work with only two postdocs and in collaboration with experimentalists. But, I think, probably from the beginning to, let's say, ten years ago, or maybe even now, my rate of publication is about the same. Also, I felt at many stages that I was changing the field of solid state physics from working on idealized models to working on real materials. So, I wanted to get the word out that this was the beginning of a new era. A revolution was taking place. And it turned out to be true. Before we revolutionized the calculations of electronic structure for real materials, people focused on simple or highly idealized models. One of the first thing researchers do today when they find a new material, a new property, or a different twist on a physical effect, they calculate the band structure of the material. For example, when high temperature superconductivity was first found in the copper oxide systems, theorists rushed to calculate the band structures of the materials. This is a gigantic help.

In the early days of my research, band structures were the purview of a small group of theorists discussing them behind closed doors and using complicated Greek letters to keep others out. The early attempts at comparisons with real material properties usually failed. I gained a great deal of support when we found that in all cases where the theoretical results based on the pseudopotential method we develop disagreed with the experimental data or previous interpretations of the data, our results were shown to be correct. The differences arose from errors in the measurements or misinterpretation of data. The focus on computation of electronic structure did change how we study solids, and it is now a foundation pillar of condensed matter physics and materials science, and I’m very proud of the pioneering work I did.

Zierler:

Is your work style such that you're able to juggle multiple papers at the same time, or do you like to work on one paper, complete it, and then move onto the next?

Cohen:

I work on several papers at the same time. More precisely, a lot of papers in “the same time period”. I focus on one research project at a given time. My approach, in physics is similar to the way I function in life, -- one thing at a time. My wife Suzy is a multitasker. She is in the art world and does many things at the same time. When something goes wrong which is common when you depend on many different people, she says something nasty, and goes on. She expects not to be able to control everything. But, my work is more focused on what I do, and I don't tolerate my own mistakes well. So, I really function in a linear, serial way. But because I work with other people, I work on several projects in the same time period. When I'm working on something, I’m fairly laser-like, then I move to the next project,

Zierler:

Looking back on all of those publications, what are the common threads that run through all of them?

Cohen:

I think being physical. In other words, one of the things that I did very early on was to refine the pseudopotential idea and to develop the empirical pseudopotential approach. When I look back at many great things that happened in physics, it seems like they started out with more empirical or heuristic approaches. For example, Einstein's theory of photoemission, the work that he received the Nobel Prize for, he called it a heuristic theory. He didn't have all the formalism yet. It wasn't until later that the theory was refined. And the Bohr model, which we all know is heuristic, explained the fundamentals of the hydrogen spectrum, and it was a tremendous advance. In many ways, it led to, the more complete quantum theory developed by Schrodinger, Dirac, Heisenberg, and others. My work with empirical pseudopotentials led to ab initio pseudopotentials, where all you did was put in the atomic number and the atomic mass, and you could predict the properties of the material given a certain structure. So, I think I would say that I like starting with simple or heuristic models or theories. I don't jump in and start formally deriving and calculating. I start with empirical ideas, and try to get the most out of them, and later I, or other people add the more formal approach.

A good example is that people often refer to density functional theory when they describe the advances in electronic structure. The truth is that we had solved the problem of the optical properties of semiconductors and their relationship to electronic structure with the empirical pseudopotentials before density functional theory came on the scene. The main physics, like the Bohr hydrogen atom, was done. This work went beyond semiconductors by providing a standard model that is used throughout the field. I've had students who were very mathematical, and they would come to me, after I’d asked them to work on something, with a thick bunch of papers with many equations and diagrams. If it was wrong, I was usually able to ask a few questions and could sort of glance at the work and tell where it's wrong using physical arguments. Also, in my interactions with people like Pauling, or Feynman, or Gell-Mann, I felt they had similar kinds of feelings or approaches. I tend to resonated with people like that.

In other words, not being totally formal in my approach to physics has been a fun way to do physics even though I believe I’ve had more mathematical training than most theorists in my area. I remember when I was working on a collaborative project in my field with Murray Gell-Mann. Somebody had set up a consulting arrangement for us so we could meet and work on certain aspects related to making diamonds and other materials. Murray was an extremely smart guy, and he could do the mathematical part easily and fast, but he really appreciated the empirical pseudopotentials. In our discussions about why we were drawn to physics, he decided that I wanted to figure out how things work, and he wanted to find patterns. I agree with that observation. Anyway, since the problem was in my area, it was interesting to see how Murray approached it. He appreciated the empirical approaches, because the mathematics wasn't what impressed him as I indicated above. It was getting those physical ideas. That's a long-winded answer to your question. I might add that Linus Pauling had “chemical intuition”. He lavished praise on me when I did my semiconductor work because the results were so close to what he expected. Other chemists were not so generous. However, at some point when Linus and I met, I told him that the methods that I used for semiconductors and insulators that he liked so much also worked for metals He was upset and said that wasn’t possible. This was at a time when I told him that I thought that the work by Dan Schechtman on quasicrystals was correct. Linus said it was wrong because “all the good crystallographers” are dead. Dan later received the Nobel Prize for that work. Linus was also upset that I wouldn’t take large doses of vitamin C. People tried to team me up with Linus. Although I had great respect for him, I felt that I couldn’t collaborate with him because his intuition, when it worked, it only worked for him.

Zierler:

Well, I want to flip it and now ask -- that's the question about the through lines -- you know, what connects all of your work. I wonder if we could flip that, and I want to ask, do you see your work as taking different forms over different parts of your career, that there were certain things that you were interested that were, in many respects, very different than your previous work? Did you jump around, or did you always stick to one area, and all of your work diverged from that one area?

Cohen:

No, I jumped around, and it's because I worked with experimentalists, and I was always in search of good thesis problems for my students. So, for example, when I worked with people doing optical properties, I’d focus on optical properties and photoemission. Same for superconductivity, and the same for nanoscience after the buckyball and nanotube were discovered. Researchers would call in and say, "Would your methods be appropriate for this or that?" So, I'd constantly switched between different areas within condensed matter physics

Zierler:

So, can you explain the process where a new technology comes online, and you get approached to collaborate, what are you offering in this relationship? How does that collaboration work?

Cohen:

Okay, well, first of all, it often starts with experimentalists who have data that they can't understand. If it “smells” interesting, then I take it on as a problem and try to explain it. Once I'm successful, then it often allows me to predict new things. So, as a theorist, my job is to explain the experimental data, and then to use these concepts to say, "But now, if you did this, and turned your experiment around, and looked for this..." For example, when the carbon nanotube came out, we started doing calculations covering many areas. This was after we showed that our methods would work for these systems. This was important. As an example of taking things beyond where they were and making predictions, I remember that I was on an airplane when it struck me that if you can make a carbon nanotube, you can probably make one out of boron nitride. So, I did some quick calculations to model the possible stability of a boron nitride nanotube and refined what I had in mind.

When I came back to Berkeley and met with my student group, which I always ran in a very democratic way, I said I have this crazy idea, blah, blah, blah. No one wanted to work on it. They said, "We have enough of our own stuff." They didn't really think it was that good of an idea. But I had one student, Jennifer Corkill, with spare time who said ok since she was waiting for her boyfriend to finish his PhD so they could get married and leave. And a postdoc, Angel Rubio, got involved too. They worked on it and showed that my calculations were right after checking them on the computer. Then we went to see Alex Zettl], an experimentalist on our faculty, and he made the boron nitride tubes. Nowadays we and others around the world use them a great deal. They have interesting properties, and we’ve learned a great deal of physics by working with them. That's a model of how we work. Regarding going further, at the same time that Alex made the boron nitride tubes, he found that he also was able to make things called boron nitride cocoons, which I hadn't thought of. So, the process is, while trying to explain data on carbon nanotubes, I got the idea that you could make boron nitride nanotubes, computer calculations verify that the idea can work and predict important details about what the tubes will be like, and then they are realized experimentally. We also successfully predicted the existence of other nanotubes made from boron, nitrogen, and carbon at that time. A next step that often occurs where something new shows up experimentally or theoretically while you are working on a problem. For example, if the experimentalists find something else, like the cocoons, we then go back and try to explain and exploit the new discovery. So, I and my group have always been very interactive with experimentalists. Often, after the physics study is over, the experimentalists go on to invent practical applications.

Zierler:

Do you see the relationship as essentially 50/50 partnership between experimentalists and theorists, that they equally need each other, or is that an imbalanced relationship?

Cohen:

It depends on the problem. Sometimes it ends up 50/50. If you can get experimental students and theoretical students to meet together, as we have over the years, both groups can learn a great deal from each other. Sometimes when the groups first gathered, the experimental students would ask us to do things that were absolutely impossible. In the beginning, they don't know what's hard or what's easy to do theoretically, and in turn, we don't know what's easy or hard in terms of making measurements. But once they're together for a while, an evolution takes place, and you hear statements like,” If you did this calculation, that would really be helpful for us. And often we'd say, "Hey, that's trivial for us to calculate." So, the idea of having groups work closely together is important. After that, the different contributions of each depend on the particular project. Sometimes, I had difficulty getting people to do experiments to test my predictions. For example, I wanted to have a measurement of silicon at high pressures to show that it went into new structures that could be superconducting without doping. I finally got it done by a friend in France, and then it was done in more detail here. And the theoretical predictions were correct.

New phases of silicon were predicted along with their properties including superconductivity. In terms of going around and collaborating, you get new ideas. So, collaborating is usually good. You asked about technology. There's the practical aspect of our work in this field. Nowadays, we can talk via Zoom because of applications of silicon. When I did my first electronic structure calculations for silicon, nobody knew its detailed electronic structure. Charlie Kittel kept saying, "You should be patenting these things." I didn't patent any of that work as far as I remember. I probably should have, like for example, using band structure engineering to make devices. All of the devices they make in Silicon Valley rely on a knowledge of electronic structure, and we did the fundamental work first on the computer. Even in the 1960s, people didn't understand the electronic structure of silicon in detail until we did it with the empirical pseudopotential, and we did many more materials, some of great practical importance, and we showed people how to use these methods.

