Thomas Witten

Zierler:

Okay. This is David Zierler, Oral Historian for the American Institute of Physics. It is September 18, 2020. I am so happy to be here with Professor Thomas Adams Witten. Tom, thank you so much for joining me today.

Witten:

I’m very pleased to be with you, David.

Zierler:

Okay. So, Tom, tell me first, to start, what is your title and institutional affiliation?

Witten:

I am called the Homer J. Livingston Prof. Emeritus, Dept. Physics, James Franck Institute

Zierler:

When did you go emeritus?

Witten:

2013.

Zierler:

What is the connection between your chair and Livingston, your position and Livingston?

Witten:

Well, it’s a financier named Livingston who endowed a chair.

Zierler:

Okay. So no particular connection with physics or with the kind of work that you do.

Witten:

It’s not even spelled quite the same as the famous Livingstone of history, although I do get jokes about it.

Zierler:

[laugh] Okay. All right, Tom. Let’s take it all the way back to the beginning. First, tell me a little bit about your parents. Where are they from?

Witten:

Well, my parents are from the rural South and the sort of semi-rural Southwest. My dad was raised in Tulsa, Oklahoma during the Depression, and my mom was raised in a little college town in North Carolina about the same time. Her father was a dentist. His father was a civil engineer. They lived in towns, and they had the regular old American nuclear family.

Zierler:

Where did they meet?

Witten:

They met when my dad had just finished medical school. This was during World War II. He had gotten a deferment to go to medical school and finish his degree and finish his training before he would go overseas, and my mom was a medical student in the same place. So they were doing rounds together, and somehow there was a spark and four months later they were married.

Zierler:

Did both of your parents have careers as doctors?

Witten:

Yes, they did. The main full-time career was my dad, who was partly a Veterans Administration doctor and partly an adjunct professor at the University of Colorado Medical Center. There was a close affiliation in those days between most veterans’ hospitals and a medical school, so you could teach residents and have fellows and do a little bit of medical research on the side. Then my mom—she wasn’t practicing for a while because I had a lot of brothers and sisters, and that sort of puts a damper on it. But then later on, she went back and got her residency in radiology and had a part-time position with a radiology practice after that.

Zierler:

Now given the generation and where your mom was from, how path-breaking did she understand her pursuit of a medical degree and a career in medicine?

Witten:

That’s kind of interesting because she wasn’t as path-breaking as you might imagine because her mom, my Grandmother Squires, had broken a path before her. She was one of the first college graduates in that region, and she was quite an educated and smart woman. But she didn't pursue a career after that. But my mom got interested in biology and wanted to work caring for people, and she asked her mom if she could maybe aspire to be a nurse and her mom said, “Would you think about being a doctor?” So that’s what happened.

Zierler:

Now where were you born?

Witten:

I was born in North Carolina in my mom’s hometown, at the hospital in the neighboring city. This was during the war. My father was already overseas, and my mom was very busy in a different city finishing her internship to finish her medical degree. She would come home when she could, and her sisters, who were staying at home, would take care of me, and so one of my aunts is very close to me because of that.

Zierler:

Did you grow up in North Carolina?

Witten:

No. Right after my dad got out of the war, which was ‘45, he got a position as an intern at Johns Hopkins, so we moved to Baltimore and lived there for the first couple of years of my life. Then he got another position in Utah, because he wanted to study neurology as a residency, as a specialty. So, we had a little house in Utah and that’s where my first two sisters were born. I should say I have lots of sisters and brothers. There were ten at the maximum, ten of us, of which the last one was adopted. So, then this neurological residency was stressful for my dad because you couldn't make anybody-- couldn't cure anybody. You could tell exactly what was wrong with them, and then you couldn't do anything about it. So, he realized that this was not for him, and then we moved to a really small hospital in western Colorado, which was a veterans’ hospital, but it was possible to do a residency there. After that, we moved to where we ended up settling down, in the suburbs of Denver.

Zierler:

And Denver was where you spent your sort of formative years as a teenager.

Witten:

That’s right.

Zierler:

Did you go to public school or private school in Denver?

Witten:

Public school all the way. Public school.

Zierler:

Tom, when did you first start to get interested in science? Was it early on?

Witten:

Very early on. I wanted to be an inventor.

Zierler:

Oh, yeah?

Witten:

My dad showed me that you could wrap a wire around a nail many times and hook the two ends of the wire to a battery, and the nail would turn into a magnet. Then I was listening to the radio all the time, and when I was like four years old, I thought there was a little guy in there and a little orchestra right inside the radio, and then I realized that couldn't be right. They told me that the sound of the guy and the sound of the orchestra were coming through the air, and that was just so amazing. I had to figure out how that worked, so that was the start of it.

Zierler:

Were you a stand-out student in math and science in high school?

Witten:

I got As. I was considered one of the smarter kids. The place where I went to school was a very kind of culturally diverse place. The town bordered a disused military fort where the officers’ quarters were taken up by doctors in residence from the nearby VA hospital, and then there was a little town where the school was-- which was what you would call a rural slum. So, we were all mixed up together. It was a really good school, but very culturally diverse. So, I had an enriched background and I was one of the smarter kids always in my class, but not a superstar, I wouldn't say. The same is true in high school. I think I graduated third from the top in a class of about 500.

Zierler:

Now your parents both having medical degrees, did you feel any pressure or did they encourage you to pursue a career in medicine?

Witten:

No. They kind of bent over backwards not to do that. They really did, and they were just as tickled that I wanted to be an inventor. I would tell them what I liked to invent, and they said, “That’s what you call a physicist.”

Zierler:

Ah! So, the first idea with physics actually came from your parents connecting your interest in inventions.

Witten:

That’s right.

Zierler:

When you were thinking about college, were you thinking specifically about physics programs?

Witten:

No, because by that time-- in high school I realized that the world had a lot more in it than these geeky invention-type things, and I felt the lack of that. I felt very narrow and constrained and I wanted to learn about the big world. The counselor asked me in my senior year what I wanted to go into after that, and I said I wanted to be a long-distance truck driver. But really what I wanted was to learn everything that was out there to be learned in college without regard to a specific field, and so I went to a place which had a strong academic reputation, but not especially in science, a little college in Portland, Oregon called Reed College.

Zierler:

Now did you specifically want to go to a small school?

Witten:

Yes, I did, but it was under the influence of my mom, who thought I wasn’t a go-getter or aggressive enough type to survive in a big place. She thought it would be better to have a smaller scale, lower intensity, lower distraction place, and I guess she was right.

Zierler:

So, Reed was a good choice for you.

Witten:

It really was. For the first time in my life, I was surrounded by people who I felt a kinship and identification with, more than I ever had in high school.

Zierler:

What kind of kinship did you feel at Reed?

Witten:

They said the same kind of crazy things and told us the same kind of crazy jokes!

Zierler:

[laugh] How much was the counterculture at Reed part of your reality at that time?

Witten:

It was. So, Reed had gotten a reputation by that time, from the McCarthy era, of being a hotbed of liberalism, and then later on from being a free-thinking campus where there were not too many restrictions placed on you and you could get dope. And they had… What did they have? They had inter-visitation, it was called. It was okay for a man to visit a coed in her room or in his room, but you had to keep the door open, and that was a little bit sort of pushing the envelope in those days.

Zierler:

Now did you declare the major in physics right away, or that came later on?

Witten:

I declared it right away.

Zierler:

What was the department of physics like at Reed? How many professors were there?

Witten:

There were four or maybe five. One guy was part-time. They really knew what they were doing, and I was particularly influenced by this charismatic theoretical professor named Nick Wheeler, who had studied at Brandeis under Schweber. He really was able to give you the deep picture of how things like Lagrange’s equations and Hamilton’s equations-- how they came about and what they meant. He told me about the rotation group and all these delicious things that you have to be a physics major to even appreciate.

Zierler:

Now obviously, Tom, you had no basis to compare it to as an undergraduate, but being such a small department, is your sense that you got a pretty broad exposure to all kinds of physics—experimental, theoretical, all of the different fields—or was it more narrow as it pertained to the areas of research that the professors represented?

