Notice: We are in the process of migrating Oral History Interview metadata to this new version of our website.
During this migration, the following fields associated with interviews may be incomplete: Institutions, Additional Persons, and Subjects. Our Browse Subjects feature is also affected by this migration.
We encourage researchers to utilize the full-text search on this page to navigate our oral histories or to use our catalog to locate oral history interviews by keyword.
Please contact [email protected] with any feedback.
Credit: Sara L. Press
This transcript may not be quoted, reproduced or redistributed in whole or in part by any means except with the written permission of the American Institute of Physics.
This transcript is based on a tape-recorded interview deposited at the Center for History of Physics of the American Institute of Physics. The AIP's interviews have generally been transcribed from tape, edited by the interviewer for clarity, and then further edited by the interviewee. If this interview is important to you, you should consult earlier versions of the transcript or listen to the original tape. For many interviews, the AIP retains substantial files with further information about the interviewee and the interview itself. Please contact us for information about accessing these materials.
Please bear in mind that: 1) This material is a transcript of the spoken word rather than a literary product; 2) An interview must be read with the awareness that different people's memories about an event will often differ, and that memories can change with time for many reasons including subsequent experiences, interactions with others, and one's feelings about an event. Disclaimer: This transcript was scanned from a typescript, introducing occasional spelling errors. The original typescript is available.
In footnotes or endnotes please cite AIP interviews like this:
Interview of William H. Press by David Zierler on July 28, August 18, September 4 & 11, 2020,
Niels Bohr Library & Archives, American Institute of Physics,
College Park, MD USA,
For multiple citations, "AIP" is the preferred abbreviation for the location.
Interview with William H. Press, Leslie Suringer Professor in Computer Science and Integrative Biology at the University of Texas at Austin. Press recounts his childhood in Pasadena and the influence of his father Frank Press, who was a prominent geophysicist, Caltech professor, and who would become science advisor to President Jimmy Carter. He describes the impact of Sputnik on his budding interests in science, and he discusses his undergraduate experience at Harvard, where Dan Kleppner, Norman Ramsey, Ed Purcell and Dick McCray were influential in his development, and where he realized he had an aptitude for applying abstract equations to understanding physical reality. Press describes trying his hand with experimentation in Gerald Holton’s high-pressure physics lab, he recounts his involvement in student activism in the late 1960s, and he discusses his involvement in computer hacking in its earliest form. He explains his decision to attend Caltech for graduate school and his interest in studying with Dick Feynman and Kip Thorne. Press describes the opportunity leading to his work at Lawrence Livermore, how he got involved with Thorne’s group of mathematical general relativists, the origins of Thorne’s work on gravitational waves, and his collaborations with Saul Teukolsky and Paul Schechter. He describes the formative influence of Chandrasekhar. Press discusses his first faculty position at Princeton where he joined John Wheeler’s relativity group, and he describes his research interests flowing more toward astrophysics. He explains the opportunities leading to his tenure at Harvard, where he was given separate appointments in physics and astronomy and where he founded theoretical astrophysics within the Center for Astrophysics. Press describes his entrée into science policy work in Washington with the NSF Physics Advisory Committee and then later on the National Academy of Science and the National Research Council, and he explains the origins of his long-term association with the JASON Study Group. He describes his interest in gravitational collapse, Ia supernovae and galaxy formation, and why the study of black holes reinvigorated the field of general relativity. Press describes the singular genius of Freeman Dyson, and he recounts his contributions to nuclear risk reduction in science policy and his service with the Defense Science Board and the Institute for Defense Analyses. He discusses his tenure as chair in Harvard’s Department of Astronomy, his experience with the Numerical Recipes books, and his collaboration with Adam Riess and Robert Kirshner. Press recounts his decision take a position at Los Alamos as Deputy Director to John Browne, he describes his education there in the concept of leadership which he never received in his academic career, and he provides his perspective on the Wen Ho Lee spy case and the existential crisis this caused at the Lab. He describes the Lab’s role in the early days of computational biology and how this field sparked his interest. Press contextualizes this interest within his conscious decision not to stay connected to astrophysics during his time at Los Alamos, and he explains the opportunity leading to him joining UT-Austin where he remains invested in computational biology. He describes his work for the President’s Council of Advisors in Science and Technology during the Obama administration, he describes Obama’s unique interest in science and science policy, and he narrates the difficulties in the transition to the Trump administration. Press reflects on what it means to be a member of the rarified group of scientists who did not win a Nobel Prize but who were advised by and taught scientists who did. At the end of the interview, Press explains that he has always been a dilettante, which has and will continue to inform how he devotes his time to science, service, and policy matter, and he advises young scientists to aspire to mastery in a specific discipline early in their career before branching out to new pursuits.
This is David Zierler, Oral Historian for the American Institute of Physics. It is July 28, 2020. I am so happy to be here with Professor William H. Press. Bill, thank you so much for joining me today. To start, would you tell me your title and institutional affiliation?
I'm the Leslie Surginer Professor in Computer Science and Integrative Biology at The University of Texas at Austin.
And how long have you had that role?
I've been at the University of Texas about twelve years now, and I've had a couple of different titles. That's my current one.
Bill, let's talk first about your parents. Tell me a little bit about your parents and where they're from.
My father was Frank Press, and my mother was Billie Kallick Press. They were both New Yorkers, raised in Brooklyn. My father was well-known as a geophysicist and was later President Carter's Science Advisor. My mother made her career first as a schoolteacher and later on as an educational administrator in various positions.
Where did they meet?
They met in high school in Brooklyn.
Tell me a little bit about your father's career trajectory. Was he the first in his family to go to school?
Frank and his two brothers were first generation Americans, the children of immigrants. All three brothers made the transition into the professional class. One became a lawyer, one became a trade-paper publisher, and my father, who was the youngest, became a scientist. My father's history is well-told in several oral history interviews by him and by my mother. He died at the age of ninety-five just last January. I love talking about his career and the amazing trajectory he had, from out of the Jewish ghetto in Brooklyn to becoming a world-famous scientist and scientist-statesman. But I think his story is so nicely told in his own oral histories that I wouldn’t add anything in telling it myself.
What extent did he involve you in his career while you were growing up? Did you have a pretty good idea of what a working physicist did for a living, even as a small child?
Obviously, I had much more of an idea than a person chosen at random in the population. But really, my father was more the canonical 1950s absent father whom I saw mainly at family dinner every night. He was totally absorbed in his work and career. I was much more influenced by my mother. To this day, people who meet me who knew my parents say, "Bill, you’re just like your mother. And you're not like your father at all!" I think that may be true of many kids of my generation, whose fathers were less an influence than their mothers. My mother was a schoolteacher as I mentioned, loved teaching, loved teaching her own kids, pushed us to excel, with results both good and bad, but certainly many that she intended.
Now, when you were seven, and your father moved to Caltech. What do you remember about that, or his earlier professional moves?
I have a fragmentary memory from when I was not quite three years old, and we spent a summer in Houston, where my father was consulting for an oil company as a geophysicist. Then, the next memories I have are all in Palisades, New York, at what is now the Lamont-Doherty Earth Observatory, then called Lamont Geological Observatory. It was brand new at the time. My father had been a graduate student of Doc Ewing at Columbia, and Doc moved with two or three of his graduate students (who became assistant professors) to Lamont. The first years that I remember are growing up on this big estate that had just a couple of years earlier been the Lamont family estate, "Torrey Cliff". We lived in a little house, one of several bungalows that were Lamont servants' quarters. These are still there, I think, and used for visitors. (They’ve been renovated, of course!)
I lived in this idyllic, rural environment through first grade at Palisades Elementary School. My mother was unhappy in Palisades, because there were no career opportunities for her. This is discussed in a diplomatic way in my parents' oral histories.
My father eventually became unhappy with being in Doc Ewing's shadow. Doc was the kind of supervisor whose name always went first on the paper, pretty much independent of how much work he had done on it. So, when my father moved from Columbia, from Lamont to Caltech, that was really when he spread his own wings professionally as a rising star. It was a big change for me too. It made me a Californian instead of a New Yorker, growing up. Although for years, the kids in school made fun of my New York accent, still somewhat present. I had to learn to say "wah-ter" and not "wau-ter." And, "choc-late," and not, "chaw-clate."
Growing Up in Pasadena, California
Now, you spent all your years through high school in Pasadena?
That's correct. And that was a good time. Pasadena was a medium-sized city. I could get everywhere by bike or bus. It had a wonderful, large public library. It had an electronics store where I could go in, and wait in line, and plunk down my dollar, and buy electronics components for my various hobbyist projects. I could bicycle to the drug store to drop off film from my Kodak Brownie camera to be developed.
My mother had gone back to work in teaching, which from my perspective, meant that I was without supervision after school every day. We had a housekeeper at home, who would prepare dinner for when my parents got home. But she and I had an unspoken conspiracy: She would pretend to be keeping track of me, and I could just be out in the neighborhood wherever I wanted.
And in the fifties, you could get away with such a thing like that, and it was okay.
Yeah, absolutely. A year after we arrived in California, we moved from Pasadena to Altadena, which is a suburb up in the foothills. We were two blocks from an open canyon space where I and my friends could run around and do anything, as long as we were careful not to step on rattlesnakes.
Now, was Caltech itself part of your reality as a kid? Did you hang out on campus at times? Did you have a sense of it before you had a sense of math and science yourself?
No, no. I think not at all. Of course, my parents’ social life interacted with the Caltech community quite a lot. My parents would have dinner parties and cocktail parties, at which we kids would run around. So, I had some sense of the social side of what a community of scientists looked like. But not really much sense of what scientific research was. I do have to say, however, although my father may have been physically absent most of the day for most of the years of my upbringing, he clearly created an environment in which it was natural for me to become a scientist. I never felt the slightest pressure for that. It was just an environmental thing.
But what do you mean by that? In what ways specifically?
Well, for example, there was a room in our house that opened to the outside, intended, I think, as a garden shed. I was maybe just nine or ten years old when my father cleared out half of it and built a workbench for me. "Bill, this is where you can do your projects." And I just grew up with the idea that I should have projects, and, in the vaguest sense, that they should be scientific.
It wasn't that either of my parents paid attention to what I was doing. It was more a benign neglect. "If Bill's busy in his workshop, he must be doing something educational." In second or third grade, I theorized that all liquids consisted of water with something dissolved in them, and that I could find their essence by putting them out in the sun to evaporate. To my great distress, cooking oil never evaporated to reveal its true essence. I think of this as my alchemy stage.
I then passed on to stages that other physicists have gone through. I built Heathkit electronics, I had an oscilloscope, and an audio oscillator and audio amplifier, and those kinds of practical things. So that's, again, back to what I mean by environmental, where it was natural for me to fall into science, but without really feeling any pressure to do so.
Bill, was your family Jewishly connected at all growing up? Were you members of a synagogue? Did you observe any traditions?
My parents were not religious in any way, but they certainly felt themselves to be culturally Jewish. My father had been bar mitzvahed at age thirteen, for example.
Now, were his parents more observant? I mean, part of that immigrant story?
No, they were good left-wing, socialist New York Jews. Their hero was Eugene Debs. Not communist, however. They were fervently anti-Bolshevik, because they had had quite a time getting out of Russia during the Bolshevik revolution. Frank’s was a secular family, but culturally Jewish enough that the boys of that generation were all bar mitzvahed.
When we arrived in California, for a while my parents went to Sunday meetings of Ethical Culture, and my sister and I were in Ethical Culture Sunday school. But, at the age of around twelve, I was yanked out of that and sent to Sunday school at the Pasadena Jewish Temple and Center, and then I was bar mitzvahed at age thirteen. In the next generation, my daughter, Sara was raised entirely secularly. I did take her a few times to Jewish services so that she would not be completely unfamiliar with them. Later in life, when she was a professional photographer, a specialty of hers was Orthodox Jewish weddings.
Post-Sputnik Reforms; High School
Bill, the long-term effect of Sputnik was enormous nationwide, and I know it had a big impact on you. Do you have a specific memory of the event itself? Or was the impact that you felt more sort of broad in terms of what it did to supercharge math- and science-learning in the United States?
Both. I have a specific memory of the first Sputnik. Everyone was looking in the newspaper to see what times just after sunset you could look in the sky and see it. My memory of seeing it in the sky may be a later reconstruction. But it certainly, as you say, Sputnik had a big effect. The immediate post-Sputnik era was an amazing time to be going through public schools in a place like Southern California. We didn't realize then how singular that time was. Even before Sputnik, it was a time when smart women became schoolteachers because there weren't other career opportunities, few opportunities for women to become lawyers or doctors, almost none to become businesspeople at a high level. So, I had wonderful schoolteachers, women, and then also I had some wonderful male teachers who had fought in World War II and then gone to college and graduate school under the GI bill. Some became, for whatever reason, schoolteachers.
Then Sputnik came along, with all these curricular reforms. For example, in seventh grade, my math teacher was a man named Arthur Benedict, who, in the War, had been on one of the islands in the Pacific, involved in exploiting the breaking of the Japanese Purple code. On Fridays, he would tell us war stories. That was a treat.
The other days of the week, he taught us SMSG math, School Mathematics Study Group, a curriculum that became famous and infamous as the "new math". It was a curriculum that had been devised by pure mathematicians. If you're a pure mathematician, and you have the empty slate of seventh grade minds, how do you fill them? The now-infamous answer was: with set theory, and fundamentals of mathematics, and all of this stuff that is of no practical use in terms of what most kids need to learn in mathematics. Richard Feynman came out very much against this kind of a curriculum. But I loved it. I ate it up. What did I know about what should be useful or not useful?
So, we did SMSG math, and then later on, by the time I got to high school, there was a physics curriculum that had developed post-Sputnik. I can't say I learned much from that- I’ll explain in a minute. Before twelfth grade physics, in eleventh grade, I had a wonderful chemistry teacher, a man named Jacob Sadoff. He was famous in Pasadena, not for being a good chemistry teacher, but because in the evenings he was a shoe salesman in the local department store. The students would go there, not including me, and make him bring out pairs of shoes, shoe, after shoe, after shoe, to try on. Later, long after had graduated, I had a chance to ask him, "How could you tolerate that? It was so disrespectful of these kids." These were not the A students, generally. And he said, "I don't care. I sold a lot of shoes." There were state and national competitions in chemistry. I scored, if I recall, fourth in Southern California. That was nice, nothing too special. But the award had the unfortunate effect of convincing me that I knew everything about chemistry, so I never took any chemistry in college or graduate school. Now, many decades later, I'm working in biology, and I have this regrettable gap in my knowledge.
Now, as to twelfth grade, and physics, my teacher was a guy named Keith Miller. He was the physical education teacher, the gym coach. In this huge public high school, there was no one qualified to teach physics. Mr. Miller volunteered, but sadly did a terrible job. I learned nothing. I cut most classes. In public school at that time and that place, you got two grades in every course. You got a grade for skills, which was the real grade, and then you got a grade for citizenship. They counted equally in your grade average. Mr. Miller gave me an A+ in skills and a D in citizenship. He was a very tolerant man to give me a D. It might have been an F. That, along with my C grades in gym, both in skills and in citizenship, meant that in my high school class of 1,500 graduating seniors, I graduated somewhere around number 100 in grade point average. All of my friends got to go up on the stage and receive awards, and I was relegated to the audience.
I wonder if that taught you any long-term rules to live by in terms of civics.
Well, this is one of the examples of: Do I take after my father, or do I take after my mother? My father was always measured, always careful in dealing with others. My mother was always a smart aleck, trying to do and say the unexpected. And I'm afraid I always took after her, and for many decades into my adult life. I think it wasn't until I was about fifty years old and had moved to Los Alamos to be a senior manager in a large organization that I first realized that maybe a more diplomatic way of dealing with others was the right way to be effective.
Bill, by the time you were in high school, did you see that you were on a path in science as a career? Were you doing sort of research, either informally or in your lab work at school, which sort of suggested that this was the trajectory that you were on?
Yeah, I think definitely. I wasn't sure whether it was going to be science or mathematics. I certainly didn't want to, "follow in my father's footsteps," as I saw them. But I tended to see them quite narrowly. I thought, "He's a geologist. I’m not interested in anything like that. I’m interested in either pure mathematics or astronomy or physics."
So, you didn't really think of him as a physicist so to speak?
No. No, I really didn't. It wasn't until decades later that I first got interested in looking at what my father had done in his early career. I went back, and I read his papers. And I was surprised that they were very mathematical, and that his way of thinking about things was not that different from ways that I thought about physics problems. But my route to becoming a theoretical physicist was not through my father at all. It was other people, starting in college, who had a big effect on me.
Which also suggests, and this gets back to an earlier question, really how little you interacted with your father at least on the basis of his scholarship.
I think that's right. I don't want to minimize his effect on my career, because as I said, you can see that my parents both channeled me into being a scientist, whether or not I was aware of it at the time. And I'd be foolish to say that his being a well-known scientist didn't affect the way I was able to interact with other famous and wonderful scientists. That was something I took advantage of, mostly unconsciously, sometimes consciously, but overall, it was a pretty subtle effect. Other people had a more direct effect.
Now, when it came time to think about colleges, first of all, of course, Caltech is right there. Did you consider sort of going to a world-class university that was literally at your doorstep? Or were you specifically looking at opportunities to sort of move far away from home?
Absolutely the latter. I was looking to get as far away as I could. In my high school graduating class of 1,500 people, only about 150 even went to a four-year college at all.
So obviously, this is a much bigger high school than just kids of Caltech faculty members.
Oh, yeah. Caltech was a very small blip in Pasadena High School. Pasadena was a middle-class bedroom community of Los Angeles, an easy commute to L.A. downtown. There were two high schools in Pasadena, one on each side of town. John Muir High School on the west side of town was entirely Black and Hispanic and had the poorer teachers and the fewer resources. Pasadena High School was all white and middle-class and was a wonderful high school.
And when you mean all white, you really mean all white.
I really mean that out of 5,000 students at Pasadena High School, there were at most a few dozen African Americans, maybe fewer. There were also a few third-generation Japanese Americans whose parents had been interned in camps during the War so outrageously.
Until the federal bussing mandates, the schools were de facto segregated, nearly one hundred percent. The realtors in Pasadena had a code. If you were a Black family, they would not even show you a house on the east side of town. This was supposed to be cured by enforced bussing. Just a few years after I graduated, the schools became fully integrated. And then the second wave of de facto segregation began, in the form of white flight. When I was growing up, Pasadena had one small private school. Ten years after I graduated, I think a majority of white families had pulled their kids out of the public schools and put them into newly founded private schools. So, it's just one of these terrible object lessons: You try to do the right thing, and the consequences are not necessarily what you wish for.
Okay. It's an important point though. Besides Harvard, what other schools did you apply to?
I applied to Harvard, Yale, Chicago, Berkeley, and got into all of them. Out of the 150 kids in my class who went to a four-year college, only fifteen of us went to a college out of state. California, then and to some extent now, is just its own universe. My high school sent, I don't know, thirty or forty kids to Berkeley; but Berkeley wasn’t far enough away from home for me, so Harvard and Yale were my top choices. Of course, the irony was that my father moved from Caltech to MIT at the beginning of my freshman year. So, my attempt at distance failed completely.
I didn't know that detail. I didn't know the timing of that. Did you see him in Cambridge at all? Did you spend time with him?
I went home once a week to do laundry at my parents' house. Right after I heard that I was admitted to Harvard, I went to Berkeley for the summer, working at Lawrence Berkeley Lab in the Powell-Birge bubble chamber group. And while I was there, I had a phone call from my mother, and she said, "We weren't telling you two kids, but, this last year, Frank's been negotiating to move from Caltech to MIT. "And we're moving there in the fall, just when you'll be going there, Bill. And, oh, by the way, all your precious possessions in the house here in Pasadena, we'll have the movers pack them up and put them into storage in Cambridge." So that came as quite a shock. But it really was convenient having my parents living there, so I could do laundry at their house. And to some extent, visit with them.
Mostly, however, it was understood by both my parents and by me that their parenting extended through getting me into college, and then paying the tuition. After that, in a friendly way, my parents and I mostly went our separate ways. My attitude towards my parents was, "Thank you very much. It's been great. I'm glad you're paying the tuition. I'm pretty much on my own now." They had launched me, and I felt empowered to then move on from there. While I may have gone home every week to do my laundry, and I did interact with them, they were both relatively minor influences in my life after that point. Still, there's no question that my father's prominence affected my interaction with other scientists because they knew whose son I was.
Adjusting to Harvard
When I got to Harvard, I learned that I was actually lucky to get in at all, probably in large part because of geographical distribution. Harvard didn’t expect me to do particularly well. I know this because, in my freshman year, I got into trouble over a prank involving phone calls made to the answering machines of deans. I was summoned into the office of the freshman dean. Across his desk, he had opened my admissions record. Very prominently, in the upper left-hand corner, was Harvard's statistical prediction of my future grade average. In large print, it said C+. Pasadena High School sent someone to Harvard maybe once every ten years, and people from big California public high schools generally did not do well in the Ivy League schools back east.
Do you have any theories on why that might be?
Well, I can tell you how hard it was for me. First of all, it was a huge culture shock, being immersed in Cambridge, Massachusetts, versus Pasadena, California. Second of all, thrown together were about half people from public schools and about half preppies—kids from private schools, especially New England residential private schools. Those kids knew how to study. Even if they weren't going to be the best Harvard students, even if they were going to be the ones to graduate with a B or C average, they had skills that, coasting through public high school as a smart kid, I had just never learned.
I arrived, and I enrolled in the highest-level pure math course that I could get into. And I did poorly. In terms of non-science and math things, I enrolled in a junior-level course in East Asian history that was being taught by John Fairbank, the famous China scholar, and Edwin Reischauer, the famous Japan scholar. And I got a C in that. This was a real shock.
It was never explained as such to entering public-school kids, but there were really two tracks through Harvard. There was the track through a great, great research university, where in your field of choice, whether it be in the sciences or anything else, you could dig deep and learn a great amount. And then there was a track for people who were going to go off and do things in the world that were not academic. There had to be a track through Harvard that let them graduate with a respectable B average, not necessarily taking the same courses as the people who would become academic scholars and leaders in intellectual fields.
So, there were courses that were "gut courses", and then there were courses that were definitely not. The catalogue didn't tell you which were which. You had to ask around to find out. Once I learned to take an appropriate mixture of gut courses, I did okay. The guts were always interesting, amusing, easygoing, and you could get an easy A, or A-.
At what point did you settle on physics?
Well, basically, when I discovered that I couldn't do pure mathematics. And that was pretty early on, so already by the beginning of sophomore year, I declared my major as physics. In high school, I had only the slightest hint of what pure mathematics was. In my senior year, I was able to take the calculus course that was offered at Pasadena City College, commuting there during school hours on my little Honda motorbike. Now, normally, Pasadena City College, a good two-year, junior college then and now, would offer a kind of vanilla calculus course. But that particular year, they had hired for the course a graduate student of Tom Apostol, the Caltech professor. Apostol had written a famously abstract first-year calculus book. That was the textbook being offered to a mixture of junior college students and a few of us high school students taking the course.
Did you have an idea at the time how out-of-the-world this was for a junior college kind of course?
I did. I did. Here's an example of my father's sort of hands-off influence: He looked at Apostol’s book, and he said, "Let me give you the calculus book that I used when I was a freshman in college." I wish I could remember what the name of it was. It must've been written in the 1920s. It had infinitesimals instead of the more rigorous delta-epsilon limits, and so on. All he had to do was give me that book, and I could eat it up. Well, that's all consistent with my discovering that I could do practical mathematics and do well in the kind of mathematics in physics courses. But I just didn't have that special abstraction of mind to do pure math. I could never have been a good mathematician. At the same time, I wasn’t finding physics courses completely easy, maybe from having skipped all the physics classes in high school. My first-semester freshman physics instructor was Dan Kleppner, who was then just a new assistant professor.
He must've been in his late twenties at that point? He must've been really young.
Early thirties. When he moved to MIT with tenure, he of course became very famous. As an example of what kind of thing just stumped me, it took me a long time to realize that physics courses didn’t teach what's true, they teach an agreed-upon set of abstractions that are useful. The first week of freshman mechanics, you learn that when you put a weight on the table, gravity exerts a downward force, and the constraint from the table exerts an exactly equal and opposite upward force. And I just couldn't understand that. How could the table know exactly what force to exert?
I think I was two months into the semester when I realized that I actually did have some physical intuition, a term of art well-known to physicists. So, I shouldn't be stumped by not understanding the equations. I should think about it physically. I thought about constraint forces, and I realized that the table wasn't infinitely rigid, that the weight sort of depressed itself into the table a little bit. We had learned about harmonic oscillators by this point. There was a restoring force that would increase linearly until it equaled the force of gravity. Suddenly, I was able to do those kinds of problems. I realized that as I used physical intuition, and also thought about the equations, I should also try to understand the nature of the abstraction being made. I think that's something that's very poorly taught in physics. Then and now.
Interesting. Would you say that this is a perspective that you've kept close throughout your career?
Yeah. I pretty quickly understood that my strength was going to be physical intuition. And I could capitalize on that strength by being just mathematical enough to understand how the equations embodying that intuition ought to come out. Then, I could either work it out myself, or, more usual in my subsequent career, find someone to collaborate with who was really mathematical. Very often I could be the impedance matcher between a physical viewpoint and a purely equation-based mathematical viewpoint.
Influences of Harvard Faculty
Particularly in the physics department, were there professors that you became close with?
I wouldn’t say I was close with Kleppner. That was just my baptism by fire in what a Harvard physics course looked like. In the second term, however, electromagnetism was taught by Ed Purcell, who was very influential on me, and became one of my scientific heroes.
I want to ask first about Purcell's style as a teacher. What was he like in the classroom?
Well, he was exactly like the book that he wrote. Which was one hundred percent committed to the idea that the physics could be described before you tried to write down equations, that the physics could be described by a kind of intuitive geometrical perspective. In electromagnetism, you could understand lines of force intuitively before you ever knew what equation they came from. Given that I had already discovered that I wasn't good at pure math, and I did have some degree of physical intuition, I really grooved on this. He became my model of what a physicist should be. Ed Purcell was always approachable. Anybody in his class could go knock on his door and come and ask him a question, and he would always have some interesting gadget on his desk that had nothing to do with the freshman course. He was a wonderful, sharing man who loved physics and loved teaching physics.
And clearly, he was formative for your own intellectual development as a physicist.
He was. You’ll see in a bit that, in graduate school, I had to unlearn some of that. I had to learn that there were other styles of doing physics that I needed to master. But if I'm listing influences, another interesting one was Dick McCray. He taught the only astronomy course I took at Harvard. McCray, became, later at the University of Colorado, one of the great theorists of his generation.
But he was an absolutely new assistant professor at that point, and he didn't have a clue how to teach. It was a course with half a dozen Harvard seniors. We were all cocks of the walk. We weren't going to take any nonsense from a green assistant professor. McCray would come in and hand out a problem set, and we would look at it and tell him which problems we wished to do and which we didn't want to do. Later, as an assistant professor at Princeton, I found myself on the opposite side of this kind of transaction. At both Harvard and Princeton, there's a fixed pecking order in which senior undergraduates outrank junior assistant professors. Maybe it’s they're going to graduate and become the donors that keep the university going. Still, Dick must have taught us something. Among the six of us, three became professors of astronomy: me; Bob Kirshner, who became a professor at Michigan and then Harvard; and Rich Gott, who became a professor of astronomy at Princeton.
Who else influenced you?
I took Norman Ramsey's quantum mechanics course. Ramsey had a set of mimeographed lecture notes. I was visiting my parents while I was taking that course, and I said to my father, "Well, I'm enjoying Professor Ramsey's quantum mechanics course. And he has these lecture notes." My father said, "Oh, can I look at them?" He looked at them, and he started laughing because he had taken Norm Ramsey's quantum mechanics course as a first-year graduate student at Columbia in 1946. And the notes were identical. They hadn't changed in an entire generation. Norm, who, later, as a colleague, I got to know pretty well, had simply perfected the art of teaching first semester quantum mechanics.
Senior year, I took Schwinger's course, called Advanced Quantum Mechanics. It was a graduate course. He brushed off advanced quantum mechanics in about a week and then started teaching Source Theory, which is something known now only to historians of science. This was Schwinger's attempt to redress the balance that his way of doing quantum electrodynamics had lost out to Feynman's way of doing quantum electrodynamics, namely Feynman diagrams and so on. And so now, twenty years after that, Source Theory was his way of thinking about, I guess, strong interaction physics. At the time, it was completely not obvious to me what that course was about. And to this day.
I dutifully attended all of Schwinger’s lectures, and I took notes, and I learned absolutely nothing. Schwinger never looked at a student’s problem set, and he never looked at a final exam. He had a grader who did those things. My problem sets would come back with grades like 0.5 out of 10. On the final exam, I looked at the questions and realized that this was going to be another 0.5, but now out of 100. I wrote on the first page of my blue book, "Dear Grader, I have been accepted for graduate school at Caltech. If I fail this course, I will not be able to go to Caltech, and I will be drafted and sent to Vietnam. Please take this into consideration." And I got a B. That's my Schwinger story.
Did you even try to approach him?
It was clearly impossible. Even his graduate students. Schwinger saw his graduate students a couple of hours per semester. Throughout his career, he delivered beautifully polished lectures. But he was lecturing to himself. If the audience were full of Schwingers, he would've been better appreciated.
At this time, I really didn't know whether to think of myself as a theoretical physicist (in the making) or as an experimental physicist. I always had term-time jobs, grading and working in the physics labs. My financial arrangement with my parents was that, for college, they would pay tuition, room, and board. I was supposed to find jobs, summer jobs and term-time jobs, to pay my miscellaneous living expenses, which at the time was possible to do.
Experience with Experimental Physics
My junior and senior years, I worked in Gerald Holton's high-pressure physics lab. I never saw Holton, although, when I came back to Harvard as a professor, I came to know him as a colleague. Gerry had inherited Percy Bridgman's high-pressure lab. The equipment was from the 1920s. Bridgman had used it for the research that led to his 1946 Nobel Prize. Now, by my time, Holton was already primarily an historian and philosopher of science. He was already involved in the Einstein Papers at Princeton, and he had founded the journal Daedalus. In any case, he was never in the lab. But the lab was still there, and he had a post-doc, a guy named Steve Hawley. Steve was supposed to get publishable results in high-pressure physics out of this ancient, literally hand-pumped high-pressure equipment. I was assigned to Steve. Steve tried to make an experimental physicist out of me because really, these jobs were what we would now call internships—supposed to be learning opportunities. I soon enough learned that I just wasn't very good at laboratory work. I mean, I was just all thumbs. But also, I didn't know when or how to ask for help.
At one point, Steve sent me off with, "Go to the stock room, buy the parts, and build a twelve-volt, five-amp power supply with less than three-tenths of a percent ripple," or some set of specifications like that. I looked in Purcell's freshman physics book to figure out how to do this. I had no idea that there were things like active components, or voltage regulators, or any such thing. So, I ordered a big enough capacitor and big enough choke (inductor) to meet the specifications that were given, calculated as a freshman physics problem. It turned out that the inductor weighed thirty pounds, and my power supply had to be moved on a cart. Hawley looked at it and decided that he would never again assign me to do practical things.
What he did instead, which was wonderful for me, was assign me to take over the data reduction for his experiments. I had to learn to use the time-sharing computer that Harvard had just bought and migrate all the data off of IBM cards. This wasn’t my first introduction to computing. The summer before college and the summer after my freshman year, I had summer jobs at Lawrence Berkeley Lab. That was computing with IBM cards.
Anyway, by the time I graduated from Harvard I knew I wasn't cut out to be an experimental physicist. But some remaining ambiguity about this helped me get into graduate schools, I think. It still looked, from my record and my letters of recommendation, like I might be an experimenter. There was a glut of people who wanted to do theory, and my experimental work, even all-thumbs, set me apart. The admissions committees probably thought that if I failed at theory, I could still be an experimenter.