Zierler:

Now, I'm curious if you dwell on such things, and if you do, that'll help me understand the larger answer. As a theoretical physicist, do you dwell much about the impact of your work on society? If you do, is the only way to measure that through your collaborations with experimentalists, or do you see your work as having a direct value to society, with or without the experimentalists?

Cohen:

I guess, both. For example, let's talk about the use of formalism in physics. We came up with theoretical ideas that are being used all around the world now for calculating properties of materials, and for doing calculations. Those ideas resulted in theories upon which many calculations related to materials are made possible. This area is probably the largest branch of condensed matter theory, and condensed matter physics is the largest branch of physics. So, in a sense, a large number of people are using things that we developed, and they are extending the methods, concepts, and applications. Years ago, we were usually about five years ahead of everybody. Now, everybody is roughly in the same time box. So, the computer programs, formalism, things like that, were given as tools to the theorists. Also, we came up with new concepts and could reveal the existence of new physical principles. For experimentalists, it was very important to help them interpret what they did. Like others, my fundamental interest was understanding the properties of nature and how things work. In some cases, the results that we got were a great benefit for society, but that wasn’t the initial goal. The fact that we have small and large powerful computers is because we understand the electronic structure of silicon.

I’m not saying that there aren't empirical approaches with very little theory that work well -- there's the Edisonian approach, and there's the Einstein approach, and we're someplace in between. I remember, for example, Charlie Townes had his office down the hall from mine, and I told him that when I was a postdoc at Bell, the only application I witnessed of the laser was to entertain our visitors. We would ask them to put on a white shirt and a black tie, and they'd walk across the laser beam, and the beam would cut the black tie in half, it would fall to the ground, and the white shirt would be fine. This was entertaining, and that was the application. I said, "Charlie, did you ever think that they'd be reattaching retinas, making computers, fiber optics, compact discs, and so forth?" He said, "No. No idea." Charlie and I had many conversations of this kind. So, did I have an idea that my research would have an impact on electronics? No. I didn't know that our band structures and methods of band structure engineering would be used in Silicon Valley, and that the methods would be used to make materials which benefitted society. So, regarding contributing to society, and to science and engineering, I'm happy about all of those things. But the thing that drives me is solving the problem I'm working on. If somebody comes up with some new effect, let's say Paul Chu at Houston, who works in superconductivity, calls and says, "Hey, I think there's a new superconductor with special properties. Can you tell me what you think?” I then rush to do the relevant calculations. That's a really exciting time for me.

Zierler:

On the question of collaborations, probably over the course of your career, I don't know how many thousands of people have approached you to collaborate on a project -- there's a calculus there where you have to decide what are the projects that you're going to take on, and what are the projects that you're not going to take on. So, I wonder if you could answer, sort of broadly, what are those projects that are easy to say yes to, and what are the projects that are easy to say no to, and what are some of the shared characteristics of each category?

Cohen:

It's really easy for me. It depends a great deal on who the person is, and if I trust him or her. I really worry about getting data that I don't trust. So, there are certain experimentalists that I really trust, and if they say that they found something interesting, I usually say yes. Then, the next thing I watch out for is importance, there may be something that's really complicated to calculate, but it's not that important. Einstein said, "Not everything that's countable should be counted." It's just what counts that we should work on. So, it's the physics. If somebody comes to you with some antigravity thing, you know it's wrong, and you don’t get involved. If I don't trust a potential collaborator, I don't get involved, and I don’t get involved for the money. I’ve done some consulting for industry, but I only agreed to do it in cases where the science was interesting or I enjoyed working with the people involved.

Zierler:

Is there really an issue where there's a lack of integrity in the field that really determines, or at least raises a legitimate concern that you're working with faulty data? Is that really an issue?

Cohen:

It's rarely a question of integrity, and more a question of ability. But there have been cases regarding integrity. At Bell Labs, despite all the care they took to screen publications from Bell that I discussed earlier, they had a terrible scandal involving fabrication of data by Jan Hendrick Schön. When he published his false data, I started working on it, because it was coming from Bell Labs and it was so interesting. I didn’t suspect anything wrong at first. In fact, I was starting to be able to explain this “made up” data, and that's pretty scary, but I began to see that things didn’t really add up, and then the truth came out. So, physics is not like biology in the sense that experiments in condensed matter physics can be reproduced comparatively easily, and it doesn’t depend on the samples as much as in the life sciences. The other thing is, if I make a prediction, and I go out on a limb with the prediction, experimentalists come with their saws to cut that limb to show that it's wrong. Usually within a year I'll know if I'm right. It's a very fast-moving field. So, when I'm questioning things, sometimes I worry about somebody making things up, but 99% of the time, I just worry about whether they are really measuring the right thing. Also, there are groups of experimentalists who I trust implicitly.

Zierler:

So, when do you have an early hunch that the data is faulty? How do you smell that out? What's that process like?

Cohen:

Well, I do some calculations; pencil and paper. I often conclude this can't really be right, and if it's right, it's really special. So, it's that kind of test. There are theoretical developments to be tested too. For example, if you have a technique, and there are many theorists who function like this, you want to apply it to many cases. I remember when studies of many-body effects techniques first became popular -- Green's functions, the renormalization group approach, and so forth -- people tried to apply them to so many different problems, and sometimes it was appropriate. One needs to test new techniques, but you basically want to work on interesting problems. I can tell you a funny story about applying a technique. Dick Feynman called me once, and he said, "How are you doing?" I said, "Fine." I was trying to figure out why he was calling. He said, "I went on a vacation." So, I said, "How was your vacation?" Dick was always on stage. So, he said, "Oh, it was good, but you're not really interested in my vacation. You're interested in why I'm calling." I said, "Yeah, okay." He said, "I went on vacation, and I had a Hamiltonian that I took with me, and I solved it. I was convinced before I went on vacation that it would solve all the problems we are having with quantum chromodynamics." And I said, "Did it?" And he said, "No. I swept stuff under the rug, and it didn't work. But I think it would be good for solid state physics. So, I'll give you the form of the Hamiltonian, and if you can give me problems that relate to it. I'll be your graduate student, and I'll solve them for you." I said “Great.” How many people get a chance to have Dick Feynman as a graduate student? He said, "Great. I’ll give you three days to find problems." So, I started thinking about it, and I went next door to see Charlie Kittel, and I told him the story. He said, “Don’t waste your time, Dick's not going to call back." However, he did call back three days later. He said, "Do you have problems?" I said, "Yes." He said, "A lot of problems?" I said, "Yes." But he noticed by the tone of my voice that I was not enthusiastic. So, he said, "Wait a minute. Marvin, are they good problems?" And I said, "Well, what do you mean by "good problems”?" And he said, "Would you work on them if I wasn't involved?" And I said, "No." And he said, "Thank you very much. Goodbye." So, I passed on my chance to have Dick Feynman as a student. 

Zierler:

Amazing.

Cohen:

See, there's an example of a case where there's a mathematical solution, and the mathematical solution can be used to explain physics. However, if the physics isn't that interesting, having a formula or method is no big deal. You can have the greatest instrument in the world, but if it's not going to be used to do anything interesting, there's no point in using it. I don't know if I'm getting my point across. 

Zierler:

No, for sure. And the question about quoting Einstein about not everything is important, how do you measure that? How do you define -- I mean, isn't there something that is essentially subjective about importance? How do you define importance, and then how do you allow that definition to determine what it is that's worth your time to work on?

Cohen:

Well, there's basically two parts to that. One is, there are things that are fashionable, and somehow you get dragged into things that are fashionable. Like, right now, graphene. We had been working on it before, and rotating layers of graphene which is very fashionable now. Then there are things that may not be fashionable, but there's something that smells good about them. There's some inherent interest -- you know, I always wanted to understand why things happen, and if it looks like an experiment may give me some idea about something different, I work on the data. So, I would say that in my case, a larger fraction of the things that I choose to work on, are things that somehow hit my fancy, as opposed to being fashionable. That's not necessarily the best way to go, because there are a lot of things that you miss out on because people get into fashionable fields because they feel that there's something important there. So, I think it’s best to do a little of both by trusting your own intuition and watching the directions that the crowd is moving into.

Zierler:

What's a good example of something that piqued your interest, that was not necessarily fashionable, but you're glad that you worked on it?

Cohen:

Well, superconductivity comes to mind, the superconducting semiconductors; finding the first oxide; Alex Müller said that strontium titanate was the grandfather of all the oxides, and that felt good since this became a great field. I was sorry I didn't stay in it and do more. There are things that pique your interest, you contribute, but you miss a big thing because of the circumstances of the time. As an example, I was visiting the Exxon lab as part of a committee, and since I was working on metal clusters at the time, I asked about their cluster work. In particular, I had a new theory of why atomic metal clusters occur in certain sizes--magic numbers. Experimentally, if you take a bunch of sodium atoms, put them in an oven, and you let them come out of a hole, they'll clump into groups of 2, 8, etc.--magic numbers. I figured out why there were magic numbers. It was an electronic effect. It wasn't because the structures fit together nicely.