Witten:

Let’s see. So, when I compare with what the physics students experience now, they get in the lab and do research. Our teachers were there because what they wanted to do was grow students. They wanted to do research, yes, but they wanted to do whatever it took to grow as students. So, I didn't get experimental research experience at Reed. I did go to summer jobs. They had the counterpart of today’s Research Experience for Undergraduate program sponsored by the NSF, and I went for a summer for one of those. We did sort of what could be called research in that program. But at Reed I was doing the Millikan oil drop experiment in instructional lab, and stuff like that. I was completely absorbed with those things and completely absorbed with the academic push to get into deep stuff, but it didn't eclipse everything else, you know? I was taking political science and I was taking philosophy. There was a big heavy course requirement at that school in Humanities, which was just a big synoptic course. It’s a lot like the Great Books course at the University of Chicago. It may have been based on that, but anyway, I really didn't feel a sense of missing out on what I could have had as a physics education. What I probably did miss out on is access to the elite of physics, you know, to have met people who were making an impact on the world of physics. But I didn't know to miss it at the time, and I don't even think I missed much now.

Zierler:

Tom, by the end of your undergraduate education, how well formulated or formed was your identity in terms of the kind of physics that you wanted to pursue for graduate school?

Witten:

Not that well. I loved physics. I really wanted to be able to pursue it and continue it, and I was enthralled by the hardest, the most unintuitive things like special relativity and quantum mechanics, which I had only the barest feeling about, but I said, “I have to get to the bottom of these things.” I started out in graduate school going into research with a professor in particle theory, but I ended up changing advisors and doing many-body theory of condensed matter.

Zierler:

What graduate schools did you apply to?

Witten:

Harvard, Syracuse, and UC San Diego.

Zierler:

That’s a fairly eclectic group of schools.

Witten:

It is. They wanted to give me-- First of all, the faculty wanted to limit how many places each student was applying to.

Zierler:

“They” is who? Reed?

Witten:

Yeah. We were marshaled. We were shepherded through this process, and they said, “We want you to be sure and have a backup school and then a reaching school that you’d have a chance at getting into as well.” So, I got an unusually-- an unexpectedly good score on my standardized test, the GRE, and that made me think that maybe I would have a chance to go to Harvard. So, I asked my professor if he would write a letter for me to go to Harvard and he said, “Sure, I’ll write a letter,” but I didn't get into Harvard. I guess I’m kind of glad I didn't. On the other hand, another school that I didn't really expect to get into was this brand new school in San Diego--

Zierler:

Right.

Witten:

--which was so new that they didn't have undergraduates yet.

Zierler:

Yeah!

Witten:

The year I got there, they had started freshman classes, and then that class would go on from year to year. As it grew, as they progressed, the curriculum would grow to meet them, and so I was doing my TA job for those first classes of undergraduates.

Zierler:

Tom, was the draft something that you had to contend with?

Witten:

No. This is a good point to bring up because if you’ve interviewed people my age, it was an enormously big deal in that era, but I was just before the cutoff age where you were at risk. In my year, you got a deferment all the way through your PhD, but after that, some of my buddies in younger classes were having to face these wrenching decisions about what they would have to do to avoid getting drafted, or maybe they would just get drafted.

Zierler:

Were you politically active at all during these years?

Witten:

Not very much. But I was in the presence of it and kind of swept away in the flow of it to some degree. So ‘68 was right in the middle of that, okay. And there was a lot of student unrest. The Berkeley students had launched a massive strike because of some mistreatment of some people in what was called the People’s Park. At People’s Park, you could kind of do what you wanted. You could squat. You could camp out. The authorities did something to clean it up. There was some conflict between the police and the students, and all the campuses went out on strike. So, we felt it incumbent on us to support our fellow students and be on strike. Then there was a rally in the middle of our own campus, and it was a rally that was sort of on the edge of going critical-- of taking off and starting to break things. This was a vivid memory for me, and it was a scary memory. Another vivid memory was in 1968. I had just driven in to have my first class on a Monday morning. I was coming into the parking lot and at the same time my buddy, George Blumenthal, was coming to the same class. This was the day after the weekend when Robert Kennedy had been shot 100 miles north in Los Angeles. George and I got out of our cars and looked at each other. I said, “Oh George, let’s just get back in the car and go to Canada.” We were so downhearted.

Zierler:

Yeah.

Witten:

So, it swept over me, but I wasn’t playing an active role in anything.

Zierler:

Tom, as you got comfortable in San Diego and you were taking classes and getting to know your cohort, your fellow graduate students, I’m curious, particularly for those students, many of whom probably came from bigger undergraduate physics programs, how well prepared you felt vis-à-vis them.

Witten:

I was kind of amazed at the cool things they knew how to do. We would be working through our problem sets in Jackson. Jackson is the canonical electricity and magnetism textbook. It’s a wonderful textbook, and we had a really good teacher. He would assign these things and my buddy Dan just knew this great approach of doing the Taylor expansion in an electric field around some point. There were many things hitting me that my cohort knew how to do that hadn't occurred to me to do, but I was holding my own, nevertheless. So, I didn't feel-- I felt it was stimulating, sometimes surprising, but never overwhelming.

Zierler:

So, given that you were sort of wide open in terms of the kinds of physics to pursue in graduate school, how did you narrow down both your research area and who your graduate advisor would be?

Witten:

Let me see. The professors would give us spiels about what they were doing, and one of the professors was a new recruit—pretty much everybody there was a new recruit—who was doing nuclear physics and he said, “Physicists have learned how to do three-body problems in quantum mechanics, and I’m studying what you can learn from those three-body methods in studying different nuclei.” That sounded like a really terrific, cutting-edge thing to do, and so me and my buddy Ron Grubman decided to work on those things. We had a joint project on this subject that we did together, and it was a great project. That’s how I got started with that, and so it wasn’t a particular penchant for a particular subject matter. I just wanted to get the deepest and most unintuitive, shocking stuff like three-body and quantum mechanics—come on! You couldn't do better than that, so that’s what got me into that problem. So although our problems were nuclear physics problems, our advisor was doing particle theory. Those were the days of Regge poles and different kinds of holistic or synoptic views of how to make sense of the set of particles that we had. So whereas he was holding his own at that, he saw that the field was not growing in scope and he said, “I’m not sure I can get you a job in that. I think the field of condensed matter is much more open and has better prospects.” So, there was a young many-body theorist named Shang Ma, and my advisor recommended Shang to me. I took a course from him. I thought he was terrific, and so he became my advisor.

Zierler:

How did you go about developing your dissertation topic?

Witten:

Let’s see. They were very open about what you could do for a thesis. It had to involve publications, but we already had a publication on the nuclear physics stuff. So my new advisor, Shang Ma, had said, “I think what we have to do is for you to just write something up and get out.” So, he proposed a modest but meaty problem in many-body quantum mechanics with Green’s functions. It was called the two-dimensional charged Bose gas, so if a gas is made of charged particles, this makes a qualitative difference in what the collective excitations of the thing are like, the elementary excitations. If it’s two-dimensional, that changes things in two different ways, so it wasn’t clear what the elementary excitations of this thing would be like. That was my job, to lay the groundwork of that.

Zierler:

How closely was this research related to what your advisor was doing?

Witten:

It was a side project. It wasn’t me helping him. In those days, he was doing things like critical phenomena, which is not quantum mechanical at all. So it was a project he gave me for me to cut my teeth, and he was helping with it. But he wasn’t really involved for his own research. It wouldn't have been a coauthored paper, nowhere near. It was up to me to do it.

Zierler:

In terms of the terminology, has solid state by this point been replaced essentially by condensed matter?

Witten:

No, they were kind of two different things. Solid state was very much based on what leverage you would get scientifically by knowing that your system was a periodic lattice-- a crystal lattice. But then there were other things that were just as interesting—like liquid helium where there was no crystal lattice, but still there were excitations to be figured out and microscopic phenomena to be figured out, especially superfluidity. So that wasn’t solid state, but it was condensed matter. So, the terms were kind of morphing during that time and for the whole next decade, really.

Zierler:

Who else was on your committee?

Witten:

Let’s see. There was Shang Ma, my advisor. The second guy I remember quite a bit was Bruno Zimm, who was a famous polymer chemist. He was in the chemistry department. At that time he was not doing anything connected with my thesis, but I liked him and I liked the stuff he was studying. Then David Wong must have been there; he was my initial advisor, and there was a fourth one, Donald Fredkin, my stat mech teacher.