And, Bill, would you say looking back that you gave experimentation an honest shot, just to see what your aptitude was in that?
Experimental physics would've been the natural outgrowth of my experience growing up, having my own workbench, and building Heathkits, and so on. And again, environmentally, my father was more an experimental physicist than a theoretical physicist, at least in the time that I was growing up. He was then famous for building new kinds of seismographs, and for early applications of computers to reducing seismic data. To this day, at home, I'm always fiddling around with peculiar projects in my shop and around the house.
But I learned that I didn’t have the concentration on detail that you need to be a good experimental physicist. Theorists are always derogating experimentalists, "Oh, their brains are in their fingertips." To me, that sounds like a big compliment, because it's exactly what I didn't have—brains in my fingertips, the way good experimenters had.
WHRB and Student Activism
Bill, before we get to the kinds of graduate schools that you were thinking about applying to, I want to switch for a minute over to the social scene during your undergraduate years. Of course, the late 1960s, there's all kinds of interesting things that are happening on campus and beyond. What was your personal relation to the various protest movements that were going on in Cambridge at the time, the anti-war movement, Civil Rights, Women's Rights movement? Did you tend to sort of get yourself in the middle of that, or were you sort of more aloof from those issues?
Well, I was involved, but in a strange way. The circle I was in was totally sympathetic with those issues. But I have to tell you more precisely what my circle was. This is the answer to, what did I actually do in my four years at Harvard, besides physics? The answer: WHRB, the undergraduate radio station. People think that you go to Harvard to take courses, get grades, get your degree. That's true enough. But almost all Harvard undergraduates- I think Harvard admissions selects for this, take up some activity, some passion, that comes to occupy most of their waking hours. I don’t mean coursework. It might be athletics, it might be the Harvard Crimson. In my case, it was WHRB. I spent by far more hours at WHRB than I did studying, or in classes, or anything related to physics. The radio station was its own self-contained enterprise. Harvard donated the space, but it was run by us undergraduates—no such thing as a faculty advisor, even. I learned a lot. Everybody did a little bit of everything. I mostly did studio engineering and engineering construction, again using my not-very-good fingertips to build electronics stuff; and I did a bit of announcing. Then I became seriously interested in classical music, and became classical music director, and did that for a couple of years. If you were a music listener in the Boston area and tuned to WHRB (one of just two classical music stations), what you heard was what I thought you should hear. That was really a lot of fun.
Okay, but back to your question. In the student protests, WHRB was very much involved, but as supposedly neutral journalism. My classmate Chris Wallace became famous later on as a career journalist. Of course, it may have helped that his father was Mike Wallace.
When students took over University Hall and ejected the deans, and later the police were called, I was in the middle of all that. I was in and out of University Hall. What was I doing? I was stringing wire out the windows because we knew that when the police were eventually called, the first thing they would do is try to cut off our broadcasting from inside. We strung multiply redundant cables through the air to freshman dorm room windows, and then from there, through telephone wires back to the radio station.
This is guerrilla journalism.
Of course. We weren't completely neutral, obviously. Clearly our sympathies were with the students. But because our broadcasts were perceived as neutral, or at least not completely given over to the students' side, we ended up being go-betweens for the negotiations between the students and the administration. Those were very interesting meetings to be in on. Still, that spring, my main worry was, would I be drafted or not? Would the Vietnam War interfere with my going to graduate school?
Bill, how systematically did you think through what you might do if you were drafted? Did it ever become a real-enough possibility that you really had to think about what your response might've been?
It was painfully obvious that the kids who were being killed in Vietnam were not the educated white kids, but the poor, African American, Hispanic, and other minority kids.
And you recognized this at the time?
Yes. Realistically, I wasn't afraid of being suddenly snatched out of Harvard and finding myself in a Vietnamese rice paddy. Looking back, I say this with some feeling of guilt. The Vietnam War had many repercussions, one of which was a deepening of class lines between those who were drafted and those who managed to get out of it. Realistically, the penalty I was facing was to lose a couple of years out of my life doing a boring desk job somewhere. Maybe that would be in Saigon, but more likely somewhere like Huntsville, Alabama. So, the worry was more about disruption than fear of danger. In my Harvard class of about 1,000 men (Harvard had not yet merged with Radcliffe) the number killed in Vietnam is very small. One or two people, if I recall. The fraction who did any military service at all must be well under ten percent, despite the draft. Al Gore was in my Harvard class. He was the son of a senator, and he enlisted.
Let’s go back to computers. At that point, as you're weighing the various options before you, are you seeing computers as part of the equation, no matter what it is that you end up doing?
I was not yet seeing that I would end up, in graduate school, riding a wave of new applications of computers to physics. But I was already in love with computer hacking. I mean hacking in the sense that you don't just write the program to do the science. You're in love with the computer and its internals, and you just stay up all night trying to get as deep into the computer as possible.
So, when you're using the word hacking at this early stage, you're not talking about disrupting systems. You're just trying to get into the machine.
Well, not maliciously disrupting systems. But, my first computer hacking was not entirely above board. We're back in Gerry Holton’s lab. I had reduced some data for Steve Hawley, and he was happy with it. But a week later, he got the bill for my computer time and he just couldn’t afford me. I said, "Let me see what I can do."
The computer was a time-sharing system, basically just a mainframe computer, but with many people on terminals working simultaneously. An individual user almost never could crash the whole machine. Time-sharing systems had to be well-engineered in that respect. These systems grew out of the Multics effort at MIT, historically. What I had discovered was that it wasn't hard for an individual user to crash his own session. You could write a program to do it intentionally. You'd run that program, and suddenly the teletype would type, "Login:" as if you had never been logged in. All your charges for the previous session got lost. This was an example of someone's bad software design: When you logged out, the system totaled up your charges and booked them. If you managed to get off the machine without logging out, you were home free.
I went back and told Steve Hawley, "It's okay, Steve. From now on, all your computing will be free." And it was. Now, obviously this raises the question, was I stealing? Was I a thief? Was I committing a crime at the time? The answer is yes, but it is also a little muddled. At the time, these kinds of pranks were seen as fooling around within the envelope of an experimental, newfangled system. It was different from sneaking into the chemistry stock room and stealing something, for example. I never, ever would've considered doing anything like that.
I saw, over the next few years, a phase change in attitudes. Computer hacking that we did would now be called dishonest, or even criminal; but it wasn't born that way. There was a change in social expectations. Was I ever caught hacking, and was I disciplined? Yeah, all the time. I would get summoned to the office of the Director of the Computer Center and get yet another verbal slap on the wrist. I was the leader of a group of hackers. They knew who we all were- we all came from the radio station, unsurprisingly. We discovered many wonderful things deep in this machine. To the extent that I'm a computer scientist today, a lot of the basic principles I learned about operating systems and machine architecture, I learned hacking into machines then.
We discovered at one point that we weren't the only group hacking this computer. This other group was destructive. They were wiping out people's files. Our group was hauled into the computer authorities. They threatened to refer us to the dean for serious disciplinary action. It's not us, we told them. We promised to find the other hackers and get rid of them. In return for that, we wanted the computer center to forgive all our hacking, past, present, and future. They didn't exactly agree to that, but they let us proceed.
We did find that other group. They were the famous MIT phone phreaks who had built blue boxes and black boxes to cheat the telephone company. That's its own whole story, one that has been told in a book by Phil Lapsley and in other places. This group of kids at MIT brought the whole Bell system to its knees, forced the reengineering of the entire Bell toll charging system. Theirs was a mixture of adolescent prank and criminal activity. They were trying to diversify and become computer hackers. Fortunately for us, they were terrible at computers.
Once we figured out who they were, then, Mafia-like, we and they divided up the territory. We would be the computer hackers in Cambridge. They would be the phone phreaks. Our two groups got together once a week at midnight for a one-hour seminar series. We taught them computer hacking, but only theoretically. They taught us phone phreaking. To this day, I have a lot of knowledge, now completely out-of-date, about how Bell System Long Lines worked.
Choice of Graduate School
Listening to your overall undergraduate experience, it certainly doesn't sound like there's an absolute linear line to pursuing a physics graduate education. You were interested in lots of things. You could've gone in any direction. And so, I'm curious how that informed the kinds of opportunities that you saw before you, both in terms of fellowships, or jobs, or graduate programs, or perhaps taking time off. Many eminent physicists say, "This is the path that I was on, and this is what I was doing, one hundred percent straight ahead." It doesn’t sound like the case for you.
That's a great question. First of all, we can eliminate taking time off because I didn't have that option, because I would've been drafted. My only hope of not being drafted was to get a graduate deferment by going to graduate school in physics. Later, even these deferments were eliminated.
I was interested in all kinds of other things, but I considered those things "extra-professional." I was immersed in computer hacking and in classical music. I spent hours and hours listening to classical music, programming classical music at the radio station, writing scripts for radio features. Some friends at the radio station became musicologists and performing musicians of note. Robert Levin, who became a pianist, harpsichordist, and composer, as well as a Harvard professor, was a friend. But I was already on a track where my profession was going to be physics. I only applied to graduate schools in physics.
Yeah. So, you never really strayed from this idea that physics was really your bread and butter throughout your undergraduate years?
I think once I discovered I was good at it, and once I became inspired by people like Ed Purcell as a scientist and human being, people like Kleppner and McCray and Ramsey as physicists, then I was just in the groove to go to physics graduate school. I got in everywhere I applied.
Which was where? We've already established that you're pretty well-rooted in pursuing theory in graduate school, right? So now that leads to, what are the programs that are doing the most consequential work in theory at the time? And, what kind of advice did you get from your professors at Harvard? And were there individuals that you were thinking that you specifically wanted to work with as well?
I applied to Harvard, MIT, Berkeley, and Caltech.
What was your specific perspective on Harvard? One school of thought is, "You're at Harvard. Why would you go anywhere else?" And the other, of course, is, "It's time for you to see the bigger world out there."
I was (and am) a total believer in the second of those. Harvard was my last choice.
But in terms of the advice that you got?
Let's go through the four because they all were somewhat different. I think that my number-one choice throughout was Caltech. It is hard to say exactly why. An element was returning to my hometown. I grew up in Pasadena, even though I had very little familiarity with Caltech while I was growing up. My father was, by this time, no longer there.
I wonder if, also, you had to go to Harvard to appreciate Caltech, in the sense that you never could've appreciated as a high schooler in Pasadena.
Well, I never would've considered Caltech as an undergraduate. I understood that it was a science school. I wanted a broader, liberal arts education. The other big effect drawing me back to Pasadena was that my girlfriend, who I knew in high school, who became my wife, Margaret Lauritsen, was the daughter and granddaughter of Caltech physics professors. She was close to her family. She was much closer to her family, in our college years, than I was to mine. So, the idea of us moving together back to the Los Angeles area, both going to graduate school there, her being close to her family, I think that was a factor.
And were you married as an undergraduate, or only when you started graduate school?
Married at the very beginning of graduate school. But we were together through all our undergraduate years. Together in a commuting sense. Margaret went to Berkeley and graduated from there in three years. She then transferred as a special student to MIT my senior year. For the three years, I spent a lot of time in Berkeley. Harvard had long vacations, and I would do an extra week on each side of every vacation and spend them in Berkeley with Margaret. So, when you asked about Harvard in those turbulent years, in the sixties, I was about equally immersed in Berkeley's turbulent scene. I was tear-gassed in People's Park by the police even though I was there only as a tourist.
Was there a particular person you wanted to work with at Caltech?
Through my parents, I had grown up knowing socially, slightly, some of the Caltech physics professors. Not Kip Thorne, because he had arrived at Caltech just two years earlier, in 1967. But I knew of Kip Thorne. He had written a famous 1967 Scientific American article on gravitational collapse that I read. Maybe it was through Margaret's father, Tommy Lauritsen, hearing about this rising superstar that they had hired, the youngest full professor in the department, and so on. In any case, general relativity and gravitational collapse was a hot subject then.
So, Kip was known as a rising star at Caltech. Besides Kip, everybody wanted to go to Caltech and work with Dick Feynman. But you arrive at Caltech, you want to do theory, and it takes about five minutes for you to learn from the rumor mill that working with Dick Feynman is a bad idea. If you could do it at all. Feynman took on very, very few students, and the students he took on tended to be students that he just got along well with personally, irrespective of whether they were any good. I wanted to do either particle physics, which was on the cutting edge, or general relativity, which was also on the cutting edge. Interestingly, not astronomy per se. The fact that I drifted into astronomy, astrophysics, cosmology came later. I went to Caltech wanting to do something very pure. Either very pure general relativity or very pure particle physics.
Before you actually made that decision to go to Caltech, what about the other schools?
A lot depended on where Margaret got in to graduate school. Margaret had been an undergraduate at Berkeley, and she was applying to graduate schools in linguistics. The linguistics professors at Berkeley knew her, and she almost surely could get in there. I could too. Berkeley took a huge graduate entering class. So that was one possible place.
MIT was another. The attraction there in linguistics was Chomsky. Margaret was there, my senior year, as a special student, and she had gotten to know some of the linguistics professors, although Chomsky not at all. But she had taken Morris Halle’s course. He was Chomsky’s longtime collaborator. If Margaret had gotten into the MIT Linguistics Department as a regular PhD student, I probably would've gone to MIT. Margaret got into UCLA, which was a very good place, if not quite as good as MIT. So, the attraction of her family living there, and Feynman and Thorne at Caltech, that decided it.
Hertz Graduate Fellowship
Let's talk about Livermore. How did that opportunity come about?
I wish I could remember where I first heard of the Hertz Foundation. This was a little foundation that, from an undergraduate's perspective, was just a notice on the bulletin board for graduate fellowships. Someone must've pointed me to it. The Hertz Foundation awarded graduate fellowships that paid twice as much as any other graduate fellowship, for example the NSF fellowships. It was not obvious who they were- just a little private foundation that gave fellowships in the applied physical sciences, valid only at a certain list of top schools.
All of the schools that I applied to were on that list. So, I applied for their fellowship. What did I know, and when did I know it? The Hertz Foundation was in fact a creature of Edward Teller. When I applied, it had been giving fellowships for only a couple of years. Teller had gone around to his rich friends, including John Hertz—no relation to Heinrich Hertz, the nineteenth-century physicist. John Hertz founded Hertz Rent-a-Car and, before that, the Yellow Cab Company, so he had made a fortune several times over. He was a staunch American capitalist who could very much groove on Teller's right-wing politics.
Hertz and his wife Fannie already had in place a foundation for their smaller philanthropic endeavors. Edward took control of it and installed a board of directors that read like a litany of extreme right-wingers of the time, especially controversial because these were the Vietnam years. J. Edgar Hoover was on their board of directors. General Curtis B. LeMay, who was famously quoted as wanting to bomb the Soviet Union back to the Stone Age. And so on. Had I known more about the Hertz Foundation I might not have applied. I had no political sympathy for them. But I had no idea. You couldn’t just Google and find out things in those days.
Anyway, I did apply. One afternoon, I was in Gerry Holton's lab when the physics department secretary came rushing in and said, "Do you know Edward Teller?" I said, "You mean the father of the hydrogen bomb? No." She said, "Well his office is on the line, and they want you to call them back at this number and then fly to New York City immediately tonight." Why New York City? Because there was an American Physical Society meeting going on. This was supposed to be my Hertz Fellowship interview.
In those days, you could leave Harvard at 2:00 in the afternoon, and you could be in New York City by about 5:00, with no reservations, flying for next to nothing on the Eastern Airlines shuttle at student fare. So many ways in which the world has changed to make those facts no longer true! Well, there was a hotel suite where the Hertz Foundation had set up camp, and they had never heard of me. They had no record of my application.
I said, "But I got a phone call from Edward Teller's office." They said, "Oh, well, then you probably applied. Okay, we'll interview you." They rounded up a couple of Livermore Lab folks to interview me, one of whom was Lowell Wood, who I later got to know well, and who later also became infamous in his way. It was a peculiar interview. The questions were just out of left field. "What do you think about gravitational collapse? What do you think about hydrogen bombs?" They didn't mean politically. They meant; how did I think the bombs worked? What was the physics of it?
Unrelated to the politics of the Hertz Foundation, the kinds of students they were looking to award fellowships to were very much of the Ed Purcell, physical-intuition type. I came to learn that that was also the Edward Teller way of thinking about physics, of visualizing how things worked, visualizing things geometrically, and so on.
I got the fellowship, and along with the fellowship came an offer of a summer job at Livermore Lab. It wasn’t a requirement to take the job, but it paid much more than other typical summer work, so I went to work at Livermore that summer. A very interesting experience. On the one hand, my interactions with Edward Teller soon confirmed that all of the terrible things that people said about him were true, that he was an extreme anti-Communist, right-winger with an unfortunate, large political influence. At the same time, one didn't have to interact much with Edward to learn that his politics were walled off from his mentoring of students. You could look at the people and the fields that Hertz Fellowships were awarded to, and there was no political coloration in it. Then and now, I think that Teller was a very evil man. Not just misguided, not just on the wrong side of history, but really evil in the sense of believing that the ends, as he saw them, justified any means.
Which ends are those, Bill?
The defeat of the Soviet Union. The implacable hatred of everything Russian. This isn't the place to try to retell Edward Teller's history.
No, but it's valuable to hear your perspective.
I'll tell you, decades later, in 2002, I read Edward's published memoirs. I knew a lot of the people that he was describing, not in his early history, of course, but the later years. I was struck by the fact that everybody already dead was stabbed in the back in his memoir, and everybody still alive was treated very gently. Why? Because the living were still around to contradict his self-serving retelling of events or to criticize him for his politicizing every possible event.
So, politically, I have no love of Edward or of his disciple Lowell Wood (I think that's really the accurate word). But they were both, especially Lowell because I saw much more of Lowell than I did of Edward, excellent mentors, not just to me but to all the people they took under their wing at Livermore. It's just one of these decisions in life. You can decide that some people are so evil politically that you want nothing to do with them. Or you can decide, "Let’s see what I can learn here, for my own benefit, while staying clear of their politics." It’s not always an easy choice, but I made it.
Just a hint of the Foundation’s right-wing origins: As a Hertz Fellow I had to sign an annual pledge that, in a "national emergency", I would devote my efforts to the "defense of the United States". These were not neutral words during the Vietnam years. Another condition was that I was supposed to do my thesis research in an "applied physical science". So, before accepting the fellowship, I called Lowell Wood, and told him that I wanted to go to Caltech and work in general relativity. Would this be a problem? Lowell said, "Well, by my way of thinking, general relativity is an applied physical science, and I feel that for the United States to be strong in general relativity is also to strengthen its national defense. But if you want, I'll go to Edward, and I'll get him to give you a call and reassure you that he also agrees." This wasn't necessary. Lowell was so close to Edward that he could speak for him in all such matters.
I was supported as a Hertz Fellow for all my years in graduate school, and I never heard a peep that I was working on something out of what they thought was applied physical science in support of the defense of the nation. But, subtly, they knew what they were doing. My Hertz experience introduced me to the world of national security and defense science, which later became a big part of my whole career, culminating when I went Los Alamos as Deputy Laboratory Director. So, in their terms, their bet on me paid off. Not just me. If you look at where Hertz Fellows have gone, a number of them have made their careers in national security, including in the military. There are a couple of Hertz Fellows who have become flag or general officers.
Teller, in person, couldn't help being a caricature of himself. Peter Sellers in the movie Doctor Strangelove is absolutely aping Teller mannerisms. But Teller was not actually the caricature that he was made out to be. He may have been an evil man, but he was a very complex evil man. I think in his calculation of keeping politics out of some aspects of what he did, namely awarding graduate fellowships, but not out of other aspects, namely the kinds of science that he pushed at Livermore Lab, the x-ray laser, Star Wars, and so on, wasn't haphazard. He had a strategy, and in terms of his interacting with young people, his strategy was very clearly, "Let's show the best young people how exciting it can be to work in defense science. Most of them will just go off and work in academic science, but a few will make careers (or parts of their careers) in defense, still a good return on investment."
I'm supportive of the Hertz Foundation even now. It has outgrown its right-wing origins and now supports the best people in a much wider range of fields and at a broader set of places. Its emphasis has shifted subtly from national security to tech entrepreneurship, though it still encourages government service. My wife and I have traveled to several of the Hertz alumni meetings.
Lawrence Livermore Laboratory, Interactions with Teller
The summer your spent at Livermore, what was it like day-to-day?
Most of the summer students at Livermore were able to get security clearances. They worked "behind the fence" and were exposed very directly to the physics of nuclear weapons. There was a series of classified lectures, including a sort-of honorary lecture by Edward Teller. Most lectures were by people who were the actual bomb designers of the time. I never got my security clearance during the summer I was there. Why? Because I was "living in sin" with Margaret, whom I later married. It was the FBI that, at the time, actually did the investigations for security clearances. The FBI had decided that anybody "living in sin" was a bad security risk. That was the actual term they used.
Even then, Margaret and I just laughed at this. I mean, this was the end of the 1960s. There had been the Civil Rights movement, the Vietnam War, the Sexual Revolution. This was just ridiculous. But there was nothing to be done about it. There was a ground-floor corridor of one of the physics buildings at Livermore that had a movable wall. At the beginning of the summer, a few offices were on the classified side of the wall. Everybody on the other side was waiting for their clearances. In the course of the summer, the wall was moved step-by-step, compressing us miserable people who hadn't yet gotten their security clearances into smaller and smaller numbers of offices.
No door in that wall, of course. The building had an entrance at one end that was inside the barbed wire fences, and an entrance on my side that went out into the open area. You would walk a quarter of a mile to get to the other side of the building, going through a badge check, fence gate, and so on. By the end of the summer, there were just two of us in the one office left on the unclassified side, me and Lou Witten. Lou was a professor of general relativity at Syracuse. He was my father's age and had been a Communist sympathizer in the late 1930s. He was never going to get his clearance. Lou is, of course, now famous as the father of Ed Witten. Some years later, I got to know Ed, who always treated me with a respect I didn't deserve, as if I was of his father's generation, because he knew that his father and I had spent a summer sharing an office.
That's very funny.
Well, it was also a wonderful coincidence, because Lou taught me a lot of general relativity that summer. I was already going to Caltech. Kip was already someone I thought I might want to work with. And by the time I was done with that summer, and Lou had taught me a lot, I knew that when I arrived at Caltech, I wanted to work with Kip.
Good. Okay. Let me tell you an Edward Teller story way out of time sequence. I don't mind wasting your time with stories that are not revealing of me very much, but are revealing of some of these other famous figures that…
I’m always in the market for a good Edward Teller story.
So, I had worked that summer in Livermore. And I had met with Teller a few times subsequently, met with Lowell Wood many times. And I was a former Hertz Fellow who had gone on to be something of a success. And it's now almost fifteen years later, the mid-1980s, and I'm at Harvard. I'm a full professor, and I'm a new department chair. An establishment figure, all of a sudden.
One day I got a phone call from Mike May, who had been the Laboratory Director the summer I was there as a student. Mike was no longer Director, but still very influential at the Lab. He was calling because he had a problem. There was a vacancy in the position of Associate Lab Director for Physics. Roger Batzel, the Lab Director, was also soon to retire. Traditionally, Associate Lab Director for Physics was the stepping-stone to become Lab Director. Edward was pushing for Lowell Wood to fill the position—eventually to become Director.
Everyone else at the Lab, although they respected Edward as their Founding Father (he was by now a pretty old man), knew that it would be disaster for the Lab if Lowell got the job. All of the Livermore Lab establishment was arrayed against Edward and Lowell, but they had to proceed covertly, because Edward was still around, and he was still very politically influential.
But Mike May had a plan. He would nominate me for the position, even though I had no interest in the job whatsoever. On the basis of my having been a Hertz Fellow, and having been Edward’s summer student, I would inevitably be one of those invited to be interviewed. I would come and give a great interview. Particularly great, because I didn't want the job, so I could be even more freewheeling than usual- I could promise anything.
In effect, I would be the anti-Lowell Wood, like particle and antiparticle. Lowell was the career Livermore person. I had gone off into academia. Lowell was politically close to Teller. I was a Harvard academic—the exact opposite. In practically every respect, Lowell and I could meet and annihilate as particle and antiparticle. The result would be that the job would go to a neutron, a dark horse candidate, a very measured, very politically adept guy, John Nuckolls, who had also been a mentor of mine, a wonderful man.
Nuckolls did in fact get the job and later went on to become Laboratory Director, so Mike May’s plan worked perfectly. But there was drama along the way. I flew out to Livermore and started doing individual interviews, which went very well because almost everybody high-up in the Lab was in on the plan. On the search committee, there must've been one or two people who were Teller acolytes, and then everybody else was establishment Livermore. They knew that I didn't really want the job, and they knew that in the end, they were supposed to pick John Nuckolls.
But I was playing my assigned part and was scheduled to meet with a lot of people, including Lowell Wood, because, after all, he had been my mentor. Lowell kept me waiting in his outer office half an hour, opened the door, looked at me, and said, "Et tu, Brute?" Slammed the door. That was all I saw of Lowell.
Lowell must've then gotten on the phone, however, because within fifteen minutes, an admin person came rushing in. "Doctor Teller wants you to have dinner with him tonight, at his home near the Hoover Institute, across the Bay." I said, "Sure. Why not?" They cut short my schedule. A car arrived to take me across the Bay to Edward Teller's house. It was just Edward and his wife, Mici, who I had never met before. There was a Mexican cook, who served the meal and then left the house so that just the three of us could be alone. During the dinner, Edward began by saying, "We cannot talk physics or policy at dinner because the lady is not a physicist. And she is Hungarian." This was a little Teller joke. Mici was no more nor less Hungarian than he was.
At dinner, Edward showed off that he had actually boned up on my research. Despite his joke about Mici, he asked very sensible questions about my astrophysics research at the time. He was very complimentary of me. That was first part of the dinnertime conversation.
But soon, the talk turned political. He started asking what I thought about nuclear arms control and such things. It became clear that he was trying to draw me out to give extreme liberal views that he felt he could later use to discredit me. Extreme meaning typical of a Harvard professor, not any more extreme than that. I thought, "Oh, wow. This is great. Two can play this game." I was effusive about how important it was to have arms control treaties with the Soviet Union, and trust-but-verify, and all these things that in normal times, would've gotten Edward angry. But no, he was just filing it all away. Maybe he even had a tape recorder running. Who knows? And I thought, "Well, this is great." I was developing the fantasy that I could be the one who could turn Edward Teller into a liberal. I would just find the right compelling and logical arguments, and he would say, "You're absolutely right, Bill. I never thought of it that way before." I won't try to do this in a Teller accent, "I've been completely wrong for the last forty years, and I'm going to announce tomorrow that I support the nuclear freeze movement, and I'm going to campaign for Teddy Kennedy." So that was the fantasy. Didn't happen.
Then after dinner, Mici excused herself, and we went to the living room, and he served me brandy, a real gentleman. He said, "Bill, I understand that you're interested in coming to the Laboratory." This was the only small lie I had to tell: I had to say, "Well, yes, I'm interested in learning about the opportunity." Teller said, "We need you at the Laboratory. I fully support your coming to the Laboratory. I'm going to work to get you a position at the Laboratory much better than the merely administrative position of Associate Director for Physics, which someone less qualified than you, like Lowell Wood, will be able to do. I want you to come to the Laboratory to be able to do anything you want, at a high salary, for as long as you want, with many post-docs as you want, fully supported and guaranteed by me. Of course, you must promise in writing that you will have no interest in anything administrative."
I said, "Thank you, Dr. Teller. That is such a generous offer. I'll have to think about it."
And it was, "Dr. Teller"? That's how that went, always, "Dr. Teller"?
You called him Dr. Teller to his face, but you referred to him as Edward. That was the way that you signaled that you were in his inner circle. I don't think he would've chewed your head off if you called him Edward, but no one did. Talking physics, he was a lot of fun. And as I say, John Nuckolls got the job.
Well, people are complex.
Going back now to when I was just entering Caltech, I should tell the story, though, of Teller's letter to my draft board. At some point, graduate deferments were ended, and I was, once again, in danger of being drafted. But draft boards, which were all community-based, could still award deferments in very special cases for people whose work was essential to the national defense. If you were an aircraft engineer at Lockheed working on military fighter jets, then your draft board could choose to defer you. Each case was decided at the local level.
Edward wrote a letter to my draft board saying that I had been working at Livermore on the national defense, and I was now continuing my work at Caltech. The letter went on, and it was way over the top. Clearly, it must've been written by Lowell and signed by Edward. My draft board was in Cambridge, Massachusetts, because that's where my parents had moved right before I had to register for the draft. The response to Teller’s letter was that I received an induction notice and was ordered to report to downtown Los Angeles the next week for induction into the U.S. Army.
I called up my draft board. There was a nice lady there who worked at the front desk. I said, "Oh, hello, I'm one of your people. I'm a graduate student at Caltech. You were supposed to get a letter from Edward Teller, and he was supposed to explain to the board…" She said, "Wait, let me get your file." She came back, she said, "I have your file, and I see this letter from Edward Teller. The board read and considered that letter. They thought it was a very impressive letter. But they didn't know who this Edward Teller was."
Draft boards were ordinary citizens chosen to serve locally. This shows that there were limits to even Edward's influence. I did manage to avoid being drafted. First by failing the physical exam on something temporary. Then, when they had instituted the lottery, I drew, by pure luck, a high lottery number.
Caltech, First Work for Tombrello
Was Caltech a homecoming to some degree? Or did it feel like a new institution that you didn't have any particular connection to?
Coming to Pasadena was certainly a homecoming. I had no particular prior experience with Caltech. While my father was there, his office wasn't even on campus. All the time I was growing up, the seismological laboratory was in the Linda Vista hills. At that time, seismometers had to be on bedrock, and the data couldn’t be remoted to another location, because everything was analog. Margaret and I were married the week before graduate school began. Margaret knew Caltech very well, had been the teenage kid there as part of her parents' social circle in Kellogg Lab, the nuclear physics lab within the physics department. At the time, high-energy physics was in Kellogg as well.
Very soon after I arrived to Caltech, I went in to see Kip and said, "Oh, hello. Can I work for you?" And he said, "Are you in danger of being drafted? Because I only accept first-year students officially if it's something that they need to tell their draft boards." I told him that I just failed my physical and had one year before the issue would come up again. He said, "Well, then I'm not going to take you on. But you can start attending the seminars in my group, and you can see if there's a research problem you might be interested in." We got to talking. I was primed by my summer compressed into the office with Lou Witten to know what was going on in general relativity research, at least at a superficial level.
A lot of what was going on was Joseph Weber's apparent discovery of gravitational waves. These are the gravitational waves that LIGO actually did first discover forty-seven years later. Weber’s experiment was faulty, something on the continuum between "just plain wrong" and "sad, self-delusional fraud". That was not obvious at the time, however. All the general relativists hoped that the discovery was real. Anyway, I was all up on Joseph Weber's experiment, and this must have impressed Kip.
Tommy Tombrello, who was a nuclear physics professor in Kellogg Lab, was thinking of doing a gravitational wave experiment. He never actually did. I think he was preempted by Richard Garwin's experiment, which definitively disproved what Weber claimed to have been seeing. But that was later. Kip said, "Why don't you go talk to Tommy Tombrello about gravitational waves?" So, I did that. Tombrello said to me, "Look, I'm a nuclear physicist. I haven't done anything like this before. I talk to Kip all the time, but it will be your job to explain to me how to build an apparatus. How does this all work, this gravitational wave stuff?" I was flattered to be treated so seriously.
I worked out a whole semi-Newtonian way of thinking about gravitational waves. Much of this seems trivial now, but at the time, gravitational waves were still considered very exotic and not even necessarily real. There had been a long historical debate about the reality of gravitational waves. Were they simply a choice of gauge, and they could be made to go away? Or did they have a physical reality? This had been settled by the time I arrived in graduate school, but not too many years before. Today it's obvious, since gravitational waves show up in the Riemann tensor, which always manifests as a physical, tidal gravitational force. But it still wasn't fully settled that there was a completely local, Newtonian way of thinking about the interaction of a gravitation wave with an apparatus. There were people who said, "Well, you can think of it that way in some approximation, but there will be effects of the same order that you’ll miss."