So, back to Exxon. They had started up this lab, which went very well for a while, and they were examining clusters of carbon atoms. In their data, I saw that carbon atoms liked to get together in groups of 60. And as I was sitting there, trying to figure out what was so special about 60, members of the visiting group I was with, said, "You've got to come out of that lab now. We've got to go to lunch and stay on schedule." I didn’t respond because I was trying to make a model for why 60 was a magic number, since “magic numbers” was the kind of physics I was doing at that time. They kept yelling at me, and finally I said, "Okay, I'll think about it later." Later was too late, because while I wondered about why carbon does that and planned a computer study, the Rice University group figured it out. It was a buckyball. I don't know whether if I had sat there for a while, I would have figured it out. I talked to Rick Smalley about it in later years. He had figured it out with models he made. So, there are things waiting to happen, and if you’ve lived in physics for as long as I have, there are important things that can happen if the situation is conducive. At the moment, sometimes you get it, and sometimes you don't. Clusters were not fashionable in physics at that time, but we made great strides for metal clusters even though I missed recognizing Buckyballs.

Zierler:

What are the drivers in the physics world that determine trends? In other words, something that's trendy in a particular given year, what drives that? Is it government funding? Is it a particular discovery? What are the things that really determine, this is the thing that everybody should be working on now?

Cohen:

Both contribute, but often, discoveries are made and become drivers. In particle physics, as I mentioned before, the idea of getting down to more elementary building blocks was a driver. In condensed matter physics, certainly, if you have an application -- let's say, you can make something that can lead to an effective photovoltaic device, that's a driver on the applied end. A new material or experiment that looks promising for finding new physics is a driver. There are also incentives coming from the government -- I mean, requests for proposals. For example, there was a nanoscience initiative, and now there are funds for studies of big data and for quantum computing. Possible new funding sources can be drivers. So, there's funding influences, but for me, it's usually the physics going on at the time. There's often something about the physics that is being done that seems to suggest that there something important around the corner. You go along, you open a door, you open another door, and then you get the idea that there's a direction. It’s an intuitive kind of thing probably based on experience. 

Zierler:

On the metaphor of opening doors, and it opens another door, and it opens another door, do you see that as an infinite process, or is there some end point where physics can open the final door?

Cohen:

No. Infinite. Infinite possibilities. It keeps going.

Zierler:

Why is that?

Cohen:

Why is that? I don't know if I can answer that question. It's just my belief or feeling. Science at some level is similar to religion in that you have beliefs. For example, belief in logic is like a religion for me. I’ve never seen it fail. Since answering your question with logic is difficult for me, I can only say I don’t know. It's like discussions about religion, or about what your beliefs are. You know, when you stay up at night having discussions in the dorms in college, trying to decide about whether there is a God, or whether you could build a God-like machine that could keep track of what everybody is doing all the time? Things of that kind. Sometimes in physics research, you just feel like you are getting a private look at some special order in the physical world. An example of “feeling this way” is when you find something new, and you realize you're the first person in the world who understands this new thing. When it’s totally new, it can give you new insights. Nobody ever understood this before, and you understand it. 

Zierler:

You've had this feeling before?

Cohen:

Yes.

Zierler:

How many times? How many times have you had that feeling?

Cohen:

Oh, probably a dozen times. You get the feeling that you have achieved something and it’s a clue to more revelations. So, what's happening is that somehow built into my DNA is the idea that once I get to something, there's always going to be something else. So, if you ask me, do I think it's going to stop, it's hard for me to think of it stopping since it's always continued in the past. 

Zierler:

Does that mean that the unified theory is not the final door, even if there is such a thing as a unified theory?

Cohen:

That's right. That would be my feeling. 

Zierler:

So, do you believe that there is a unified theory?

Cohen:

Well, it's not my field, but if there's a proposed unified theory, you've still got to put a box around it. There may be a proposal for a unified theory of the fundamental particles and their interactions. Something like the standard model has been a unified theory up to a certain range. Newton's equations work in a certain range. For quantum mechanics, there's never been anything that people found wrong with quantum mechanics. It's hard to interpret, but it's the real world. Do we know that it was appropriate at the time of the Big Bang? I don't know. The point is, it seems like theories get expanded. You look deeper and there's more needed. So, I don't expect that I'm going to see in my or my grandchildren's lifetime, somebody construct a unified theory that explains everything about everything. I mean, just think of biology. There's so much to explain like the emergence of life and consciousness. 

Zierler:

When you consider over the course of your career, major new discoveries, or experiments or theories that really and truly pushed the ball forward, how much of that is because of advances in technology? How much of it is political, and funding support? How much of it is strokes of genius? How do those things work together in various ratios to create those true leaps in understanding?

Cohen:

What was the first one again?

Zierler:

Let's see. It was -- I've got to remember myself. Basically, it's genius, it's funding, and it's technology. Genius, funding, and technology. 

Cohen:

Technology is a fabulous component, because if you build a new instrument that can see further, or see smaller, or whatever in more detail, it almost always opens up a new area. If you can get to lower temperatures or higher energies, you usually find something interesting. If you invent a tool like the scanning tunneling microscope, you find new things. Also developing faster computers with larger memories can have a big effect on the progress of science, Then, the strokes of genius, like the Josephson effect, the fact that he thought of that. The question is, can anybody else think about it? Sure. Eventually, somebody probably would have thought of it. There are discussions about genius which emphasize that if Beethoven had never lived, we'd never have had the Beethoven symphonies. No, but we'd have something else. Science may be different, and this has been discussed by the world’s greatest scientists for well over a hundred years. In my view, it’s a question of belief and semantics when you ask if Nature has these properties waiting to be discovered by anyone, while only Beethoven could have written his symphonies. Bottom line: genius is good for science.

Now, in terms of funding, sometimes you need the institutions. I'm thinking of funding, not so much as just being able to go out and buy certain equipment, although you have to, particularly for cutting edge experimental physics. The best experimentalists have to have enough in the way of state- of- the- art equipment if they're to stay in competition with each other. Using the best probes is very important. The other thing is the institution. Bell Labs was probably one of the best, if not the best, physics lab of all time. The fact that it's gone is sad. We've lost a national treasure. Just look at the applications like the solar battery, the transistor, and they had a big part in the development of lasers and computers. People trained there went on to universities around the world and educated other scientists. Their scientists received the highest honors including Nobel Prizes. If you ask about the funding there, I believe that because AT&T was a monopoly, they were able to have and support this laboratory which was amazing. So, funding makes a difference-- all three--genius, technology, and funding are important. Getting back to my role, of course, as a theorist I don’t need a lot of funding, but having graduate students who are geniuses and good computing facilities are a real plus. Fortunately, I had both. But an experimentalist does need good funding, because, like I said above, once you make a new microscope that goes beyond previous microscopes, there's always something new to see. Anyway, this seems to be generally true. It’s been true so far.

Zierler:

In what ways has technology specifically advanced your field and your own career?

Cohen:

For me personally, lasers and synchrotron radiation along with better spectrometers and derivative spectroscopic techniques are examples, because they can see those little peaks that we predicted caused by electronic or vibrational transitions; better methods for making new and purer materials, so that they can find materials that we predicted that never existed before or were not previously made in the laboratory ; and high pressure systems, like diamond anvils, to find and characterize predicted materials and predicted properties. It’s easy for me to simulate higher pressures, I just changed a parameter to set how far apart atoms are in my models. It's trivial, but for experimentalists, it’s hard. They used to need huge buildings for their presses. Now, they do it with diamond anvils which you can carry in your pocket, and they can get to very high pressures, close to what you'd find deep inside the earth. Other very useful tools are the scanning tunneling microscopes, where they can see electronic processes and watch atoms move. For example, we're working on some problems now looking at how atoms and molecules move on graphene. Experimentalists can actually use probes to push atoms and molecules around and watch them interact. So, work like this has a big effect on theory in general, just the fact that experimentalists can see things on this size scale is very important. Also, as I mentioned, good computers help.

Zierler:

Are there theoretical problems that have just sort of nagged at you over the course of your career that you never were able to resolve or solve to your satisfaction?

Cohen:

Yes. For example, in the field of high temperature superconductivity, Phil Anderson and I had a paper on the maximum transition temperature that you could get for certain conditions. I would like to extend that work to other conditions to see how high a superconducting temperature is possible theoretically. To explain the oxide superconductors, the so-called high temperature superconducting oxides, would be wonderful. There are a lot of theories about these systems but no consensus on the fundamental origin of the properties. So, problems like that that are around that don't have solutions yet.

Zierler:

Why is there an assumption that there is a solution? Is it possible that there is not a solution, or is that not how it works?

Cohen:

That's not how it works. When Kammerling Onnes asked Einstein to explained the superconductivity that he discovered, Einstein supposedly said, "Oh, it's some complicated molecular interaction. You'll probably never figure it out." I'm paraphrasing, but I'm paraphrasing from things I read about what Einstein said. But Einstein would have loved the BCS theory, because it explained a certain class of superconductors with a fundamental theory. Unfortunately, Einstein died a year before BCS was proposed. As for electronic structure, developing new theorems about how electrons move in crystals and new physical and mathematical approaches like topology, are being developed. I think there's a lot more to understand there, too. I believe that there are solutions to problems like these. Also, they may even lead to even more interesting problems.

Zierler:

Have you ever thrown up your hands and just given up on a particular problem, and never looked back, or do you sort of hold these theoretical problems close to you because you're never willing to do that?

Cohen:

The latter. They're always there. 

Zierler:

They are always there. Have you ever come back to a problem that you couldn't solve, in say, a given decade, that you were able to solve, or resolve, later on?

Cohen:

Yes. When I was younger, longer than a decade, sometimes when I was really stuck on something, I would think deeply about it before I went to sleep and wake up in the morning with new ideas and solutions. This idea of letting it go in, and having it turn around in my brain works sometimes -- for example, if somebody asked me your name in a week, I'll probably say, "Wait, wait, wait..." And then two hours later, I'll come up with your name. The brain keeps working even when you are not trying. Now, the thing that happens with a decade or several decade intervals is that you learn new techniques and gain new perspectives, and with these you may be able to go back and solve old problems Let's say I have something on a back burner, and much later I develop a new technique. Often it works out that I can then solve a problem that I previously couldn’t solve.