Zierler:

What did you want to do after you defended? What were some opportunities available to you?

Witten:

Well, I had a lucky opportunity because the last year of my thesis work. Shang, my advisor, decided to leave for a semester and go work at the Institute for Advanced Study with Roger Dashen. Shang had taken these leaves before. He had gone away from campus to work for some senior person, and so this time, when he said he was going, I said, “Okay, that’s good. I’m coming, too.” He said, “What?” I said, “That’s right. I’m coming,” so he said, “Since you're coming, I’ll try to get you a place in the physics department at Princeton.” So he did and I got to know some physics graduate students that way. I struck up a friendship with different people there, and one person that was involved was Prof. John Hopfield. John Hopfield and his collaborator had money for post-docs, and they ended up hiring me as a post-doc, but they wouldn't have done it if I hadn't been sitting there in the carrels with the graduate students and talking to them months earlier.

Zierler:

Tom, to what extent was the post-doc an opportunity to go on to new physics, and how much of it was an opportunity to continue on in the field that you had developed for your dissertation?

Witten:

It was almost entirely new. Hopfield had some very general ideas about condensed matter that he suggested to me, but they were so general I didn't see how to go anyplace with them. The things I did mostly there were to absorb what he was doing in biophysics, where he was explaining how the hemoglobin molecule is able to be especially smart and efficient at taking up oxygen. Then I was hanging out with the experimental condensed matter people and trying to understand what their experiments meant. Then I had a chance to interact with Freeman Dyson. He was teaching us a graduate course on elementary excitations. He wasn’t required to teach at all—he was at the Institute [for Advanced Study]—but he felt the need. So, he introduced me to a problem-- well, to two problems. Really all I did in graduate school that led to papers was inspired by Freeman Dyson. But another big element in this was that we had a visiting professor who had come to teach us about a new subject. The professor’s name was Ken Wilson, and the subject was something he called the renormalization group. Now, Wilson was a field theorist; he was a particle theorist.

Zierler:

Yeah.

Witten:

Now, field theorists had noticed that you have these funny fudge factors in the field theory equations called renormalization factors and that they were not uniquely specified by the data. You could change one of them, but if you compensated by change of another one, the physics would be the same. So there was a group of transformations, a group of tradeoffs you could make that would keep the physics the same.

Zierler:

Now how exotic was renormalization at the point that you entered the scene?

Witten:

Well, it was quite exotic. The going-in picture received by Wilson was the most abstract thing in the world. I mean, they wanted to get to the deepest understanding of everything with the particle theory, and one of the things you had to settle and get worked out was the significance of the group of physically equivalent renormalization factors. But Ken Wilson applied that in a completely bold and mind-blowing way—

Zierler:

Yeah.

Witten:

—to study second order phase transitions and critical divergences.

Zierler:

What was his insight from your vantage point, sort of watching this happen before you? What was his insight that made this so mind-blowing?

Witten:

Well, we had thought that renormalization factors were some kind of dirty laundry of what you had to do to calculate quantum electrodynamics and the Compton effect and stuff like that. It was bookkeeping. It was not an appealing thing or an insightful thing at all. Then in his hands, you got to see that it was a way of abstracting what occurs on large scales from the basic source of it on the tiny scales, and that you could imagine going to such large scales that almost everything about the basic source didn't matter and you could predict a lot of things without knowing anything about the basic source. To have that leap of insight of what these fudge factors were really good for, what they were really telling you—that was mind-blowing. I mean, I was scrambling all through that course. The course was mostly full of post-docs, high energy post-docs, and they were kind of getting it and glomming on to it, and I just had my tongue hanging out trying to get some sense of it. But I knew enough to be blown away.

Zierler:

Yeah. What were some of the immediate theoretical implications of this advance?

Witten:

Well, the main thing was that it enabled you to predict the values of the so-called critical exponents that described e.g. the disappearance of the internal magnetic field as one heated a magnetic material. So, what Wilson cooked up was to figure out a way to make real calculations out of it by finding a small parameter in a perturbation procedure starting from a calculable theory and then extending it to the regime where the perturbation was small but nonzero. This was the famous epsilon expansion where the small parameter was the dimension of space minus four. His Phys Rev Letter is called “Critical exponents in 3.99 dimensions.” You might even know that paper.

Zierler:

I’ve heard it.

Witten:

Did you ever run across that paper?

Zierler:

Yeah.

Witten:

So, he was showing us on the blackboard how to study physics in 3.99 dimensions. But you asked me what were the immediate implications.

Zierler:

Right.

Witten:

The immediate implications were that in this domain of critical phenomena, in which measured quantities were diverging, these divergences seemed to be described by non-integer power laws. Wilson’s theory both explained these things and explained why the power laws had to be “universal.” Even though the exponents were such improbable values––not integers or simple fractions––they had to be the very same weird values in a whole continuum of systems as long as they had this divergence at all. So, this idea of universality in this very concrete and inescapable set of phenomena in real tabletop experiments—it was a hugely powerful synthesis and at the same time a hugely powerful thought process that would allow you to make the synthesis, a completely new way of thinking about it. So that’s what I would say was the impact.

Zierler:

As you were developing your own interests in soft matter physics, what were some of-- What was exciting about this to you personally, these developments? In other words, where could the field go where it might not have gone before as a result of this?

Witten:

So, I have to say that at that time I wasn’t thinking about this. I said, “I don't know if this has to do with soft matter or anything I’ve ever heard of. I just have to understand this.”

Zierler:

Yeah. It’s that exotic, essentially.

Witten:

It was that powerful. It was as though you were in 1920 or 1915 and somebody says, “You can turn space into time.” This was the reason I was doing this business. It wasn’t till my second postdoc that I began to be able to use the renormalization group. I had managed to wangle a second post-doc with one of those post-docs who was sitting in the audience for Wilson’s lectures Edouard Brézin—this was in Saclay, France. They were writing a big treatise to nail down the methodology and generalize the methodology of doing these renormalization groups in condensed matter. And people were seeing how renormalization explained phenomena beyond phase transitions. Another French scientist named DeGennes had come to the realization that you could study something that doesn't seem like a critical phenomenon at all—namely, long-chain polymer molecules—via the Renormalization Group. This had not dawned on me at all, but people at Saclay were applying the methods to polymers. I thought that would be a good place to push and find a new problem to do.

Zierler:

Just so I understand, your appointment was with the department of physics at Princeton, but you were mostly at the Institute?

Witten:

Yes, my appointment was in the department of physics. My advisor, my supervisor was John Hopfield, who I would see occasionally, but I never did that much with him. But my main intellectual involvement was with the experimentalists in the physics department and with Freeman Dyson, whose home base was the Institute, but who came to the physics department all the time.

Zierler:

I never tire of hearing Freeman Dyson stories. What was he like to you? What were your impressions of him at that stage in his career?

Witten:

Oh, he was wonderful. He had a really good rapport, both with me and with other people, because he was bigger and broader than the scope of physics allowed. He was teaching us a course in spin waves and the condensed matter things that he had discovered, but he would wax philosophical all the time. At the beginning, there were a handful of students in the course—there were a dozen or a half dozen graduate students. By the end there was just me and another post-doc, but he kept on with the class. We would just have the class in a conference room. He was telling after the class one day that he had been watching a play on TV. It was a Samuel Beckett play called Krapp’s Last Tape. Dyson said, “It was just so remarkable. This actor played this Krapp beautifully!” [laugh] Then he realized what he had said, and he lost it and burst out laughing and we all got into it with him.

Zierler:

[laugh] That’s great.

Witten:

That’s pretty good, huh?

Zierler:

Yeah!

Witten:

So, I kept up with him after that. I had another couple of encounters with Freeman Dyson, but mostly, after I left that post-doc, I didn't see too much of him. But I would say his letter got me my next job.

Zierler:

In Saclay.

Witten:

No. After Saclay. I was a year in Saclay with Brézin after Princeton. But then I came back to the US to try to get an academic job, a real job. I got a job at the University of Michigan, and Dyson’s letter got me that job.