So, I went off and, learning some general relativity, thought of a way of expressing it in a way that made clear that a Newtonian stretching force was not only there, but it was all that was there. The local Newtonian picture described the wave’s interaction on masses, its interactions with photons and mirrors, and all that stuff. I was pretty proud of myself that I had developed all this.
But when I went back to Kip to tell him about this, he didn't like it at all! I can say this safely, now that Kip has won a Nobel Prize and been recognized as the great physicist that he always was, but Kip was not an Ed Purcell kind of physicist. He was not an Edward Teller kind of physicist. Kip starts with the equations. You start with the equations, and then you work out the physics, and then maybe you say, "Oh, I could've thought of that." But it's irrelevant that you could've thought of that. It came from the equations.
And that wasn’t your approach?
Well, I was doing for gravitational waves what Purcell does for electromagnetism in his freshman textbook. It starts with physics, not mathematics. He draws lines of force. I was the first person to think about lines of force for gravitational waves. Is that an important discovery? No, not at all. It's just a perspective that lets you visualize things more concretely. I was taking relativistic equations, carefully calculating Newtonian limits (so that I wouldn’t miss any effects) and showing that you could cast the equations in a way that looked pretty much like electromagnetism, except it's quadrupole instead of dipole, implying some obvious changes.
Any modern researcher in the field of gravitational waves or general relativity will say, "Well, of course. We've always known that viewpoint. That perspective has been around since, oh, I don't know, the mid-seventies." Exactly so. These viewpoints get established because someone actually describes them and evangelizes for them; in this case, me. And, pretty soon, more effectively, Kip.
It’s a different approach?
There's a kind of doing science where your single-minded goal is to discover things. You want to have things with your name on that you've discovered. Okay. I've been lucky enough to do a little bit of that, but not much. It’s not even what I'm proudest of. What I'm proudest of are instances where I've succeeded in developing, feeding into a field, new, different viewpoints, or perspectives, or ways of thinking about things. And the first example of that, in my career, really was gravitational waves.
Towards the middle of the year, Kip scheduled me to give a talk to his group on gravitational waves. It was an honor- me, a first-year graduate student. John Wheeler, who had been Kip's graduate thesis advisor at Princeton, was visiting at the time. Wheeler was also Feynman's graduate thesis advisor. So, I gave my talk, and it was terrible. It went over like a lead balloon. Wheeler fell asleep, and woke up towards the end, and gave me (maybe I'm over-dramatizing) a look of sheer horror at the fact that I was still talking. Normally at the end of Kip's seminars, people applaud politely. This time, no one said a word. People filed silently out of the room. Richard Price, Kip's senior graduate student at the time, came to my office that afternoon and said, "Bill, it's not as bad as you think." I said, "How bad is it, Richard?" "It's not as bad as you think. If Dick Feynman had been there, he might've led you to the right path."
This forced me to some serious introspection, because I was pretty sure that what I had done was good work. It was my first realization that the different styles of physics are different not just cosmetically. They're profoundly different. And a field advances, in part, because people in different places have different ways of thinking about things. Kip’s was a group of mathematical general relativists. I had completely failed to speak their language. My talk would've worked for Ed Purcell, bless his heart; would've worked for Edward Teller; and surely would've worked for Dick Feynman, because Feynman was so smart that he could understand, speak, and translate among all the styles. Feynman might have thought the work was trivial, but he wouldn't have been mystified by it.
The last opera written by Richard Strauss is Capriccio. It’s a wonderful, long, musical debate about, who is more important, the poet or the composer? Which has the greater primacy, poetry or music? Since it’s Strauss, you know how it has to come out. In physics, the similar debate is, which is more important, physical intuition, or equations? After my talk, I felt like I had been cast as the poet in that opera, on the losing side.
Caltech, Social Milieu
I want to touch a little bit more on this idea that both you and your wife had come from Caltech families, and you sort of had arrived back home, at least intellectually. And so, I want to get a sort of broad-based view on how your intellectual pedigree as a couple influenced, at least from the beginning, how you thought you might pursue your graduate education at Caltech.
Well, it was a funny situation. In my first year in graduate school, also Margaret's at UCLA, we kind of lived dual lives in the Caltech community. Socially, we were "that cute young couple" at Kellogg Lab parties. And Kellogg Lab was a real partying place. We partied with Willie Fowler, Dick Feynman always came to the parties, and Margaret's father, Tommy Lauritsen often hosted them. Tommy was a heavy drinker at these parties, not always to good effect. So that was sort of our social milieu.
But then, by day, we were struggling graduate students. I would limp in, hung over, to Willie Fowler's 8:00 am course on nuclear astrophysics, and, of course, it was a different Willie Fowler and a different me. I was the lowly graduate student; Willie was the professor who always insisted on lecturing at 8 o'clock am. So that was just a strange thing to get used to, or for Caltech and that community, the Kellogg Lab community but that included Kip Thorne's group in those early days, getting used to that.
In terms of classwork and graduate school life, that was just the hardest year of my whole life. Harvard had not prepared me for Caltech. Of course, as we’ve discussed, my version of Harvard was spending time at the radio station and trying to spend as little time on physics as I could. You could do that at Harvard. Caltech was, and I gather still is, the Marine Corps of graduate schools, where the courses were very intense, and all of the preparation was for the candidacy exam that came at the end of the first year. You took your candidacy exams, they were graded, and you were told what your rank was in your class. That rank enormously affected, especially for theorists, which professors were willing to take you on as thesis students.
But you embraced that duality?
Here's what I would say about that. I could not escape the reality that I was Frank Press's son, and Margaret was Tommy Lauritsen's daughter. We were a kind of junior- very junior- power couple. But there was also in front of me the example of Tommy and his father, Charlie Lauritsen. Charlie was, by almost any judgment, the better physicist, notwithstanding his having gone off in World War II and never quite coming back to pure research. Tommy was always in his father's shadow, professionally and personally. Margaret and I enjoyed the Kellogg social scene, and we didn't mind being the cute young couple. But I didn't want that social life to become defining of who I was professionally, a struggling first-year graduate student. I especially didn't want to end up like Tommy, in his or my father's shadow.
You mean you wanted to be taken seriously on your own terms.
Not just wanted to. Expected to. I was planning never to play the social connection card. Now, realistically, you can say that I didn't have to play it. It was just there. Maybe so, and there was nothing I could do about it. Still, I was determined never to play that card intentionally. So, yes, it was a little uncomfortable sometimes, palling around with the older Kellogg Lab folks at these parties, which-
But you and your wife had to know this coming in. This could not have been a surprise to you.
Yes. Not even a Faustian bargain. We didn't have to sell our souls. It was a choice to go back to Pasadena. I have a good friend, just my age, a very distinguished physicist. He really fought hard to get where he is. He comes from a working-class background, from a family that discouraged his going to college, and so on. And it's just amazing the stories he has about his struggles to become, as he has, quite a well-known physicist. He and I were joking around about how maybe each of us should write his own memoirs. I must have been telling him some Dick Feynman story at the time. I said, "No, no. You should write your memoirs. Because yours is an amazing story, what you went through to make it in life. The trouble with my writing a memoir is…" I didn't even have to say it. He finished the sentence for me. "Your story, Bill, is about how you had every possible advantage in life and nevertheless managed to succeed."
(Laughter) Well, on that note, let's end it here.
This is David Zierler, Oral Historian for the American Institute of Physics. It is August 18, 2020. I'm so happy to be back with Professor William H. Press. Bill, thank you so much for joining me again.
My pleasure, of course.
Joining Kip's Group, Gravitational Waves
Let’s talk more about what led to your interest in trying to join Kip's group. What was going on there at the time that attracted you?
Well, I think I was no different from most graduate students with only the vaguest idea of what was really going on, what was exciting in physics. But at the time, Kip was the rising young star in the department. I knew that I wanted to be a theorist. We already discussed how, as an undergraduate, I pretended to be an experimenter; but I just wasn't good at it. So, I went to Caltech like everybody else, thinking that if I wanted to do particle theory, I wanted to work with Feynman. And if I wanted to do something different, well, gravitational collapse and gravitational waves were very much in the air. Weber had just done his experiment, which seemingly detected gravitational waves.
But in terms of working with somebody like Kip or Dick, how much overlap is there really in terms of what the student him or herself is interested in? In other words, you have the kinds of physics that you want to pursue. How do you decide who to work for?
Well, I wouldn’t say that you have a lot of agency in this regard. You arrive as a lowly graduate student. If you're lucky enough to do well on the candidacy exam, then you make the rounds. Now, I had more independence than most because I was a Hertz Fellow, so I was financially independent. I didn't have to find a thesis advisor with the funds to support me. Kip, when I first went to him, said that, after a year, he would see how I did on my candidacy exams, and see how many open slots he had in his group.
What did you do with Kip's response? How did you respond to that?
Well, that was fine with me. Of course, I didn't know whether I would do well on my candidacy exams. Like all first-year graduate students, I had forebodings. As I said, Harvard had not prepared me. So, it was a rough year. I think all graduate students have to have fantasies about what are they going to do when they flunk out. Margaret and I were going to open a small restaurant and become short order chefs or something like that. It's questionable how much we truly believed that, and how much you just need something to grasp in graduate school to get you through. And, luckily, I did do well on my candidacy exams.
Let me mention something ugly about those times. I came in number two in my class in the candidacy exam. The person who came in number one was a student from China. But the lore in the department was that a Chinese student always came in number one, and that he should be discounted. So, the perceived top of each class was the non-Chinese student who came in as number two.
Bill, let's talk about the sort of origin story of LIGO. When Kip talks about LIGO, he prominently mentions the review article that you did with him on gravitational wave astronomy. Where do you situate yourself specifically in that broader origin of LIGO and gravitational wave physics?
We already talked about my work for Tombrello, and how badly it went when I tried to give a seminar on that. Kip, at first, didn’t like my perspective at all. But he did come around. By the time I was Kip's student, I would say ninety percent I was learning from Kip, as one should as a graduate student with a thesis advisor. And maybe ten percent, Kip was learning from me, because I was giving him this sort of more physical viewpoint.
So to the extent that I contributed to that field, I was an early influencer in understanding that gravitational waves were like electromagnetic waves: When you think of them as metric perturbations, they may seem esoteric, but locally within the wave zone, there's the equivalent of an electric field, there's a magnetic field. On time scales short compared with the frequency of the wave, they're quasi-static. The electric field makes charges move; the changing magnetic field makes currents flow. Exactly that viewpoint goes over to gravitational waves. You can understand essentially all modern gravitational wave experiments in this kind of quasi-Newtonian way that we now take for granted.
What were some of the bigger questions that Kip had in mind about gravitational waves?
From the very beginning, Kip had a vision not just of gravitational wave physics, but of gravitational wave astronomy. That was the title of our review article, and it was Kip's title, "Gravitational Wave Astronomy." It was a controversial phrase to put on the title of a review article. Gravitational waves had not even been detected yet except possibly in the Weber experiment, which had fallen into grave doubt.
And why? What was the doubt behind that experiment that didn't make this a done deal?
Weber’s supposed gravitational waves were hugely, many orders of magnitude, stronger than any astrophysical source that anybody could dream of. When I started working more mathematically on gravitational waves with Kip, people were still trying to find a way that maybe Weber could be seeing a real effect. So, for example, Charlie Misner at Maryland and his group had an idea, really just a phrase, "gravitational synchrotron radiation." Could there be astrophysical sources close to black holes that could beam gravitational radiation?
Richard Price was working on that, and Kip assigned me to a kind of an apprenticeship with Richard, I think to train me into being a more mathematically oriented physicist. My physical intuition was all well and good, and it did find its way into the review article and into Kip's textbook with Misner and Wheeler. But, fundamentally, if I was going to be in Kip's group, I had to learn to be more of a mathematical physicist. Richard Price was supposed to teach me. Well, that didn't lead anywhere in terms of the Weber experiment. The Weber experiment disappeared when Garwin and others definitively failed to confirm it.
One of the problems that Kip assigned to Richard Price and me to work out, it could be done in perturbation theory was a speculation due to Jim Bardeen, that if you shot a small particle into a black hole, you could get gravitational waves whose wavelength were characteristic of the size of that small particle, or the velocity of that small particle, or something that could give hope of generating these relatively high-frequency gravitational waves that Weber was claiming to see.
What Richard and I showed was that the gravitational waves that you got out always had the characteristic scale of the black hole itself. That, I think, was the first hint to me that I should use these perturbation equations to understand black holes as dynamical objects. When you banged on a black hole by shooting something into it, what you got out was characteristic of the black hole, not characteristic of what you shot into it. When you shoot a BB gun at a bell, you get out the sound of the bell, not of the BB.
Every scientist has something in his or her head that they consider their first discovery, their first understanding of something that the rest of the world didn't understand yet. For me, it was my paper, "Long Wave-Trains of Gravitational Waves from a Vibrating Black Hole." I wouldn’t claim sole credit. I was immersed in a research group and in a whole milieu, but I would say this idea, that you should talk about vibrations of a black hole and not simply gravitational waves from things external to the black hole, was a viewpoint that I originated. Kip, and others in his group, and soon others in other places, adopted this, and it became the standard view.
Nowadays, when LIGO talks about analyzing events, they talk about dividing the event into an in-spiral stage, which is semi-classically the gravitational waves from two-point objects circling around each other and spiraling in. Then, the system goes over to what they call a ringdown stage, where you're no longer seeing the object spiraling, but you're seeing properties of the black hole itself. Seeing, in fact, normal modes of the black hole. I think I was the first to use the word "ringdown" for a black hole. Or "normal mode." But who can know now?
And in terms of your workflow with Kip, did he essentially give you a problem to work on that he needed worked himself? Or did you have sort of the inclination, or the initiative, or the independence even, to come up with that problem for yourself?
Kind of in-between. Kip was great at understanding when problems were ripe, when the equations existed, the technology existed. And then Kip would assign problems to his graduate students or to teams of his graduate students. Kip always believed that there should be collaboration within his group, that no one should be hiding results from anybody. In terms of communication between his group and other groups, I think we would depend on Kip for telling us when we had results that were far enough along that we should share them with other groups, even well before publication.
Niche in Computational Work
One major thing that I brought to these collaborations among graduate students in Kip’s group was that I understood what you could do with a computer. I understood that better than my peers because I had been a computer hacker at Harvard, which really got me into the depths of the machines. And I had worked at Lawrence Berkeley lab and learned to program on what were big computers at the time. And then, at Livermore Lab, I had picked up the culture of, "Here's a physics problem. How do we compute it? How do we map it to what a computer can do?" This was a culture that was then was quite new in theoretical physics.
Richard Price was two- or three-years senior to me. He would say, "Try to derive this result from these equations." And I would say, "Well, why should I derive it? I can just take your equations, and put them on the computer, and see what they do." These were ordinary differential equations. Nowadays they would be trivial to calculate, and even then, not too hard. And he would say, "Really? Really, you can do that?" So that became my niche, this kind of interface between the analytic calculations and what could be done on computers at the time.
That was a niche that continued throughout my graduate career and beyond. I always needed a more mathematical collaborator to work with in relativity. I was never a truly facile mathematical relativist. My principal collaborator soon became Saul Teukolsky, who was one year behind me in graduate school, who had come to Caltech from South Africa (which, at the time, was considered very, very exotic). With Saul, it almost was the inverse of with Richard. I would go to him and say, "I'm about to program this on the computer. Would you like to collaborate with me on it?" And Saul would say, "Why work it out on the computer? I can do it analytically." So, I think in both directions, that was a very fruitful interplay.
Who was on your PhD committee?
Kip, of course. Dick Feynman. Jim Gunn, because by then, I had actually done some work with Jim Gunn on gravitational lensing. And Tombrello. Generally, in Kip's group, it was never clear when Kip was going to let you get out. And that was really the way we phrased it. Kip's view was, "Don't think of yourself as a graduate student, think of yourself as a scientist. Do research. Publish papers. When the time is right, I'll tell you that you’ve done enough. Then you’ll bind those papers together and schedule your committee."
In my case, I had several papers pretty early in my graduate career that had gotten me a reputation outside. By the end of my second year in graduate school, I already had, if not a formal assistant professor offer from Berkeley, certainly a very strong feeler. The Berkeley people were saying, "Why don't you get your degree right away and come to Berkeley?" Essentially preemptively. George Trilling and Charlie Townes at Berkeley were behind this. Kip didn't think I was ready.
In what sense?
I don't know. Certainly, I had done enough work, publishing two or three papers on different subjects. I think it was a question of scientific maturity. May also have been a question of emotional maturity, because I always had the reputation in graduate school as a kind of troublemaker.
Well, you're the hacker from Harvard, right?
Yeah. I think Kip just didn't see that I would make a good transition yet, that I would suddenly appear at Berkeley, be an assistant professor, be teaching, never having been a post-doc anywhere else and gaining experience there. Kip told me, "You're not ready yet. Put off Berkeley."
Bill, if you'll permit me, perhaps also there's the possibility that you were just continuously useful to him, and he didn't want to let you go.
I don't think so. That may have come into play a little later when I actually did leave Caltech. At the time, in general relativity, there was going on what some of us later called the "Children's Crusade." There were a few groups producing lots of PhDs: Wheeler at Princeton, Kip at Caltech, Misner at Maryland, Chandrasekhar at Chicago. The assumption was that those groups were going to scatter PhDs to every physics department in the world. Everyone was going to be hiring a relativistic astrophysicist or a mathematical relativist. So, I don't think that they were trying to keep their graduate students. They were trying to spew them out into the world.
Scientific Heroes: Feynman and Chandrasekhar
Bill, on that point, when did you meet Chandrasekhar?
I was lucky, early in graduate school, to meet and come under the influence of two people who became my scientific heroes. Not necessarily my heroes as human beings, but certainly my scientific heroes. Those were Dick Feynman, who we've already mentioned, and Chandrasekhar. I got to know Feynman because he acted, within Kellogg Lab and beyond, as a kind of a godfather to Kip's group. He had been instrumental in bringing Kip to Caltech.
And Kip and Dick were close personally?
Yeah. A generation apart in age but got along very well. Feynman made a point of regularly, one or two days a week for years, eating lunch with Kip's graduate students in the Caltech cafeteria. I think everyone who came through Kip's group at the time felt that Feynman was somebody they could talk to at lunch. More likely there than by going to his office, at any rate.
Now, Chandrasekhar, that was an interesting story. After most of his huge contributions to astrophysics in the 1930s through 1950s, Chandra had become the Editor-in-Chief of The Astrophysical Journal, a position that he held for twenty years. In that role, he hugely affected the whole direction of astrophysics.
And he was self-conscious of the power of this perch. He used this to advance the kind of science he wanted to advance.
Absolutely, and a lot's been written about it. And I think he himself has written a lot about it. Chandra came from what was a tiny tradition of very mathematical astrophysics. Eddington, his PhD advisor was very mathematical. But the idea that you could do astrophysics at the same mathematical level as other areas of theoretical physics, Chandra deserves the credit for introducing that, first via his own works, and, in the mid-fifties to mid-seventies, as editor of The Astrophysical Journal.
When, after twenty years, Chandra retired as editor, he decided to take a year's sabbatical in Kip's group at Caltech. His office was two doors down from the office that I shared with a couple of other graduate students. Chandra loved the fact, and he kept track of this, that the average age of his collaborators had remained unchanged—the number was something like twenty-seven- from the time that he was twenty-seven to the time that he was now in his early sixties.
So, Chandra also made a point of eating lunch with all of us graduate students, and haunting our offices, and trying not to act like the major founding father of modern theoretical astrophysics that he was, but to act like a graduate student. He really didn't impersonate a graduate student very well. He always wore a dark suit and tie; he always had a very formal manner. And he was very self-conscious about exposing his ignorance on things. But he had come to Caltech to learn new things, in fact, to learn general relativity. Chandra's later work on black holes, his famous book on black hole perturbations, originated completely from his time in Kip’s group.
Chandra would lurk in the hallway, and he'd peer into an office that might normally have three graduate students, and if you happened to be alone in that office, he would sometimes come in, and shut the door (which we generally never did) and he would sit down. That meant that he wanted to ask a stupid or naive question, because he really wanted to learn this stuff. So, in this way, I got to know Chandra quite well.
You got to know him, but in what ways was he influential on your thinking?
Of course, I knew of Chandra before I met him personally, from his books, on stellar structure and so forth. So, it was interesting to see a bit of the process behind those mirror-perfect works. I came to realize that success, in theoretical astrophysics (and probably other fields) had three axes. The three axes were: (1) choice of problem, how to pick just exactly the right ripe problems to work on. And (2), we've discussed already, physical intuition, being able to understand physical systems intuitively. And (3), sheer technical ability, including mathematical ability. That might also include computational ability.
What I realized was, you didn't have to master all three. You could figure out where your talents lay. You could be a pretty good scientist if you were really good at one of those three axes. The number of scientists in history who were really outstanding at all three was only a handful. The only one of those I ever met was Feynman. What was interesting in getting to know Chandra, was to see that he was good in exactly two. Chandra was utterly off-scale in technical ability, and outstanding in choosing problems to which to apply that technical ability. But he really didn't have much physical intuition. I don't think he even cared much about physical intuition. He relied on others to bridge the, for him, small gap between deciding to work in an area and then digging into it with his incredible applied mathematical ability.
With me, when he would shut the door and ask stupid or naive questions of a graduate student, he was bridging that gap. He already understood the big picture, and he was preparing to dive into the mathematics. He came to me for physical intuition. It was very flattering. It certainly influenced me. I realized that my greatest strength was in physical intuition. I was not hopeless at choosing problems, but, better, I should listen carefully to others for that dimension. This will come in throughout the story later on. As for the third dimension, technical ability, I could do well on the candidacy exams at Caltech, at that level, but I always needed to seek out collaborators who really excelled technically. In graduate school and later, that was Saul Teukolsky. Later at Harvard, it was my colleague George Rybicki, my student Phil Marcus, and others.
Now, of course, another direction at this time is Kip's friendship with (Yakov) Zel’dovich. I'm curious where you fit in in terms of that friendship and what influence you might have felt as a result of Kip’s friendship with him.
I don't know what the origins of Kip's connection to the Moscow Zel’dovich group were. They were well-established by the time I arrived on the scene. Kip had traveled to Russia when that was still quite uncommon. He visited Moscow a couple times a year.
Did the Zel’dovich group come to the States as well?
Never, because, at this time, they couldn't travel. Not just Zel’dovich himself, but even his young people could rarely, rarely travel. If they could get even to Western Europe, that was amazing. Kip spoke a bit of Russian. I had studied Russian in college. Kip's Russian was a better than mine but by no means fluent. When I arrived in Kip's group, one of the things going on was translating the two-volume set of books by Zel’dovich and Novikov on relativistic astrophysics from Russian into English. The way that worked was there was a translator, who was an American, I believe, who had turned out to be not very good. Kip had agreed to be the scientific editor of this translation but not the translator himself.
So, chapter by chapter, the translator would produce manuscripts and send them to Kip. Kip would mark them up in his characteristic red pen, very densely, fixing all the science, and then he would provide a Xeroxed copy of each chapter to each of his graduate students, the six or eight of us in his group, including me as a first-year hanger-on. Then, we had a weekly book club where we would discuss each chapter.
In keeping with my wise guy reputation, I would complain, not about the science, which I was only then learning, but about the quality of the English, which was terrible. It was bad even after Kip fixed it up, although Kip is actually a very good writer. My guess is that he planned to go back over it at a later stage and edit the English. It was things like "Russian writer of humoristic stories" when it should have been "Russian humorist". Well, the result of my complaining was Kip said, "OK, why don't you copy-edit each chapter and fix the English?" I took on this assignment, not only reading the book chapter by chapter for content, but also copy-editing.
Anyway, the first time I met Zel’dovich, I was just finishing up as a graduate student. There was a conference in Warsaw, I think the 500th anniversary of Copernicus, actually a series of conferences in Poland. Zel’dovich was not allowed to travel to the West because of all of his involvement with the Soviet nuclear weapons program. He was rarely allowed to travel even to the Eastern Bloc. But he was allowed to travel to this meeting.
I had just done this work with Paul Schechter that later become known as, it's about the only thing named for me, the Press-Schechter formalism. The question was, how does structure grow in the universe from the smallest scale, maybe a first generation of stars, then star clusters, then forming galaxies, then forming clusters of galaxies? (By the way, that whole picture, which was called the "bottoms up theory" turned out to be somewhat misguided.) I wanted to sell Zel’dovich, the great man, on this work that I had done with Schechter. I approached him at the conference, and I said, "Can I come talk to you about this work I've done?" Well, it turned out Kip had already talked to him about my work. He said, "Your paper is wrong. Go talk to Doroshkevich."
Doroshkevich was a younger member of his group, a very dour guy. He looked like he could be a KGB agent when, in fact, no one in Zel’dovich's group was a KGB agent. Zel’dovich kept his group absolutely free of political influence, and, as the father of the Soviet A-bomb, had the clout to be able to do that. I spent an hour with Doroshkevich, and I managed to convince him that my paper was not completely wrong. He checked back with Zel’dovich and announced, "7:30 tonight in Yakov Borisovich's hotel room." I arrived at this hotel room, and it was like a dentist's office. Zel’dovich had scheduled appointments every fifteen minutes with everybody who wanted to talk to him from the West. But it was just his hotel room, so we all sat in order on the bed waiting to be called over to the other corner of the not-very-large room for our fifteen minutes with Zel’dovich.
In my fifteen minutes, I managed to convince him that the paper wasn't completely wrong. Zel’dovich said, "I will think about this. Talk to me at the next conference." That was going to be in Cracow. I had achieved a high status, to be invited back to talk to him a second time. Zel’dovich was a bulldog of a man. He had a crew cut; he was built like a wrestler. Not necessarily heavy, normal-sized, but just very muscular and very dismissive in that Russian style of, "I will tell you when you're wrong, and after I tell you, don't bother me again."
At that next conference, I went up to Zel’dovich, and I said, "You said we could talk again." And he said, "Yes. Come to dinner with us." The us was Zel’dovich, Doroshkevich, and a couple of other Russians. We went to a Mleczny Bar just around the block. Mleczny means milk, so it was a dairy restaurant. It was a strange experience because all Poles hated all Russians. The only people the Poles hated more than Russians were Germans. The Russians treated Poland like a conquered colony. We sat down at the table, the waiter came over, and the Russians immediately spouted off their orders in Russian. The Polish waiter said, in Polish, "I'm sorry, I don't speak any Russian," which was ridiculous because of course he did. I then said, "Hello. I'm American. Do you speak any English?" And the waiter said, "A few words." The Russians then repeated their orders in Russian to me and I, from my halting Russian, translated them into English. It didn’t matter if I got it right, because the waiter understood their Russian perfectly the first time.
Over dinner I got Zel’dovich interested in some of the other things I was working on, and Kip must've told him that I was one of his best students. So, after dinner in the mleczny bar, Zel’dovich said, "Let us walk around together." And, of course, that was very Soviet, that you could only do business by twos. A walk in the woods, right? And we talked about my work, but he was also very interested in the Nixon presidency and what was going to happen to Nixon, who had not been impeached yet.
We made a bet. The bet was on whether Press-Schechter would turn out to be correct. Later that year, 1973, Nixon was in fact impeached. A rare Soviet visitor showed up at my office at Caltech and presented me with an autographed bottle of Vodka, the real Stolichnaya, autographed by Zel’dovich. It was inscribed "To Press, in paying off our wager." He had misremembered. He thought we had bet on whether Nixon would be impeached. He was sure that Nixon would not be impeached because he thoroughly subscribed to the theory that America was governed by a group of plutocrats in smoke-filled rooms, that Nixon was their man, that they were punishing Nixon for something, but in the end, they would save him.
So, that was very interesting. The other time I got to know Zel’dovich in a personal way was a couple years later, there was a summer workshop at Protvino outside of Moscow that was two weeks long, mostly with the whole Zel’dovich group there. Our group of Americans got to know all of them then. On that occasion, because he had mistakenly paid off the wrong bet, I brought a bottle of White Horse scotch with me to give to Zel’dovich. In the original bet, he had specified White Horse. I presented him with the scotch and explained that this was really the bet.
The part of Press-Schechter that we were betting on, and which was wrong, was not the part that later became well-known. It's interesting how papers can have right parts and wrong parts, and if you're lucky, it's the right parts that get remembered. Peter Goldreich came to me a few days later, and said, "Zel’dovich wants your glow-in-the-dark frisbee." I had brought a glow-in-the-dark frisbee because otherwise the only thing we were allowed to do for recreation would be to play soccer, where the Russians always won. I said, "Peter, why doesn't he just say to me, 'I want your glow-in-the-dark frisbee?' He knows that I'll give it to him." Peter said, "No, he's embarrassed because you already gave him a gift. You already gave him the White Horse scotch. So, he sent me as an emissary." Peter and I both thought this was quite funny.
Thus, I had the privilege of taking to Zel’dovich's room at this conference my glow-in-the-dark frisbee and saying, "Yakov Borisovich, please do me the honor of accepting this additional gift, a glow-in-the-dark frisbee." And he was gracious about it. He took the frisbee.
I also knew other things about Zel’dovich in an interesting way. This is a bit out of time sequence. One of my friends, a colleague at Livermore, was George Chapline, who also did work outside the fence and is well known as a physicist. In Zel’dovich's last years of life, Chapline visited Moscow and talked to Zel’dovich very extensively, then came home and shared things he had learned with people behind the fence at Livermore. It was clear that Zel’dovich was sharing with Chapline things that- this was before the fall of the Soviet Union, the Soviets would've been very disturbed at him sharing.
What was going on? He wasn't revealing any state secrets that would actually have been useful to the United States. Zel’dovich was a totally loyal Soviet. Interesting, because by ancestry he was Jewish, which might've given one pause. But the system had treated him well, and he was totally loyal. His beef was that the lore in the West wasn't crediting him with everything he had done on the Soviet nuclear weapons program. Zel’dovich wanted to leave a record, at least an oral record, with his peers behind the fence at Livermore, a record giving him proper credit for his bomb work. Specifically, the lore in the West was, roughly speaking, that Zel’dovich was responsible for the A-bomb, but Sakharov was responsible for the H-bomb. That is to say, the fission bomb and then the latter two-stage fusion bomb.
This is only hearsay, Zel’dovich to Chapline to me over a period of years, but my recollection now is that Zel’dovich's claim was that he had also been responsible for the key ideas on the fusion weapon, the Teller-Ulam idea, for example. He credited Sakharov with important contributions, but nothing like the magnitude of contributions that Sakharov was credited with in the West. Sakharov and Zel’dovich were close friends, but Zel’dovich never publicly spoke up to support Sakharov politically. That would have crossed the line of his loyalty to the state. Zel’dovich credited Landau more than we do in the West with being the other important figure in the development of the Soviet fusion weapon. With the fall of the Soviet Union and the declassification of many of those records, historians of the Soviet nuclear weapons program may have come to different conclusions. But that was the story I heard.
Graduate School, Other Interests, Job Offers
Let's zoom out a little beyond your thesis life at Caltech. What else were you doing during those years in the department or beyond?
Well, I was learning to channel my tendency to be a general troublemaker into more productive pursuits. The perceived wisdom was, "You're a graduate student. You do publishable papers. Write them up into a thesis and get out." After a certain early point, I didn't have any doubt that Kip would eventually let me out.