Zierler:

Have you ever turned to one of your colleagues for help on a problem that you couldn't solve yourself, or do you determine that it's for you to solve on your own?

Cohen:

No, I talk to people all the time. 

Zierler:

So, over the course of your career, who have been some of your go-to collaborators for issues that eluded your ability to solve them?

Cohen:

First of all, I went to whoever my mentors were at the time when I was working under somebody and to fellow students. Then, collaborators, going to conferences to see people, and fellow theorists here at Berkeley. But let me give you one example as a more direct answer to your question. I had a friend in graduate school, David Penn. We are still close friends. We went through our PhD years together, and then he worked at NIST, which used to be NBS. I'd try things out on him in graduate school, and this habit continued afterward. He was one of these bright but nonjudgmental types that are hard to find. The type of questions he would ask are the “emperor has no clothes” kinds of questions. So, I would say, "Dave, I have this thing, and I think it's right, but it doesn't quite fit together, and I would blab on." And he would come up with, "Well, wait a minute, why do you think that assumption is right?" And I’d say, "Of course it's right." And he'd say, "Really?" And often I’d find out that it wasn't. You know what I mean? Sometimes talking to yourself in a mirror, or if you can talk to somebody who says, "Well, why that?" helps, because often you overlook things. You assume something sometimes, and it screws everything else up. Dave’s help was often much more than that, as he would sometimes suggest new approaches and new ideas. So, talking to people often helps you get the bugs out of your theories. The other thing that's not appreciated is the advantage of working with foreign students, and with foreigners in general. The language mismatch is such that you'll say something and they misinterpret you, or you misinterpret them, and then you realize that the gears are not working right together, and that takes you out of your mindset which can be very helpful. I understand drinking does that too. Supposedly a lot of good physics was done in the last century in the German beer halls.

Zierler:

On that question of foreign students, you've collaborated with scientists all over the world. Have you detected value in basic cultural differences, and how those differences might be productive to advance your field?

Cohen:

Well, certainly there were cultural differences, but we always had the common languages of physics and mathematics which unified our thinking. However, there are sometimes differences in the style. For example, let's say I have a visiting student or postdoc coming from abroad. Often when they first join my group, they seem to be shocked at the fact that even my youngest students sometimes tell me I'm wrong or my idea sounds crazy. I'll do something at the board, and my American student may say, "No, that can't be." A foreign student thinks lightning is going to strike. Another example, postdocs from abroad that I've had would come in and call from the airport when they arrive, and they would say they want to come to campus and start work on their assigned problem right away. What I used to do is to string them along for several weeks, sometimes even a month, to get them to relax. I imagined what they were thinking when they would sit in our group meetings while we were having what seemed like low key conversations. I think they asked themselves: Why is this a good group? I don't understand. Nobody is really working. Then they'd get their alpha waves down to a level where we're really talking about some of the basic things, instead of jumping in and calculating. It really makes a difference. The reduced pressure has a good effect on people coming from some countries.

Nowadays the differences I saw in the past have diminished a great deal. The amazing thing is how diverse my group has been. I remember coming to a group meeting on my birthday. When I came in, I was greeted with a Happy Birthday and the students and postdocs had written happy birthday in their native languages on the wall. I think there were 12 different entries. The only redundancy besides English, was Chinese at the time. And yet, the differences between us seem to go away once we started doing physics and putting math on the board. Same thing with gender. I had the highest percentage of women PhDs in our department. There was never any gender related or diversity problems in my groups. Everyone was treated in the same way. However, I will say that one gender difference was that the women's opinion of themselves was usually below my opinion of them, compared to the men where I sometimes had the opposite feeling. In other words, with some of my students, because they're doing theory and they're alone working on problems much of the time, they tend to evaluate themselves too often, in my opinion. These are bright achievers. Often from the day Mommy takes them to kindergarten, they're the best person in their class for their whole life. Then, all of a sudden, they are in a research group and there's a student sitting next to them from another country who's way above them scholastically, or whatever, and they're not the fastest gun in the West, and it really gets to some of them. The women didn't have that problem as much. But in terms of the physics, our work was gender blind.

Zierler:

I remember, I was talking about Mary Gaillard about being the first woman to be tenured in the department. Obviously, in the course of your career at Berkeley, the faculty and the student body are much more diverse. I wonder if you can talk about the impact of diversity on physics.

Cohen:

Well, in terms of producing physics, in a theory group at least, I don't think diversity matters, but it’s almost always there. For example, in the condensed matter theory group, we always had diversity in the sense of having members from many countries, women, men, and one of the men actually changed gender and became a woman. Everybody in the group was always very tolerant about their differences. I came here in 1964 to join the faculty, so I've had a lot of students and postdocs--probably 100. The groups differed in character depending on the nationality makeup in terms of how they acted, but they were all tolerant of each other and got along well. The main thing they cared about, really, was doing the physics. So, I don't think that the fact that they were from different countries, male or female or transgender, affected the physics. You know what I'm saying? But this is theoretical physics. Maybe in a lab, it might be different. Perhaps people do things differently in different places. But here, it was always just the physics, and the atmosphere was always tolerant and open. I’d expect that the study of physics is pretty much the same everywhere. It’s not like the arts where there are national styles or approaches which perhaps can make a difference.

Zierler:

I want to talk about some of the prizes that you've won over the course of your career: The Buckley Prize, the Lilienfeld Prize, the Dickson Prize. What are some of the awards that meant the most to you personally and professionally?

Cohen:

I felt that receiving the National Medal of Science was an important recognition of our work. I was impressed by being in Washington, having it being awarded by the President, and having a long conversation with President Bush. The Franklin Medal festivities were also very impressive. I was really surprised and moved when my wife and I were getting out of the taxi in front of the Franklin Institute, to see these flags flying, and one of them had my picture on it. So, in terms of being impressed, I was very impressed by both of these honors. At the time, I had the feeling that I would like to share these honors with my collaborators. I’m sure many others have had this feeling when they receive honors or recognition. The honorary degree ceremonies were wonderful. The Weizmann Institute of Science festivities were spectacular. It was such a nice occasion. I had a similar experience at the Hong Kong University of Science and Technology. And the University of Montreal, which is a French speaking institution, so I didn't understand it all, but it was great to receive this honor in the city I was born in. So, the honorary degrees were special. My first major prize, the Buckley Prize, made a big difference in my career. It had a large impact on me because it was the first major national recognition of my contributions. 

Zierler:

So, the others, because it was early in your career, that one actually had an effect on your career, as opposed to being something that was nice for you to be recognized. 

Cohen:

Yes.

Zierler:

In what ways did it change your career?

Cohen:

It was just that everywhere I went, people would mention that I had a Buckley Prize. But I should add that I was always very appreciative of all the prizes and honors I’ve received. An unusual recognition which I wasn’t aware of, was that Jorge E. Hirsch came up with this h index. All of a sudden, people were calling me offering me deanships, and actually a presidency of an institution. It turns out that there was an article in Science that I hadn’t seen that said that Ed Witten and I had the two top h numbers, and these were higher than Stephen Hawking’s, and people on that level. I was really surprised and embarrassed by that. I was not thrilled about it since, I felt that in a sense it meant that people were going to really put a lot of faith in measuring scientists by a particular number. So, that was something that worried me at the time and still does. 

Zierler:

Why did that not sit well with you, to have you reduced to a number like that? What was problematic?

Cohen:

Well it happened -- I mean, if you look at how things have evolved in this area. You go into a faculty meeting now, and they're talking about hiring, and you've got five people to consider, and then someone mentions their h numbers. If somebody writes an incredible paper, and it doesn't have a lot of citations, it may not count as much. How do you measure these things? An underlying problem is that nowadays people can't understand all the subfields of physics, or don’t take the time to read the research papers, so they start relying on things like the h-index. It’s similar for rankings of universities and departments. These comparisons are very complicated. Anyway, I am not happy with the idea of reducing scientific accomplishments to a single number. 

Zierler:

Do you see that there's a disconnect between the h index, and the relative importance of a given scientist's work?

Cohen:

No. I never looked at it deeply, but I'm just saying that should not be the only or perhaps even a major factor. I think it may have too big a role. I was happy to learn that I have a high number, but I don’t really think about it. Some people keep checking their own h number. I haven't done that, but they ask me, "Is your h still going up?" That type of enquiry just doesn't sit well with me, and I usually don’t answer those questions.

Zierler:

We can debate how valuable it is or not, but why do you think that your h number is so high? Why do you think that is?

Cohen:

I think the fields that I’ve worked in, and the calculations I did, were of use to many experimentalists. Also, the work was usually relevant for both theoretical and experimental studies, so I got a lot of references. 

Zierler:

It was irrelevant?

Cohen:

It was relevant. You know, we're in a big field. We have a lot of people, and I guess some of my work was fashionable. 

Zierler:

Well, I don't want to let you dodge that with too much humility, because the obvious next answer is there are a lot of people who are contemporaries of you, who work in the same field as you, whose h numbers, however valuable, are not nearly as high as yours. So, clearly, there's something more to it than what you were doing, and when you were doing it, that explains this. 

Cohen:

Well, we were first on a lot of things. When you do something first, you're often reference number one or at least referenced, then after that you get a lot more people who keep referring to you. We did set the tone for a large part of the field of the electronic structure of materials and contributed some major breakthroughs in that field. And the field had a big effect on condensed matter physics. It’s a lot easier to study a material if you know its band structure, and it’s not that long ago when people didn’t have that tool.