Zierler:

Mm-hmm. But let’s go back to Saclay. How did that come together for you?

Witten:

It was… Let’s see. I wanted to study with the Renormalization Group in Saclay, and Brézin, a staff member at Saclay, thought enough of me to hire me, and he had post-doc positions to give. He had a duty to fill these posts with post-docs, and he thought I was good enough to be suitable. That’s how I got there, and Brézin was studying exactly what I wanted to study. It was a two-year post-doc, but I only stayed one year. When I got there, I saw that I wasn’t going to be helping my advisor, and I wasn’t even going to be able to see him very much because he and two other senior colleagues were working like crazy to develop this theory. I would do the same thing in their place, but I was sort of marooned. I had people to talk to, and the stuff where I was later to make my own mark was born there, and a lot of relationships and friendships that go on to this day were born there. Nevertheless, I saw that I wasn’t advancing in my career and that I was in danger of being forgotten in the States where I would need to get a job.

Zierler:

But you were there only for a year.

Witten:

I was there only for a year, and halfway through that year, when it got to be November, December, I said, “I’m going to have to go back and go job hunting.” Then I announced to the director that would leave after the year was over, per my postdoc contract.

Zierler:

And this was not a great job market at this point.

Witten:

No, it wasn’t. So, I didn't leave there because I had gotten another job. I left there because I was scared I wouldn't get another job, and I was feeling very much that this was a make-or-break moment and I wasn’t going to be able to manage it from France. So, the director was a little bit shocked. He said, “Is there something wrong? You're leaving after a year,” and I said, “Well, wasn’t that the deal? The appointment was for a year, and we would continue it on mutual agreement?” They allowed as how that was true. It wasn’t acrimonious or anything; I was just feeling like I wasn’t going to be able to continue if I went on as I was.

Zierler:

What were your impressions of Ann Arbor and Michigan when you first got there?

Witten:

I liked it. I liked it, and I felt a kinship with many people. I liked the atmosphere. But I didn't feel like it was uniformly welcoming. When they decided to hire me, it had been a contentious decision. It was the turn of the particle theorists to hire somebody their field, and there was a good candidate who later came on to do a strong career in Ann Arbor named Marty Einhorn. But they didn't hire Marty that year. I guess because of Dyson’s letter people thought this was an opportunity they should go for, they asked me to interview. This was during the days when the big physics meeting was in January, and so I had organized a series of flights to go to different places including this meeting. During the meeting Michigan asked me if I could rearrange my schedule and come back to France via Ann Arbor, which I did. Now this was a rocky time-- job hunting during that trip. But when I was in Ann Arbor, there were some people I kind of knew, and I wanted to see the labs. I had a great time talking to people, and so that probably helped.

Zierler:

Now in terms of the department, did you feel like you were filling a particular niche on the faculty in terms of the kind of work that you were doing?

Witten:

You mean whether I had sort of a portfolio that I was supposed to uphold in the department. Not so much, no. I mean, there was a group of condensed-matter theorists like me. There were two other theorists who kind of felt a natural affinity with me, and they kind of mentored me or wanted to be supportive of me. One of them became a close collaborator. This was Len Sander. We did a famous paper together. But apart from this kind of informal affinity with each other, I didn't feel like I was pigeonholed into some slot. In fact, I talked to a lot of experimentalists to try to see how their experiment worked, and they liked talking to me. I would rephrase to them what they were saying to me. They’d say, “I never thought of it that way. That’s a good way to ask the question.” So, I didn't feel like I was encapsulated with my group.

Zierler:

What kind of research projects were you undertaking at this time when you got to Michigan?

Witten:

My collaborator from Saclay, Lothar Schaefer and I were trying to push this renormalization business into polymers. So, the big thing I was trying to work on was a class of quantities that you’re supposed to be able to predict besides the exponents I mentioned above. You’re supposed to be able to predict ratios of the coefficients in front of those exponents. Let’s say you’re looking at a magnetic susceptibility. Then that susceptibility diverges as a power as one approaches the critical temperature from above, and there’s a number in front that tells the strength of the power law. That coefficient can be related to another coefficient describing this divergence below the critical temperature, and the ratio between those two coefficients is locked in place by universality. If it’s a particular class of phase transition, the ratio between those two coefficients has to have a certain value. Now, in polymer physics, there is no critical temperature. There’s no phase transition. In fact, there’s only one giant molecule, and yet there is one vestige of correspondence between this polymer case and the phase transition case: compared to the constituents, there is a length scale of the phenomenon that it can be arbitrarily bigger than those constituents. In the polymer case, length was the size of the whole polymer molecule compared to the size of a monomer. So we were trying to find some kind of analog of this critical amplitude ratio, as they call it, that would have predictive power and would tell you something about polymers that would not be at all obvious from just trying to read it across from the phase transition case. So, we found one of these things and we predicted a value for it and it sort of worked. That was the main thing I did when I was there. Then let’s see. This was my first three years. I was struggling with this problem. I was struggling with trying to teach undergraduates and trying to look for some further path to take this research along and not doing very well. So, then I went to see my buddy Lothar Schäfer, the German post-doc at Saclay with whom we had started the whole polymer project. He was still a junior scientist in Heidelberg. So, I went to visit him and I went to some seminars. One of the seminars was by the famous fractal man, Benoit Mandelbrot. I said, “This critical phenomena stuff and these geometric things that Mandelbrot is studying are the same thing.” I didn't quite see how, but I saw that there was something there. So, this got me on a new track which turned out to be fruitful, and that was to look for individual physical instances of the fractal property that Mandelbrot called “self-similarity.” Our polymers clearly had this property. I started looking for other things that might have this self-similarity in the lab. There was an experiment at Michigan that showed apparent self-similarity, in clusters of soot particles. We were trying to justify why these clusters might be fractals and how this could generalize the known properties of critical phenomena.

Zierler:

Did you take on graduate students right away at Michigan?

Witten:

Yes. Yes, I did. I had one graduate student only in those early days––Jeff Prentis. Jeff went on to get a post-doc with my PhD advisor. He made it in academia. He’s still a professor at University of Michigan at Dearborn. He’s probably retired now. But that was the only one.

Zierler:

What courses did you teach undergraduates at Michigan?

Witten:

We had courses for undergraduates with many, many sections, like freshman physics and I would mostly teach these sections. I wanted to concentrate my attention on getting something going in research, so I was not ambitious about what I was teaching undergraduates. I would just take a section or two of those discussion sections. Later, I was offered to give a big, advanced course on the renormalization group where I would try to present my own synthesis of the subject. I wanted to show why you didn't need critical points to have this formalism apply; you only needed the scale-invariant symmetry that the critical points have. So that wasn’t an undergraduate course at all, but it was a major course that I was struggling mightily to teach.

Zierler:

In what ways did the field advance with renormalization during your time at Michigan?

Witten:

This was a time when Wilson had figured out how to apply renormalization to the Kondo effect. The Kondo effect is some big, large-scale structure in electronic states in a certain kind of magnetic solid. Besides that was the stuff we were trying to do to make a bridge between the conventional physical chemistry that polymer physicists had done and the new renormalization approach. Then there was other stuff to do with polymers that were not especially renormalization. They were more primitive than renormalization, but they shared the feature that you could say something universal about polymers because there were so many random subunits in each polymer. It was a little bit like the premise of statistical mechanics. If you have arbitrarily many independent degrees of freedom and you just know that they are some kind of maximally disordered or maximally random, that property alone can give you robust predictions. So, as with traditional stat physics problems like ideal gases, so too with indefinitely large molecules, one could gain insight by imagining the situation where the number of degrees of freedom had grown arbitrarily large. One needed the renormalization group to determine the exponents. But even without knowing the values of these exponents, one could deduce how the functional relationship between different measured quantities must simplify when the system grew arbitrarily large. For example, a polymer’s size inferred by x-ray diffraction and its size inferred by its diffusion constant had to have a ratio that remained constant however huge this size becomes. This fixed ratio must then be independent of what the polymer is made of. It is universal. This was deeply appreciated by Pierre de Gennes whom I mentioned above. He was trying to play the polymer phenomena without so much emphasis on the formalism of renormalization. I was coming to recognize the power of this “scaling” approach.