At Harvard, I got a lot of use out of problem books in the physics library. These are published books of worked physics problems at graduate-course level. In fact, I'd gotten through Harvard copying more solutions out of problem books than I had worked myself, always crediting them to their source, of course. At Caltech, Kip had just completed, with Wheeler and Misner, his monumental book Gravitation, which is still in print after fifty years. I organized a team with Richard Price, who had by then graduated, Saul Teukolsky, and Alan Lightman. The four of us, graduate students or recent graduates, together wrote what became the Problem Book in Relativity and Gravitation as an adjunct to Misner, Thorne, and Wheeler’s Gravitation. That was in no way helpful to my thesis or my scientific reputation or anything like that. It just seemed worth doing.
But I wonder if, in some ways, you were sort of operating above and beyond what a graduate student was capable of doing.
I guess I was. In that respect, the world recognized me as precocious. And that put me in demand. But it also gave me license to work on oddball things. There existed a tradition of work by physicists computing the characteristic size or mass of things in terms of the fundamental constants. For example, in terms of the h-bar, G, c, and the mass of the proton, why are stars the mass that they are? How high are mountains? And so on. Viki Weisskopf, MIT’s famous theoretical physicist, liked this kind of thing. I wrote a paper deriving the size of man, nowadays one might say the size of Homo sapiens, in terms of fundamental constants. How could one possibly do this? Well, you could figure out, from very general considerations, how big a planet had to be to hold an atmosphere, and for that size planet, how big could a life form be, such that given the strength of materials, when it fell over, it wouldn't break. Those kinds of arguments. There may have been a little more to it than that, but not much. It was published in the American Journal of Physics, which is a pedagogical journal for physics teachers.
Kip thought I was crazy to be working on something like this, because after the fun of working it out, it takes a long time to write something up as a paper, and deal with referees, and so on. Kip just scoffed, until one day, Viki Weisskopf happened to come to town. As a very junior physicist, I was invited to the dinner in his honor. Feynman was there, and Murray Gell-Mann, who had been Weisskopf's PhD student. And in this crowd, when he was introduced to me, Vicki said, "Oh, you're William Press. I was very impressed reading your paper that arrived the other day." It was my paper on the size of man, but thankfully Vicki didn't explain. Murray was hovering and his eyes bugged out. He never imagined that I was capable of writing an important paper. I don't think he ever found out the truth about what paper it was.
Still, however, I was a troublemaker. I never thought highly of the Caltech honor code. Caltech undergraduates were very strongly into this. All discipline among undergraduates for things like cheating was done by a court of peers, and without any faculty supervision. I thought this was crazy. I thought these so-called peers were more likely than sensible grownups to mete out excessive penalties. And, an oddity was that graduate students were thrown into the same system as an afterthought. So, we were judged by a kangaroo court set up by undergraduates.
The physics department graduate students were allowed one representative on this "honor court". I ran for the position on the platform, "I will vote to acquit in every case regardless of the merits." I won handily, since I was the only candidate, and in fact voted to acquit in all cases. Surprisingly, they all merited acquittal on the merits anyway. They were generally sad cases of foreign students who didn't understand what the rules were. Mostly, however, despite my vote, they got crucified.
Just in terms of the intellectual tradition, in terms of Kip working with Wheeler, did you sense that it was sort of preordained that you would go work with him at Princeton?
No, not at all. By the time that Kip decided that I was ready to get out, I thought I was over-ready to get out. Kip did then let me get my degree and converted me to a post-doc for about a year. That was just a seamless transition having to do with when in the calendar year degrees were awarded.
What about the transition from post-doc to assistant professor?
That's the more interesting one. By this time, I was the flavor of the month on the job market. I don't know how else to describe it. You know how these things work: When there is a number one candidate in a given subfield, he or she gets all of the first offers. I had an offer from Yale, that was Jim Bardeen, who had moved from Yale with a strong mandate to try to build a group in general relativity. In the end, that didn't work, and he went back to Seattle.
My assistant professor offer from Harvard was the strangest thing. Paul Martin, who was chair of the physics department at that point, called me and said, "We'd like to offer you an assistant professorship." I said, "Oh, wow. I'm very grateful. I didn't even apply." And he said, "Well, we've already voted it for you. But the only paperwork we can find is your undergraduate transcript. Could you send us some paperwork?" It was clear that this both was and wasn't a serious offer. The Harvard Physics Department voted this because Ed Purcell was still my fan, and the physics department would vote for an assistant professorship for anybody who Ed Purcell thought was good. Assistant professors at Harvard virtually never made tenure in those days, so they weren't giving away much. It was not a good position to go into.
And you fully appreciated this.
Yes. I knew Harvard pretty well. So, then I had an offer from Utah, where Richard Price had gone, and I had the offer from John Wheeler at Princeton. And Berkeley was still in the mix. The job I wanted, however, probably for all the wrong reasons, was to stay at Caltech.
Even though you just said that you were quite over-ready to leave.
Well, I didn't mean I wanted to leave. I meant I'd clearly done enough to get my PhD and leave if I wanted to. Part of wanting to stay at Caltech was clearly for the wrong reasons—the baby duck patterning on and following the mother duck, even when it would be better for the duckling to swim off on its own. Part of it was that Margaret still had a year to go in graduate school. Even if I didn't want to stay at Caltech forever, it would certainly help Margaret for me to stay there for a year or two. I decided I shouldn't feel any guilt if my secret plan was staying for a couple of years, then leaving. It's a free market.
But it was also relevant that Caltech almost always promoted its assistant professors to tenure. So, an assistant professor offer from Caltech was tantamount to a tenure offer, unless I subsequently failed completely. Caltech, as a result of its promotion policy, had a higher percentage of deadwood in its department than some of the other first-rate physics departments. But those people who were deadwood did useful things, just not research. They taught courses, they did service, and Caltech was still one of the top two or three physics departments in the country.
Bill, just to be clear, your offer at Princeton was assistant professor.
And so, the emphasis of going there where the idea is that you would work with John Wheeler, that sounds a little post-doctor-y. But just to be clear, you were joining the faculty proper?
It was an assistant professor offer to join the faculty. However, Princeton, like Harvard, essentially never promoted people. I had ruled out Harvard, but not because I never wanted to go back there. Obviously, I did later go back to Harvard. But it didn't feel like the right time, and Paul Martin had made clear to me that it wasn't really a serious offer in terms of my future career. It was, "Hey, come and spend three years here because Ed Purcell likes you." For Princeton, it wasn't that personal; but Princeton's reputation was that they never promoted anybody.
So why did Princeton win out?
You’re jumping ahead. Caltech looked very good in terms of being a place I could stay, if I wanted to, indefinitely. Tenure was clearly in prospect. Indeed, when Kip came to Caltech, he had been promised another slot for a tenured professor in his group, and that slot remained unfilled. (It was eventually filled by Roger Blandford.)
So, for all these reasons, I wanted the assistant professor job at Caltech. My case came before the Division of Physics and Astronomy (astronomers and physicists participated in each other's positions), and they voted me the job. I was called into the office of the chair, who was Bob Leighton, an experimenter who later became my ex-stepfather-in-law (but that's a different complication). Leighton said, "Congratulations on getting the job. There's still the formality that your offer has to be approved by the Council of Deans. But the vote in the division was unanimous, and so it's unthinkable that the deans would turn it down."
Not quite. One of the deans was Jack Roberts, the chemist. His daughter Anne was best friends with my sister Paula in elementary school. They were Campfire Girls together. Jack saw in me the perfect test case to stop, once and for all, these appointments from within, these inbreeding appointments where Caltech hired its own PhD students.
I think Jack figured me the perfect case because his philosophical objections wouldn't get confused with the actual merits of the case, whether I deserved to be an assistant professor. He could raise a pure point of principle: Here was the place to draw the line, on an appointment that was otherwise a good appointment. Thereafter, everyone would understand the line had been drawn in the sand. No department would dare to propose promoting their own graduate student to assistant professor anymore. I have to admit that, abstractly, his position was reasonable.
I think Caltech did do too much inbreeding. It was just my unfortunate position to be the test case. Jack convinced them all, including Caltech president Harold Brown and provost Bob Christie. Since they were physicists, and my offer was in physics, that made it ironclad. So, several weeks after Leighton had given me his congratulations, he called me back into his office. "A terrible thing has happened. You don't have the job." Okay. So that was very depressing. And at that point, I had to go back and look at what were my options.
Well, wait a minute. What's the story there? What was pulled out from under you exactly?
The assistant professor offer. I no longer had a job offer. Well, when I was originally voted the job at Caltech, word had spread throughout the community, and places like Yale, and Utah, and Harvard had already moved on to their next candidates. It wasn't clear that I still could choose among those jobs.
But, when the deans decided to take away my job, Willie Fowler, not a dean but a very powerful figure at Caltech, was out of town—at a National Science Board meeting, I think. Leighton, the physics chair, telephoned him with the news. And Willie just went berserk. Willie had started at Caltech as an undergraduate and been promoted through every rank. He was the poster child of promotion from within—of inbreeding. Willie felt, correctly or not, that Jack Roberts was personally attacking him. In the highest circles of Caltech, which is a pretty small place, it became a battle between Jack Roberts and Willie Fowler. I was completely the pawn in this battle. It went on for something like a month, after which the Council of Deans reconvened and without further ado voted to offer me the job again. Unanimously except for Jack Roberts. Willie was that powerful at Caltech.
So, what was my reaction to all of that? I think I was clear-eyed about it. The episode convinced me that I could not stay in the long run at Caltech. I would always be tarnished by this with, "Why is Bill here?" "Oh, he seemed great at the time, but he was really there only because Willie Fowler…"
I went to Kip and to Willie and Bob Leighton. I said, "Look, I’ve got to get out of here. Can I stay one year as an assistant professor and then leave? Willie said, "Absolutely. After what you've been through, I'd be surprised if you did want to stay here." I thought that was a double-edged comment. I wasn't quite sure what he meant by that. So, I was an assistant professor at Caltech for a year, and, during that year, decided to go to Princeton, again, pretty clear-eyed that I would be an assistant professor there for three years, or maybe five years, and then move on. I think the whole sordid experience of job-no-job at Caltech woke me up to the fact that I should not try, at this early career stage, to settle on where I wanted to end up.
Assistant Professor at Princeton
And after that year you went to Princeton?
The Caltech fiasco increased my flexibility in this funny way. I no longer cared about my chances for tenure at Princeton. I could go there, do good work (hopefully) and then move on.
Bill, what about broader considerations? 1973, 1974 at Princeton is an incredibly exciting place to be. David Gross and Frank Wilczek, for example. Were you thinking more broadly? Or did you have tunnel vision with your field in your career?
Definitely tunnel vision. I went for Wheeler's relativity group, which turned out to be bait and switch, and I already knew that I was moving from mathematical relativity more into astrophysics, so it was just the general reputation of Princeton in astrophysics. The reason I say bait and switch is that, in the two years or so from the time that Princeton first appeared on my horizon to the time after spending this extra year at Caltech and then finally arriving at Princeton, Wheeler had quietly started to dismantle his relativity group.
I saw Wheeler a number of times during those two years, either his visiting Caltech or my making visits to Princeton. The way John Wheeler was, he would never give you negative news. He would only hint at anything negative. You had to learn to read between the lines. I was slow on the uptake for that. I'm sure that if we could ask Wheeler, "You knew Press was coming to Princeton. When did you first hint to him that you were going to wind down your group?" I'm sure he would say, "Oh, I told him right away." But that was not how it felt. It felt that I was within months of actually arriving there. I had already engaged the moving van before I fully realized that Wheeler was cutting down the size of his group by a factor of two, and that I would kind of be on my own there. I was supposed to be John's assistant professor. That concept of indentureship existed at Princeton more so than in other places. But, in fact, I was going to be on my own.
Which answers my question, how are you both working with him and also being a member of the actual faculty? And therein lies the answer.
Yeah, very much so. At Princeton, assistant professors are the lowest of the low. Watching out for you is supposed to be the responsibility of the senior faculty member that you work for. Wheeler pretty much abdicated that role. Partly, we only learned this after the fact, because Wheeler was already secretly negotiating with Texas to move there, as he did. Larry Smarr, with me at Princeton but with connections at Texas, figured this out about a year before it was announced.
Now, about this transition to astrophysics, being at Princeton for only two years, of course, is pretty short. But as far as the transition was concerned, wasn’t it rather formative in that regard.
Yes, very much so. Still, it wasn't what I expected. I expected to go there and interact closely with people like Jim Peebles, Jerry Ostriker, possibly even the great paternal figures in astrophysics at Princeton then, Lyman Spitzer and Martin Schwarzchild. Well, I actually got there and found that none of those were people that I could really work with. Peebles was a great scientist, and exactly in the field that I was interested in, that is to say physical cosmology. But he was a diode. You could go there and talk to him, and he would listen with interest to everything you were doing and give nothing back. I think he was (and is) a somewhat reserved and maybe even shy individual, and that partly explains it. But I think also he was jealous of his territory. Peebles's record with students, how and whether he's gone to bat for them later in their careers, is quite mixed. He was a great scientist, and fully deserved his later Nobel Prize, but not somebody I could go there and work with.
Jerry Ostriker was different. Anybody could come in and work with Jerry, but you worked with him on his terms. And I was just prickly enough that that was never going to work. Or maybe I should say each of us was prickly, and Jerry had the advantage of being quite senior to me and prickly. So, Jerry and I had many scientific discussions, but few if any collaborations. The high point of every week was the group discussion at Tuesday Astrophysics Lunch at the Institute for Advanced Study, presided over by Schwarzchild, Spitzer, and John Bahcall. This is where John enters my story. He became, in practical terms, my greatest career mentor. Two aspects of mentorship go side by side: First, scientifically: John was the astrophysicist in Princeton who I interacted with most, along with the smart people in his group at IAS. I learned an enormous amount of how to think about astrophysics from John. Second, John became a champion of my career at every subsequent stage, until he died at the young age of seventy in 2005.
What clicked between the two of you that made for this formative relationship? Did you have similar styles? Or did you really complement each other with quite different approaches?
I wrote only a few papers directly with John. Where John and I are both coauthors, he would call me up, "Bill, you have this special skill to do X, Y, Z. You're part of this collaboration now. Here are the parts that you're going to write." Those collaborations were not his major influence on me. The major influence was seeing how John functioned as an organizer within astrophysics. Sometimes you learn science by doing science with someone. Sometimes you do it at one remove, watching them, watching how they approach science and seeing what works and doesn't work.
Relating back to the three axes of success, John was superb in the axis of choosing ripe problems and understanding how to attack them. And he was pretty good, say comparable to me, in physical intuition. And he was not good at all mathematically. Earlier in his career, he must have tried harder and done better mathematically. But that wasn't his strength.
Transition to Tenure at Harvard
I was at Princeton just two years, another example of "failure" in my career. By the time I was there for two years, I was already being looked at closely by Chicago. The impetus there was Chandrasekhar, and also my near contemporary Dave Schramm, another person who was influential in my career, and who also died young. So, there was Chicago; and then Harvard was also interested in, I later came to understand, complicated institutional reasons. Princeton went through the motions of considering me for tenure, but I was ultimately vetoed by Lyman Spitzer, on the grounds that I wasn't a "real" astronomer. I think he had sought out Jim Gunn's advice on that. I can't really argue with their conclusion.
Harvard’s offer this time around didn’t come from Ed Purcell and the physics department, although my connection there didn't hurt. It came because the Center for Astrophysics was just getting going. George Field had been brought to Harvard to create a large enterprise joining Harvard and Smithsonian in a joint institute, the Harvard-Smithsonian Center for Astrophysics. I imagine that there was a discussion along these lines: We're going to build a major center. Well, what fields should it have? One of those fields should be relativistic astrophysics. And who is the flavor of the month in relativistic astrophysics? Oh, that's Bill Press. Let’s make him an offer.
At least to be fair to yourself, minimally, you've been the flavor for two or three years at this point.
I was at Princeton just two years, and I have to say, those were not tremendously productive years.
By what standard?
I just think the quality of papers that I was writing in collaboration with students of my own and collaborators. I'm not judging by things with hindsight like, did they turn out to be right? I just wasn't doing the same quality work at Princeton that I had been doing in Kip's group at Caltech.
Part of that perhaps was the two-body problem with your wife and her status in graduate school?
Well, in the extra year we stayed there when I was an assistant professor at Caltech, she did finish graduate school. But in Princeton, she was unemployed. We had our daughter there. But Margaret was miserable.
Well, I think that explains it. A newborn and an unhappy wife, that's not going to help anyone's productivity.
Well, let's go through the whole list: A newborn, an unhappy wife, a lack of collaborators, joining a Wheeler group that was in disarray and being dismantled, and being an assistant professor, lowest of the low at Princeton.
And culturally, Bill, on top of that, coming to Princeton, you must've looked back at Caltech as a freewheeling, no-holds-barred kind of place. And now you're at Princeton, which is the apotheosis of formality and convention. That must not have been great for your creative impulses either.
Okay, the reasons may or may not reflect badly on me, but still these were not very productive years. It’s clear that I was given tenure at Harvard (or it might equally well have been Chicago) on the basis of promise left over from my years at Caltech.
So, you're saying that you rode into Harvard on the vapors of Kip Thorne. Princeton is almost beside the point.
I wouldn't say only Kip. I think I did a lot of good work at Caltech. Kip's group and that special time in relativistic astrophysics were extraordinary opportunities. I was able to work in areas that Kip wasn’t even interested in. For example, let me claim a little bit of credit for what I think of as the "premature pre-discovery" of gravitational lenses.
Gunn and I wrote a paper that said essentially, "Look, on the basis of a cosmological calculation, gravitational lenses ought to be all over the place there, waiting to be discovered." Gunn was a truly great observer, and he contributed a lot to that paper, but I had the basic ideas. I went to him and said, "Jim, I don't know enough observational astronomy to write this paper." That may have been courageous for a graduate student, but we've covered that. If Jim had not only written this paper with me but also said, "Wow, that's interesting. I should do an observational program to find these gravitational lenses," he could have found them, and he and I could have been famous for it. I could have ridden on his coattails. Maybe he and Spitzer would have considered me a "real" astronomer.
I discovered, by the way, that Gunn was a very elegant writer. Kip took scientific writing very seriously, and he is a very straightforward, very well-organized writer. Jim may not be as organized, his handwriting is almost illegible, but he is very elegant. Even as a graduate student, I appreciated that.
What actually happened with our effort was that our paper disappeared into the pages of The Astrophysical Journal without making a ripple. Six years later, completely independently, gravitational lenses were discovered. Now, did people go back and say, "This amazing discovery was predicted in the theoretical paper of Press and Gunn six years ago?" No. That's never the way it works. Discovery in astronomy is something that observers do. In the end, our paper did get referenced, I think it was 300 times, among the thousands of papers that were subsequently written about gravitational lenses. So, I'm not crying in my sour beer or anything.
I was using this as an example of my being in an environment that was so creative; and Princeton was just the opposite for me. So yes, on the vapors of that environment, as you put it, I rode into the Harvard-Smithsonian Center for Astrophysics as a tenured Harvard professor. My appointment was in both astronomy and in physics.
Physics and astronomy were separate at Harvard?
Very much so. From the physics department, it is a twenty-minute walk up Garden Street to Observatory Hill, hardly a hill, really, but in the 1840s it was a wooded hill with an observatory dome on top. By the 1970s, Observatory Hill housed, in a series of connected buildings, the astronomy department, which was a teaching department; the Harvard College Observatory, which was a research institute with its own endowment whose staff was about ninety-nine congruent with the astronomy department; and the Smithsonian Astrophysical Observatory, which was a completely different entity that, for convenience, had decided to build its building on Observatory Hill, leasing the land from Harvard. It had been there for twenty years, since the late fifties maybe.
It was only with George Field coming to Harvard just a couple of years before I did, so '74-ish, that these two institutions actually got joined at the hip to form the Harvard-Smithsonian Center for Astrophysics. Harvard and Smithsonian each contributed new appointment lines to the new director to make it work. Smithsonian, they're not called lines. They're called billets, because they're government positions; Harvard’s are faculty lines.
My real job was as a professor of astronomy, occupying one of George's lines in the Center for Astrophysics. I was to be one of the founders of theoretical astrophysics within the Center for Astrophysics. But, if it was only that, I probably would not have accepted the appointment of Harvard. I would've gone to Chicago, which I always had good feelings about.
And that was because?
Because it wasn’t clear to me that the Center for Astrophysics would succeed, and because Spitzer was right: I didn’t know if I was an astronomer in any sense of the word. So, early in discussions, I said, "I'd like to have a joint appointment with physics," The physics department knew me, and had previously offered me an assistant professor job, so they were willing. That made me take Harvard’s offer seriously
Right. Because also at this time, the distinctions between astronomy and astrophysics were probably more pronounced than they would later become.
Well they're still pretty distinct. Let me illustrate that, although it means skipping ahead to 1982, when I became chair of the astronomy department. In almost every good astronomy department, two cultures have to be co-managed. You're managing a physics culture, and you're managing an astronomy culture. People in either culture can call themselves astrophysicists. But they know what they really are. Once, just for fun at our faculty meeting with a dozen or so Harvard professors around the table, I said, "Let's go around the table. You’re to say the first thing that comes into your head, a forced binary choice: are you a physicist, or are you an astronomer?" And we went around, and no one had any self-doubt. It was, "Physics." "Physics." "Astronomy." "Physics." "Astronomy." "Astronomy." "Astronomy." "Physics." "Astronomy." "Physics." "Astronomy." It wasn’t a question of what topics they worked on. It was about what tribe each was in.
Bill, what about the other divide: observational, versus experimental, versus theoretical?
That's a more fluid boundary in modern astronomy, I think. It wasn't fluid in the 1930s for Chandrasekhar. But times have changed. Anyway, the physics department offered me a joint appointment. I had my interview with Henry Rosovsky, the famous dean of Harvard. You have this interview before you have the offer. It's one of the final stages.
The interview was in Henry's beautiful corner office in University Hall, overlooking Widener Library to his right and Memorial Church to his left. Henry swept his arm, and he said, "This has been the Harvard dean's office for almost a century. Have you ever been here before?" And I said, "As a matter of fact, Dean Rosovsky, I was in this office in 1969 when it was occupied by students, and there were slogans spray-painted on the wall." I assured him that I had been there as a staffer for WHRB, not as one of the radical students, however. Henry wrote a great book on how to run a university. It's out of date now, but it's still worth reading. I even appear in that book, thankfully not by name, in the cameo role of an arrogant young physicist- those are his words for me.
Anyway, I mentioned to Henry that I was concerned that it would turn out that I was not an astronomer. I would try my best, but it might just turn out that way. Henry then explained a subtle thing about Harvard joint appointments. You can find it to this day in the Harvard catalogue. If my title had been "Professor of Astronomy and Physics," that would mean that I was a voting member in astronomy, but with only a courtesy appointment in physics. My title was instead "Professor of Astronomy and of Physics," that second "of" indicating that I was a full voting member in physics, and that at any time, I could inform the then-Dean that I wished to change my home from one department to the other.
Did other Harvard faculty understand this?
Rosovsky must have had to give this speech to a lot of recruits, because if you go through the Harvard catalogue, you find a lot of joint appointments, some of which have the second "of", and some of which don't. I did actually spend ninety percent of my time on Observatory Hill at the Center for Astrophysics, but I had an office in the physics department. I attended physics faculty meetings, and I was a full voting member of physics for the more than twenty years that I was there.
Let's talk about George Field. When did you first get involved with George Field on a substantive level, and what were your feelings about his leadership?
In the beginning, I didn't understand that leadership was a thing. Actually, I probably didn't really understand that my whole twenty years at Harvard. I assumed that George would be a good leader, because he was an eminent theoretical astrophysicist who had led some important national committees and commissions. And, he was a likeable man. Now, he was given this job of building the new Center for Astrophysics. Anyway, this was my recruitment phase, when everybody was treating me nicely. The whole astronomy faculty took me to lunch at the Harvard Faculty Club, and incidentally, we all came down with food poisoning, which perhaps I should have taken as a sign.
Harvard-Smithsonian Center for Astrophysics
It was only after I arrived at Harvard as a tenured full professor that I saw that the Center for Astrophysics was not one big, happy family. The actions by the Harvard and Smithsonian administrations to create the merger, and to invite George, were actually quite controversial. George was in effect dropped into a viper's pit, and he was trying to survive. It was a complicated situation. The Harvard Astronomy Department as such, that is to say, the people who had the Harvard title of Professor, were largely deadwood. The department had worked its way into this position over a generation. Basically, Don Menzel had gotten his own students appointed to professorships. All of these people were nice people. I socialized with them, no problem. They were just not very distinguished astronomers. David Layzer, Chuck Whitney, Owen Gingerich (who did distinguish himself when he switched fields and became a historian of astronomy and made his reputation on Copernicus). Bill Liller was Menzel's grand student. Bart Bok was not deadwood, but he managed to escape and was replaced by one of his students, Ed Lilley. Ed was famous mostly for carrying a gun to work.
Harvard takes a very long view of everything. Its procedures are supposed to prevent making bad appointments, but they had gone dramatically wrong in astronomy. Only after many years, when a department has become truly awful, does Harvard put it in receivership. Receivership is not an official action. I don't think there's ever a letter that says, "Your department is in receivership." What happens is, the department finds that the dean and president will no longer consider appointment recommendations initiated by the department. Tellingly, the department is asked to recommend members for an outside committee, to be brought in to recommend new appointments. It really is like a receivership where a court takes over. The astronomy department was in receivership. George Field was told that he would have a certain number of lines, I think it was three or four. These lines would become available on certain dates over the next small number of years. Sure, the astronomy department would take pro forma votes, but really, George would control those appointments, and the dean and president were supposed to back him up.
The first of those appointments had been Al Cameron, who was a nuclear astrophysicist. Quite an unconventional appointment. I don't know who had been asked to become a Harvard professor in astronomy before Al. I'm guessing that there were at least several who wouldn't consider it, seeing the sorry state that the department was then in. Al had an unconventional history because he was Canadian and had spent a lot of his career at the Chalk River Canadian government nuclear laboratories. Almost on his own, he had become, in nuclear astrophysics, a peer to the famous, Caltech-Cambridge group of Willie Fowler, the Burbidges, Fred Hoyle. Cameron was a bit younger than those, but he was really, in terms of academic science, a self-made man. So, you can understand that he might not have been put off by the seamier side of the Harvard department.
I wasn't put off, because I was oblivious to the issue. But, because of my age, I was also a very unconventional appointment. I was appointed at age twenty-seven and arrived when I was age twenty-eight as a full professor. I later learned that this had caused a lot of heartburn with Dean Rosovsky and President Derek Bok. George Field had first come in with Cameron, whose only teaching position was at Yeshiva University in New York City, the graduate school. Well, if you're in the field, you know that actually, Yeshiva University graduate school has some areas of excellence. But to go to the Harvard Dean with, "I want to appoint someone who's at Yeshiva University"-
Lenny Susskind made that quite clear to me, how strong the differences between the programs and how strong the graduate program is.
Yes, in some fields. Now, I was the second candidate George Field wheeled into the dean's office. "I want to appoint this arrogant twenty-seven-year-old."
Who's not really even an astronomer.
Yes! "I'm going to use up my astronomy slot on someone who's never been in an astronomy department and never written an astronomy paper." Almost. I guess my paper with Gunn might possibly count. So, you can understand that that was viewed with some suspicion. It's amazing that they agreed to the appointment. It actually set the department back, because after those two appointments, the dean and president became very conservative about who they would let us next appoint in astronomy. It took a long time for us to appoint an optical observer, which was the next slot to be filled. I think it took three or four years.
The other situation brewing at the Center for Astrophysics was on the Smithsonian side, where, thanks to Fred Whipple, they had also fallen into a torpor of not-very-good scientist appointments. A new sign of life was that the Harvard and Smithsonian higher-ups had come up with the idea of moving Riccardo Giacconi's entire x-ray astronomy group to Smithsonian. The group was already in Cambridge at AS&E, a for-profit company that made, among other things, x-ray equipment for airports. AS&E had been founded by Bruno Rossi and others at MIT. Giacconi, brought in from Italy, was one of their early hires. The Giacconi group was one of the two or three top x-ray astronomy groups in the world, and they all moved to Smithsonian.
It was a great coup. It made Riccardo king of the roost on the Smithsonian side. Riccardo was a very forceful figure. He was not happy when told that his group would be part of the Center for Astrophysics, and that George Field was now his director. As part of this merger, Riccardo became a full Harvard professor. But Riccardo being Riccardo told Harvard that one vote in the astronomy department was not enough for him. He wanted two votes. So, his sidekick, Herb Gursky, who was a terrific sidekick, but not a great scientist, was also made a Harvard full professor.
So, let’s total it up for the Harvard astronomy department: we had the five or six deadwood; we had Giacconi and Gursky; we had George Field; and then there was Al Cameron and me. That was the astronomy department that I walked into. It was an interesting education. George was a wonderful man. Later, after he stepped down as director, his office was right across the hall from mine. We could wander into each other's offices. We'd talk about all manner of things. Just great guy. But he wasn't a great leader, or even administrator, for the new Center for Astrophysics, given the complexity of it all. There were always crises springing up that would (the way we used to say it) roll down Observatory Hill and into the Dean's office. And the Dean was not pleased.
Bill, how much of those crises were inherent in the partnership between Harvard and the Smithsonian? What built-in frictions or turf issues might there be just from the inception?
I do think that almost all the problems were structural. I also think that if the leadership had been better, more thoughtful, the early years could've been less rocky. If George Field had been a better leader. It would have helped if he had been an observer or experimentalist—because those just never respected theorists controlling their resources. And Giacconi was a fine leader of his group, but he was an autocrat. A different Giacconi might have understood that he could help build something larger than his own group, but Giacconi never thought that way. He one hundred percent cared about his group and zero percent cared about either Harvard or Smithsonian.
That became obvious in all kinds of laughable ways. There was a time when Riccardo asked to be on the graduate student admissions committee. We thought that was strange. It's an important committee in the department, but far beneath Riccardo. Why would he want to waste his time reading graduate student applications? The reason was, he wasn't getting enough slave labor in his group, and he wanted to admit graduate students who would become slaves at low levels in his enterprise. So, he served on the committee. Gursky probably read the folders for him. The committee made an ordered list to present to the department. At that meeting, Riccardo announced his position, that we should peremptorily reject the top ten names on this list. They would get offers from other places. They would be just fine. We should admit numbers ten through twenty on the list without further discussion. We said, "That's crazy, Riccardo." "No," he said. "My group will get more recruits in numbers ten to twenty than from numbers one to ten." Unabashed.
Yeah. And no one was going to tell him otherwise.
Well, we didn't take his suggestion. Even with his two votes, we didn't do that. Still, looking back, I think there are some valuable lessons in science about this. One of them is, the prizes don't go to the nice guys. I'd like to think that discoveries are at least uncorrelated with how nice or not nice you are, but I suspect not. They're correlated positively with being not nice. Riccardo deserved all the honors that were given to him, including the Nobel Prize. He got them by fighting for them, tooth and nail. You didn’t have to like Riccardo as a person to recognize that he was a very, very successful scientist.
In the astronomy department, another structural difficulty was that our faculty slots were all funded by the historical endowment to the Harvard College Observatory, which was now controlled by George Field, as director of the Center for Astrophysics. When the astronomy department had a vacant slot, even from an immediate retirement, we had to go hat in hand to George and say, "Will you fund this position?" I saw this mostly coming into play at the junior faculty level, where the deadwood would pull every which way, and the live wood would want to simply appoint the best people independent of field. Pun intended—independent of Field. George Field as director would find himself torn among pressures on him to make junior faculty appointments in a certain area. Maybe Riccardo was pressuring him. Or maybe it was some other kind of political balance. The astronomy department often wouldn't go along. And that would be a crisis that would roll down Garden Street into the dean's office. "The department has voted eight to three to make an appointment to so-and-so, and the Director won't come up with the money." That never really resolved itself the whole time I was there, even under the next director, Irwin Shapiro. It was simply structural. Irwin was even worse than George on this issue.