Zierler:

So, let's go through them. Do you view your major contributions, generally, in terms of firsts, or are those two separate considerations? In other words, if I were to ask you, what do you see as your top x number of contributions to the field, would that answer accord pretty perfectly with what were the major areas in which you were first, or are those separate considerations?

Zierler:

Yeah. So, the question was the correlation between the contribution and being first to be involved in it. So, then, the question is, what's the break down in terms of being first? How much if it is insight, how much of it is luck of being in the right place at the right time, and what are the other factors that go into determining those firsts in the course of your career?

Cohen:

I think for theory, it's mostly insight. That’s true also for experimentalists, but sometimes Nature helps them out. I remember, Don Glaser said that people told him that he stumbled onto something and received the Nobel Prize. And he said, "You have to do a lot of walking before you stumble." So, in terms of my career, as I said, people need to know the electronic structure of materials both for fundamental studies and for applications. Conyers Herring gave a talk about how little we knew about the electronic structure of silicon in the early 1960s, and right after that, I had worked out the electronic structure of 14 semiconductors. So, that opened up the field, and a lot of people would refer to me for that. I remember that when I did the first charge density plots that showed where electrons are in solids. When it came out, the chemists said it was clearly wrong, and some physicists did too, and now charge density studies are “everywhere” so to speak. In fact, Phil Anderson wrote a paper saying the charge densities I calculated were caused by the truncation of a Fourier expansion used in the calculation and weren’t physically meaningful. He sent me a preprint, so I called him and asked him not to publish it. He was surprised when I said, “I’m asking for your sake, not mine, because the calculations are correct.” He withdrew the paper. Let me explain that these are contoured plots showing the density of electrons at various locations in a crystal. An important first result was the prediction of the electronic density pile up between two silicon atoms to produce the covalent bond.

Experiments were done later for silicon using x-ray scattering and these data verified the theory. Another example is when I came up with the idea that there was a way to look at how electrons behave on the surfaces of semiconductors by theoretically constructing a supercell which contained many atoms representing the bulk crystal along with regions representing space or vacuum. So, the result was a super-crystal made of slabs, and each slab had a surface. This theoretical super-crystal could be treated with our methods as if it had an infinite number of surfaces, and the properties of these surfaces could now be calculated. Supercell calculations have become common, and that was a breakthrough in surface physics, so I got a lot of references for the work we did on surfaces and interfaces. Then, together with a student, we figured out how to calculate the total energy of a solid. You consider a bunch of atoms, and you put them in various positions and calculate their total energy, and then you change the position and do the calculation again.

For Mother Nature, the arrangement that has the lowest energy always wins, and that tells you the structure of the material. It also tells you how the structure changes with pressure. That was another thing that got a lot of references as it led to calculations of vibrational properties, and electron interactions with vibrations. The various contributions in superconductivity were breakthroughs, and they got many references, but not on the level of semiconductors, basic methods, and nanoscience studies. When we started research on nanostructures, the first question was whether the methods that we used for bulk crystals and surfaces would work for nanosystems-- nobody was sure at that time. Now it's implicitly assumed, but we were the first to show that our methods worked in this regime, and we followed up with important calculations. And so, it goes. It was more insight than stumbling. In other words, insight is needed when figuring out what you want to know and then asking whether you can find a way to achieve your goals.

Zierler:

On the question of insight, I wonder if you can explain. I wonder if you could talk about the drama of it. In other words, when you realize that you have made an insight, is it a eureka moment, or it is the culmination of a lot of work that is not necessarily dramatic, but that when you step back and assess all that you did, you realize that's the insight in total?

Cohen:

That's a fun question, because in the end, it's more the second option--a collection of eurekas. You look back and say, "Did I do all that?" But at the time, it's a bunch of eurekas. In other words, the first time that I figured out, that I can do all the surface physics with this structure factor building supercells, I was very excited. Or with the boron nitride nanotube, when I went to my group meeting and I said, "Hey, you guys, I think this really could work.." Or superconducting semiconductors, or strontium titanite, an oxide -- they're sort of little eurekas, but the important thing is when you look back and you put them together to look at the body of work they add up and it’s a good feeling. The strange thing is - and other people I've talked to have had the same experience - when you go back and read one of your old papers, you start reading the details and you think, "Did I really understand that?" Because, you remember it sort of generally, but you don't know whether or not you really understood every one of the details back then.

Zierler:

How much do you value elegance in an equation? Does an equation have to be elegant for you to be satisfied with it, or can a messy equation be satisfactory also?

Cohen:

I'm okay with messy equations that have good physics in them. I appreciate elegance and economy in communication. I learned some of this from Charlie Kittel. I had written a paper after I'd been at Berkeley for just a short time, and I gave it to Charlie to read and comment on. He said, "The physics is good, but think about how you're writing up this research. Why don't you go back and read Dirac's papers?" I had read Dirac's papers before, but when I read them again, I saw the economy of words and elegance in a way. Here Charlie was talking about writing style, and he was right, I was too verbose. However, I also know that Dirac felt that it is important that an equation look good and be beautiful. But in the end, I feel that the elegance comes through only sometimes in an equation, and that can be wonderful, but sometimes it's coming through more in the physics, and the equation could look a little, perhaps, not as beautiful as other beautiful equations that are empty. Also, when people use computers, the equations they solve are often ugly and approximate, but the solutions can be important,

Zierler:

To the extent that theory serves as a facsimile for nature, would that suggest that nature can be messy as well?

Cohen:

Yes, look at biology. To me, biology is messy. Whenever I talk to bio- people, and they try to hook me to work on it, I have the following thought. It seems that in biology if you've got a process, you've got a little widget that is responsible for it. If you've got another process, then you've got a different widget that does that. But getting to general principles, is hard. So, it bothers me when they say that last century was the century of physics, and this century will be the century of biology. I often reply, "It'll be the century of biology if biology turns into physics." 

Zierler:

What do you mean by that?

Cohen:

I mean, if they get theories that explain collections of phenomena. If they get a deeper and more general understanding, instead of a little piece of this and a little piece of that. So, for me, even though many advances in biology and the life sciences have been fabulous and important work, for a theoretical physicist, the general principles of biology are scarce. Again, there has been enormous progress in some areas, and people have done important things like the structure of DNA. Look at the difference that knowing the structure of DNA has made. Hopefully they can achieve more advances like that. For DNA, it became physics. I mean, Crick was a physicist.

Zierler:

Do you see yourself for your work as part of a particular intellectual tradition in physics? In other words, when you study the work of physicists of past generations, do you connect particular lines that go back several generations that you feel a particular affinity toward, or an intellectual heritage with a particular set of physicists? If so, who might they be?

Cohen:

Yes, I think I do. For example, the Franklin Institute did some wonderful research going through and connecting the medalists to the past medalists. I'd connect myself up with theoretical physicists who were close to experimentalists, close to the underlying physics. For example, I wouldn't connect with Freeman Dyson. Freeman, who I knew fairly well, was mathematical. I envied his mathematical skills. Fermi, I would certainly connect up with. For example, Freeman talks about his interactions with Fermi, and how he had worked out these complex theories, and Fermi said, "No, no, no, it can't be that because there was some physical aspect that Fermi didn’t like." And Freeman said that Fermi was right. So, I identify with people who get their hands dirty. which means getting closer to experiment. I do feel that there's a tradition of doing theoretical physics with an eye on the physics as opposed to over emphasizing the theoretical formalism. My students have traced our actual roots back to the first American Physical Society President, Henry Rowland and before that back to researchers in Germany. I believe the family tree is on the web in various places.

Zierler:

And Fermi really embodies that for you?

Cohen:

Yes.

Zierler:

More so than his contemporaries?

Cohen:

I'm not sure. 

Zierler:

I mean, I guess my question is what makes Fermi unique that you so strongly identify with his process?

Cohen:

Well, first of all, he did experiment and theory. It’s almost impossible do that at the level he did nowadays, but the point is that he really got into the nitty gritty of what was happening physically.

Cohen:

Emilio Segrè told me about the times when they'd have group meetings, and Fermi would bring up these problems that were very physical, and how Fermi had his standard methods or approaches as first attempts to solve the problems. In other words, he had favorite techniques that he used on many different problems. Density of states is one and the Fermi “Golden rule” which is perturbation theory, and pseudopotentials. I told you previously that I worked on a problem with Murray Gell-Mann, and Murray would continually ask about stories that I heard about Fermi that I got from Segrè. He was also interested in Fermi. It was the physical insight, getting down to the fundamental physics, and getting a feeling for what was going on. After you did that, then, you could develop the equations. Dick Feynman was also very physical. He said that he felt that if he solved a lot of problems, this would be useful because he would recognize the essence of them in other research problems. Also, he made a relatively famous remark about Greek versus Babylonian physics. For Babylonian physics, you jump in the middle, and there's a relationship between this and that, and you just keep going until you’ve solved problems and understood Nature. The Greek approach builds on basic assumptions and keeps building to higher levels. This is beautiful in some sense. However, when one of the lower levels in the structure breaks down, it all falls apart. Whereas the Babylonian web of knowledge was more robust. A break doesn’t destroy the whole structure. I think that Dick also had more affinity for physicists who think physically.

Zierler:

Did you see Fermi as -- you know, in describing his style, and his approach, did you see that as sort of like he created this on his own, or is there an intellectual tradition that he came from where you can see how he developed as a physicist?