Zierler:

Yes, exactly. Tom, how did Exxon come about for you?

Witten:

So as you have gathered, my success at University of Michigan was not the smoothest, and when it came to be tenure time, the famous discovery that Len Sander and I had made linking aggregation processes like clustering of soot particles with fractal structure was only just emerging. Its significance was not known, and it was regarded with some skepticism in many quarters. So, I didn't look like a good bet for tenure, and I didn't get tenure. So, I had to find another job. Fortunately, I had this emerging discovery about aggregation which looked kind of exciting, but unfortunately, I was not in a good position to be on the street, so to speak.

Zierler:

Tom, I’m curious if even before you got the writing on the wall about the tenure decision, if you were thinking about possible industrial applications for your research, if that is something that occurred to you even before you thought about Exxon as a viable next move.

Witten:

I was not actively thinking about industrial applications to pursue. I was keenly aware of broader scientific applications than just critical phenomena. So, we had a handle of understanding things about polymers that others did not have—and polymers were, of course, important in industry. You could even learn that by watching movies like The Graduate! [laugh] Let’s see. So, I had an appreciation that there ought to be some value that I could add in an industrial setting, and I was keen to do it insofar as it was possible. I wanted to make it have a real-world impact. I wanted it to have a real, concrete impact in people’s thinking. It wasn’t that I had to find a new job and therefore this was my big opportunity to get a job in industry. I wanted to have a new job where I could do physics if at all possible, and industry would be just fine if that’s the way the cookie crumbled. So that’s the way it did crumble.

Zierler:

Now with corporate research labs, Bell being the gold standard in terms of being a place that supported basic science with very little regard for how it might affect the bottom dollar—on that spectrum, where was your understanding of Exxon at that time? Was it a place where you could pursue just pure basic science, or was there always an expectation that the research needed to be in some way related to the profit of the company?

Witten:

When I was hired there, they were at pains to be as little tied down as possible by the practical needs of the company. They wanted to be the Bell Labs of energy, and the vice president for research had come from Bell Labs. This man was the managerial father of the Unix computer operating system.

Zierler:

Who was this?

Witten:

A man named Ed David. So, they were bound and determined to be a research-driven and publication-driven and scientific-impact-driven place and resisting the pressure to be relevance-driven.

Zierler:

There were some really good people at the lab by the time you got there.

Witten:

Oh, yes. Yeah, there were really good people. Who can I name? A father of hydrocarbon catalysis, John Sinfelt, was there. Another prominent person in the scientific world was this guy who did molecular clusters: Andy Kaldor. He wasn’t in my field. He was not far from my office. And there was a big guy in scattering from Bell Labs called Peter Eisenberg. And then David Weitz was there, who has gone on to be very well-known. David Pine also was there.

Zierler:

Harry Deckman—was he there at that time?

Witten:

Ah, you know Harry Deckman.

Zierler:

Yeah.

Witten:

How do you know him?

Zierler:

I’m speaking to him next week!

Witten:

All right! Yeah, he was right down the hall from me! He was our mad scientist. He was full of neat ideas and awash in incomplete thoughts. He was one of these people who had a cluttered office that if you walked sideways, you could just about get into it. Right in the center of things. Remember him to me when you see him.

Zierler:

I will. I will.

Witten:

I haven't seen him in a long time, but he was at the heart of a major project to capture natural gas deep in the ocean floor of the Pacific in Indonesia, and another thing with the oil sands of Canada. So, he’s quite a guy. I’m tickled that you're interviewing him.

Zierler:

[laugh] Tom, as you got settled at Exxon, how well was the reality-- How well did that accord with the ideal of Exxon wanting to be, as you say, the Bell Labs of energy, to be a place of pure research, basic science?

Witten:

I think they succeeded well on that. They did have to have a change of course because this idealistic guy, Ed David, got fired. The corporate budget (for the corporate research lab, I mean), got a third cut off of it, which meant they had to cut half of the staff because there were a lot of fixed costs that could not be cut. After that there was a sea change where you were supposed to make friends with somebody in the more practical part of the company and try to help them. This was not a drawback to me. It was a boon to me. The engineers were so fearless and daring. You know, they weren't afraid of the boss because they could really make something happen that the bosses needed. So, it was wonderful talking to those guys and trying to hash things out and make some difference in their problem. I really liked that. But what I didn't like about Exxon is that it became clear that as long as I was in Exxon, I would be a hired hand.

Zierler:

Yeah.

Witten:

I came to realize that I didn't want to be like that forever. Then after they had this downsizing and a lot of my buddies had to leave just because they were too little coupled to the practical side, I said, “Well, I’m going to keep my eyes open and see what I can find.”

Zierler:

Yeah. So, what were you able to accomplish on the research side at Exxon? In what ways did you capitalize on the opportunities both in soft matter and coming from this basic science ideal, at least from the beginning?

Witten:

One of them was about polymers. It was about polymers that had sticky places on them. You can design polymers that are uniform, and you can design polymers that are modulated and you can design ones that have these discrete sticky places on them. What I mean by “sticky” is they attract each other. They can make some structure in the polymer that’s not usually there, and we found out that just by changing the placement of these sticky places, or changing the statistical distribution of them, you could go from an overall macroscopic interaction which was self-repelling so that it would try to get bigger because of its repulsion. Or just by tweaking where the little stickers were, you could make it go continuously from self-repelling to self-attracting. So that was a real solid piece of science. Exxon didn't exploit it and get patents on it and stuff like that, but it really gave them an insight of what kind of power you had when you could control a polymer that way. The other thing was a thing called a polymer brush. So if you stick polymers to a surface, these polymers make a buffer between that surface and the outside and it’s helpful for a lot of things. But it’s possible to make polymers that stick to a surface at one end so strongly that there’s not room for them all and they’re obliged to stretch out in order to make room for their friends. We found out that the environment of that stretched state was a unique environment that people didn't know about and that you could control by controlling the chemistry and the different chain lengths of the chains that were attached. You could control the environment by introducing branches in the polymer chains, and so forth. So that was a big impact that was relevant to Exxon, and Exxon recognized that relevance. The other thing we did was more of a sociological or cultural nature that we had these little Friday lunchtime gab sessions. These would mostly be about an experiment, a result someone had gotten in the lab that didn’t make sense. So, they were stuck; they didn't know what to make of it. It was just a frustrating puzzle. We would hash out the puzzle and people would give all kinds of crazy ideas. We did it every week and everybody liked it. It became kind of a touchstone of cooperativity, you know, of doing science together.

Zierler:

Tom, how well connected were you to the academic physics world during your time at Exxon? Were you collaborating? Were you presenting at conferences? Were you writing articles?

Witten:

Yes, I was doing all those things. As far as collaborating, I was mostly collaborating with the people at Exxon. But was I getting invited to a lot of talks? More than I had been, but I wasn’t enmeshed with the academic world at all. I knew academics. We had a lot of academics coming to us and giving seminars, and you know, I interacted with them around that. But were there a lot of coauthored papers? There was one with Sidney Nagel. So, Sidney was an old friend from the time when I was a post-doc at Princeton. He was a graduate student there who became an assistant professor at the University of Chicago. Sid would tell me on the phone about some puzzles that he had discovered, and I would try to help him figure it out and sometimes I got to be a coauthor of the paper. Another one was Phil Pincus. He had been at Exxon and then he had left. Sometimes he would call me up and say, “Here’s something funny to think about,” and that would sometimes lead to a paper, too.

Zierler:

At what point did you read the writing on the wall and see that it was time to leave Exxon? When did that happen?

Witten:

Well, the big downsizing came in 1986, and that was when I realized that I should find options other than staying at Exxon. I was there for three more years and having a happy time and doing exciting things, but with the feeling that the best times had been behind us.

Zierler:

Did you know specifically that you wanted to reenter academia?

Witten:

Yeah. Definitely.

Zierler:

What was available to you? What were your options at that point?

Witten:

I have to say I was not really beating the bushes. I was kind of keeping my eyes open, but the bushes were beating me a little bit. One place that was interested in me was the University of Florida, and they were looking at me for a full professor position. They made me an offer and invited me to come down. I got to talk to the guys—it was the theory group, and they were mostly particle theorists. I realized I kind of admired their work from a distance, an ignorant distance, but I didn't really have that much to contribute to them. It didn't look like a great prospect to me, so I didn't take the job. Then later this University of Chicago job came up, and there was nothing wrong with that at all. It was a wonderful opportunity, so I took it.