Bill, I wonder if the Smithsonian connection at all played any role in your beginning involvement of working on science policy at the national level. Did it serve as an entree in any way?
No, the only thing the Smithsonian ever got me was a card getting me for free into their museums. (That's a joke, because the Smithsonian museums are all free. My card did get me a twenty percent discount in the museum gift shops.) The barriers between Harvard and Smithsonian came down only very gradually. They are still there to a significant extent. The cultures of the two places are very different. The current director, Charles Alcock, with whom I went to graduate school at Caltech, has done a better job of lowering the barriers than his two predecessors, I think.
But look: For all of the internal friction within the Center for Astrophysics, it's been a very successful institution for a generation one now, one of the leading couple of astrophysics centers in the world. And it’s more than a thousand miles from any large telescope, which is saying quite something. Sometimes, you see the frictions only from the inside.
NSF Advisory Committees
So, if not Smithsonian, what first connected you to the Washington science-policy scene?
I guess my general reputation as a young scientist on the way up. John Bahcall probably played a role. I think I had established myself as someone with broad interests within physics and beyond. I had moved from mathematical relativity to relativistic astrophysics and I was still broadening.
Which would lend itself well to the kind of aerial view of science that would be useful to an agency like the NSF.
The first thing I was asked to do was to serve on the NSF Physics Advisory Committee. At that time, NSF program directors had a great deal of latitude in awarding grants. Of course, there was peer review by mail. But that peer review did not produce a strict ranking that the physics program officers had to follow. The peer review gave a qualitative sense of the merits of each proposal, and the physics program officers could then weave together a portfolio that they wanted to fund.
In those times, NSF could fund a significantly larger fraction of proposals than now when the success rate with NSF is just unsustainably low, so low that U.S. science is now really being damaged. The check and balance on the program officers' discretion was the Physics Advisory Committee, made up of outside people, including now me, who would come in three times a year and review the individual decisions that had already been made by the program officers. This, not with a view of reversing them, but with a view of guiding: You're emphasizing too much this area. There are these weaknesses that you need to be more alert to. What a fantastic initiative that you've dreamed up! (The advice wasn't all negative.)
At each meeting of the Committee, we reviewed one sub-area. At one of my first meetings, the sub-area was experimental low-energy nuclear physics. At the time, this was being funded by NSF at something like fifty separate small facilities at universities around the country, small university groups that might have a Van de Graaff machine or something like that. Even the words now sound so archaic! Or a very small accelerator. And normally, these reviews, as I said, would result in adjustments only on margin, readjusting the balance of things. This one, however, was a significant review, in part of because of a man I came to know as a great friend, Hans Frauenfelder, a distinguished nuclear physicist, Swiss, emigrated to the United States in 1952.
Hans had by this time changed fields to work on biological physics, but still had enormous street cred in nuclear physics. He felt strongly that the time had come to simply terminate most of these fifty groups, a very radical step, because it was clear that that the future of the field lay in a smaller number of large facilities already being built. LAMPF at Los Alamos. Brookhaven. So, Hans made an eloquent speech to a normally pretty conservative committee. In memory, I hear this in his Swiss accent: "There is a great arc of life. There is a time to be born, a time to live, and a time to die. And for small facility, low-energy nuclear physics, we have come to the time to die." I loved the idea.
This wasn't sprung on the full committee completely by surprise. There had been a report by the field's supporters, predictably rebutting that idea that their funding should be decreased. In their rebuttal, they conceded that yes, one or two groups out of the 50 could be terminated. Groups that had become too applied, they said, and were no longer looking at fundamental questions. Their one named example was the Caltech group in nuclear physics, which now mostly specialized in nuclear astrophysics, Willy Fowler's Kellogg Lab. Since Kellogg Lab was elite in the whole world, this was manifestly mice voting to bell the cat. Or maybe "gold watch-ing," where you offer up something sure to be preserved. But, in either case, it was throwing down the gauntlet to Willy who (recall my inbreeding job offer at Caltech) could get quite aroused. Willie simply showed up at our committee meeting in Washington. Officially he could only be there as a member of the public witnessing the meeting, but no one was about to stop Willie from speaking. He in effect just appointed himself to the committee.
Hans Frauenfelder, Willie Fowler, and I became the leaders of this faction to wipe out most of the low energy experimental nuclear physics groups in the country. I forget whether we succeeded at that one meeting. It may have taken longer. But it did happen. Hans was absolutely right. It simply had to happen, and the question was only when had that become completely obvious.
Finding in me an unexpected ally, Hans took me out to dinner in Washington. Hans is a great gourmet. It's in his Swiss background. He's originally from a distinguished professional family, his wife Vreneli from a very rich Swiss family. The Hotel Hassler at the foot of the Spanish Steps in Rome, one of the most famous hotels in Europe, was owned by Vreneli's family. So, Hans took me out to dinner in Washington to a restaurant that I never could have afforded. He specified in advance that it was his treat.
It was one of the most wonderful meals I had had in my inexperienced, young life. At the end of the meal, he called the waiter over. He said, "That was a truly wonderful meal. I would like to give my praise and make a few suggestions to the chef. Could you ask the chef to come to our table?" I'd never seen this before. I probably didn't even know that restaurants had chefs. I was shivering in my seat. The waiter, with complete aplomb, said, "I'll convey your thoughts to the chef, sir. He is far too busy to come to your table. I so apologize for that." Hans said, "Could you tell the chef that I am Escoffier's great grand-nephew?" It was true, but who knew? The chef was there like a shot. He and Hans had a long conversation, far above my level, about the preparation, the presentation, and so on.
That was my first experience with Hans Frauenfelder. Fast-forwarding twenty-five years to when I was at Los Alamos, in 2001, the entire town was evacuated as the Cerro Grande fire swept through the town. Hans had by then retired from University of Illinois and was at Los Alamos- still working full time. The Frauenfelders took us in, my wife and son and I, and our dog, and we sheltered for a week in their beautiful house in Tesuque, just north of Santa Fe. Because it seemed somewhat like the end of the world, with forest fires burning everywhere, Hans for the whole week we were there brought up from his wine cellar the best of his vintages. It was a wonderful experience, except that we thought that our house had burned down. There was a map showing all the areas that had been burned, and our house was inside that area. It wasn't until the end of the week that we discovered that the map was wrong. But let's go back to the late 1970s.
Bill, did serving in this committee give you a newfound appreciation of just how vital the NSF was to physics across the board?
Certainly. An example that turned out to be important, where I had, I think, a significant influence, came a few meetings later. We were to consider whether there should be a program in experimental relativity. The very notion that relativity could be its own experimental science was new. There were just a few modern experiments: Bob Dicke at Princeton had improved the Eötvös experiment by orders of magnitude, and Bill Fairbank at Stanford was proposing experiments in space. I was on the subcommittee of the full committee to look at this. Our report to the full committee led to the chartering of an NSF Program in experimental relativity, which later led directly to LIGO. So, listing my small contributions to LIGO, there was my review article with Kip, and then, in the policy realm, my service on the NSF committee. I was also a perennial anonymous reviewer of the LIGO renewal proposals.
Another area where I was able to make a policy contribution was to scientific computing, my first experience with chairing a national committee. This is also interesting as an illustration of how technology gets adopted: Minicomputers first became available to researchers around 1980. These were nothing like today's PCs. Think of them as things the size of a small refrigerator, but cheap enough to be owned by a single investigator, by a single research group. That was a huge change from each university having a single computer center with a single mainframe computer. That had been the only paradigm in existence up to then.
Many research groups in all fields, but especially in physics, were abandoning their university computer centers, and buying these minicomputers and asking NSF for money for them. It was a trend that was starting to make university computer centers financially nonviable, they were losing business. NSF management in physics division, especially Richard Isaacson, who was the Program Director in Relativity and Gravitation but also played a big role across the division, saw the problem, and saw that there was no national plan. What was going to happen? Would they just be shelling out all this money to buy everybody a minicomputer? Some researchers could use minicomputers efficiently, others couldn't. Some computers that NSF was paying for would be idle most of the time.
At Rich's instigation, a committee on the future of computing in physics was chartered. Now, really, he knew that this was a committee on the future of computing in science, not just in physics. But his span of control was officially only physics. He asked me to chair this committee. It was the first time I had ever chaired something like this. He gave me my marching orders: "You're going to write a report pretending to be only about physics. I'll then take your report to the Director of NSF [at the time Dick Atkinson], and say, 'Shouldn't this be across all of science?' The Director will then charter another committee, who will come to exactly the same conclusions you will have come to, I guarantee it. And then the NSF will proceed."
We had a good committee, with representation from a bunch of different subfields in physics. Ken Wilson, an interesting person, was on it. Ken was already famous for a purely theoretical contribution, renormalization group theory. But he had decided that computing was the future of physics. Steve Orszag, Larry Smarr, Claire Max, were all on the committee. Anyway, our report essentially said, "You're going to have to balance things. You are going to have to pay for some of these lower-end computers." We called that "capacity computing". But we said there should also be "capability computing," which meant that there should be national computer centers. This was a concept that had never existed before, could not have existed before, because data communication rates to national computer centers previously would have been inadequate, were just then getting big enough.
Basically, Rich Isaacson's vision came true. There was a subsequent committee (that I was not on, as it happened). After that, the NSF chartered the first round of national computer centers. I think five. Several of those first-round national computer centers were directed by people who had been on my committee in physics. They grabbed the ball and ran with it. The one I knew best was Larry Smarr, who I'd known since he was a graduate student in relativity. Larry became the founder of the National Center for Supercomputer Applications at Illinois.
Most committees, most of the time, are a waste of time. But if one has the zitzfleisch, which I gradually developed-
We should translate. That means flesh on your rear end, but it really means fortitude to be able to sit through things.
-fortitude to sit through a lot of boring committees, then, every once in a while, you are present at a cusp, where you can help accomplish something that changes the direction of science, whether it be enabling LIGO, enabling national computer centers, whatever.
So, Bill, what were those fortuitous moments, where sitting through all of these meetings was really worth it?
Well, I think those three that I mentioned. The other place where I could exert an influence in certain areas was the JASON group of advisers to the government, largely advising on defense matters.
JASON Study Group
Were there any women on the JASON group at that point?
At that point, no.
I believe Claire Max was the first, but this is before her time.
How did you become connected with JASON?
It was the Princeton Physics connection, oddly enough since I was already at Harvard, people like Sam Treiman and Curt Callan, who were in the JASON group. John Wheeler was not a member of JASON but had been active twenty years earlier in its founding. JASON wasn't dominated by Princeton, but there was a strong Princeton contingent who knew me. They knew that my interests were broader than strictly my published papers might indicate, they knew that my style of physics was often physical intuition and not simply equations. That was useful within JASON. It probably also helped, oddly, that they knew that I was a troublemaker. JASON is very tolerant of troublemakers, because often we're challenging the views of our government sponsors, or the generally accepted wisdom.
Bill, had you thought about national security issues in any sustained way before joining JASON?
Well, yes, because of my Livermore connection. I knew a lot about nuclear weapons, and a little about arms control issues. In that context, I already knew that if Edward Teller was against something, I was probably for it.
Well, there's politics and there's science.
Strictly speaking, JASON looks only at the science; but one can't help being aware of the political implications. JASON was one of those post-Sputnik innovations. It was formed first in 1958, and it was a bunch of whiz kids. I guess the term "whiz kids" came in with President Kennedy, and McNamara, and so on.
Who stands out in your mind among those founding fathers as the most whiz-kid?
Dick Garwin, of course. The original founding father, and as you pointed out, they were all males, unfortunate but true, was Murph Goldberger, who had been my department chair at Princeton. The situation (by the mid-seventies) was that between 1958 and about 1975, the membership was almost completely static. These were the smartest people in the world (they thought), so why should they let anybody else in?
JASON had become a club that met every summer, worked hard, and had some great successes in advising the government to do sensible rather than un-sensible things. I know there have been a number of oral histories taken of the JASONs. I've heard two contradictory versions. The version that I think is probably true was explained to me by Henry Foley, an original JASON, one of the more realistic and cynical of them. Henry's version of it was that the sponsoring government agencies in the Department of Defense, mostly ARPA, looked at the aging JASONs and said, "Either you renew yourselves with new membership, or we're going to wind you down."
The other version is rather different. The Vietnam War was a real challenge for JASON institutionally. They lost a lot of people who didn't want to work for the military during the war. JASON's own history working for the military in the Vietnam War is checkered. A lot has been written about this. It was all before my time. The JASON old soldiers (I shouldn't call them that) felt, "Well, Vietnam was a time of stress for the group, and coming out of Vietnam, we realized that it was time to renew our membership. And it was our own decision. Sponsors had nothing to do with it."
In any case, the group had started, maybe three years before I joined, inviting a new, younger generation of people, then in their 30s. The original JASONs were then in their roughly fifties. I was maybe the fourth or fifth of those people to be invited in. All of this has been quite accurately written about by Ann Finkbeiner in her book on the JASONs. So, we don't need to spend time on it here.
But, Bill, particular to you, given your broad range of interests, what intellectual assets did you feel like you were bringing to the group?
Okay, great question. They were exactly the same assets that I had brought to relativity. I understood what computer modeling could be done, in a time when the people who worked with a pencil and paper and wrote out equations just had no idea what was easy to model, what was difficult to model, and, in any case, how to do the modeling.
And how was this relevant in a national security advising framework?
Well, I can tell you about the very first JASON project that I really worked on. I worked with one of the original JASONs, Hal Lewis. Hal himself had an interesting history. In the years 1948 to 1952, when Feynman, and Schwinger, and a few others were the rising stars in theoretical physics, Hal Lewis was often included in any list of the half-dozen stars. Now, you look back and say, "Hal who?" I liked Hal as a human being. He never achieved his promise as a physicist. I don't think it was because he gave all his time to JASON, and therefore didn't do physics. Rather, I think he discovered that his promise could not be achieved, and then found various useful things to do. Hal became a professor at UC Santa Barbara soon after its founding, and did a lot of teaching, and spent a lot of time on JASON work.
Anyway, at the time we're in, 1977, the Air Force was facing a decision of how capable its next generation of fighters should be. There was a performance versus cost tradeoff. They could have cheap, low performance fighters and have a lot of them- some of the Air Force thought that was a good idea because the Air Force always likes to have a lot of planes. Or they could go for a much smaller number of much more capable aircraft, truly state-of-the-art. The avionics would cost more than the structure, by a lot. Capable in every way, in the radars, in the performance, in the Gs it could pull. That was the live debate.
This is a credit to the Air Force: They had done actual experiments, a series of air combat games, one would call them, at Nellis Air Force Base in Nevada. They used surrogate aircraft, an existing, relatively low-performance aircraft and an existing high-performance aircraft. Result: The greater numbers of lower performing aircraft always won the engagements. This drove the people on the other side, who wanted the Air Force to be high tech and at the cutting edge, crazy. Basically, I built a little computer model for air combat, something that nowadays you could do on a laptop. Each plane had a certain range at which it could see the other planes, and it could maneuver a certain way, and it would just play out these engagements in a way that was, in modern language, a hybrid between a physical model and a statistical model, but it was a model.
Bill, you called it a little model, but I can only imagine that this caught on in a big way about what the broader implications of being able to do this was.
I wouldn't say I had much to do with that. I would say it was sweeping the world, and mine was one instance.
Yeah, but it was you who introduced it within this advisory framework.
Oh, within the JASONs, it was me. In fact, the JASONs had a rule, it was written into their charter, that there could be no computers or computing at a JASON summer study. When they had written this rule in the 1960s, they were afraid of being distracted by high priests who would descend on them and speak a language that they didn't all speak computers. So, we didn't have any computers, only portable terminals, I had one of those. You can find pictures of these antiques on the web, computer terminals where you put it your telephone handset into these little earmuffs and communicate via an acoustically coupled modem back to your home computer center. I used, of course, the Harvard computer center.
In my early years at JASON, that was my main niche, but I did other things also. JASON did some of the very first climate-related studies for the Department of Energy. An unlikely agency, you would think, but there was a far-seeing program officer in the Department of Energy, Ari Patrinos, who later became the far-seeing program officer in the Department of Energy for the Human Genome Project. Ari later moved to the private sector. I consider him one of the unsung heroic government bureaucrats whose names never appear- they never want them to appear, but they know how to make important things happen, including, in Ari's case, twice within the Department of Energy.
So, under Ari's sponsorship, the JASONs did some of the early, one-dimensional atmospheric models of climate, how much warming do you get for how much CO2. In those days, there just wasn't a recognized field of climate science. There were a few people in meteorology doing it, but not very many. JASON was an early home for it, and I was involved.
On another front, I got involved in what the intelligence community called "enigmas". These were puzzles where they had observed things by national technical means. (Nowadays, we can say "from satellites" but in those days, you couldn't say the word "satellite".) These were typically facilities in the Soviet Union, or facilities being built elsewhere in unexpected locations. The intelligence community's bread and butter was to observe these things and write reports saying what they were. Every once in a while, some were puzzling enough that they would bring them to the JASONs. We would pour over the evidence and try to figure out what they were. I found this fascinating, and it was often something that physicists could contribute to. You could start with, "What are the physics parameters of this facility? How much power is going into it? Does it have domes on top?" (That was always a favorite.) Those kinds of things.
This was obviously a lot more fun, a much nimbler advisory environment than your NSF work.
It certainly was, and to this day, is fun. By now, the JASONs have gone through, it depends on exactly how you count, two generations beyond when I was brought in. The very last of the founding JASONs are dying off. Freeman Dyson died earlier this year. Dick Garwin, happily, is still alive in his nineties. We hope, irrationally, that he'll stay around forever.
Let's stop here for today.
This is David Zierler, Oral Historian for the American Institute of Physics. It is September 4, 2020. I'm so glad to be back with Professor William Press. Bill, thank you so much for joining me for round three of our talk.
My pleasure, yet again. By now you know that I like to talk!
So, we've been toggling back and forth so far between a chronological focus and a thematic focus. Let's continue in that vein today, talking about some of your main scientific accomplishments in historical order. Let's go back to your work in general relativity, and how you got to know Saul Teukolsky.
Modes of Science, Wheeler vs. Chandrasekhar
Let me first take a cut through the subject that is not so chronological. I've thought a lot, over decades, about, what are multiple ways of contributing to science. I've never been good at building an empire: picking a subfield, making it my own, diving deeper and deeper into it. I couldn't do this even for my first love and first field, general relativity. Then, also, I was torn between what I thought of as the Chandrasekhar view versus the Wheeler view of science. The Wheeler view was, work on the most fundamental problems and make contributions that are so clear that your name gets attached to them. Maybe not just for the ego trip of it, maybe to crystalize the ideas involved.
And the opposing (in my mind) Chandrasekhar view, that your goal should be to weave things into the fabric of science, advance science without necessarily getting any credit for it at all. Maybe anonymously, maybe through your students, maybe in other ways. Chandra quoted some fable about an Indian guru to illustrate his belief. Of course, it is ironic that Chandra espoused this view, because it was not the story of his own career, with so many things named for him!
Ultimately, I always came down on the side of the Chandra view, maybe simply because that's what I was better at. Only very few of my contributions have been "discoveries" per se. So now, to my collaboration with Saul Teukolsky-
Now, I want to interrupt because I want to go back to Wheeler and Chandrasekhar. Just to historicize your coming to grips with their representative viewpoints, is this happening sequentially? Is this happening concurrently? When are you sort of grappling with their different approaches in your career?
From the time that I knew each of them, from graduate school on. The decision points for me were when there was some interesting, offbeat thing that I could pursue that wasn't going to lead to a "discovery" but might lead to advancing the general scientific appreciation or understanding. Should I do it or not? I typically decided yes, I think to the distress of many of my mentors, who perhaps had a higher opinion of me than I had of myself and thought that I could accomplish more if I put aside all these extraneous things and concentrated on depth.
Steve Weinberg, a good friend, more than once in my career has taken me aside and lectured me along the lines of, "Bill, I wish you would stop being a dilettante. Pick an area in which you've already made some contributions, really dig in, and make the important discoveries." I always felt a little badly after those interactions, since Steve was obviously right from his perspective. But from my perspective, I knew I could never do that. I knew that wasn't me, and it wasn't the kind of science I wanted to do.
We were talking about my work with Saul. There was a case where we really did dig in (as Weinberg would have wanted) to questions of perturbations of Kerr black holes, and we proved that rotating black holes are stable. Saul discovered something that got his name attached, the Teukolsky equation, still used when people do calculations on black holes in linear perturbation order, only now being replaced by the ability to do things numerically on the computer, fully nonlinearly, to all orders. At the time we did it, I thought of our proof of Kerr stability as a discovery, something I would be remembered for. I thought I was on the Wheelerian side of that one.
But no. People who work in black holes today never think about the question of whether Kerr black holes are stable. The fact that we observe rotating black holes in nature and there's good evidence that they are rotating, makes it a moot issue. It's completely settled science. So, Saul's equation lives on as a Wheelerian example, but its applications, like Kerr stability, have just become subsumed in the fabric of things Chandrasekharian, if that is the word.
I should say while we're talking about Saul, that he has been the most influential collaborator in my career. I already mentioned that I learned early on that I needed more mathematical collaborators.
Where did you meet Saul?
Saul was one year behind me in graduate school at Caltech. So, he arrived when I had been there a year, a fresh young face in Kip's group from South Africa, which at the time we considered very exotic. Soon, I guess after his first year, Saul went back to South Africa for the summer to marry Ros Teukolsky. Saul and I were very close not only professionally, but the two couples, the Teukolskys, Margaret, and I, were close socially.
Wheelerian vs. Chandrasekharian is not a complete taxonomy. I already talked about my work on gravitational lenses with Jim Gunn. That was a different mode, "premature pre-discovery". I don't know who to name it after. Maybe Gregor Mendel? The effort with Gunn taught me that there was such a thing as a premature pre-discovery paper. A paper that could be absolutely right, but if the times weren't right, especially in an observational field like astronomy, it would just get published and sit in the literature, and then years later—just hit or miss—either it would be discovered and seen as founding a whole subfield (not our case), or it would be found only by historians of science and pointed out as a little quirky thing. "Oh, look, Press and Gunn did these six years before the first observation." The discovery paper on gravitational lenses didn't even reference our paper.
And, Bill, just to zoom out a little bit, a lot of people talk about the ebb and flow of general relativity over the decades. Some decades, it was very much trendy and others it was not.
Well, when I worked in general relativity as a graduate student and for a few years after that, it certainly was trendy. Kip, in his popular book, Black Holes & Time Warps, describes this as the first golden age of black holes, or something like that.
So, it was black holes that brought it back from the abyss essentially of the fifties and the sixties.
Gravitational collapse, yes, and gravitational collapse had to lead to black holes. Discovery of quasars, maybe that was the first. The discovery of pulsars and realization that they were neutron stars. Also, things like the ability to model supernovae. So, between, on the observational side, quasars and neutron stars, and on the theoretical side, the PhD-producing "schools" of John Wheeler at Princeton, Charlie Misner at Maryland, Kip Thorne at Caltech, there really was a renaissance. Some of us in the field referred to it not as a renaissance, but as a Children's Crusade, because the market at the time couldn't absorb the large number of PhDs in relativity and relativistic astrophysics that were turned being turned out. In the historical Children's Crusade of the 13th century, a group of children set off to free the Holy Land and ended up being sold into slavery. That was the analogy. The overproduced relativists didn't end up in slavery, but they had to find alternative careers. Most came out okay.
What about Press-Schechter? Did Schechter also fill out the mathematical side of things?
No, there, it was the opposite. Schechter was a graduate student of Gunn's, I guess. I'm not sure whose student he was: Paul tends to work very independently. It's very hard to steer Paul into going in other than the direction that he sees himself as wanting to go. At the time, he was assembling, largely from published data, a lot of evidence that clusters of galaxies had a very particular form in their mass function. That is to say, how many galaxies of each mass were there? That form was, at the low mass end, a power law, then rising to an exponential cutoff at a characteristic mass. That general idea is another example of something that's now just in the fabric of science. But at the time, it was quite controversial because the senior observers of the time, Allan Sandage, for example, were in love with fitting functions like two power laws, with a break between them.
Paul, my peer as a graduate student, kept showing me these beautiful observational results, and I kept thinking, "There just has to be a simple explanation of that." At the time, the places to find such things were in the works of Jim Peebles, specifically, his recent book on physical cosmology. The book was not well organized, a lot of different results brought together, but it was really our bible. I could almost find the page in Peebles' book where Schechter's result should've been derived, and it wasn't. So, I worked it out.
Were you in contact with Peebles at this time?
No. I probably had met him because he probably visited Caltech. But no. Press-Schechter has gotten over the years perhaps more credit than it deserves, because it could've just been another two pages in Peebles. But we were filling a gap. Okay, that's yet another mode of doing science, being alert enough see other people's blind spots.
So, what exactly was the gap that you filled?
Just taking the formalism of cosmological perturbations on large scales with power law initial conditions and showing how, even with scale-free initial conditions, you would naturally get an exponential cutoff at any given time in the mass function of clusters of galaxies. When talking about results now almost fifty years old, it is hard to make them seem important, because science moves on.
Once, when I was at Harvard, I went to the Houghton Rare Book Library. I did all the things you have to do to get permission handle the rare books and went and read one complete volume of Proceedings of the Royal Society, from the 1690s, I think. I chose a period where Newton was writing papers. Most of the papers in that volume weren't even what we would now call science. We might call them engineering data. People measuring the properties of India rubber or something like that.
But they were not deductive.
They were not papers that in any way affected the course of science. And then, every once in a while, there'd be a paper that I'd read and think, "Wow, that was really important." It might be by Newton; it might be by someone else. But there were also seemingly equally good papers on things that were completely wrong. Phlogiston, or things like that. Theories about nature that turned out to be completely wrong. It was clear that, if you lived at the time, you couldn't possibly have distinguished the wheat from the chaff.
So, with that in mind, how was your work with Schechter received?
Oh, pretty well. Once one pointed it out, it really was obvious that there was a gap of a certain size in the field that needed filling.
Exploring the Branches of the Tree of Discovery
But here's a different way to think about it: Some kinds of discoveries in science are the golden apples on the apple tree, and you're climbing the branches of the apple tree, trying to find a golden apple and reach out and take it. But the apple tree has dense foliage. You can't see through it. You don't know in advance which branch to climb. There's this intuitive feeling that choosing the most important problems to work on, that will be more likely to find the apple. But in the end, they're just a lot of possibilities.
A lot of my career, especially the time that I was at Princeton and the first years, maybe even almost decade, that I was at Harvard, I was looking for the apples in what turned out to be wrong branches. I'm explaining here why nobody reads my old papers anymore. But those are papers that, when I look back on them, I don't feel badly about at all. They were on branches that needed to be explored, and you can just ask, was I doing a capable and even creative job of exploring them.
Let me give some examples. Biggest example, there were a lot of us climbing around in that tree, was the solar neutrino puzzle. It was clear from the start that either there would be an astrophysical explanation, something in the sun that changed the central temperature of the sun by just a little bit to suppress the neutrino flux, or there'd be a physics explanation, something very exotic that astrophysicists weren't qualified to think about, but that some deep theoretical particle physicist would eventually find.
That bifurcation was clear early on. John Bahcall, who had a lifetime commitment to finding (or facilitating the finding of) the solution to the solar neutrino problem, John played both sides of that divide. He encouraged others to work on both sides of the tree. Well, with collaborators, I put an enormous effort into possible astrophysical explanations: Models for turbulence in the sun that would have larger turbulent overshoots than people expected and would carry heat away. Or the idea that there could be weakly interacting massive particles that, at a density pervasive in the universe, pervasive in the solar system, could be captured by the sun and cool it down a little bit. This was the first work I did with David Spergel, who has turned out to be an outstanding theorist of his generation in astrophysics, not especially thanks to me or that first work.
And all of that stuff was just wrong, wrong, wrong, in the same way that the papers in the Proceedings of the Royal Society, the papers on the phlogiston theory of heat were wrong, wrong, wrong. But I don't think it was bad science. I wouldn't discourage young scientists from exploring the branches of the tree that need to be explored, because every once in a while, you'll explore the right branch. I'll give two more examples, and the second one will be where I was lucky enough to get on the right branch.
My next wrong branch was this whole question of galaxy formation, how did galaxies form and how did clusters of galaxies form. There were two theories, "bottom up," that small structures formed first and then clustered into large structures (like Press-Schechter), or "top down," that the cosmological perturbations favored the formation of very large structures like clusters of galaxies that then hydrodynamically would fragment to form galaxies. An enormous amount of work by many, probably hundreds, of astrophysicists was exploring the wrong branches of the tree. Why were these branches wrong? Because we didn't yet know the fact that dark matter dominated ordinary matter in the universe. And you would never get agreement with observation without there being much more dark matter than there was baryonic matter.
And later on, there was dark energy as well- I'll come to that. For dark matter, the pioneer was Vera Rubin, who, as an observer, had made it her career's business for decades to patiently measure (with Ford) the rotation curves of galaxies. It took time for Vera and a few others to bring the world of astronomy and astrophysics around to understanding that the data could only be explained by large amounts of dark matter. Once that was grasped by the theorists, then problems of galaxy formation, large structure formation, and cluster of galaxy formation became readily solvable as technology, primarily computer technology, developed. The more computer power that came into existence, the better the models became. By now there's little doubt that we understand that picture down to even very fine details. So, my papers on solar neutrinos and galaxy formation, several pages in my list of publications, are irrelevant today. But I don't feel too bad. They were interesting branches to explore, and they required exploring.
Standardizing Type Ia Supernovae
Well, a branch that I helped explore- played a role in exploring, that did pay off was the branch that led to the discovery of the dark energy and the accelerating universe. We are skipping forward in great leaps in time here to the mid-1990s. I happened to be at a meeting in Aspen, a winter physics conference in Aspen, when the observers were talking about type Ia supernovas, and how depressing it was that they weren't turning out to be so-called standard candles (that is, all have the same intrinsic luminosity).
The desire for standard candles was because they could be used to map the universe and its rate of expansion out to great distances. The general tone of that meeting was one of disappointment. The evidence was becoming conclusive that these were not standard candles, that some of these supernovae were brighter, and some were dimmer. That spoiled the whole enterprise. There were many talks showing light curves, how the light rises in a supernova and then decays over time.
I wasn't at that meeting particularly for this discussion. I was there for some other topic, also on the agenda. Probably galaxy formation. But, at this supernova session, Bob Kirshner may have had the first seed of the idea. "Well, maybe there's some way we could standardize these light curves. Maybe someone mathematical like Bill Press could find some way of standardizing these." I raised my hand and said, "What a great problem! I'll think about that when we get back to Harvard."
Bill, what was compelling to you about this?
There was a lot of data, there was a clear need, there was a big payoff, and it was fundamentally a theoretical astrophysics problem, that is to say a mathematical problem. There wasn't just some simple fudge factor that the observers could figure out on their own. So, this, of course, led to Adam Riess's thesis, on how to construct templates from observations in multiple colors. It turned out that, to very good approximation, while the type Ia supernovas were not all alike, they differed along only a one-parameter family. Basically, the only free parameter was the mass of the collapsing supernova core.
But it is an important point that it didn't have to be so simple. Maybe rotation could've turned out to be important and the multicolor supernova light curves would form a two-parameter family. That would've been more challenging to standardize. By standardize, I mean this: The observer observes a distant supernova and measures its light curve in multiple spectral bands. And they want to know, how intrinsically bright is it? If there's more than one physical light curve possible, they have to know which one it is, which template it fits. Then, each separate template in the one- or two-parameter family can be calibrated in absolute luminosity using local supernovae.
In the end, there were two approaches. At the time, and maybe even still, there was a lot of rivalry between the two groups, the two observer groups. The group based at Harvard was Bob Kirshner, Adam Riess, Brian Schmidt, and collaborators in Chile. The group based at Berkeley was led by Saul Perlmutter. There was useless controversy over who originated ideas, ideas that were talked about at conferences. Who lifted good ideas from one group and "stole" them for the other group. In my view, that's the way science advances, by cross-fertilization. It's an empty exercise to assign blame to cross-fertilization.