Cohen:

I don't know the answer to that. I heard that when he was young, he was self-taught and would write down the essence of what he learned about various subjects in notebooks for various fields. These notebooks contained a distillation of many studies, and he would refer to them for new problems. For me, I think of Fermi as more of a Mozart. Although Mozart was trained by his father, his contributions emerged from him in a process that’s hard to understand. I don't know how Fermi's mentors influenced him. I know he had excellent colleagues. There are schools of theoretical physics that nurture geniuses. I remember Phil Anderson and I sat together drinking and discussing this, and we got into a competition. He said the best theory school of all time was at the University of Chicago, with Fermi and his people. I said the best school was created by Max Born when he had his famous group in Gottingen. We started matching man for man. I said, "Well, Born had Fermi in his group for a while. He came through the Born group, and Heisenberg came through, etc." And Phil kept saying, "Well, Gell-Mann came through Chicago, and there was Yang and Lee”. We went back and forth. It was a fun game, and it brought out a lot about what we each valued. I value Born's ability to attract those people and to teach them, and for having such high standards. The point is that these were all great people, and the atmosphere was nurturing. If you line up 100 violinists, you can pick the top 10, but between those 10, one is going to have a better technique, another's going to have a better tone, one's going to have a better interpretation of the music, one's going to have another whatever. So, it becomes a question of taste as to who is best. 

Zierler:

So, on that question of taste, in terms of all of the incredible scientists that you've worked with, who most stands out in terms of, when you were in their presence, you were most in awe of them?

Cohen:

I'm not sure. I'm always very impressed by how good people can be. I guess, we're talking about theorists.

Zierler:

Or anybody. Anybody you've worked with, or anybody you've learned from. 

Cohen:

I was very impressed with Bardeen. He had this steady way of going at things, but you had to have patience talking to him, because you'd say, "Good morning, John." And eventually he'd say, "Good morning." After he thought about what you said. He was so persistent. When I talked to him about superconductivity in semiconductors, or surface physics, or whatever, he was so slow to talk. But when he did, he seemed to always be right. He was doing things in his head “really right”. Soon he assimilated everything I would say, and he kept going developing new ideas. I did publish a paper at one point claiming he was wrong in some work on excitonic superconductivity, and he was famously wrong on the validity of Josephson’s early work, but he got two Nobel Prizes in Physics for two incredible contributions. He was special. When I worked with Murray, Murray was fast. But it was a very different kind of thing--very smart with insight about everything. I thought Phil Anderson was deep and thought broadly about things. He also had a unique way of looking at puzzles. I was always impressed with Jim Phillips, my thesis advisor, because he was so creative, and broad, and fast, and brilliant. Jim is an excellent game player, and these skills were evident when he did or discussed physics.

Another impressive mentor was Morrel Cohen who would come up with many creative and often unique ideas. Leo Falicov once said that “Morrel sparkles.” A good description. Also, there was Charles Kittel who brought me to Berkeley, and Charlie had the style of turning an important physics study into a smaller physical problem that he could solve elegantly. He was clear, and had clean solutions that just had to be right, and he was physical. And so, his people, like De Gennes, learned from him, and so did Hubbard, and Phillips. All three were his postdocs too. He had this array of brilliant postdocs and students, like Morrel Cohen and Al Overhauser, and they learned to solve problems.

I remember Segrè telling me that he felt the same way when he worked with Fermi. He said that he would ask him questions, and Fermi would reduce the problem to its core. If Fermi didn't have the answer right then, he would say so and then take time to work it out. But when Emilio came to Berkeley and asked Oppenheimer about some experiments he was doing, each time Oppenheimer would give him an answer almost instantly. Oppenheimer was notorious for being lightning fast. Once Segrè said, "I don't think you're right. I think you should think about it." And Oppenheimer blew a gasket. He was like that, from what I hear. I interacted with him a little when he was at the Institute for Advanced Study, but I never really knew him on a personal level. These very fast thinking people are interesting, but they're not always right. I found that Edward Teller was very clever and quick. I was brought up to think of him as Dr. Strangelove, but he wasn’t like that in my view. He was very interested in understanding physics, and he grasped things quickly. Edward would call me and say, "I have to see you." And he'd come to Berkeley with a million questions, particularly about superconductivity. He seemed to develop an atmosphere of urgency whenever I was around him. For example, at one point, he had a theory of the oxide superconductors. When he described it to me, he'd went on, and on, nonstop. Every time I'd try to interrupt him, I’d get a few words in, and he'd interrupt and say, "Please, Marvin, we don't have much time. Let me finish this." And he'd go on, and at the end, he'd say, "What do you think?" And I'd say, "I’m sorry, but I think you're wrong." And he'd say, "I'll have to think about it." And then he immediately repeated to me what I said in every one of my interruptions. In other words, when I said, "But, Edward, you left out the q-dependence.”, he was listening, and it would go in. He was very bright, and I was impressed by how quickly he caught on to things, but he was involved in all kinds of activities at the same time and not focused just on physics.

In contrast, Hans Bethe impressed me as really focused on physics, because even when he was very senior, when I heard him talk, he approached problems sort of “like a graduate student does”, putting in all the details. I remember meeting with him at Cornell and having wonderful discussions related to his questions about the colloquium I had just given. However, talking to him in Berkeley much later stands out in my mind. This was just about the time when nanoscience research started “happening”, and there were all these buckyball experiments being done. In particular, it was found that they could be superconducting. Hans was visiting Berkeley to give a colloquium, and he came to my office and said, "Marvin, I have two small questions regarding what you are working on. I'm not in your field, I don't know much. Will band structure work for these things, and is C60 a BCS superconductor?" And I said, "Yes, yes." And he said, "You're speaking so softly." Then, I saw him about a year later, and I walked up to him, and I said in a loud voice, "Hans, yes, yes!" The point was that in the beginning things were still developing. I believed what we were doing, but more experiments were needed to verify the theoretical calculations. However, my main point is that here is this very senior guy working in another field, and he asks two major fundamental questions in my field. These are the kinds of theorists that impressed me. They go right to essence and asked the important questions. I think Fermi and Bethe were like that.

Zierler:

Now, the question of intellectual heritage goes in both directions. I asked you about the past where you trace your lineage. What about some of the students that you've had over the years? Who do you see as -- you know, as many people that you want to name, but some prime examples of students that you've had that have carried that flag, in terms of your intellectual heritage, and the way you as a mentor have tried to pass that tradition onto your students?

Cohen:

Counting up my PhD students and postdocs, I get about a hundred individuals, so I can’t discuss them all. I will send you a list of each group. I’ll mention a few, and make gross omissions because I’m not doing this systematically. An obvious person who carries the flag is Steven Louie, who is at Berkeley. Steve and I wrote a graduate textbook together, “Fundamentals of Condensed Matter Physics” based on a course I began teaching in 1965. Since he and I had taught the course numerous times over an almost 50-year span, I thought writing the text would be relatively easy. However, I underestimated the amount of time and effort needed to complete this project by a huge factor, but we did it. Steve runs his group in a similar style to the way I ran mine. They're doing wonderful things, primarily in nanoscience. When Steve was a student, I was impressed by how he would get past stumbling blocks or mountains in the way when solving a problem. If he needed better math or deeper thinking, he would come up with the resources and the answer. He seemed to “tunnel through” obstacles. The former student who picked up on what I think of as my style, with regard to training students and added to it, was John Joannopolous. John is a physics professor at MIT. John is an incredible teacher, and his style of teaching students and running his group was similar to what he had here. His former students, which includes a Nobel Laureate, Bob Laughlin at Stanford, are at major universities -- one of his students, Dung-Hai Lee is a professor at Berkeley; Karen Rabe and David Vanderbilt are professors at Rutgers; Tim Kaxiras at Harvard; Gene Mele at Penn: and his students are professors at Cornell, Yale, etc. He has had one of the most successful group of students in condensed matter physics in my opinion. These are all my grandchildren, and I’m proud that they are on the faculties at major universities across the country.

So, I think that some of the intellectual heritage you mentioned will be carried on. I had brilliant students and postdocs, like Jisoon Ihm who went to Seoul National University. He has won prizes and is considered a national treasure in Korea. A TV crew from Korea came to Berkeley to interview me about him and to see where he did his work in the US. He developed a school of theoretical physics at Seoul National University. Before he came, I had the impression that all Korean students were workaholics because of their rigid training. However, Jisson was much more relaxed than most of the students I’ve mentored, but very effective. He'd work at his desk, and he'd fall asleep, and he'd wake up and move on. I remember when my son was a teenager. He came to visit me on campus and he said, "You should be ashamed of yourself. You make your students work so hard." I said, "What do you mean?" He said, "They're so tired; they fall asleep working." I told that is the way Jisoon worked. I heard that Jisoon had taken a national exam at the high school level in Korea, and he was number one out of something like 35,000 students. Then, when he was at Berkeley, he was number one. There are many Korean students that worked under him who did very well.

Another excellent student from Korea was Kee Joo Chiang who had a marvelous thesis which was one of the major achievements of my group over the years. He also developed a theoretical school at KAIST. Of course, there were impressive American students too: Phil Allen at Stony Brook and Ching Fong at Davis were early students. Phil was always extremely clear, careful, and therefore easy to work with. Everyone loved working with Phil. Ching was extremely sound and careful. His work was thorough and always trustable. Later there was Vin Crespi who not only did brilliant work, he connected us to the experimental groups. He went to Penn State, and contributed to their Physics Department in a major way.