Zierler:

What was the offer? Was this a tenured offer?

Witten:

Yeah, so it was a full professor offer.

Zierler:

Which clearly was in recognition of your record of scholarship while you were at Exxon.

Witten:

Yes!

Zierler:

In other words, this wasn’t just like you were picking right up from being an assistant professor and then you were off in the wilderness.

Witten:

No. I mean, I was able to be scientifically prominent at Exxon and go into whole different fields that, given my record earlier, would have been unthinkable. It was good for me internally. If I had stayed at Michigan, I would have never been that person. So, I often said the luckiest thing that happened to me was to get fired from Michigan.

Zierler:

[laugh] Tom, I wonder if you could reflect. Of course, you were prominent in the field, you were active in the field, but perhaps when you got to Chicago, it clarified your appreciation for some of the advances that were happening in the field from an academic perspective. What were those kinds of things that might have been apparent to you or things that you thought about in a way that you might not have within the research environment at Exxon?

Witten:

Yeah. In other words, how did it grow my perspective to be in this rich university atmosphere? I really do have the feeling that it did grow, but I have to say the dramatic growth in perspective came at Exxon. There were more things to learn and grow from when I came to Chicago. So now almost all my research was motivated by the experiments of my buddies here and outside of Exxon. The main discovery that I made while here, is this business of thin sheet singularities, and how did that happen? It didn't happen because of research at Chicago. Have you ever heard of a buckyball?

Zierler:

Yeah. Sure.

Witten:

Did you know that buckyballs come in all sizes? There’s the well-known sixty-carbon buckyball, but if you grow the buckyballs in the right conditions, you can get great big ones that have hundreds and hundreds of carbons in them and they’re still roughly spherical. So, I got hooked on the question of what would be their shape if they were really big, because we know that the molecular structure of them requires there to be a certain number of fivefold rings instead of the majority of them that are sixfold rings. I said, “Well, what do those fivefold rings do to the sphere? Do they just make a little, local pointy place on it or a more far-reaching deformation? What do they do?” These points turned out to be the thing that dominates the whole shape. They dominate the energy of the structure. So we went through some labyrinthine process that led to that result and we were shocked and amazed. I’ve been working different aspects of that ever since. So that was the single biggest thing that I did while I was here. I definitely benefited from the environment, the supportive part of it, but I didn't get the idea from here. Much of my work now concerns pattern formation that comes from capillary flow. This idea did originate at Chicago. It started from a puzzle posed by Sid: “Why when a drop of solution dries in the surface, do all the solute molecules end up at the rim instead of being dispersed uniformly as they were initially?” Another big subject I worked on is the distinctive properties of granular solids. They are not like ordinary solids. They hold their shape but the way they deform when forced is qualitatively different from ordinary solids. Something I’m very fond of that did come from Sidney’s lab was—it didn't make a very big impact, but I think it’s still a very neat thing. It’s called a myelin figure. It’s a self-assembled worm-like structure that comes when you mix lipid soap powder and water. These worms spontaneously grow that are hundreds [of] times the size of a lipid molecule. We figured out how that could be. Here’s another way the world opened up more for me at Chicago.

Zierler:

Where is renormalization at this point? How had this continued to influence the field?

Witten:

Let’s see. It was a-- it became sort of a background that you accept. Like in the ‘20s, people figured out how to accommodate the rotation symmetry of basic atomic interactions with quantum mechanics to say how they would constrain a wave function and that you’ve got to have these quantum numbers that depended on angular momentum. You’ve got to have rules that could tell you how to combine this angular momentum and make other kind of states. Some states you could not combine; some states you could. There were theorems about this. It was a very active field for a while, and then it was just an accepted part of the landscape of the discipline. That’s kind of the state that renormalization is at now.

Zierler:

Can you talk a little bit about the relationship between the physics department and the James Franck Institute?

Witten:

Yes. That’s a point of pride for this place.

Zierler:

Yeah.

Witten:

And it’s a point of pride with a resonance with Exxon. So go back to the Manhattan Project. The James Franck Institute was really inspired by the Manhattan Project in the following sense. Chicago was a place whose people were steeped in and recognized by their accomplishments in an academic discipline. There was a body of experts who would validate your accomplishment in that discipline—for example, atomic spectroscopy or quantum mechanics or whatever—and this was a powerful engine for discovery. But then with the Manhattan Project, they had certain urgent needs, and this wasn’t going to just happen by itself in academia like the early discoveries such as nuclear fission had done. Instead, they had to really combine the expertise of complementary disciplines to make the Manhattan Project happen. They realized that this jumpstarted the intellectual growth and power of those people who had been in their own discipline for all this time. Physicists were learning stuff from the chemists. The astrophysicists were learning something from the condensed matter physicists, and nuclear physics and the metallurgists, and so on. So, they realized that they would like to put that genie in a bottle, and that’s how the Institute came to be founded. Other wartime institutes were founded in this same spirit, like the Lincoln Lab at MIT. But with that motivation and that inspiration, it was realized that the best way, the best environment for an academic discipline would be in the context of a cross-structure. So here are the academic disciplines. [stretches out the fingers of one hand vertically] Then here [overlays the fingers of the other hand horizontally] are the interdisciplinary institutes based on a certain body of phenomena such as atomic energy or such as metals, including chemists and engineers. Then by having this blurry connection where you had allegiance to your discipline but also allegiance to your interdisciplinary body of phenomena, you could not only discover things better, but also you could train students better. So, we still go by that philosophy.

Zierler:

And in terms of intellectual environment, not necessarily where you spent your time, but in terms of collaborations, post-docs, working with graduate students, what has been sort of more your home, the Institute or the department, or is it really sort of like a 50/50 kind of arrangement?

Witten:

It’s hard to say. It’s not 50/50; it’s 100/100! The art of this is to believe in two contradictory realities at the same time. It’s like some piece of music where you have two themes going, or you have a picture that you can look at as a woman at a dressing table or as a skull. So, your mind just shifts a little bit and you see a completely different thing. The coexistence of the Institute mindset and the departmental mindset needs that same tension, and I would be hard-pressed to say whether I was in one mindset more than the other.

Zierler:

And that worked to your advantage, obviously.

Witten:

Yes.

Zierler:

Did you find at Chicago that you were… I guess what I’m trying to get at is the way that Exxon may have changed the kinds of physics you wanted to do and teach, and perhaps an easy barometer of that would be the kinds of classes at Chicago that you sort of wanted to teach or were expected to teach.

Witten:

Yeah. I got to blossom out in what could be taught and what I would have to bring to awaken people. That’s where our book came from. So, I had an idea of how to think about polymers and critical phenomena on the same footing and some other ideas like that. At the same time, Fyl Pincus had similar kind of thinking that both of us somewhat got from de Gennes, but it was a way of stripping to the essentials the common features of a whole bunch of soft matter phenomena. There would be some principles that emerged that would allow you to take on a whole range of those phenomena. So, Fyl called me and he said, “Let’s write a book.” I said, “Okay, you're on. We have a great premise. We can write this book.” So, I started teaching classes in the soft matter phenomena, and class after class I would grow the notes in the direction of a book. Fyl was working to some extent on it too, but he wasn’t regularly teaching classes like I was, so I was making more inroads to this book thing. But that’s really an example of what you're saying, that I could bring this body of thinking that is now called soft condensed matter into more of a discipline.

Zierler:

Yeah.

Witten:

More of a coherent body of work.

Zierler:

So at this time, the idea of there being a discrete discipline, soft condensed matter, from a theoretical perspective this was still very much an intellectual pursuit in the making.

Witten:

Yes. That’s right.

Zierler:

Why would that be? I mean, soft matter as stuff had been around for a long time.

Witten:

Yes! There was a polymer science ever since the ‘30s. People were cooking up these tar-like goops in their labs. And there were lasting scientific understandings coming from that. But I would say polymer science was a pre-existing discipline as you implied. There was some kind of a discipline in colloid science as well, dating back from Derjaguin in the early 20th century. But I guess if I would try to say what changed, it’s that these different forms of matter were not thought of as being part of one discipline.