Well, the Nobel Committee certainly has its version of events on how these things played out.
In the end, there was a healthy interchange, even if it was, at times, fraught, between these two groups. But still, there were differences. Adam's work on templates was very general. If the light curves had been a two-parameter family, Adam's work would have found that. I'd have to go back and carefully read the Perlmutter papers to be sure, but I think they assumed a one-parameter family from the start, and maybe even assumed that the light curves differed only by a stretching in time. Well, that turned out to be true as a pretty good approximation, but there was no particular basis for assuming it a priori. It is possible that it was already in the air that Adam, and I, and Bob Kirshner had found a one-parameter set. Maybe the Perlmutter group got that from us. Or maybe they didn't. It doesn't much matter. What matters is that the two groups both confirmed an accelerating universe. That made it easy for the Nobel Committee.
Effect of Nobel Prizes on Science
The Nobel Prize was conceptualized long before there was such a thing as Big Science or major collaborations. Do you think it has a corrosive effect on the willingness of people to share information with each other?
Yes. But it also depends on field. Biologists, I've found, are much less willing to share things across labs before they're absolutely nailed down and accepted for publication. It may be that biology experiments are simply harder, and that the field advances more reliably if results don't get out until they've been refereed and are more certain to be correct. Theoretical physics, on the other hand, can advance with many cockeyed ideas in circulation at the same time. Eventually they get sorted out.
I dropped out of the Kirshner collaboration right after this methodology of standardizing the candles, standardizing the type Ia supernovas, was established in Adam's thesis and in the papers that we wrote together. I didn't participate, therefore, in the actual discovery of the accelerating universe, but only in a key technological advance that led to that discovery.
After the discovery had been made, it really hit the headlines. It was the Science magazine discovery of the year, if I recall. I then had an interesting new role. I was perceived as a neutral party in the rivalry between the two groups. Although I had worked with the Kirshner group in the early stages, I was also on good terms with Saul Perlmutter, who had been a student of Rich Muller, who was my colleague in JASON.
There were several conferences that hosted panel discussions- debates, really- between the two discovering groups. Whose data was better, who did a better job of analysis, which group jumped the gun in announcing the discovery? Each group was trying to pick away at the weaknesses of the other's data and interpretation. It was a very personal thing, almost ad hominem, but healthy for getting to the truth. I was invited to be the moderator of several of those panels, because I could be seen as having been close to the discoveries, but without necessarily having a dog in the fight. To this day, I am friends with the leaders of both sides. To this day, there's some prickliness between them.
Bill, just to foreshadow a little bit, I wonder if you had some interpersonal skills that were on display here that might've occurred to you or others that, "These could be put to good use in a policy realm."
No, I don't think I've ever been admired for interpersonal skills. But you raised the issue of whether the Nobel Prize skews science. Well, much talked-about is the fact that the Nobel Prize can only be divided three ways. And that means that when there is no way to divide the baby into three or fewer parts-
LIGO, for example.
Well, with LIGO, in the end, they found a way and awarded the prize. But there are discoveries where there can never be a Nobel Prize because the discovery is too distributed.
But I'm saying even with LIGO, thousands of people were involved. This is not to take anything away from Barry Barish, or Kip Thorne, or Rai Weiss, but again, thousands of people made this happen.
Sure. I think it's a new-ish mode for Nobel Prizes, trying to parse out who are the leaders of large collaborations in a way that they can find three or fewer. I'm sure the Nobel Committee must be wrestling with the question of awarding a prize for the sequencing of the human genome. Everybody I know believes that there are three people they could award the prize to for that—but they're not always the same three people! My choice would be Francis Collins, Craig Venter, and Eric Lander.
Not to mention that you have these fantastically enormous bureaucracies like the DOE and the NIH that are at the center of all of these things. It's much different than somebody like Rai Weiss plugging away in obscurity for decades in his laboratory. It's a much different proposition.
Yeah, yeah. In the case of the discovery of the accelerating universe, it was easy to get down to four names. Those four would be the three who were ultimately awarded (Riess, Schmidt, Perlmutter) plus Bob Kirshner. Kirshner was absolutely the leader of the collaboration that included Riess and Schmidt. It was such a clear and profound discovery at the time that many people, including me, said, "Wow. The Nobel Committee's going to have a hard time with this one because it clearly deserves a Nobel Prize. It even deserves one soon." (Although in the end, it took quite a number of years.) But how to get it down to three people?
It's an underappreciated detail that after a couple years, Kirshner put up a blog post, one can still dig it up somewhere on the web—that was his version of the history of the discovery. Bob is a great friend, but I'm still going to express a negative personal opinion here: Parts of his blog took aim at the Perlmutter group in a way that was personal, emotional, and not productive. But interestingly, other parts of that same blog were read, by me and others, as Bob's stepping back from the discovery and offering the lion's share of the credit to his two students, Riess and Schmidt. It's possible that this blog post broke the ice and allowed the complex mechanisms of the Nobel awarding, it's not just one committee, but a whole set of hierarchically arranged committees, that allowed them to decide to award that prize to the three that they did. I've gone back to Bob and complimented him on the generous thing that he did, withdrawing to allow the prize to be awarded. He said, "What are you talking about? That's not what I meant at all!"
It was interesting, the way Berkeley Lab got into astrophysics. And it certainly wasn't regarded as a powerhouse in astrophysics.
No, but they certainly were a powerhouse in instrumentation and in supernova searches. Rich Muller was a very early figure in that and probably hasn't gotten the credit that he deserves. No, I don't think that the basis of the rivalry was anything institutional. It was scientific competition, pure and simple. Neither group expected to discover anything as amazing as the accelerating universe. They did hope to make definitive measurements of whether the universe was open or closed, whether it would continue to expand to infinity or whether it would re-collapse. That was a big enough prize to motivate real competition.
Bill, where is inflation in all of this? Where is Alan Guth in all of this?
Inflation and acceleration are both exponentially expanding phases of the universe, but at completely different times. Inflation happens very early on. It has to terminate at some point by a change in the nature of the underlying particle physics, whether it's decay of particles, or a different vacuum state, or something like that. We know that it terminates, because, otherwise, we wouldn't be here. There'd be no formation of structure, no galaxies, no stars, no planets.
Once inflation terminates, it's no longer relevant in the universe. The physics is the universe that we know, with matter and radiation. Up to about now, the dark energy that is fated to power the accelerating universe is negligible compared with the matter and radiation. But after the matter and radiation becomes sufficiently dilute, then you start seeing Einstein's "cosmological constant," what we now call dark energy.
Working with Freeman Dyson
We talked about Dyson a little last time in the context of the original generation of the JASONs. Is that where you got to know him?
Yes. That's also where I got to know- it sounds a little awkward- how to harness Freeman's genius on projects that he might not consider important enough to work on. "Freeman, here's a little thing. I wonder if you could just do me a favor and think about this for ten minutes," and if it's the right problem, you get access to genius-level mathematical ability.
And how did his genius manifest in interacting with him? How is it obvious to you?
It's obvious to everybody who knows Freeman. Freeman can take a problem, well-posed but seemingly intractable, and simply solve it. That was often his role in JASON. I learned this early on. Like for my science, I always needed more mathematical collaborators, but could often frame and pose the problems to make them amenable to be solved.
I'll give an example, a fairly early project that I did as a member of JASON. It was some problem where you had satellites observing the earth, and it took a certain amount of time for them to observe a patch on the earth, and then it took a certain amount of time for them to slew around to observe a different patch on the earth. And meanwhile, the satellite was moving in its orbit, so a point of interest could only be observed for the interval that the satellite was close enough to. Then the question was, if you had a big list of points on the earth that you wanted to observe, how do you optimize that slewing and observing and slewing again, while the earth is rotating, and you're going around in orbit? It's a complicated combinatorial optimization problem akin to the knapsack problems in computer science. For the particulars of this problem, I was able to abstract it down to a kind of a math-y problem, a graph problem of connecting points and lines. It still looked impossible to me. I took it to Freeman: "Freeman, if you have this kind of a graph with these points and lines, and they're distances that maybe represent the slew times, and this and that, what can you do with it?"
Freeman said, "Well, I'll think about that." And a few days later, had a beautiful whole theory of this, with theorems and proofs, and suggestions for how to do it on the computer. I don't think this was ever published. Over the years, that must've happened in my collaborations with Freeman at least half a dozen times. Of course, we were often working together as part of larger groups within JASON.
But, still about Freeman Dyson, let me jump ahead many years to the work that I did with him on the Iterated Prisoner's Dilemma. For a number of years, every Christmas holiday vacation, I pick a problem that I would otherwise never have the time to think about, ideally something far outside my scientific expertise, and work on it for the fun of generally understanding it and rediscovering well-known results of other people. Some of these I post on my website as little research notes that aren't quite real science, but neither are they not real science. One holiday vacation, the problem was the Iterated Prisoner's Dilemma problem. That is a problem in game theory that has puzzled people for about half a century. It's so well known that I won't try to define it here.
Bill, did you work at all with political scientists on this problem? People like Bob Jervis, for example?
No. My problem was a particular aspect of the existing muddle. You could define a game in which the strategies of the two sides, Alice and Bob, were points in an eight-dimensional space. That is to say, Alice could set four parameters, and Bob could set four parameters. And then for each point in this eight-dimensional space, you could, on the computer, by simulation, evaluate who the winner would be, or what the score would be in that game.
Now, in game theory, there's something going all the way back to the beginnings of the field that even I had learned about, something called Nash Equilibrium. Named after John Nash, of course. The computer problem I posed myself was, could I write a sufficiently efficient optimization code to answer the question, is there a Nash Equilibrium for Iterated Prisoner's Dilemma? That would be an objectively winning strategy. Or perhaps more of an objectively stalemate strategy, where neither Alice nor Bob could do any better.
You'd think that that would've been known years before, because this game had been studied for half a century before this particular Christmas. But no, the literature on the game was all heuristic. People just proposed various strategies. The most successful one, invented by Bob Axelrod, was called Tit for Tat. These strategies didn't have any mathematical basis. I thought, "Oh, this is going to be really interesting. Maybe this will converge to Tit for Tat, and I'll be able to write a little paper saying, 'Axelrod was right all along, and here's a proof.'"
But instead of finding that, my computer program kept crashing. I could see exactly where the point was in eight dimensions each time it crashed. The crashes lay exactly on planes, or rather hyperplanes, in this eight-dimensional space. Completely mysterious. There was no reason my program should crash at all. I'm a great programmer. It couldn't be me. It must be the game itself.
I thought, "This is a problem for Freeman Dyson." Because completely abstracted from the Prisoner's Dilemma game (which Freeman, of course, was familiar with), I could say, "Here's an eight-dimensional space, here's a function in it, here's where I'm trying to find the minimax point. Why is it crashing on planes?" Within days, Freeman wrote back with what turned out to be the mathematical content of this paper that we wrote, which has now been referenced many hundreds of times and has changed that little subfield and produced very counterintuitive strategies in this game and maybe in other games. We called these Zero Determinant or ZD strategies.
It was a nice example of a collaboration where I could be the phenomenologist, finding something interesting that I could specifically characterize, and then finding someone more mathematical than me, in this case, vastly more mathematical than me, to solve the problem. I think our paper together was Freeman's last published mathematical work. He was 88 at the time.
Nuclear Winter, Role of Scientists in Policy
Bill, the NRC involvement, was that separate from JASON in terms of the circles that you were running in, or was that all connected?
It was a completely different set of activities. But the activities were connected through the old boy (now thankfully old boy and girl) network. If you're involved in some of these things, you tend to get involved in a lot of them. Let me tell you about the first NRC study that I was a part of. The NRC is the National Research Council, the arm of the National Academy of Sciences that does studies for the government. And we're back in 1983, when the topic of "nuclear winter" had come to public attention.
Carl Sagan was among those who thrust nuclear winter into the public view. What was nuclear winter? Recall, we were still in the middle of the Cold War. It was not impossible that there could be an all-out nuclear war between the United States and the Soviet Union, an Armageddon that would kill hundreds of millions and possibly billions of people. Nuclear war is bad, but the particular idea of nuclear winter was that, in addition to everything else bad that it did, the fires set by this nuclear war would put so much soot into the atmosphere that it would cause an immediate climate change.
A global cooling.
A global cooling, a deep freeze that would wipe out any remaining vestiges of human civilization. Well, it's a scientific question: would that happen or not? It's not an easy thing to calculate. The National Research Council chartered a committee and getting back to your question of relationship to JASON, there were actually three young JASONs on that committee. Doug Eardley, Jonathan Katz, and me. Wherever we were politically as individuals, our membership in JASON branded us as the nuclear warriors on the committee. We were the ones who were supposed to know about of how nuclear wars might be fought. There were also some others in that camp. Spurgeon Keeny represented the old guard of arms control. There was someone from Los Alamos, and so on. And there were also, on this committee, the proponents of nuclear winter. I hate to say proponents. I don't mean they were in favor of it happening. The people who put forth the possibility of nuclear winter in a series of papers that had been published.
So, your entree to the NRC was very much a national security, JASON connection.
Yes. JASON is sometimes written about in the blogosphere as some deep conspiracy, because you can find members of JASON on all kinds of other committees. It's not a conspiracy. It's that there are very few academic scientists who are knowledgeable on national defense matters, and JASON is a prime source of those people. There are other sources, for example the Defense Sciences Study Group that's run by the Institute for Defense Analyses; that program selects assistant professors in academia for summer crash programs in national defense. Their alumni have also been very influential.
Do NRC studies allow for classified information? Or it's an unclassified environment?
There are classified NRC studies. This one was not.
It does beg the question with nuclear winter, if we're going to analyze exactly what's going to happen, why not have this be done by JASON and not the NRC?
There was indeed a separate JASON study on nuclear winter that I was a part of. Also, one by a task force of the Defense Science Board, chaired by Hal Lewis, that I was invited to join. That led to my later being invited to join the DSB as a full member. Studies of nuclear winter could be unclassified. While weapons design is highly classified, weapons effects are predominantly unclassified.
But isn't it a continuum between weapons design and weapons effects?
Not really. Fundamentally, for purposes of military utility, a nuclear weapon is a point deposition of energy at a certain location at a certain time. It's almost irrelevant how that deposition of energy is achieved. Anyway, the inventors of the concept of nuclear winter represented on this committee were led by Rick Turco. Carl Sagan was not on this committee. There were also people on the committee, this is the way the NRC operates, who were felt to be in the middle, felt to be good scientists from a variety of fields who would, in a sense, cast the deciding vote on whether these scenarios were realistic and what, if any, policy recommendations could be made.
Well, this was my first introduction to what I think is an issue in science that's both important and problematic. And that is, what should scientists do when their scientific work leads them to conclusions with urgent policy implications? It's very easy for those policy implications to cloud the objectivity of one's own research. My introduction to the phenomenon was arriving at these NRC meetings as a then-quite-young scientist and realizing that the three nuclear winter folks, (the other two were Brian Toon and John Birks) already knew what answer they wanted. They believed that nuclear winter was an apocalyptic threat to civilization, and they wanted the National Academy to say so. That would justify a strong stand on nuclear arms limitations, nuclear disarmament, and so on.
Now, I happen to agree with those goals, that those are good positions for scientists to take as citizens. But it bothered me enormously that they clearly wanted to influence the study to come out in this policy-specific way, and they didn't seem to be very open-minded about whether it was scientifically justified, or what scientific evidence there might be on the other side.
I'm not saying that scientists should not pursue, as citizens, the implications of their work. But I am saying that it's human nature that the objectivity of science can be lost when you do that. So, what can we do? How can we get most rapidly to the correct results of the science, and also most rapidly to the right policy decisions on important issues? Well, it's now more than forty years later, and I would say I'm no closer to understanding the answer to that than I was when I was first faced with the problem. I'm certain that scientists have to engage on public issues, and I'm also certain that doing so clouds the objectivity of their science. It's just one of these tightropes that we, who want to work in both those areas, have to walk. This is where we rely on the self-correcting nature of the scientific enterprise. Scientist Y may spot Scientist X's lack of objectivity and call him or her out on it. We have to keep an incentive structure that rewards Scientist Y.
In 1983, you're working with Reagan, a president who was genuinely horrified by the prospect of nuclear war.
Yes. Now, I'm using nuclear winter as sort of a micro-example because it happens to be the one that first engaged me. The much more important example is, of course, climate science and climate change. I think that there was a lot of early research in climate science that was not of high quality. Why is that? It was a brand-new science, it had people coming into it from all kinds of different fields. Sometimes the first to enter a field are not the deepest or best. There was simply a lot of variability. And, many of the early climate scientists had very strong and very visible policy agendas.
I was never a climate doubter in any dogmatic sense. In the early years, I was more, "Let's wait and see, and let's try to do good science on this." But, how did things turn out? It turned out that the consensus of the early climate scientists was right after all. It turned out that almost all the evidence subsequently much better evidence now, I think, by anybody's standards, has shown that climate change is real, that with very high probability it's anthropogenic, and that policy recommendations are well-founded and necessary, both for mitigation, that is, trying to slow down the changing climate, and for adaptation, that is, trying to save our civilization from the effects of what is already inevitable in climate change.
So, climate change is overall an attractive example because in the end, the science and the policy recommendations are concordant. Any lack of objectivity from its early days has been overtaken, as it should be, by the preponderance of solid scientific evidence. But, at the very end of Prokofiev's Peter and the Wolf, the narrator says, "Well, if Peter hadn't caught the wolf? What then?" Could there be pathologies of interacting science and policy where the scientific method somehow fails? Where the science is so much in conflict what people would like it to be that the science remains permanently skewed by a lack of objectivity? Luckily, I don't think that's happened yet. But there are examples like ESP research in the 1950s where wishful thinking held on for far too long.
Bill, administratively, when you're part of the JASON group, you're always a JASON, right? With the NRC, is it more ad hoc than that? Is your affiliation project to project? It is like that?
So, just because you were involved in one NRC study doesn't automatically mean that you'll be drafted for another.
Formally there is no continuity. This is an ongoing issue within the NRC. Committees get put together by NRC staff, generally PhD-level scientists who have gone into policy work. There's a natural bias that if somebody has been effective on a committee, helped get the report out and so on, the staff will try to get them onto other committees. There's a continuing, conscious effort to broaden the base of people involved in NRC studies, but there's a tension, especially after any study where the members are too inexperienced and don't know how to frame a consensus.
Now, obviously, it's not a fair comparison because the groups are so different. But on the baseline goal of good science, to good advising, to good policy, what's more effective, JASON or the NRC?
These two sources of science advice are not in any sense rivals. Their scale is hugely different. The NRC does about $300 million a year of business with the federal government. JASON is more like $5 million. NRC is much broader in scope than JASON.
Although the counter to that is that JASON is much more elite.
I think if you look at the membership of NRC studies in any particular area, they bring in pretty elite people. JASON is able to do reports much faster than the NRC, because the work gets done face-to-face in a summer study; the report gets written and out the door. The NRC has a much more deliberative process that is much less likely to make serious scientific errors. For example, there's a review process independent of the consensus committee that may bring in half a dozen, a dozen, or even more reviewers drawn from a series of different places, industries, perspectives.
I think the JASONs are always a little bemused that we're so often mentioned in the same breath as the NRC. We see ourselves as a maybe talented, maybe immodest, group of scientists trying to give the federal government good advice. Nothing like the whole ponderous machinery of the NRC. Anyway, in subject areas that are relevant, JASON members are often on NRC studies. At any given time, about half the members of JASON are members of the National Academies of Science or Engineering.
It is interesting, however, that there aren't many more science organizations with established mechanisms for advising the federal government. The American Physical Society from time to time charters studies that have been influential. I'm sure some other professional societies do also. The American Association for the Advancement of Science does studies but has somehow never achieved the level of influence it ought to have. The American Academy of Arts and Sciences does studies, but more influential in scholarly than in government circles, I think.
Department Chair in Astronomy
We're up to a point in your life where you're spending more and more time in Washington, and then you become department chair. So how do all of these things work together?
I wasn't physically in Washington that much. It's more a question of how much time I'm spending working on these other activities back at my office at Harvard. I actually become department chair a little bit by accident. I'm sure I would've wanted to become department chair at some point. I was an arrogant Young Turk in the astronomy department, and I wanted to make changes that my senior colleagues didn't always go along with.
But I did become department chair three or five years sooner than I might've expected, when Al Cameron, who was the department chair in astronomy, was not renewed for a second term. This requires a bit of explanation. I've already given some, how at the Center for Astrophysics that there was always tension between the department chair, who represented the Harvard faculty in astronomy, and the Director for the Center of Astrophysics, who had line management control over the whole Smithsonian side, and a lot of financial control over internal funding of research on the Harvard side. George Field was Director when I first arrived, and then, later, Irwin Shapiro.
Well, when Al Cameron came up for renewal as department chair, he was as usual locked in some battle, I don't even remember over what, with George Field. One day at a faculty meeting of the astronomy professors, Al said, "I'm adjourning this meeting of the Department of Astronomy and calling to order a meeting of the Council of the Harvard College Observatory." "What?" Everybody said, "There's no such body." "Well, there is now," Al said. "And as the first order of business, we need to elect a chair." We elected Al, of course. Al then wrote a letter to the Dean saying that since he had been elected by his faculty, he could not be replaced by the Dean, and that he would now control the HCO budget instead of the Center for Astrophysics director.
This became known as Al Cameron's Declaration of Independence. The Dean took a dim view of this, naturally. Not too long after that, I was summoned to the Dean's office, this was Henry Rosovsky, who said, "Bill, I've been looking at all the candidates who could be the next chair of Astronomy," and he went through the list. Henry was very diplomatic. He had many synonyms for the word "deadwood", meaning a person who he would never make chair in a million years, because they were not active in research. When he got to Al, he said, "Of course, I could reappoint Al-" Long pause. "-Bill, why don't you become the next chair?"
The way it turned out is that I convinced Rosovsky to reappoint Al for a year, but name me as the heir apparent, so that Al and I would sort of co-rule. That worked out very well. I was named heir apparent in 1981, and I became chair in 1982.
Yes. Heir apparent, age thirty-two, chair, age thirty-three. Not quite a boychik, as we say. But young.
What did you want to accomplish as a department chair?
My goal as chair was to take a department that was aging, which had brought me in as the vanguard of a younger generation, and get appointments made, especially at tenure levels, of a new generation of Harvard astronomy professors.
Bill, before we get to your tenure as chair, obviously you don't wake up one day and say, "I'm going to jump in with two feet on policy, and I'm not going to do these fundamental real hard science collaborations that I was doing earlier." And yet, there is this shift, right? Was there anything conscious about it in your mind?
Yes, absolutely consciously. Real life is complicated, but I think the best way to impute a causal chain to events is to say I was already very involved in JASON and a bit in other similar things, so I already knew that that world existed. I don't think I thought of it as science policy, I think I thought of it as science for national security science, broadly construed.
So, the pull of policy was greater than the push of not doing science on that level?
Well, let me try to connect the dots. I became chair by accident, as I described. The intermediate few dots, which you haven't mentioned but turned out to be a very important in my life, were the Numerical Recipes books. That was a very conscious decision: When I became chair, I found that I could do the whole job in less than two hours a day. Maybe it was only an hour a day. But that hour or two came as ten or twenty five-minute interruptions during the day, this or that little decision to be made, or a student to be seen, and so forth.
That did interfere with my ability to do the kind of science that up to then I had been doing. Theorists sit and think and need concentration. Or maybe we sit and code and we still need concentration. I get addicted to coding. When I'm coding, I get to my office at 8 a.m. in the morning, and when I look up, I've missed lunch, and it's 6 p.m. As chair, I didn't have blocks of time like that. I was consciously thinking, "Well, what can I do during the day that doesn't require that kind of intensity?"
Numerical Recipes Books
The answer was to write the Numerical Recipes book and what became a whole series of books with coauthors Brian Flannery at first, Saul Teukolsky mostly, and then Bill Vetterling. This was work that didn't take deep thought. I could be writing a paragraph in the book, put down the pencil, deal with some departmental thing, and go right back to where I left off. So, yes, I did give up the kind of science that I had been doing—not completely give up, but significantly so—partly by becoming chair, partly by the pull of things that JASON led to in other policy-related things. Numerical Recipes ultimately sold half a million copies and had, I think, a big influence, even if textbook writing was looked down on by the science research establishment.
Did you have those ambitions for Numerical Recipes? Did you see what that was going to become? Or was this a pleasant surprise?
A little of each. I talked earlier about gap filling. When you recognize a gap in science and have the flexibility to jump into it with both feet, you can accomplish something. In the world of graduate teaching and textbooks, there was such a gap in the early to mid-1980s in computational science. Computational science is different from computer science: it's the application of computers to different fields, for example, astrophysics. Courses in scientific computing were springing up everywhere, and there were no established curricula and no good textbooks. Numerical analysis was a pretty field of mathematics dating back to, say, the 1920s- or back to Carl Friedrich Gauss, if you really want. But that field's theorem-proving was mostly irrelevant for these courses that were starting to be taught in physics departments, astronomy departments, chemistry departments, on how to really do computing.
Brian Flannery, as an assistant professor at Harvard, developed a very successful course in the astronomy department, but of course with small enrollment; there just weren't that many astronomy graduate students. Brian left Harvard to become a scientist at Exxon research labs. On his way out the door to Exxon, he handed me his lecture notes with, "I bet you'll be teaching this next year." That became true. The obvious thing was to convert a much-expanded version of his lecture notes into a book.
Unfortunately, Brian, as a scientist in the industry, didn't have the luxury of having time to spend on writing a book. He's been a listed coauthor through all the different editions and versions of Numerical Recipes, because in a sense, he was the founder of the enterprise. Anyway, this was a niche that needed filling. We knew that the community of people who were teaching these courses would be grateful for such a book. We never guessed that, over time, we'd sell half a million copies.
Less Science, More Time in Washington
During my time as chair, and while writing Numerical Recipes, I began to realize that those branches of the science trees that I was exploring weren't going to turn out to be the correct branches. For example, it became clear that the answer to the solar neutrino problem was a physics answer and not the kind of thing I had been working on.
Did fundamental science give you less of a thrill than previously? Had that sort of thing become less exciting to you?
When it clicked, it was always exciting. And since I'm a theorist, it can click even if it turns out to be wrong!
But things like sitting at the Aspen Institute, working over problems on the blackboard. You removed yourself from that environment where things could click. So that's really the nub of the question.
I plead guilty to the charge that I knowingly chose, in the 1980s, a direction that was less doing science and more doing these other things. It was not completely a free choice. Some of it was just the ebb and flow of what was going on in science, what was I interested in, where my talents lay. I did have, in this period, several of my most successful students. I've already mentioned David Spergel. David could have succeeded, as he did, with any advisor, the way that John Bahcall, a great astrophysicist of his own generation, had succeeded despite having David Layzer as his PhD thesis advisor at Harvard.
David Layzer, long before I got there, was already the deadest of deadwood. He did a little teaching, he did a little advising, he lived a good life, he played a lot of tennis. Layzer invited undergraduates, always female and attractive, to his home to hear his string quartet rehearse. He was first violin, of course. I asked John once, "What was it like being David Layzer's graduate student?" John said, "He gave me the most valuable lesson that any graduate student ever got: Get out as fast as you can." Robert Brandenberger, who is a professor in particle astrophysics at McGill was my student in these years. Ronald Kahn was my student. Ron was one of the "rocket scientists who went into Wall Street," and he's now head of quantitative equities at BlackRock. Quite a successful career.
But what I'm leading up to here is, it wasn't a free choice in some other ways. Irwin Shapiro had become director of CfA, replacing George Field. Irwin, if not actively hostile to theory, certainly put theory at the Center for Astrophysics low on his list of priorities. We didn't have the same ability to make new appointments that we had had under George. I didn't feel this was personally against me, but it was clear that theoretical astrophysics at CfA was becoming increasingly marginalized under Irwin. Some of my closest personal friends at CfA were observers: Bob Kirshner, Margaret Geller (who by this time was fully an observer), John Huchra. We may have been friends, but there was no room for me in their collaborations.
It's a tough decision for observers, whether to include theorists organically in their collaborations, or whether theorists simply exist beforehand to propose interesting things and afterwards to interpret published data. Earlier in my career, I had been involved with observers who were more welcoming. I've already given Bob Kirshner credit for putting me onto the enormously interesting problem that led to Adam Riess's thesis. But after that, I was specifically excluded from their collaboration. There are exchanges of email around the time: "Bill, if we need you again, we'll let you know." This didn't sour me on these people personally. I understood where they were coming from. But it did distance me from the kind of work I had been doing previously, with the notable exception of the work with Riess in the 1990s.
Bill, when you were in D.C., did you catch the bug a little bit? Not specifically to pursue political ambitions, but surely, you knew people who became heads of enormous government operations and were tremendously politically influential. Was any of that attractive to you?
A bit, yes. I don't think it was a consuming ambition, to become NSF director or anything similar. I was only offered a full-time job in government once, and that was much later, and I didn't take it.
Did you have to pull back from the advisory work during your time as chair?
No, not at all. Those things all just sort of fit in. If you had spied on me to see what I was doing at my desk, it would be a mixture of chair business, JASON business (of course, only unclassified), other studies like National Academy's, some graduate student advising, and Numerical Recipes, which took up a big chunk of the 1980s.
Bill, you emphasized before that you do quite well in collaborative endeavors. I wonder if, with these other things going on, you were perceived during these years as sort of not open for business for people who otherwise might have wanted to collaborate with you?
Well, I'd frame that a little differently. At the end of the 1980s, I'm in my early forties. By that career stage, most scientists with any degree of success have built an empire of one sort or another. They have a field, they have a lab, they have a regular flow of graduate students. "I'll take three new students a year," or whatever. Kip certainly had that. Kip took at least a couple of new students every year for many decades. I never built that kind of an empire. I was always more opportunistic looking for impromptu collaborations where I could join with a group of people and accomplish something, sometimes by leading, other times by just joining, as I did with Bahcall and would happily have done with the Kirshner, Riess, Schmidt collaboration.
I was once bemoaning this very issue to Larry Smarr, who, as my contemporary, had, by this time, built a great empire. He was director of the National Center for Supercomputing Applications, NCSA, at Illinois. His was the most successful NSF national computer center, in my opinion. I said, "Larry, you have an empire. I don't have any empire." Larry said, "Yes, you do, Bill. You have an empire like the Phoenicians. It's a floating empire with no land holdings at all, just trade routes." I think he was trying to buck me up, but extending the metaphor, yes, I was at sea: I was becoming too senior to just drop in on people's collaborations, yet I didn't have my own empire to pursue.
Harvard being Harvard, any department there sets a tone for its field nationally if not internationally. Did you ever see what you were doing in the department as reverberating well beyond Cambridge?
Not me personally. The Center for Astrophysics, despite its growing pains, despite my personal feeling that it undervalued theory for a decade under Irwin, turned out to have enormous national influence. Avi Loeb deserves credit for the later rebuilding of theory. There are national fellowships at the postdoctoral level, such as the Hubble Fellowships, that are given out on merit. The people who receive them can choose where they want to go. The Center for Astrophysics does enormously well in these, along with a few other elite places like Caltech, Princeton, and the Institute for Advanced Study.
What do you see as your principal achievements as chair at Harvard?
As I said, bringing in, as tenured professors Bob Kirshner, Josh Grindlay, John Huchra, Margaret Geller. Some of these were joint appointments with Smithsonian, so my role was only partial. These people became the core of a new generation of astronomy at Harvard.