Except for Steven Louie, who is my colleague here, the former student that I’ve interacted with most and resonated with most over the years is Jim Chelikowsky. Jim is a professor of physics, chemical engineering, and chemistry at the University of Texas at Austin. He has had a distinguished career. I continue to collaborate with him. We seem to be in sync on most subjects. I think that although many former members of my group function in a similar way, many branched out broadly and so did their students. For example, Bob Laughlin, worked on the quantum Hall effect. Jay Sau works on topology, Renata Wentzcovitch on earth and planetary science, etc. I've been lucky. Berkeley attracts excellent graduate students. I’m often asked about my PhD students who were women. The list is: Mei-Yin Chou, the late Jennifer Corkill, Amy Liu, Yvonne Tsang, Carmen Varea de Alvarez, and Renata Wentzcovitch. All have had exceptional careers. Regarding the postdocs, I had very good postdocs. Michael Schluter had a big effect on my group in the 1970s. He went on to Bell Labs and had a big effect on the theoretical physics effort at Bell too. Unfortunately, he passed away at a young age. In looking over my postdoc/visitors at the postdoc level, I’m struck by how successful they became around the world. They are active in Japan, England, Korea, Spain, Germany, Portugal, and France. Permit me to list some of them: Choi, Coh, Giustino, Grossman, Lischner, Martinez, Martins, Miyamoto, Petroff, Pickett, Richardson, Rubio, Saito, Son, Vanderbilt, Yndurain, Zhang, Zunger, and others. I keep in touch with most of them and even collaborate with some like fairly recent work done with Susumu Saito.

Zierler:

You've named a very diverse list of students. What are some characteristics that they share, that might serve as a through line to understand their success as a group?

Cohen:

I think one thing is that they all work closely with their own students and with colleagues. They tend to collaborate. The ones who went into the academic world, developed groups. Another thing is that they coupled well to experimentalists. For example, Steven Louie works with a lot of experimentalists, and John Joannopolous does too. John actually got into things that are also applied and into new interesting fields like photonics. He started three companies or more and wrote books. Jim Chelikowsky has a large group of students, covers several fields in science and engineering and also writes books. So, I think many people from our group went on to be very collaborative people. 

Zierler:

You mentioned earlier that you had sort of an aloof relationship with patenting your work, with developing patents. I wonder, where have you seen economic opportunities as a result of your work, and why does it seem that you generally stayed away from that?

Cohen:

I think it was because I was from a different era. In other words, I think that for many theorists in my time, patents were almost dirty. You were not supposed to be motivated by stuff like that. You're in the theory convent. Your service is to Mother Nature. It's basically not a business. However, things really changed. My early students would never think of saying anything about patents, whereas more recent students have said, “Can we get a patent on this?" When I was at Bell, any patents that came out of our work went to Bell, but when I came to Berkeley, the profit on patents were shared with the University. As I mentioned before, Charlie Kittel had brought up the subject. He once said, "You ought to patent the computer programs that you developed." I said, "What? Why?" I did not think deeply about this. Here's one story. I went to a conference in Paris, and it was all about the computation of electronic structure and related science. Out in the hall near the lecture auditorium, there were people selling computer programs. I walked up to one, and he was selling one that was essentially a copy of the empirical pseudopotential code I developed very early on. I said, "How much do you want for that program?" He said, "Only $300." I said, "Only $300! Do you know how much time I put into writing that thing, and you're selling it for a lousy $300! " Now, there are many companies in this business. It’s getting to be a fairly large industry. One of my students, who was from the former Soviet Union actually started a small computer business using some of the skills he learned at Berkeley. I don't know if he'll be a success. But if I had patented some of those computer programs, as Charlie suggested, that would have been lucrative. If I had patented some of the methods, ideas and materials related to our studies of the electronic structure of solids, like band gap engineering, that would have been very lucrative. I don't know if I could have gotten these patents, but I never tried to find out. There were people who wanted to come in and help me patent things, but I just said I didn't want to get involved with it. I really felt, throughout my career, I was fortunate to have the job that I had, I liked doing what I was doing, and I wasn't looking for much else.

Zierler:

Do you think in physics, it's possible to achieve greatness if you don't have a good work ethic? In other words, if you don't work hard, is it possible that somebody can be a genius and really achieve great things without necessarily working hard at it, or is grinding it out day in and day out truly fundamental to pushing the field forward?

Cohen:

I think the answer is somewhere in between your limiting cases. I think being tremendously creative in coming up with brilliant ideas and applying them without putting in the effort and time, and then making major contributions must be rare, because I just haven’t seen examples of that kind. Just like you don't see mad scientists, it’s a myth. The image that most scientists are weird or crazy is inappropriate. They're just like everybody else in most ways as far as I can tell. I do see exceptionally brilliant students, but they have to work to succeed. Again, going back to Feynman, one of his favorites was talking to the freshmen class at Caltech -- these are young people who have succeeded their whole lives -- he revealed that he has something terrible to tell them, and that is that half of them are going to be in the lower half of the class. Some of them express surprise. You know, there was always this illusion of smart people not working. Some people would say, “Oh, he's so bright, or she's so bright, and they never work”. It's not true. I think anybody who does anything important works at it.

Josephson, before he developed mental issues, was brilliant and productive. As an undergraduate, he pointed out that a famous experiment testing general relativity had been interpreted without including an important component. His thesis ended up being Nobel Prize winning research, and he did it himself. But he worked. I got a Sloan Fellowship just after I joined the faculty at Berkeley, and since Brian and I had discussed working on superconductivity together, I decided to use the fellowship to go to Cambridge. I thought it was a great opportunity since we were both about the same age, and he had done such great things. So, I went to Cambridge, and when I got there, he wanted to work on higher levels of consciousness, and he was clearly having mental issues. Hence, I had difficulty when I tried to collaborate with him, so I didn’t stay long. From the standpoint of physics, I should have stayed since I was successfully interacting with Volker Heine, and Nevill Mott asked me to see him each day around 4pm. In the beginning, Nevill and I discussed various problems mostly associated with strontium titanate and the two-dimensional systems that Abe Yoffe was doing experiments on. Then Nevill asked me to collaborate with him on amorphous semiconductors. He felt that combining my knowledge of the electronic structure of crystalline semiconductors and his explorations concerning amorphous systems could lead to a theory of amorphous semiconductors and beyond. I was interested, but I wanted to leave cold, damp Cambridge. Mott called me a “foreign wife”, because when visitors came to Cambridge to do research, they usually adjust and love it, but their wives didn’t. Anyway, Nevill received the Nobel Prize in Physics in 1977 (11 years later) for his work in this area. So, I should have stayed. Getting back to Josephson, you get someone like that who may give the image that all you need is to be a genius -- he's clearly a genius, but when he was productive, he worked hard, and when he worked hard, he accomplished great things. The problem I had in mind to work on with Brian was solved years later by other researchers much in the way I envisioned that we would solve it.

Zierler:

I feel like I have to ask now -- do you feel like there's a possibility that supernatural phenomena exist, or is that, by definition, not possible? 

Cohen:

I don't believe in it. For the fun of it, when I was in high school and college, whenever I took English or speech or debate classes, I always took the side opposite to what I believed to test my debating skills and to show how easy it is to be deceived. For example, I argued that flying saucers are real, and then after I made my case and won the debate, I would give all the reasons I was wrong just to show how garbage can slip into the media. This is now called fake news. But no, I don't believe in the supernatural. 

Zierler:

So, there's never been anything in your research that has made you question some fundamental principles of how the world works -- things that you might have thought about, or seen, that don't make sense within the given framework?

Cohen:

For me, for my whole life, logic always worked. If something goes wrong, I made a mistake, or I don't understand the problem, or someone gave me bad data, but supernatural--no way! I marvel at what nature has created, I marvel at our natural world, but I don't think of anything supernatural as causing the working of the system. However, I don’t pretend to know how we got the system we have.

Zierler:

I think I have two final questions, and they're both very broadly conceived. One is backward looking, and the other is forward looking. The one that's backward looking is I'd like to know what you think about the concept of mystery. You could answer that either personally, or as a representative of your field. So, the question is, at the beginning of your career, what were some examples of things that were fundamentally mysterious, either to you personally, or mysterious to condensed matter physics, that over the course of your career, either through your own discoveries, or through the discoveries of your colleagues, are no longer mysterious? Things that you have seen in the course of your professional lifetime really transition from mystery to understanding.

Cohen:

I think superconductivity is one. I just couldn't believe that the resistance in a superconductor is absolutely zero, that something could go around in a circle forever. 

Zierler:

What does that mean?

Cohen:

If you took a ring, and kept it cold, and got a current going, as long as you kept it cold, the current would go forever. It would never decay. The calculated time it takes for it to decay is of order of the age of the universe. And then I understood it. Also, I couldn't understand how electrons got down wires without bumping into things, and how currents would move, and then I understood it. Let me emphasize that there's levels of understanding. I could understand how a theory like quantum mechanics works and how to use it correctly. I've fed my family by doing quantum mechanics, and yet, there are fundamental questions of quantum mechanics involving going from the quantum level to the macroscopic level, the questions that people are asking nowadays, that I never understood on a level where I felt I had the physical intuition to explain some things to myself successfully. I have physical intuition about what a quantum state will do, because I worked with that my whole professional life. I talk to my students about wave functions, and so on. We can think about things like that, and that's the real world. It's not the world that most people think of. I have no trouble thinking about things at that level, but understanding it and having a feeling for it like I do for classical physics isn’t there. These are the types of questions that Einstein was asking when he said that quantum mechanics was spooky. He said, "God doesn't play dice." But God does play dice. I get that part, but quantum mechanics is still spooky.

So, answering those kinds of questions and mysteries are important. Now, Einstein made a big deal about mysteries. He said it was one of the things that keeps you going. It's one of the reasons you do what you do, and I would agree with that. So, if you ask about the role of mysteries, yes, in a way, I would say I was always driven by puzzling behaviors, and I would try to explain them logically. I’ll give you a trivial current example. Alex Zettel came to me, and he said, "If I take a nanotube, and I fill it with iron. The iron atoms crystalize inside the nanotube. I can then use an electric current to move the iron, and as the iron keeps moving it passes through narrow sections of the tube, but it remains a crystal. How could the iron atoms remain a crystal? That drove me crazy. I said, "It melted." He said, "Nope. Definitely not melted." I thought it was melting, and re-crystalizing, melting, and re-crystalizing. This was not the case, but we did figure it out, and it was the mystery that motivated us. Seeing pictures of iron atoms moving through pinched areas of a tube and remaining crystalline and understanding the mechanism was very satisfying. Anyway, things like that happen when you work with experimentalists, figuring out mysteries can be a strong motivation.