Zierler:

So, what changed? I mean, the fact that there has been polymer science since the ‘30s—what had changed intellectually, technologically, theoretically that had allowed for this thing that we now understand to be theoretical soft matter physics to become sort of its own discipline distinct from its predecessors?

Witten:

Here’s what I would say. You're good at posing these sharp questions that I haven't explicitly thought about, but now that we have been talking this long and I’ve realized what I’ve been saying. I would say the thing that jumpstarted it and made it go with this extra bump of kind of unity or synergy is Wilson and de Gennes’ insight that the same principles that made universality happen in critical phenomena worked also with polymers: you had the right to expect analogous things to happen throughout these phenomena where you can control some length scale where you're looking at phenomena to be arbitrarily bigger than the constituent units.

Zierler:

That’s a great answer!

Witten:

I’d buy it. [laugh] But if I think about that, I’d probably say, “No, that’s not right.” But at the moment, as I think about it, I really think that’s it, that you had the right to expect universality.

Zierler:

You had the right. What do you mean by that?

Witten:

There was every basis-- as much rationale to expect it for this broader set of phenomena as there was to expect it for the well-established for the critical phenomena. The rationale was that if you are exploring phenomena at arbitrarily larger scale than the particles causing it, the specific details of those particles must become irrelevant.

Zierler:

Yeah.

Witten:

What can I say analogously where you have the right to expect… You grow to realize that you have the right to expect something? Well, in the turn of the 20th century, Einstein would say, “We have the right to expect nature to play fair.” He would say stuff like that. “Nature is subtle, but she is not malicious.” If you had said that in the 16th century, eyebrows would have been raised, you know?

Zierler:

Sure.

Witten:

It wasn’t so clear. So that’s what I kind of think it would be.

Zierler:

Tom, just to take that same kind of question, not from a retrospective perspective, but from when you got to Chicago to the present, in what ways has theoretical soft matter physics gone from this fledgling intellectual proposition to be what it is today, of course, which is a card-carrying member of the larger physics community as we understand it to be divided by subfields?

Witten:

It is a real thing.

Zierler:

Right!

Witten:

It’s a real thing.

Zierler:

It’s a real thing! So what’s the transition or the narrative that got from this early propositional point in the history of the subfield to, as you say, the real thing that it is today?

Witten:

[Pause] Let’s see.

Zierler:

And specifically, maybe, where does experimentation play its part and where does theory play its part in this development.

Witten:

Well, let me think of a few things. Did you happen to interview David Weitz?

Zierler:

It was my great pleasure. Yes.

Witten:

So, if you think of people like David Weitz or David Pine or a guy in Minnesota called-- a polymer guy—Frank Bates, they were demonstrating year after year and paper after paper that they could control the individual constituents of a polymer solution or a colloidal dispersion—the things on the invisibly small scale—and make dramatic and predicted shapeable and tailorable properties on the macroscopic scale. A big example I didn't even mention yet is block copolymers. These are chain molecules that have alternating segments of two or more types of monomer. By changing segment lengths and positions in the chain, you can create a fluid that structures itself into precise periodic interlinking patterns. You see one coup like that after another, and still, you might say, “This isn't as impressive as the transistor.” But then it would start to have practical implications like controlled rheology of tires or multi-layer ketchup bottles and stuff like that. It was a drumbeat of those demonstrations of control and power over the material that made soft matter a real thing.

Zierler:

Also a great answer.

Witten:

[laugh] Okay! Again, having said it, I kind of get seduced by my own words and I believe it. So I’ll write you an email if I have to take anything back. [laugh]

Zierler:

Tom, who have been some of your most successful graduate students from Chicago?

Witten:

Okay. The one student who did the most impressive work with me was a guy named Alex Lobkovsky. Alex wanted to be his own man and live life on his own terms, and so he hasn’t made a huge mark in the world after that. But I still count him as the graduate student who made the biggest impact as a graduate student. Another one, I would say… If you let me count post-docs… Will you let me count post-docs?

Zierler:

Of course.

Witten:

John Marko comes to mind. He went on to do experimental work on DNA, and now he’s one of the main figures who are understanding how DNA organizes and compacts itself when it’s got to be jammed all together inside of the cell nucleus. You can imagine what a mess it would be if that were just a hopeless tangle like a ball of yarn. So, John I count very highly. He was a post-doc with us at Chicago. There’s also a guy at UC, San Francisco who is a very deep molecular biologist figuring out the statistics of things like PCR and decoding the human genome and like that. But I wouldn’t say, if I compare myself with other physicists, that my main impact was to produce great students and great post-docs. I’d love to take credit for Michael Cates, now the successor to Newton’s chair at Cambridge. I did work with him when he was a student, but I don't get to take any credit. The things he did were all him. Matthieu Wyart found his calling while he was visiting Nagel and me at Chicago as a student. He made a deep insight about the puzzle I mentioned above: how can a granular pack be solid yet not have the properties of a solid, like a stress field? He has continued to build on that insight to the present day. Who else? You have in mind some people that I might have chosen to mention, don't you?

Zierler:

Probably. [laugh]

Witten:

No, it’s on your screen. Just look a little bit to the right. [laugh] Let’s see. You know, I’ll think of half a dozen people as soon as we get off, and then I’ll tell you what--

Zierler:

[laugh] We can always go back and edit the transcript. That’s what goes into the archives. That’s fine.

Witten:

Okay.

Zierler:

Tom, to get back to you, over the past 20 years, can you reflect on some of your own significant research achievements? What have you been able to do, both during your time when you were full-time on the faculty, and, you know, physicists never retire. You're certainly no exception to the rule. Since you went emeritus in 2013, what have been, in broad brushstrokes, some of the most significant things that you’ve been involved in?

Witten:

Okay. There are three things. One of them is about granular material. One of them is about thin sheets, and one of them is about driven rotational response of colloids. So I’ll start with the granular material. Back a decade or so, a Matthieu Wyart came to work with me and Sid Nagel, He wanted to work at Chicago because he was puzzled and interested about granular material and how the granular material crosses the threshold from flowing to being solid or jammed—it seemed to happen continuously. You could make this happen in simulations and you could compact your granular mass together and with hardly any free space, yet the particle could still move like a fluid. But then as you squeezed further, there would come a point where they couldn't move. People had tried different ways of understanding this, but in the experiment, in the simulations, it was too continuous compared to these other pictures and it didn't fit with the other pictures. This student Matthieu came to us, and he figured out a way to make an implication between that threshold mutual forces between the particles that would happen right at that threshold. It would be different from the elastic forces of any ordinary solids just because it was at the threshold. It was an amazing thing that that he did, and Sid and I were sort of the midwife of it. So that was one big thing, and that led to several other things that we did. The most recent one that I’m kind of really excited about has to do with this act of cyclic annealing. Let me explain that. You have the box of granular material that’s all disordered, and you start doing this to it. [horizontal hands one spaced above the other, hands moving left and right in a shearing motion, then back to the center] Every time you do this, the little grains tumble over each other and they get into a denser situation. Then after a while, the thing stops changing with time and does more or less the same thing every time you go back and forth. What it does in each cycle of going back and forth a motion made of irreversible steps. The particles tumble over each other, and they get into a new configuration. If you would try to go back, they wouldn't tumble back the way they were. They have done something irrevocable, again and again as you keep pushing. You have dozens of these tumbling events as you go through a shearing cycle. What happens, strangely, is that after a while, you look at the sequence of tumbling events and you look at that same sequence the next time you do the cycle and it’s thee sequence. You can do this on the computer where you know exactly where every particle is and exactly how it moved during one of these irreversible events. Each particle moves through the exact same sequence of tumbles. We’re trying to understand how that could come to be, that they would be able to do the very same sequence and at the same time having each sequence full of irreversible changes. That’s the granular project. As far as the thin sheet project, it’s a little bit harder to explain and more technical, but there’s… When you start to crumple a sheet, one of the things you make is a point-like thing like this. I’m going to move. See this point-like thing? [bends a sheet of paper so that it makes a buckled conical shape]

Zierler:

Yeah.