In the end, I served only one three-year term as chair. I didn't send the Dean a declaration of independence like Al Cameron. But in my own way, I wasn't very good at being dean-ly or chairman-ly. Later in life, I regularly got phone calls: Dr. Press, would you be interested in being the Dean of X at the University of Y. I always said, politely, no. I didn't understand until much later in my career, maybe even after I got to Los Alamos, the techniques for getting what you want, yet having everyone around you think that it came from a consensus. Al Cameron was good at that, and he tried to tutor me, but I was too young to listen. I accomplished a lot for Harvard astronomy as chair, but I also alienated a lot of people. I was succeeded as chair by Josh Grindlay.
Defense Science Board, Institute for Defense Analyses
Was the kind of work you did on the Defense Science Board or with the Institute for Defense Analyses, were these substantively different than your other advisory work? Or it was kind of more of the same?
They were a little different. Defense Science Board was and still is a misnomer for what it actually is. The Board largely brings together defense-industry high executives, often CEOs of major aerospace companies. These are people with a technical background, usually engineering, somewhere in their past; but they are by no means research scientists. And then just a few of us were people that you would recognize as scientists doing research, or administering research, or in some other way knowledgeable about research.
The years I was on the Defense Science Board were Reagan years with Casper Weinberger as Secretary of Defense. Those were years in which industry was listened to a lot more than scientists were, so I wasn't terribly effective. The Defense Science Board was my introduction to a different side of the Washington bureaucracy, the industry side, the lobbying side. Not so much a side that I wanted to get a lot closer to professionally. There were a few exceptions. Norm Augustine was a wonderful human being with whom I continued to interact; and there were others whom I happily added to my (at least distant) network: Bill Perry, Fred Brooks, George Heilmeyer, Bobby Inman, Ivan Sutherland. At least they would subsequently take my phone calls. Later, Anita Jones, who I happily crossed paths with many times afterwards. I already knew Charlie Townes and Josh Lederberg from their involvement with JASON. Admiral Ike Kidd was quite a character and full of salty jokes. His father, also Isaac Kidd, also an admiral, was killed on the bridge of the USS Arizona in the Pearl Harbor attack.
My involvement with the Institute for Defense Analyses was quite different. IDA, as it's called, is an FFRDC, a Federally Funded Research and Development Center, what is popularly called a think tank. It's not in government, it's outside government, but it works exclusively for government. When I first was involved with the Institute for Defense Analyses, it was with just a single center of theirs in Princeton that worked on cryptomathematics in support of the National Security Agency. My involvement with them came through their advisory committee, populated mostly with pure mathematicians, some computer scientists, and a token physicist or two.
I was one of those token physicists. That was very, very interesting stuff, and it led to my getting "promoted" to be a member of the board of the parent Institute for Defense Analyses. I still am on their board. IDA does a wide range of studies for not just Department of Defense now but many federal agencies. You can think of IDA as the full-time, professional counterpart to JASON's seasonal, academic piece. In fact, historically, JASON was originally a division of IDA if you go way back to 1960. It branched off from IDA only later. IDA President Larry Welch, a retired four-star Air Force general who had risen from airman to Air Force Chief of Staff, quickly became my model for what thoughtful leadership at the highest level should look like. I also got to know retired Army General Andy Goodpaster, who was Eisenhower's military aide as a young officer, and later himself Supreme Allied Commander in Europe.
We haven't talked about your father, I think, since much earlier in our discussion. I can see how you might have wanted to keep him at arm's length in your own career. But I wonder if drawing on his perspective was too irresistible when you were starting to operate at these fairly high levels in Washington.
I often discussed things with him. That became easier when he was no longer in the White House, after the Carter administration. I had to be careful while he was in the government. Everybody in any administration's White House has a certain degree of paranoia, I think. I would raise some issue with my father, and he would say, "Who put you up to telling me about that?" That showed me that when people are in government at a high level, you have to tread very cautiously, even if you're family.
After he left government, he became President of the National Academy of Sciences. Then we had many good interchanges. I often had a worm's eye view of the same issues that he saw from the top of the pyramid. As you and I discussed, my father and I were not particularly close when I was growing up. These may have been the years when we were closest.
Election to National Academy of Sciences
There was always looming the question in his mind, inevitably, would I, Bill, ever be elected to the National Academy of Sciences. I had quite consciously given up on that-put it aside. It was part of my decision to diversify, work on many different things. That meant that I had no single constituency, in or out of science. And that's simply not the way to get elected to the National Academy of Sciences, which is divided into classes like physical sciences and then sections like physics or astronomy, which justifiably elect people who are the top of their fields in those defined areas.
I wasn't supposed to know, but I did know that John Bahcall, my longtime mentor, and Dave Schramm, my near-contemporary but also longtime mentor, kept trying to get me elected. I wasn't supposed to know anything about that. But both John and Dave were people who just couldn't resist keeping me informed as to what was going on. For a period of three or four years, one or the other would call me, to have me write a few paragraphs describing how wonderful I was, which they would then feed to people to write letters of recommendation for me. This was all against the rules. John, who always took great delight in political maneuvering, would (when I was visiting Princeton) take me in the office and say, "It's looking good, Bill. It's looking good. We've made it to the next level. It's looking good."
And then, maybe a month before the election results would be announced, he would go radio silent. He never said, "By the way, Bill, your election failed again this year." But a year would go by, and it would be, "Can you update your paragraphs?" and then "Looking good, Bill. Looking good." And then radio silence. So, to that extent, I understood the situation, and I, as I say, frankly never expected to be elected.
Were there nepotism considerations with your father being president? Do you think that that delayed things? The timing couldn't have been coincidental in terms of him stepping down and your election. Or was it?
I think that cuts both ways, and it's just impossible to know the balance. I think in the long run, surely, it must have helped me, since many Academy members in other fields than mine knew me as Frank's son. They might have voted for me on that basis. But I also think that there was probably a reluctance in the physics and astronomy sections to elect the son of the NAS President. The year after Frank stepped down, I was elected. This time, John Bahcall didn't have to go radio silent. I was lucky, I think. In most parallel universes, I wasn't elected. My wife and I sometimes muse about the few key events that changed the direction of my (and later our) lives: Applying for a Hertz Fellowship. Being invited to join JASON. Being elected to the National Academy. Deciding to go to Los Alamos- we're about to come to that. Looked at this way, it seems all very haphazard.
Do you feel like your election happened in part because of your advisory work or despite it?
Despite it, for sure. I think that that was what John and Dave were up against, namely the very valid objection, "Sure, Bill did some good work, but look at what he's doing now. It's not astronomy. It's all over the map. That's just not the kind of person we elect in the astronomy section or in the physics section."
So, your service in these advisory committees is not something that would be recognized by the National Academy?
No. Nor Numerical Recipes. "Oh, that Bill. He's written a very successful textbook. He tells people that he's made enough money to pay off the mortgage on his house. That's not the kind of person we elect to the National Academy of Sciences, a person with no mortgage!"
But your work with Riess and Kirshner, that does sort of push back against this idea that you had left hard science.
That hadn't happened yet. I was elected to the Academy in 1994. And the Aspen meeting was just the year before. The first paper with Riess and Kirshner was later. That I was doing this work wouldn't have been known at the time. It hadn't borne fruit yet.
Just to interject, in the last three or four years, the Council of the Academy, on which I serve, with a confirming vote by the membership, have almost doubled the quota of Academy members elected every year. That's proving to have a very positive effect, because the number elected was previously so absurdly small compared with the number of truly qualified, first-rate scientists. A lot of the election process was coming down to the irrational magnification of small differences instead of the big picture on who are the people who ought to be elected. And I think now, just in the last couple years, we're closer to that ideal.
When did you step down as chair in astronomy at Harvard?
A long time before these events. I think 1985, 1986, something like that.
So, you were essentially a civilian in the department for the intervening years.
Yes, and as I've described, a somewhat disenchanted civilian because of the institutional directions that the Center for Astrophysics was going, now under Irwin Shapiro.
Leaving Harvard for Los Alamos
Well, let's get to Los Alamos now.
Yes, that's the perfect setup. We're now approaching the end of the 1990s. I've done this fun, and turned out to be useful, maybe even important, work with Riess et al. They've gone on to make their great discovery, announced in 1998. I was approaching my fiftieth birthday. One interesting event, one of those events that could've been trivial but turned out to be important was that George Field, at age seventy, was retiring as a professor, becoming emeritus. By this time, George and I were good friends, close professionally, because after he stepped down as Director of CfA, he had moved into the office right across from mine. We shared a secretary, and we saw or chatted with each other several times a day.
George's retirement party was one of these predictable occasions where everyone said good things about the guest of honor; but, in this case, George was truly loved. George got up to make his own speech, and he said, "I look back over my whole career, and let me tell you what I'm proudest of. I didn't stop doing science when I turned 50. I stuck with it. And in the twenty years between when I was 50 and today, I've had two or three students, and I've published at least a few papers a year." George is a wonderful man, and has a good sense of who he is, and felt that that this was an achievement worth mentioning, because most scientists do most of their science when they're young.
When I heard that speech, I had an immediate sinking feeling. I thought that twenty years from then, if all I could say was that I had written a couple of papers a year and had a couple of students and probably my numbers wouldn't be as good as George's because of our different personalities, at age seventy, I wouldn't feel good about it. I wouldn't be able to stand up and, like George, say that I was proud of my record. I would think that it was twenty years down the drain. That really set me thinking.
Then, about six months before I turned fifty, the decimal system has a big effect on our lives, my wife, who by this time was Jeffrey Howell, said, "Bill, turning fifty is a big milestone. I want to invite all our friends from all over the country. Not exactly on your birthday, but at the beginning of the summer when people can travel, let's have a big party to celebrate." I said, "Sweetheart, stop! I may normally be a robot, but I'm having a rare emotional reaction. I have to go off and think about this." Because it exactly resonated with the sinking feeling at George Field's retirement. I came back to Jeffrey, my wonderful wife, and said, "Look, I have a counterproposal: I don't want to spend the next twenty years at a plateau as a Harvard professor like George. Let's not have a party. Let's spend a year looking for something completely different we can do with our lives. No preconditions, just find something different to try."
Jeffrey is an adventurous person, and she agreed. It was kind of fun. We put out feelers to everybody we knew. "Hey, we'd like to move somewhere, try something different." My network had enough connections that I started getting feelers back. Not exactly offers but feelers. Would I be interested in going into government in a certain position? Maybe, but it wasn't one that I wanted to immediate jump at. Then I had a very interesting feeler from Sid Drell, and very soon a phone discussion with John Browne, who had just become Director of Los Alamos National Laboratory. John invited me to think about coming to Los Alamos as his Principal Deputy Laboratory Director.
He must've been attracted to the idea that you were not just a Harvard physics professor, you knew your way around the national security world at this point.
I was in JASON, I knew my way around national security. It was not lost on Browne that I had started my professional life as an Edward Teller protege, although by no means his acolyte. And also, John was in a kind of a funny situation. The previous director, Sig Hecker, was a good director in many ways, but stayed on longer than he should have. He grew stale. Some things had decayed, in terms of having a well-run lab. Community relations had decayed very badly, especially after Sig was blamed for what was known as "the RIF."
In a year when the Lab's budget was cut, it had to RIF, meaning "reduction in force", a large number of people, something on the order of 200. These people would simply lose their jobs. There was a process in place to figure out, supposedly rationally, which 200 people should be fired, somehow based on their documented performance. But that process had gone off the rails, or, rather, neglected an important boundary condition. Of the people fired, all except a tiny number were Hispanic. Only a tiny number, like one or two, were Anglo. No one person was responsible for the decisions, but it was a clear example of implicit racism. In Los Alamos, "Anglo" and "Hispanic" are much-used and highly freighted words. New Mexico is a place where three ethnic groups, each with a very strong sense of identity, have to work together, ideally without attacking each other. (Native Americans, who in New Mexico refer to themselves as "Indians", are the third group.) Things exploded when Sig released the RIF list.
In the end, the people RIF-ed went to court, and the Lab had to rehire almost all of them, because there was no statistical way to defend the Lab's list. Well, that did it for Sig, despite his having been, in most ways, a good director. He's an excellent scientist, one of the world great experts on plutonium chemistry and physics. Sig stepped down, and John Browne became director, chosen in part for his stellar people and community-relation skills. When I became John's deputy, I started flying around the country with him, going to meetings in Washington. He knew every flight attendant, every ticket agent. They would come and hug him. "Dr. Browne!"
There are people like that! Sadly, I'm not one. The other thing about John's being appointed was that, in the final stage of choosing a director, when they were down to about half a dozen candidates, he was the only internal candidate. In Los Alamos and Livermore history, that had never happened before. The idea of having a director other than someone who knew the Lab and had come up from inside was anathema within the Labs. Maybe it shouldn't have been, but it was. When John became director, he took this close call as a mandate. His thinking was, "Look, we almost had an outsider forced on us in an uncontrolled way. I need to bring in some outsiders to work for me, so we can't, in the future, be seen as so out-of-touch." So, I was brought in by John as part of that program. About half of John's team of senior executives were brought in from outside. Tom Meyer, a distinguished chemist at Chapel Hill, was another example, someone who I worked closely with.
And brought in as what? This was a one-year position, you're saying?
No, mine was a permanent position—as permanent as these management jobs ever are. I took a two-year leave of absence from Harvard, which was the maximum that Harvard would allow. The idea was that after two years, we would see. If it worked out, I might stay in Los Alamos. If it didn't work out, we could go back to Cambridge.
So, with your wife, did you commit to one year or two?
We had a two-year safety net because that's the way Harvard operated. I could have gone back at any time during that period. A lot of bad things happened at Los Alamos in this period, we can talk about them. But for us personally, we liked living there. It was clear within a year that we weren't going back. Still, there was no reason to resign my professorship at Harvard prematurely. Everyone at Harvard was convinced that I was coming back because how could a Harvard professor ever leave that position? Isn't it just the pinnacle? Maybe you go from Harvard professor to Supreme Court Justice or President of the United States, but surely not to anything else. J. K. Galbraith was Kennedy's ambassador to India and then famously came back to Harvard when his two years were up.
Our close friends at Harvard knew that we liked our new life and were not returning to Cambridge. And, as two years approached, I told my department chairs in physics and astronomy that I had decided not to return. But I never communicated to Harvard officially. I eventually started getting form letters from the dean's office: "Here's the checklist for returning from leave: You have to contact your chairs. You have to figure out what your teaching is going to be, blah, blah, blah. Please advise us the date you will be in Cambridge." I just threw those letters away. Just tossed them into the trash. We liked it in Los Alamos. Just after the two-year deadline, I got a letter saying, "Dear Professor Press, has it escaped our notice that you have returned to Cambridge? Please advise as to your activities, blah, blah, blah." Threw that away too. Finally, about six months after that, I got a letter, written by lawyers, that said, "We haven't heard from you in two and a half years. Herewith, you are no longer a tenured Harvard professor. Good luck in your endeavors. Goodbye." So that was how I left Harvard: I never resigned. The lawyers terminated me. A couple of years after that, Harvard sent me a pewter plate commemorating my twenty-five years as a professor, because the office that sends pewter plates had never been informed of my departure, and, for them, my clock was still running. I'm hoping that they send me a bigger, fifty-year pewter plate in 2026.
Bill, what did you like so much about Los Alamos?
Well, I liked the challenge of the job, and it was a big challenge for me.
And what was the initial job?
John Browne was the Lab Director. The structure when I first arrived was to have two deputy directors. In the Lab today, they would be called principal deputy directors; under John's structure, we were called deputy directors. One was the operations guy, who ran the operational, but not science or program, side of the Lab. And I was the deputy director for science, technology, and programs. It was a novelty that John put both science and programs under a single deputy director. About half of the work at Los Alamos is unclassified, is published in normal scientific journals. The Lab is a national treasure, working on a wide variety of problems. That's what was meant by "science." The other half, for which the euphemism is "programs," is classified defense work, largely the Lab's nuclear weapons design mission, the reason it was founded in the first place. John could instead have had three deputy directors, me for science and a separate one for programs. Towards the end of my tenure, the structure was rearranged somewhat, and it became more like that, although the weapons guy never got the title "Deputy Laboratory Director."
But John's thinking was, he wanted outside perspective, not just on the open science done at Los Alamos, but also on the weapons side. Going all the way back to my summer with Edward Teller, that wasn't something I was ignorant of—at least not foolishly ignorant. I could be trained. The hard part, for me, was suddenly being a senior manager in a large enterprise, large meaning 12,000 people came to work every day, 3,000 of them were science or engineering professionals with either PhDs or master's degrees. That was just a different world.
Learning to Appreciate Leadership
John understood that it would be a challenge for me and tried to be very mentoring. Very early on, I had gotten myself into trouble in a few minor ways, annoyed people who I shouldn't have annoyed in decisions that I had made. John called me into his office and said, "Bill, you don't know anything about leadership, do you?" Not only was that true, but I didn't even realize that there was a word Leadership, with a capital L, meaning not just a vague concept, but a discipline that you could actually set out to learn.
So, department chair didn't cut it, huh?
No, that's not leadership. Depending on the person, that can be either bullying or else being "deanly". Neither is leadership, per se. They teach leadership at West Point or Annapolis every bit as consciously as they teach physics or math. It's a skill that, in a high executive or managerial job, you really need. For sure, some people are innately better at it, but much of it can be learned.
So, John had the Lab send me to a two-week program at Colorado at an organization called the Center for Creative Leadership. It's still around. They do various kinds of programs in management and leadership. They are distinguished from many other such organizations by claiming to be, and largely being, data-driven. My program was called "Leadership at the Peak." Most of the participants were from industry and were typically in their forties, already senior vice presidents being groomed for the CEO position of the firm or a big subsidiary. A couple were sons of the CEOs of family-owned firms being groomed to take over from Daddy.
They didn't do things like have us fall out of trees and be caught by our teammates. Instead, three hours a day, we took every known "objective" psychological test, test of abilities, test of aptitudes, test of this or that. Then in the afternoons we met individually with our coach, who would go over the previous day's test results. We also did war games and tabletop exercises, which they observed through one-way glass. Then the next day, your coach would critique you on how you had contributed.
I learned a lot about myself, even at the advanced age of fifty. One of the tests was supposed to be a creativity test, whatever that means. Well, the industrial psychologists think that they can measure that! It was things like, "You have sixty seconds. Write down everything you can do with a pencil eraser." So, I did that and went through that test. Next day, my coach looked at me with what I took to be an expression of awe. He said, "Bill, we've been giving this test for twenty years. No one has ever scored as high as you scored on this test." I said, "Aw, shucks. I must have been having a good day…" I was trying to sound modest. "No," he said, "I don't think you're understanding me. We track outcomes. No one who's scored this high on this test has ever succeeded as a CEO or lead manager in an enterprise. This score is completely disqualifying."
That got my attention. I didn't always respect these people, but they were professionals in their field. I said, "Tell me about this." He said, "To lead a large organization, creativity up to a point is a plus. And after that point, it interferes with good communication, with good strategic planning. In your case, one week, you'd be telling your direct reports to do one thing. Then the next week, you'd have a great idea that would be even better. You'd order everyone to stop doing what they were doing, and to do this new, better thing. Am I right?" He was right. He pointed out that (from what I had already told him) as a department chair, I had never really made much of an effort to get buy-in from the people I was supposed to lead. I just did things or gave orders.
His strong advice: If I ever had the opportunity to be a CEO (for example, Lab Director at Los Alamos), I should get a very down-to-earth operations person as my deputy. In all communication, they should be the filter between me and the organization. I could hash out ideas with that person, be as creative as I felt like in private, but the responsibility for distilling and communicating the results would be theirs. If I tried to do this myself, I would fail at the job.
I absorbed that lesson. I think it was entirely correct. I absorbed it not by becoming a CEO and finding that perfect impedance-matching deputy, but by losing all my ambition to be anything like a CEO, or national lab director, or university president. Those are important jobs, and I now know that I'm ill-suited for them.
Los Alamos Laboratory in the 1990s and 2000s
Bill, obviously you weren’t yet there, but what was the effect of the post-Cold War environment on the Lab- the Clinton “peace dividend”?
Well, that was what led to the Sig Hecker RIF, of course, the big budget cut in the Labs that occurred after the fall of the Soviet Union. By the time I was there, the Lab had adjusted to operating at a lower staffing level and a lower budget level.
But not just budgets. What about nuclear weapons, research, and design? In what ways might that have changed as a result of the end of the Cold War?
I wasn't a direct line manager of the nuclear weapons work. John Browne had a senior team of about eight people. We were called the Senior Executive Team. We normally met every day at 8:00 am for half an hour. That was the high-level coordination of Lab management. The team consisted of John as Lab Director, myself, the Operations Deputy (Dick Burick during most of my tenure), and then there were associate directors, who didn't report to me or Dick. On the org chart, the associate directors reported directly to the Lab Director, but they did have dotted lines to either Dick or me. One of those associate directors, by far, the first among equals, was the weapons programs associate director.
So, if your question is, "What was happening in nuclear weapons in the first five years of the century, 2000 to 2005?” I wasn’t the line-manager responsible, but I was involved indirectly, so here's my summary: “Stockpile Stewardship” was a term of art in nuclear weapons, meaning the transition to detailed computer simulation that allowed the United States to give up underground nuclear testing and, with high confidence, maintain its weapons. That was underway by the time I was there, but it was not yet a done deal. This was a time when the three labs, Los Alamos, Livermore, and Sandia, were bringing in, year by year, larger and larger state-of-the-art computers, moving to more advanced simulation codes, and generally gaining confidence that Stockpile Stewardship would work—that we would not reach a point where the United States or the Labs would suddenly say, "We no longer have confidence in our nuclear weapons, and we're going to have to return to testing."
The architect of Stockpile Stewardship was Vic Reis, who had been head of DARPA, then moved to the National Nuclear Security Administration within DOE—the Labs reported to him. Stockpile Stewardship contributed to stability in the world because it allowed the United States to stop testing, to feel confident about not doing nuclear tests.
The flip side of the coin was Weapons Modernization. Even if you don't intend to design new weapons, there are components in the existing weapons that wear out. Well, plutonium is always going to be plutonium. But the specific metallurgical state, what it's coming in contact with, what are all the other things wrapped around it that make a weapon a weapon—these, from time to time, need modernization. As an example—not a nuclear example, but a typical example at the Lab—when I arrived, one of the, "gee whiz," things that they took me around to see as a new deputy director was the explosives range, where they would set off few-pound charges of high explosives for various experimental purposes. Only a few pounds, but enough to get your attention! You're in the bunker looking through the glass, watching it go off. And the computer that ran the explosives range was a Radio Shack TRS-80. You're not old enough to know what that is.
I know what Radio Shack is though.
It was a Radio Shack TRS-80 hobbyist computer from 1980, then twenty years old. When it broke from time to time, they would go out on eBay to find a used Radio Shack TRS-80 replacement. Our nuclear weapons program depended on this ancient piece of junk.
Weapons modernization, other than replacing ancient components, was politically controversial. One program went through a number of names. The one I remember was RRW—Reliable Replacement Warhead. This was a proposal to design a completely new weapon (or small family of different weapons, maybe one for Air Force systems, one for Navy systems, and so on) which would be specifically designed for long lifetime. In other words, we could use the lessons from the Stockpile Stewardship program, we could find all the things that were the most problematic to model, and we could design a weapon that just didn't have those difficult features. By the standards of the Reagan buildup of fifteen years earlier, which developed very sophisticated weapons, Ferraris and Lotuses, the RRW would be more like a Ford F-150 pickup truck. You could put it up on blocks in the garage, and it would still be a nuclear deterrent one hundred years from now.
Personally, I thought that was a good idea. I thought that in the long run, a country with those kinds of weapons would have a more secure and more credible nuclear deterrent than a country that was constantly remanufacturing or trying to upgrade its weapons. But the RRW and several successor attempts at a program failed because, to the Obama administration, it looked incompatible with our position on nuclear proliferation. We wanted the rest of the world to stop building nuclear weapons, at a time when we would ourselves be building a whole new generation of them. Sure, we could tell the world, "Oh, these are just going to be primitive weapons to park in the garage. They're going to be Ford pickup trucks." But who would believe that?
To get back to Sig Hecker and talking about plutonium and metallurgy, was that one area where he left a legacy at the Lab? And did you have any role?
I was in charge of overseeing science in the Lab, but I also had to be a kind of Godfather to the weapons side. Fundamentally, Los Alamos is two things: It's an engineering lab, and it's a materials science lab. A lot of its best openly published science is materials science. The science behind nuclear weapons is largely material science. So, Los Alamos is surely the outstanding lab in the United States for plutonium, not just its materials properties (which are fantastically interesting because of the unfilled 5f and 6d orbitals), but also its engineering in nuclear weapons. The other pieces of weapons science, involving radiation, and high temperatures, and so on, look a lot like astrophysics, as Edward Teller always pointed out. And the Labs do have different specializations: Los Alamos is more the materials science lab, Livermore is more the laser and photons and radiation lab.
Not just subject matter, but also style, the Labs are different. This was something John Browne explained to me very early on, probably in my first week. He knew that I had come out of the Livermore culture. He had spent ten years at Livermore early in his career and understood the differences. Is the science done in a sort of a confrontational academic way—that would be Livermore—or is the science done with a kind of organized, engineering approach—that would be Los Alamos.
Part of the reason that John sent me off to leadership school was that I had arrived thinking, "Oh, this is just going to be like Livermore. You go into somebody's office, and you start insulting them about their science because that, then, leads to a lot of shouting and a great technical conversation." That works at Livermore. Supposedly it also worked under Zel'dovich in the Soviet weapons program—a legacy of the Landau style of doing physics. But it did not work at Los Alamos. Not at all. Partly because it's a small town. The person you're insulting may take it wrong, and they may be the coach of your kid's soccer team, and they may tell everybody in the supermarket about you—it's a very different culture, and there was only one supermarket in town. I'm a little lost. What question was I answering?
The question about Sig Hecker and plutonium.
Sig made large scientific contributions in plutonium science. There's a two-volume set of books, Challenges in Plutonium Science, I and II, that came out in 2000, near the beginning of John Browne's term, that are kind of monuments to Los Alamos plutonium science at the time, especially Sig's contributions.
Wen Ho Lee Case
Now, it may be impossible to separate out in your mind what you knew about Wen Ho Lee at the time versus looking back. But I'm curious if you can isolate these things. Was the security environment that would allow something like this to fester something that people were generally aware of?
The Lab took a bad rap over Wen Ho Lee as well as some other espionage cases. There were also espionage cases at Livermore. The cases at Livermore were probably more serious in terms of loss of national security information, but they somehow got less attention than the Wen Ho Lee case. The notion that the Labs had gotten lazy, and that security was lax just wasn't true. But it was one of those narratives that, once the press gets ahold of it, can't be stopped. I'll come back in a second to what I know about Wen Ho Lee but let me tell you what I know about the New York Times. I grew up reading the New York Times from the time I was a teenager. In California, my parents got the print California edition of the New York Times—one even existed back then. I believed everything I read. The Times is a great national newspaper. But during the Wen Ho Lee scandal, it was a real shock to me to read the Times’ detailed investigative stories that were just plain wrong, just full of wrong facts and slyly implied wrong conclusions.
Where were they getting their information?
Well, that's the thing. The accepted standard of reporting sets the bar at having two independent sources. We hear that from journalists all the time. But when a strong narrative takes hold, you’re in an echo chamber. Sources aren't independent anymore. A lot of not-so-well-informed people love being an inside source to the New York Times, exaggerating their access. The process just goes awry. I would be there at meetings at which things were discussed, which then, in substance, got reported all wrong in the Times. Probably not directly leaked by anyone at that meeting, probably not even leaked by a chain three deep, but ultimately, through a chain of recounting, coming out just all wrong. And the Times printed it as fact.
Wen Ho Lee was an object lesson on how to do things wrong on a lot of different fronts. The week I arrived at the Lab, literally, John Browne called together the Senior Executive Team, and said, "None of you know anything about this, but I have finally obtained, after a year, permission from the FBI to tell you that we have a probable Chinese spy working in the Laboratory." What was I supposed to say to that, especially if it's my first week on the job?
Here was the situation: The FBI had started building a case on Wen Ho Lee, as John Browne said, at least a year earlier. They had gone to John as Lab Director and ordered: You have to keep this person in place. You cannot cut off his access to classified materials or do anything that might indicate that he is the subject of investigation. We're the FBI. If you do anything different, we will charge you with obstruction of justice.
I've since learned- all of Los Alamos has since learned that this "obstruction of justice" is a term that the FBI throws around very casually. It's a bluff almost always. But here we are in Los Alamos, where we build nuclear weapons, we believe in national defense, we believe in law enforcement, and we even believed in the competency of the FBI. John simply did as he was told. What's wrong with this picture is that there was no adjudication of the relative harm of letting more secrets go out the door through Wen Ho Lee, versus the harm of tipping him off so early that the FBI couldn't build a good case to prosecute.
Also, isn't there a jurisdictional issue here? I mean, where is the DOE in this?
Exactly. Now, if this happened today, the right thing would be to say to the FBI Special Agent in Charge, "We'll both be on the redeye flight to Washington tonight. You go up your chain of command, and I'll go up mine. My chain of command goes to the Secretary of Energy, who, by the way, outranks your FBI Director, unless you want to go farther, to the Attorney General. And the four of us, you, me, and two Cabinet officers, will meet tomorrow morning at 9:00 am in the Forrestal Building. And I'm sure that the Secretary of Energy will want to make time for this, rather than let nuclear secrets go out the door." That's the right answer. But we were naive. We didn't know how things worked.
Bill, what about general counsel at Los Alamos? The legality of this threat of charging John with obstruction.
I think John was so compliant that, when the FBI said, you can't talk to anybody, he interpreted it very strictly. You'd have to ask him. It took us all a long time to understand that the FBI guys had an incentive structure that put no weight at all on the loss of nuclear weapons secrets.
Yeah. There's a disconnect though. If John needs to have such a close hold on this that he can't even talk to his work lawyer, the Los Alamos general counsel, doesn't that presuppose incorrectly that no one at the Lab has these suspicions about Wen Ho Lee?
There were people inside and outside the Lab who had suspicions and whose suspicions resulted in the investigation starting. John mentioned the notorious Notra Trulock in his interview with you. But I think the insiders were then cut out of things.
But the whole basis of this threat is that Wen Ho Lee must not be made to understand that he's the subject of an investigation. How can you prevent something like that happening if there is general knowledge and suspicion about what he's doing during the course of his day?
Well, there's a whole series of things Wen Ho Lee did that probably did reflect his increasing understanding that he was coming under scrutiny. But let's get to the key question: Was he guilty? And, if so, what was he guilty of? In the strictly legal sense, in the end, Wen Ho Lee pleaded guilty to one count of mishandling classified data, and he was convicted of that. So, strictly speaking, you can't say he was a spy. You can say he mishandled classified data. But after we, the Senior Executive Team, were brought in on this, after we got to see all the evidence, including classified, it was pretty clear to us that there was a lot more than that going on.
Wen Ho Lee seemed to have amassed, in his home and elsewhere, on magnetic tapes and other media, a complete compendium of what it would take to design nuclear weapons. This included drawings and blueprints; it included scientific data about materials, equations of state, opacity; it included computer codes for the simulation of nuclear weapons. During the investigation, certain people in the computer side of the Lab were sworn in, to be able to go back forensically and look at his computer accounts and such. His systematic removal of classified material, having it leave the Lab, wasn't just casual sloppiness. He was making a toolkit for designing nuclear weapons.
But that doesn't prove he was spying for China.
Correct. It comes down to intent, and whether he was tasked by another country. My opinion is that he was never spying for China at all. His actual masters, if they existed, were more likely Taiwan. He was Taiwanese; there was evidence of money flowing from Taiwan into bank accounts that he controlled. There was a fascinating series of events that could be interpreted as China, that is, Beijing, trying to expose his activities to the United States. Why would they do that? To impede Taiwan. It is generally believed that China has complete visibility into all of Taiwan's secrets. There are too many Taiwanese with cultural loyalties, or who think that reunification is someday inevitable.