Zierler:

Is there anything that you can think of that remains mysterious, either to you personally, or to your field, that are as mysterious, possibly, as when you started your career?

Cohen:

Well, there are specific problems that I've always wondered about. Like, can you make hydrogen into a metal using pressure? This is a controversial subject now. Experimentalists argue about whether they've been able to do it. After squeezing it into the metallic state, would it be a high temperature superconductor? I am working on this problem now. So, that's a mystery, and it’s a mystery I've thought about for a long time, and I worked on it at one point. But I had to put it away for a while. 

Zierler:

What's standing between that theory and understanding it? Time? Insight?

Cohen:

We'll get it figured out at some point. So, it's just time. There are fundamental things about physics that come up all the time. In terms of having enough mysteries, I feel that the experimentalists will keep finding things and asking questions, and we theorists will be asking new questions too.

Zierler:

On the question of fundamental concepts in physics, are there fundamental concepts in physics that, for you, remain close to you day in and day out? Things that really shape how you see the world, or inform how you do the work that you do? Any theories, or laws, or concepts, that stand out, that are unique among others that you feel a particular affinity towards, or just close to you?

Cohen:

Well, generally, thinking mathematically, like other physicists, I probably have a different approach to viewing the world than many lay people do. Also, I would think that knowing quantum theory and understanding phenomena around me in terms of quantum mechanics, gives me a different view than people who don’t know the theory. I'm not sure I'm answering your question, but I think that knowing what physics I know does color my view of the world. As I said previously, I like mathematical explanations, and I've never found logic to fail. If something's not logical, then I don’t accept it until I figure out what went wrong. Also, something unexplained, will gnaw at me, and really bother me, and it'll stay in there until I get it worked out, because I can't pass it off as a miracle, or I can't pass it off as supernatural. In life, you witness amazing coincidences. Seemingly strange events happen. A lot of people would say, "Spooky," but these kinds of things can be acceptable probable occurrences if you carefully compute the probabilities. So, again on the supernatural, no. On the mysteries, yes. Anytime something is not understood, or something doesn't fit, it will stick with me. In terms of the beauty, I look to music and art and literature for beauty, as opposed to physics except for the beauty or whatever it’s called of understanding Nature at a deep level. And spirituality. I keep my spirituality separate from my science. 

Zierler:

Well, to the extent that you would have to merge them at some point. The question of your unshakable allegiance to logic, and your unwillingness to accept the existence of supernatural phenomena or miracles, right? It seems that the question boils down to is, is the concept of God logical? Not the God that has anthropomorphic characters, or that speaks to people in any of the way to humanity relates to the concept of a God, but is the idea that creation has a creator a logical concept to you?

Cohen:

It's okay with me, and I believe in the God you describe. Another story, and then I'll let you go. My wife Suzy said, "The Rabbi is going to call you." I said, "Why?" And she wandered off without answering. So, he calls, and he says, "I want you to give a talk on science and religion." I said, "I'm sorry, I can't. I've never thought much about the subject. I'm not an expert, please get someone else. There are wonderful people who think about these things." He answered, "I have an opening in the middle of October." I said, "No, I really can't." Then he tells me, "Well, at that time I'll be talking about the tower of Babel, and you could work it in." I knew there was no way I could get out of it. No way. So, I went to the internet, and I looked up what Charlie Townes said about the subject. Charlie was a very religious man. Not Jewish, but very religious and a great scientist, so who could be better. It turned out that Charlie had said all kinds of things about the subject which I was fine with, but I felt that he got too close to creationism. I always respected Charlie and his views in science and human events. He was a bright, nice, fair, and wonderful man, but for me he went too far in some areas on this subject. So, I gave the talk, and I made Suzy promise never again. After the talk, I sat there, and people came with questions, and we talked for a long time. In my talk, I mentioned Townes, and I also mentioned that I consulted a colleague (who I didn’t name) when I was asked to give the talk.

It was Michael Berry who I think you know. Well I went to a party, and Michael who is an old friend was there. I see him every year at the Weizmann Institute and we have long discussions about physics and other things. His brother is a Rabbi. So, I said, "Michael, I have to give a talk about science and religion. Any ideas?" He said, "I'll give you my talk." I said, "You gave a talk on science and religion?" And he said, "Yes." I said, "Fabulous!" So, he sends me his talk, and in his talk, he says that the worst thing that ever happened to the world is religion. I was so disappointed since I thought I had a set talk. Charlie Townes felt that a scientist had to believe in God at some level, and I’m sure that Michael didn’t share that view. Although I didn’t mention Michael’s name, I assured the congregation that he was an exceptional scientist, and I tried to explain to them that scientists have really different beliefs about this subject. Some, like Charlie, believe that God did it all, and it’s important that you believe that. Others don’t have that view. For me, I don’t think about it when I think about nature, however I do find myself saying or thinking Suzy's favorite expression, which is "Oh, my God." She says it several times a day. For me, there are things that somehow evoke that, "Oh, my God. How can this be? How did that get put together? You see a little dot, and it's alive." I mean, things like that. But in thinking about things, I never try to rationalize religion in scientific ways, like how an ocean can split apart, or how biblical stories can be true. I’m not fond of miracles.

I knew Al Overhauser quite well. He was a devout Catholic, and he was with a group who investigated miracles for the Vatican. He would tell about how he went out to check on reported miracles. I said, "Did you ever find a miracle?" At least at the time I talked to him some years ago, he had never found a real miracle. So, I don't look for miracles, but I do have an underlying faith. When people say, " Why do you fast on Yom Kippur?" I say, "I've done it since I was a kid." "Well, why?" "That's when God writes the names of who's going to live, and who's going to die for the next year." They say, "You believe that?" I say, "Well, I’m here." I had a colleague from our Philosophy Department whose father was a cantor, but he was a flaming atheist. He asked me the question above and he asked me the following question. "When the Pope gets into his pajamas, and nobody else is around, do you think he really believes that the Angel Michael was in The Church of the Holy Sepulcher?" And I said, "Yes." He answered, "What do you mean?" And that's when we got into what you mean by belief. You vote with your feet. I do what I do. Mostly, I do whatever I did as a kid.

Zierler:

Last question. The one thinking forward to the future. What are you most excited about personally in your own career, and then broadly for your field? What does the future hold, and what excites you about that?  

Cohen:

Well, I'm usually mostly excited about the problem I'm going to do next. I'm excited about the problems I'm working on, but what's the next problem? I have three, or five, maybe ten problems going on now with my two postdocs. I get excited about them. I don't look far ahead. I look at solving these problems, and then I’ll get other problems, and so on. So, for me, I feel that I’ll always have interesting problems or phenomena to work on. For the field, I think computers, artificial intelligence, and the rest, have limitations, but they are going to be great tools. I'm in a position to say something about that, because I was one of the first to really start using computers in condensed matter physics. The Russians were very much against it. I was a fan of Lev Landau and Landau's school of theoretical physics was outstanding. I knew many of his former students and heard a great deal about how he educated his group. I also knew of the questions he would ask, and his students tended to mimic whatever Landau said, so they often asked the same questions he did.

On one occasion, I gave a talk, and there were a lot of Russians in the audience, and this was at a time when using computers in physics was controversial. When I finished my talk, one of the Russians asked a standard Landau question, "What is the small parameter of your theory?" Landau would set up some problems with a small parameter so he could use perturbation theory based on that parameter and solve the problem approximately. I said, "I don't need a small parameter. I have a computer." They were just floored. And I remember, at another conference, somebody got up at the end of my talk and said, "Don't you think that working with computers is mindless research?" There were a lot of jokes about physicists doing these expensive calculations and coming up with an answer 3.14, i.e. coming up with Pi, and they could have done it simply without a computer. I remember saying, "I can do mindless research without a computer." It was those things that changed attitudes. It was that kind of response that made a bigger impression than a detailed rebuttal. I used the same approach when I was President of the American Physical Society. This was in 2005, the International Year of Physics and the 100th anniversary of Einstein’s great year. I travelled a lot that year and spoke about physics and Einstein around the world. I had to testify for congressional people to support having foreign students.

They said, "Well, sure, you educate all these foreign students. They go back to their country, and then they beat you in research, and you want us to give you more money so you can compete with the researchers you created." I said, "Okay, you're right. I'll just tell you that right now, the international language of physics is broken English. If you want it to be broken Mandarin, then don't support our proposals." Those kinds of things have a bigger effect than showing detailed arguments or calculations. Regarding the overuse of computers and their use by people who don’t understand the problem at hand, I like what Jim Chelikowsky said, "Given the computers of today, and our knowledge of the physics in the 1970s, compared with being given the computers of the 1970s and our knowledge and the physical insight we have today, I would take the latter." In other words, the fact that the computers helped to develop our physical insights and our physical models is the important thing. But you have to be careful. There are computer packages available to all, and people show beautiful pictures of their results, but they don't know what it's about. It's a “garbage in, garbage out” kind of thing. Understanding the physics and making sound physical models -- are essential for getting really good insights into what's going on in physics, that's the bottom line.

Zierler:

Well, Dr. Cohen, this has been an incredible discussion. I want to thank you so much for your time. There are going to be so many people that get so much value out of your insights, and your experience. I really want to thank you. 

Cohen:

Okay, thanks. What happens to our recording?