Witten:

Now suppose you ask yourself, how tightly is this wrapped? How small is this thing? [points to the top of the cone] Anyway, the distance between this finger [at the cone tip] and this finger [at the boundary of the paper] is [roughly] the size of the sheet if you measure a place [?], and another length scale is the thickness of the paper. So, which is it? Now look at the tip here. Let’s see if I can… Can you see it?

Zierler:

Yeah.

Witten:

You can see it has a distinct size there—this little crescent is far bigger than the thickness and also far smaller than the whole cone, and the question is how did that crescent know how big to be? This has been a puzzle in this field for over a decade because it’s neither a fixed multiple of this thickness, nor is it a fixed multiple of this size. It’s an intermediate length scale that’s much bigger than the one length and much smaller than the other. So we’re trying to figure out where that length scale comes from and how to account for it, and we have a new angle of how to do that. My graduate student is just finishing his thesis on that, and we’re excited about where that can lead. It’s just an early step. It’s clearly a whole new way of thinking about it. It’s not all the way to explaining that size scale. Then a third project is the colloidal rotational project. Suppose you have particles that are not round but they’re some asymmetric shape—and often in biophysics, whatever made them have a funny shape made them all have the same funny shape. So, each one of them could respond the same way to some external forcing, but the only thing that stops them from responding the same way is they’re in different orientations and their response depends on the orientation. So, what we’ve been working on and looking at is how to organize these things so that they do respond the same, and into the bargain they all give the same orientation. This can be done by forcing them in a time-dependent way. So this is a new kind of way of creating order that people hadn't understood. You know, it wasn’t a field. It wasn’t a problem that people were trying to solve, but there are principles behind this and so we’ve been working on these principles. Now there’s a big review article that is in the proof stage that explains the ins and outs of this whole business. So those are the main things, and each of them has the potential to grow. They’re not-- There’s a pretty good future in all of them, so I’m hopeful.

Zierler:

Going back to your work at Exxon, in what ways have you remained sort of connected with some of the industry applicability of this work?

Witten:

My industry experience is a repository of things that I know you could do, that I know you could exploit. Let’s say, if you could image stress—what kind of materials could you make, or if you could control the relaxation of stress in some material point by point, what could you make happen? So, I would say it was a really good library of potential things I see could be done, but as far as how much fruit this library has borne… Let’s see. It hasn’t borne that much fruit. I’ve looked into working out applications to especially the rotating colloid project, but this has not led to anything. But you asked me how does having been connected with Exxon inform what I see as being able to be done in science, and I would say it’s with me all the time. My knowledge of Exxon phenomena informs the questions I might ask at a seminar, for example. It brings up parallels of the speaker’s experiment that might not be recognized.

Zierler:

Tom, how much has technology influenced the theoretical aspect of the field?

Witten:

Do you mean in the means of doing the theory?

Zierler:

Correct. The way that experimentation might confirm the theory or the way that experimentation might cause new avenues of theoretical pursuit.

Witten:

The main thing is fast video. So, you can look at a video picture of, let’s say, actin entanglement, actin fibers that are all driven and pushed past each other by molecular motors, an extract gotten from cells. It makes a complicated swirling pattern. It’s amazing to look at, but you can't make sense out of it just from saying, “Oh, that looks amazing.” But with detailed image analysis, you can extract all kinds of quantitative things that you can go back and try to predict and verify and so on. So the world of validating theory by fast video analysis, and computer simulation analysis is powerful. As I was kind of mentioning before, you can try to get to the nub of something and say, “If that’s what’s important, I can simulate it in a simple way,” and then see if that nub was sufficient to explain the real system that you're trying to understand. Those are the main things. Those are really very much field-changing.

Zierler:

Well, Tom, at this point, for the last part of our discussion, I want to ask a broadly retrospective question and then one that looks to the future. Do you see for your career any particular research that you’ve been involved with that is most significant, either in terms of its impact on theory, in terms of its impact on the way that it spurred additional research, or even the way that it might have had impact in terms of applications? What are the things that you’ve been involved with that really stand out in your memory as having that greatest impact?

Witten:

Okay. So, I had just a few big hits. One was this diffusion-limited growth, a finding that happened at Michigan. One was the polymer brush thing that I told you about where the polymers are stuck to a surface at one end, which gives the basis of a lot of phenomena. And one was the thin sheet. That crumpling was a thing you could explain, and it made materials you could understand on a universal basis. Now, what would be-- If I were to try to gauge the impact of each of those… Let’s just take them in turn. The first one kind of established the idea that a kinetic process could obey the same kinds of universality that had been recognized in more tame phenomena. It was fascinating to the field for a while and now it’s on the back burner.

Zierler:

Yeah.

Witten:

It’s not on the back burner because it became understood like the angular momentum rules I told you about earlier. It’s not like that. It just came to a dead end because people didn't know what to do! There’s some important principle that whose potential impact is unclear, but it’s kind of an unfinished part of our science that needs to be finished someday. I wouldn't have said that’s a big impact, but it’s kind of a known bone in our craw. To be an instigator of that bone in our craw, that means something and that can have an impact down the line. The second one was with the polymers. I think that will-- That’s been kind of assimilated into the daily life of people using polymers. Nowadays people can make polymers much more flexibly. They can harness cells to make polymers for them in very explicit structures. So I think this facility, with being able to make what you want in tailored polymers to get materials that you want, was being done by many people. The stuff we did was kind of one piece of one chapter of the textbook of that whole field. So that’s a lasting impact, but it’s one of many in its field. As far as the—what else did I say? —the crumpling thing, that’s kind of a niche subject. People are interested. There’s this subfield called extreme mechanics that’s a niche subject where people are trying to get this particular controllability inherent in a thin sheet and the ability to like, change the topology of a thin sheet by kind of just doing that, [twists a sheet into a rigid, structure with conical vertices and a narrow ridge joining them] like I just did that. I made this topological structure that makes it a cone, and I can monkey with this in a lot of different ways. The macroscopic properties, material properties that you can put into this material by knowing what a thing like this [points to the joining ridge] will do to it, that’s going to be a nice thing that people control more and more. It’s going to be a kind of augmented science of origami to supplement the preexisting science of origami, and the long preexisting art of origami. This is a different kind of thing because it’s not all geometry; it’s somewhat mechanics. So that’s going to be a nice thing, I think.

Zierler:

Tom, looking to the future for my last question, right, now that the field is mature and you’ve been involved with it so long, using your powers of extrapolation, both in terms of the future possibilities of technology, the future theoretical implications, the future possibilities of continuing to look for ways that this will have societal implications, where do you see all these things headed generally and where do you see yourself specifically within these developments?

Witten:

I have to ask you did it happen that you ever interviewed Gordy Kane from Michigan?

Zierler:

No.

Witten:

Do you know who he is?

Zierler:

No.

Witten:

He’s a particle theorist, a sort of elder statesman of particle theory, and I overlapped with him. He’s a neat guy. But later, at the millennium, at 2000 the APS held a contest to predict-- to write a news story, an annual review kind of retrospective news story from the perspective of 2020. Gordy Kane wrote one and it was so optimistic and so rosy. It just was very imaginative and very heart-warming, and I am honored to be asked to do that for you. [laugh] You could say… I don't feel full of visions of a whole different world—only glimpses. I think we just, in the last couple of decades, have a blossoming of a whole different world, and we have that to assimilate and consolidate. We have the world made possible by large-scale computing, by unprecedented ability to interact with each other, and then the possibility of manipulating quantum states so that all of their phase information is harnessed and made into qubits or who knows what else. Those things are, to my mind, the emerging things. So in summary, I don't claim any special originality or insight about those things. I think most people would say the same thing. As far as dark visions of whether this whole thing would get eclipsed by some breakdown of civilization or some—what’s the word? —disillusionment with science, I have never gotten so pessimistic as to believe that. But it could really happen.

Zierler:

Something to stay tuned to for sure.

Witten:

Indeed, and we will.

Zierler:

Tom, it’s been great spending this time with you. Thank you so much for doing this with me.

Witten:

It was great on my side, too, David.

Zierler:

Excellent.

Witten:

You're a good interviewer. You made me think, and you probably made me want to follow up with you on some things.

Zierler:

I’ll welcome that any time.