It's a plausible hypothesis that Wen Ho Lee was an explicitly tasked spy for Taiwan. It's another plausible hypothesis, maybe more likely, that he was not yet a spy, but was preparing to become one. There were many signs that he felt undervalued by the Lab. He was a headache to his management. He may have been preparing to go to Taiwan (or even China) with, "I have a whole bunch of stuff. How much is it worth to you?" He may have been preparing to move to Canada, or return to Taiwan. I don't think there's anyone at the Lab among those who have seen the actual evidence who believes that Wen Ho Lee was a complete innocent, unjustly persecuted by law enforcement.
Still, there's the unresolved question: If he was guilty, was he already a spy, or was he only a spy in the making? We'll likely never know this, now, because of the extraordinary gross misbehavior on the government side. The FBI investigation was incompetent, and it was then followed by gross prosecutorial misconduct. Wen Ho Lee was initially charged with things that they had no way of proving in court, part of, it's sad to say, a wave of China-phobia that was sweeping not so much the Labs but was sweeping Congress. He was chained in solitary confinement with the lights on twenty-four hours a day for months. He was held for nine months without a trial. The judge rebuked the government and dismissed fifty-eight of the fifty-nine counts. So Lee is surely a victim of government misconduct. Does that make him innocent? No, unfortunately.
There are sad aspects to this story. Wen Ho Lee's daughter, Alberta, was at the time, and I'm sure to this day, convinced of her father's innocence. She led a campaign in the Chinese American community to exonerate him. Somewhere in the middle of all this, I had a phone call from a very senior American scientist of Chinese ancestry, active in the Chinese American community in fighting racial prejudice. He knew me as an academic scientist, not as a Los Alamos person. He told me that a group of very distinguished Chinese Americans were debating whether to take a public stand on Wen Ho Lee's innocence and the government's persecution of him. What was my advice on this? It was, "Don't do it." The damning evidence that might come out would discredit this well-meaning effort. This was just not the right case on which to take a stand.
Bill, was there evidence that Taiwan had an appetite for a nuclear weapons capacity?
I'm not an expert. I think it's generally believed that Taiwan has from time to time had, or considered having, a nuclear weapons program. Probably not today.
Well, that is a very nuanced and useful sort of overview on this very complicated story with Wen Ho Lee.
That was my first week on the job at the Lab.
"Welcome to Los Alamos."
This is David Zierler, Oral Historian for the American Institute of Physics. It is September 11, 2020. I am so glad to be back with Professor William Press. Bill, thank you so much for joining me.
My pleasure, yet again.
So today, we are going to pick up on your lessons learned on-the-job at Los Alamos. Where do you want to start?
Broader Lessons from Cerro Grande Fire
Coming to the Lab, my title was Principal Deputy Director, reporting to Lab Director John Browne. I was his only "Principal" direct report. Somewhat later in my term, John reconsidered this, and he made the science deputy (me) and the operations deputy (Dick Burick first and later Joe Salgado) more like peers. I suppose that was a demotion for me. But in 2001, when we had the Cerro Grande fire, I was still Principal. So, oddly, it was me, rather than the operations guy, who had the primary responsibility of bringing the Lab back into operation after it had been closed by the fire for two weeks. That was a real learning experience.
What were the origins of the fire? How did that start?
Originally, it was a controlled burn by the Forest Service. Somebody made a really, really bad mistake by trying to do a controlled burn on a very windy day. It was soon out of control. Burned across the whole lab. The larger buildings of the Lab were enough separated from anything that could burn that there wasn't that much physical damage. There was a lot of minor damage, however, smoke damage, for example. In the town of Los Alamos, 400 homes burned to the ground. The beautiful mountainsides, the Los Alamos scenic vistas, were turned to black ash. More than 40,000 acres of forest burned. It's now twenty years later, and, looking out my window here, those mountains have finally turned mostly green again, with ground cover and the beginning of aspen trees. It will take a century for the pine forests to grow back.
And what was your role in this?
Well, pun intended, it was a baptism by fire for me. It turns out that an organization as complex as the Lab is easy to shut down—there were procedures for putting things in a safe mode and sending everyone home, and very hard to restart. You and I happen to be talking now during the COVID pandemic of 2020, when the whole world economy has in some large measure shut down. There's only speculation about how hard it's going to be for the world economy to recover. The lesson I learned at the Lab doesn't make me particularly optimistic. Something as complicated as the Lab, or the world economy, has all kinds of cross-connections and dependencies that are not documented and that you don't expect. In our shutdown, everybody had an emergency procedure, every building had an emergency procedure, every line management organization. They executed them. Everyone left safely.
There were not corresponding procedures for restarting. There were unanticipated dependencies. You couldn't reopen a building until there was water and power. There couldn't be water and power until something else. There could be circularities that made things deadlock. The building had a storeroom containing a piece of equipment needed to get the water going, but without water for firefighting it was marked as unsafe to enter. An exception was needed. The person who could make the exception had evacuated and could not be located. Whatever! My days were spent chairing large meetings around tables at which all constituencies were supposed to be represented: operations, security, program. The program people badly wanted to get things restarted. That, rightly, wasn't the highest priority for the safety and security people. It was a continuous negotiation of compromises: okay, we'll open that building "unsafely" without water, so that we can get the equipment needed to get the water going.
It was a terrific educational experience for me. Whether it was my fault (I hope not) or simply intrinsic in the structure (I think so), I don't think the Lab distinguished itself greatly in restarting. Then, on top of everything, we had the apparent disappearance of two classified computer hard drives. This became a big national news story, a crisis almost as big as the Wen Ho Lee crisis that we had gone through less than a year previously with its repercussions then still continuing.
Lost Hard Drives Incident
The Lab had a team of scientists and engineers whose volunteer job was immediate response in the unlikely event of an emergency involving nuclear weapons anywhere in the world. They had a whole kit of suitcases, equipment, computers, and so on, the things that they would need to respond to the emergency. At the time of the fire, all of that stuff was moved to a secure remote location. After the fire, it was all brought back. An inventory was taken. Two hard drives containing classified material were missing. This was an operations issue, a security issue, not directly under my span of control. But I was part of the Senior Executive Team and thus involved in the events.
Interestingly, it was a guy who did work directly for me who in the end found the missing hard drives, a scientist named Miles Baron, who sadly died a few years ago at a young age. Genetically, Miles was a full-blooded Native American. He was very handsome. He could have been the model for a classic nineteenth century painting of an American Indian. He had been given up at birth for adoption and was raised by an Anglo family in Farmington, New Mexico; had a PhD in physics from Texas Tech, I think. Miles was a very productive scientist at Los Alamos. He was extremely athletic, ran marathons and ironman contests or whatever. I had created a one-year rotating position to bring promising younger scientists into the Director's Office fairly early in their careers, to give them a perspective on how senior management operated.
Bill, was there a diversity component to this program?
Not explicitly. Certainly, diversity was always something on my mind. In personality, Miles was very persistent. He got it into his head that the hard drives could not have been lost with the other stuff that was carefully evacuated and guarded, that it must have been some slip-up during the evacuation process. He believed that they were still in the protected area of the Lab. That would've made the incident an administrative failure, but not necessarily a security failure. Instead of using his lunch hours to train for marathons, he started using them to do his own, unsanctioned search in the classified "exclusion areas" of the Lab. (He had the clearances for that.)
I was eating lunch in the main cafeteria at Los Alamos. I was the only member of the Senior Executive Team who made a point of eating in the cafeteria with other scientists and staff, rather than having my admin bring me my lunch at my desk. My senior executive colleagues thought I was crazy, or not serious about my job, because I was losing 45 minutes a day of valuable work time. For me, it was because, deep in my heart, I still believed myself to be an academic scientist, not a manager, and I wanted to have the perspectives of other scientists. So, I was eating with some of my lunch buddies in the cafeteria when Miles found me. "Bill, you need to come with me." I said, "Well, Miles, I'll find you in your office after lunch." He said, "I really think you have to come with me right now."
He took me into the secure area where he had found the hard drives in a little room, behind the Xerox machine there, as if they had fallen or been dropped there. He had drafted someone to guard the room—she happened to be a member of the same canine search-and-rescue team as my wife and had some training in forensics. At this time, the FBI was still crawling all over the Lab. They were grilling dozens of people as if they were criminals. Instead, they should have just been searching, as Miles was. Miles and I went to the room where there were a half a dozen FBI agents lounging around, and we led them to the Xerox room.
It was obvious to us that, with forensics, you could have figured out what had happened. Were the drives intentionally hidden behind the machine, or had they slid and fallen there when somebody lifted up the lid? It was dusty back there. You could have traced the path in the dust. You could have looked for fingerprints. That was not the FBI way! A burly, obese actually, FBI agent put his arms around the Xerox machine, and yanked it away from the wall. We heard the hard drives clatter to the floor. He reached behind, no gloves or anything, picked them up, brushed the dust off of them with his hands, and said, "I think I've found them!"
He then turned to Miles. "We're going to question you now." It was an incredible miscarriage of justice. Miles should've been a hero, and instead, thanks to FBI incompetence, he suddenly became a suspect. He had to mortgage his house to pay for the lawyers he needed to defend himself. The University of California (which managed the Lab) was supposed to pay his legal expenses as a Lab employee, but they uncourageously delayed until long after he had been exonerated. It took something like two years.
Anyway, I mention all of this because none of this got reported in the newspapers. The practical result of the hard drive fiasco was that it was made to look like a leadership failure of John Browne as Director. So, the combination of Wen Ho Lee, the fire, and the hard drives incident—and the heightened scrutiny of the Labs after 9/11 (which they, of course, had nothing to do with), led eventually to John Browne's being fired. This occurred suddenly over Christmas vacation in 2002.
Bill, I wonder if, given all of your time in D.C. and your familiarity with the policy process, and even specifically national security policy processes, if John Browne leaned on you in terms of managing the tensions with DOE and up on the Hill.
I have to say no. The DOE—by then NNSA, the National Nuclear Security Agency—they're their own thing. I certainly was active on the program side of some large NNSA projects at the Lab. Los Alamos was responsible for the front end of the Spallation Neutron Source at Oak Ridge. That was a $200 million project at the Lab whose immediate project manager reported to me. I was senior executive on a comparably sized project to greenfield a former tritium processing facility. That also had high visibility, and I learned a lot about environmental releases of tritium. One of my main jobs was supervising LDRD, the Laboratory Directed Research and Development program, which distributed about $100 million a year to Lab scientists.
Bill, I want to go back to the FBI investigation, finding the hard drives. To what extent is the Wen Ho Lee issue shadowing this investigation?
I think the main effect was simply that the FBI Field Office in Albuquerque believed that the Lab was just rotten. That the Lab was a nest of pointy-headed scientists who didn't care about security. That it was badly run. That it had let a spy flourish. That it had let these hard drives be stolen. Turned out, I want to emphasize, that during the entire time of the fire, the drives were in a completely secured exclusion area. They were misplaced, which is bad, but not insecure at any time.
Personal Reactions to Crises
I wonder if, in the middle of this, you were starting to think the people who'd thought you were crazy for leaving Harvard maybe had a point, that this wasn't such a bright idea.
The years that I was in management at the Lab were without question bad years for the Lab. I was part of what, in hindsight, you have to say was a failing senior management team. I don't think that I especially contributed to that failure. My view, and this may be just an excuse, is that my job was to keep the quality science high, to keep the best scientists enabled to work on important things and not be distracted by these things going on. But it is a fact that I was one voice among eight in the director's office every morning at 8:00 am in a team that, seen objectively, failed. Whether for reasons beyond its control or not.
But to refine the answer a little bit, you emphasized that these were bad years at the Lab. But I was asking more if they were bad years for you in terms of having made this decision.
No, that's the funny thing. My wife and I were very aware of that issue. I would come home after some particularly horrible thing had happened at the Lab, and we would say to each other, "Knowing what we know now, should we have stayed at Harvard?" And the answer was always a resounding "no" from both of us. It was sometimes horrible, but it was always interesting; and so, broadening an experience for me, as once a scientist, but now a manager, learning firsthand about large-organization leadership and management.
Certainly, it speaks to your ability to manage stress as well, if you were able to compartmentalize these things and not say, "What have I done? Why did I leave the Academy? Why did I leave Harvard for all of this?"
Well there was stress, to be sure. I thought that the firing of John Browne showed cowardice and mismanagement on the part of the University of California. I think that if they had looked closely at what had happened at the Lab, and at whether John was responsible for it—sure, the captain has to go down with the ship, but this ship wasn't actually sinking. It was being portrayed that way by people who had agendas. John Browne never tried to avoid responsibility. What was really going on was that Bill Richardson, who was Secretary of Energy, and Pete Domenici, who was senator from New Mexico and the real power behind both New Mexico Labs, Los Alamos and Sandia, they read the New York Times, panicked, and lost confidence in Lab management and UC. UC then in turn panicked, thinking that they would lose the management contract of the Lab if they didn't just fire its Director, whether it was deserved or not.
So, they fired him. Now, in fact, ironically, UC did lose the contract anyway. You'll still see University of California's name on the masthead, but they now play a very minor role in the management of the Lab. Today, the Lab is managed by an LLC that is owned by an industrial partnership, with UC's participation just a fig leaf.
So, Bill, you read into this that they would've lost the contract whether or not they fired John, that was sort of irrelevant.
Not completely. I think that if UC had supported its Lab Director, not just after the troubles occurred but before, things could have gone differently. UC was the sole manager of the Lab up to this time. I think they abrogated their responsibilities. There were times when the Lab director needed UC's backing in Washington, in the Office of the Secretary of Energy, and the UC Office of the President would just weasel out of it. "Well, John, you go on your own. You see what you can do. We're just a university." (That's a made-up quote, but I imagine it as pretty accurate.) UC did a bad job of managing the Lab in the years I was there, and for that reason, it was right that they should lose the contract. Maybe if they had been more activist, more engaged as managers in these crises, they wouldn't have.
Anyway, John was fired and in a very ugly way. He was in essence told to get out of town, literally to move his family within a matter of weeks, and not even accorded the normal respect of a formal Director. Panic and fear had overtaken the University of California Office of the President and the UC Office of Lab Management. Dick Atkinson, as president, could have done better.
Bill, how active, if at all, were you as a JASON during your Los Alamos years?
I had become chair of JASON before I got the offer to go to Los Alamos. I was in the middle of what probably would have been a five-year term. I resigned as chair of JASON when I went to Los Alamos, but I stayed a member. Steve Koonin succeeded me as JASON chair.
Now, is that just because the responsibility was too great? You could have theoretically stayed on?
Not just. There would have been a conflict of interest, because JASON did too much business for the Department of Energy and had too much interaction with the national labs. While I was in management at Los Alamos, the rule within JASON was that I could not work on any Department of Energy-sponsored problems.
So, you could retain your membership, but you could not be chair.
Yeah. These are one-off decisions made by the JASON steering committee depending on the circumstances. I certainly had less time to be active in JASON. While I was a manager at Los Alamos, instead of going out to JASON for four to six weeks, I probably went out for just two weeks.
Bill, did you consider resigning as deputy at Los Alamos in solidarity with John Browne?
No. My job was to protect the Lab's science base. As the new Director, first acting and then permanent, UC chose Admiral Pete Nanos. Pete was a retired three-star in the Navy. In my work in Washington, I had by now met, interacted with, worked with, a good number of active-duty and retired three- and four-star flag officers. They were an impressive group. Without exception, these were thoughtful people who radiated a calm quality of leadership. Pete was different. His style of leadership reminded me more of Captain Queeg in the Caine Mutiny novel. I couldn't understand where he was coming from. In Los Alamos, the Lab Director is a very public figure. Pete quickly became the most hated man in town. It was an amazing thing to see. I felt a certain degree of reserve, but my wife, for example, really hated Pete.
Our senior management team still met every morning at 8 o'clock. Nanos would stand up and stride around the table, yelling obscenities at people, at his senior managers. It took us a while to understand that his language wasn't directed against us personally. My interpretation was that he was sort of acting out his vision of what a forceful leader should be like. And it was just nuts. I was, by this time, on the board of the Institute for Defense Analyses in Alexandria, Virginia. IDA's president was a retired four-star admiral, Dennis Blair. Denny became famous, in some ways notorious, in his later career as Obama's Director of National Intelligence; but he was and is a highly respected former commander in chief of U.S. Pacific Command.
I said to Denny, "Do you know Pete Nanos?" And he said, "Oh yeah, I know Pete." I said, "Why does he seem to know nothing about leadership?" Denny said, "Well, Nanos, he's not really an admiral." That floored me. I said, "Oh, you mean he's not a four-star." Pete retired from the Navy as a vice admiral. Three stars. Still very high position. Denny said, "No. That's not what I mean. He came up through procurement. He never had a real at-sea command. In my book, he's not an admiral." Now, Denny's armchair analysis may not be exactly right; but there may be an element of truth in it.
I think that Pete was the rare instance in the military of someone reaching flag rank without having learned and internalized the intangibles of what it took to lead. Even so, I felt that it was my duty to try to support him. I thought I could still carry out the part of the Lab mission that I was responsible for, even if this meant being verbally abused from time to time. Pete's behavior was so extreme that I didn't take it personally. I carried on in this mode for about a year.
In 2014 there were a couple of small safety incidents at the Lab, including an eye injury from a laser beam. Well, no injury is too small to take seriously, and the supervisor involved was truly negligent; but in normal times, this would've been dealt with by a process. Likely the supervisor would have been fired. Now, however, under Pete's leadership, the Lab was in a state of constant fibrillation. As I heard the story, a government GS-12 safety guy in the Los Alamos office of the NNSA called up Pete and told him to shut down the whole lab over these safety concerns. GS-12 is the civilian equivalent of a Navy lieutenant, so this is a lieutenant giving orders to an admiral. But Pete somehow saluted, and said, "Yes, sir," and he shut down the whole lab for a week.
It was just crazy. What made it crazier was the punitive way he did it, absolutely the wrong way to instill a safety culture. Lab professionals were ordered to do eight hours a day safety training for the week. Well, maybe OK, but far excessive. But a larger number of Lab personnel were crafts and contractors, the people who really kept things running. They were ordered to report to the high school parking lot and made to stay there for eight hours in the hot sun with no bathrooms and no water. That was the only way they could get paid. It was unclear how this was supposed to make them take safety more seriously.
With the shutdown, I realized that Pete was doomed, that he would eventually be fired. The shutdown made his unfortunate style of leadership visible to high levels in Washington. You don't shut down the nation's nuclear weapons design capability because a GS-12 tells you to. At that point, I decided that I had had enough of Lab management. I'd worked for two Lab Directors. I didn't think I was going to get a chance to work for a third- Directors usually appoint their own deputies. So, I went into Pete's office one day, and I said, "Pete, I've been doing this about five years. I come from academia. I'm not really, deep in my heart, a senior manager. I want to go back to doing science, here at the Lab, in fact." My hiring letter from John Browne said very explicitly that if for any reason I left my management position, I would become a senior scientist at the Lab. The lab doesn't have tenured positions, but senior scientist position is somewhat the equivalent.
With the same job security, or no, that's not part of it?
No. But, de facto, no senior scientist at the Lab is ever fired except for gross malfeasance. So, I said to Pete, "I'd like to step down from management and be a scientist again." Pete pointed his finger at me and said, this is engraved in my memory, "You rat. You're a rat leaving a sinking ship, and I'm going to make sure that you drown." And first of all, I thought, "Wow. This guy has more self-awareness than I gave him credit for. And a naval metaphor, even!" I didn't say that. What I said instead was, "Oh, no, Pete. This has nothing to do with management at Los Alamos. This is just my desire to get back to doing science." Pete tried to have me fired. His first thought was that he would call security, and they would, that day, escort me off the property, and that would be it. My paycheck would stop.
Bill, did you sense personal animus leading up to this moment? Or did this sort of hit you like a ton of bricks?
Ton of bricks. When he wasn't cursing at you, Pete was a genial guy. So, yes, then I did get a bit nervous. Back in my office, I called my friends in the University of California Office of the President and the University of California Office of Lab Management. John Birely was a wonderful person and a good friend. I said, "I think your director has gone berserk; and unlike John Browne, I'm not going to go quietly. I'll make sure that this becomes very, very public." There was then a period of a week or two where I was cut out of the loop. Eventually, cooler heads prevailed. The compromise was that Pete got to cut my salary by twenty-five percent immediately, which he did; and I was given three years of funding as a scientist at the Lab, with the understanding that, if Pete continued as director, I would figure out somewhere else to go during that time. Of course, I didn't think he would be continuing.
My prediction came true. Pete was fired—officially, resigned as Director in May, 2005. The Lab, since that time, has had very good management. Things calmed down and the next several directors, Bob Kuckuck, Mike Anastasio, and Charlie McMillan, all of whom I knew well, were very sensible people, very aware of both the Lab's missions, the basic science mission and the nuclear weapons mission, and also the imperatives of safety and security.
Return to Science, Moving Towards Biology
Bill, you used the term flavor of the month for graduate school when you were really highly sought after in many places. Did you start to think, when things turned south at Los Alamos, that it was time to put out feelers for any number of departments that would have been thrilled to have you if you wanted to sort of get out of Dodge yourself?
I didn't then. Once I was secure having three years of LDRD funding, I planned to stay at Los Alamos for a while. I already had a plan to join the Lab's Statistical Sciences group. That choice, instead of joining the Theory Division (which everyone expected) came about in an odd way, as the result of 9/11.
How was that?
On September 11, 2001, I happened to have been in Washington, D.C. I had been in some meetings and was visiting my parents, whose apartment at the Watergate was, as the crow flies, two miles from the Pentagon. We watched the smoke pillar rising from where the plane crashed into the building.
It turned out that there were more than a hundred Los Alamos Lab people in Washington on 9/11. That wasn't unusual out of the 3,000 professionals at the Lab, 12,000 total staff. On average, that number might be in Washington for one reason or another. The Lab travel office was very good. They knew where everybody was staying, everybody's contact phone numbers. The big question was how to get us all back home. There were no planes flying for several days. And then, it took a week to clear up the airline backlog.
During that time, the only way you could get back to Los Alamos from Washington was being lucky enough to have a rental car and driving. The Lab managed to pair me up with Sallie Keller, and with somebody else's rental car—that person was going to wait for a flight. Sallie was, at that time, the group leader of statistical sciences. I knew her only slightly. There were 300 groups at the Lab, including operational groups, so there were about one hundred scientists at the group leader level, first line management. Later in her career, Sallie was President of the American Statistical Association, and she was later Dean of Engineering at Rice. We became good friends driving thirty-two hours over two days from Washington to Los Alamos. The Interstates were crowded with a rainbow of license plates from different states. Everyone was trying to get home. At rest stops and restaurants, everyone was trading news and rumors.
Now, jumping back to Pete's trying to fire me, I went to Sallie, who was still the group leader in statistics, and said that I needed a safe house, a place to spend the next three years flying invisibly below the radar. So, I joined the Statistical Sciences group at Los Alamos and became invisible. In fact, I'm still an adjunct member of that group, and in the summers, when I'm in Los Alamos, I go in and meet the new people, and I usually give a talk on something that I'm working on. I liked getting to know the statisticians. They're a different tribe, and I learned to feel at home with them and speak their language.
But, for me, the big decision in going back to science was, what should I work on? The natural thought was that I should go back to being an astrophysicist. But by now, I hadn't really done astrophysics, started a new research program in astrophysics, for almost a decade. In contrast to that, at Los Alamos, I was, administratively, very much involved with the biology research. One of my accomplishments at the Lab was to triple the size of the Life Sciences Division. I was involved in all of the senior hiring decisions in biology, which meant that I had gone to dozens of job talks by candidates, seminars, and so on. It seemed to me that this was a great opportunity to just switch fields. I was pretty current on what were interesting problems in biology and (especially for me) computational biology. So that was the direction I moved.
So computational biology was already being done at Los Alamos?
In years well before me, the Lab always did some life sciences, but in a small number of very directed areas. The technology of flow cytometry, microfluidic cell sorting, was developed at Los Alamos. But the Lab was also the home of some of the largest supercomputers on the planet, so it was natural that that new biology efforts should tend towards areas requiring large-scale computation. So, genetics for sure. When I arrived at the Lab, it was already an important part of the Human Genome Project.
You were quite the hacker in your undergraduate days and afterwards, so I imagine that on the computational side, you were comfortable in that. But to what extent, as you were getting more involved in biology, were you an autodidact, you were learning things on the fly? Were you just learning what you needed to answer the questions that you had in front of you? How did you approach that?
Partly just reading books. Bruce Alberts' Molecular Biology of the Cell was my bedtime reading for a year. I knew Bruce from when he succeeded my father as President of the NAS. But, more importantly, I sought out collaborators who knew more than I did. I made several long visits to the Institute for Advanced Study in Princeton, a familiar and friendly environment, where, with John Bahcall's connivance, I pretended to be visiting the astrophysics group. IAS was making a serious attempt to create a group in theoretical biology. It had recruited Arnie Levine, formerly President of Rockefeller University, a very distinguished biologist, always on everybody's list as a likely candidate to win a Nobel Prize, but he hasn't yet.
I spent, over the course of a couple of years, what must've added up to four or six months in Arnie Levin's group, collaborating with a very smart guy named Harlan Robins. Harlan had started life as a string theorist in physics, and then, much earlier in his career than my late career move, had transitioned to become a biologist. Harlan later became the head of computational biology at the Fred Hutchinson Cancer Research Center in Seattle, and founded a biotech company; but at IAS, Harlan was my tutor. And when Harlan didn't know the answer, I could wander down the hall and ask Arnie.
Here's an example of the difficulties of a physicist trying to drop into biology: One day, I decided to browse all the biology books on the shelf in the Institute for Advanced Study library. There weren't that many of them. I opened a book on population genetics, it was Ewens' Mathematical Population Genetics, and right away, it clicked for me. It had equations, it had differential equations, it had partial differential equations, a diffusion approximation for how favorable genetic mutations, alleles, spread through a population. This all made sense to me.
I took the Ewens book and went running to Arnie Levine's office. "Arnie, finally, there's a part of biology that makes sense to me! Why didn't you tell me about this?" Arnie took the book and paged through it and said, "Oh, yeah. They made us take a course in this in graduate school. This isn't biology. You're just wasting your time with it."
So, for you, it was like, here's an astrophysicist who's starting to get interested in biology, and now I'm just simply going to jump into biology with an astrophysicist's sensibility?
Well, Arnie guided me as to what was real biology, versus things that were vaguely in life sciences but dominated by techniques from other disciplines: biophysics, biochemistry, mathematical population genetics, and so on. Important fields, but not what I wanted to get into. Arnie also taught me, maybe this is obvious, but I had to learn, that there was a "biological intuition" that good biologists all had. Now, I certainly knew what physical intuition was. We talked about that earlier. My pretty good physical intuition was responsible for a lot of my physics career.
Well, biologists, maybe even more so than physicists, have an intuition, and it is different. I would bring to Arnie some computational genomics idea. "Arnie, I'm thinking of spending a couple of weeks looking into this." He would say, "You're wasting your time. You're won't find what you are looking for." And I would say, "Well, how do you know that?" And he would say, "Well, life just doesn't work that way." And he was generally right.
I definitely internalized that lesson. In biology, I try to collaborate only with real biologists. The test of "real" isn't how big is their lab, or how adept are they with pipettes. And, certainly, it's not how mathematical are they. The test is, do they have this biological intuition about what's going to work and what's not going to work? I'll never develop that myself, I started too late. And the next test is, then, for any particular project, will the results be interesting to real biologists, not just to people like me with a physics sensibility?
Bill, of course, this idea of pursuing "real biology" is a highly subjective term of art. So, I wonder if you can give an example of a research question or project that would illustrate the point as you define it.
Well, I've worked at the University of Texas with a talented- I think he would self-identify as a biophysicist, but to me he's a biologist, Ilya Finkelstein. Ilya figured out a way of repurposing the microchips that are used in so-called next-generation sequencing so that they could, completely unexpectedly, enable massively parallel experiments in DNA-protein binding. That was something that the chips' designers never thought of or expected.
The big immediate application of this technology is to CRISPR, the technology that allows editing a genome almost as if you were on a word processor, deleting, inserting, correcting, and so on. And CRISPR all depends on enzymes, originally natural enzymes found in bacteria. The CRISPR protein complex runs around the genome and finds the place that you want to edit and binds there. But where else might it unintentionally bind? What are all the so-called off-targets, the hundreds or thousands of other places in the genome where it might bind a little bit and cause some damage that you didn't expect. Before Ilya's platform, it's called CHAMP, was developed, you guessed what might be another off-target location in the genome, and then you assigned a graduate student to spend a day or a week measuring that one location. CHAMP looks simultaneously at a million or more locations of the genome and assesses quantitatively how large the off-target binding is.
How did I get involved? Ilya happened to sit next to me at a biology seminar, and he said, "Bill, my lab's been working for three months trying to make sense of this data, and we just can't make it work. Do you think that's something you could have a look at?" I love being invited into collaborations that way. My graduate student John Hawkins and I dived right in. The data comes in the form of a splotchy photographic image with millions of little dots on it. You have to align those dots with known locations on the chip of various genetic sequences in the presence of all kinds of noise. In some ways our contribution was prosaic, just a kind of bespoke image processing. But the same time, we contributed something essential to the technology.
So, for me to be interested, the underlying biology should be cutting edge, should be of interest to other biologists (in the Arnie Levine sense)—and should make an advance. At my career stage, advancing a technology that will let other people make discoveries is fine. I don't expect to win any prizes in biology, although I hope that my collaborators do. Jennifer Doudna's lab was an early user of the Finkelstein lab's technology, so I'm a minor co-author on a paper whose authors include her, and she just won the Nobel Prize.
So, you are never guided by your sensibilities as an astrophysicist, and how much is it just something that's fascinating to you, and you're good at it? In other words, can you detect your physicist fingerprints in this work?
Okay, you got me on that! Sometimes I do see a little problem in biology that I know is not going to be of interest to the biologists, and (if I can do it by myself, without ruining the career of a student) I do work on it anyway, in violation of all the principles that I just outlined. One of the first projects I worked on in biology, with Harlan Robins, is a phenomenon in mammalian and bird genomes called isochores. The isochores are blocks, sometimes quite big blocks, where the genome will, for a stretch, be very rich in C and G and poor in A and T. And then a block the other way around. This alternating block structure marches irregularly down the chromosome. It's been observed, and not explained, since sequencing first became practical. Harlan and I wrote a couple of papers on this, despite Arnie's not finding it interesting.
Another biological mystery that any physicist would instantly be interested in are features called ultraconserved regions, UCRs. Across a wide range of vertebrate genomes, there are regions that are exactly conserved over hundreds of millions of years of evolution, without even a single base-pair mutation. Why? How? It is the fate of all papers written about UCRs to disappear into the literature without leaving a trace. Real biologists just don't find the topic interesting. I spent months on this and found some interesting things, but never published anything.
Reasons for Leaving Astrophysics
Bill, at any point did you self-consciously see yourself walking away from astrophysics and sort of jumping with two feet into biology?
I made the decision early on that I would not go back to astrophysics. After a few of these visits to the Institute for Advanced Study, I was committed to working at least at the interface between biology and computation, if not literally as a biologist.
Now, let's just explore that a little bit. It's not like there was nothing left for you to accomplish in astrophysics. So that does leave open the question, why? What were you looking to accomplish as a scholar, as a scientist, that was more invigorating, more compelling in jumping into biology self-consciously?
First of all, I think it would've actually been harder for me to go back into astrophysics. I would not have had the advantage of cross-fertilizing a new field with things that I knew how to do, that they didn't know how to do. In the almost ten years that I had been away from astrophysics, the field had moved on. Computational models were much more detailed. The modeled physics was different. One could speculate on whether I even could have gone back into astrophysics. This gets back to my discussion of not wanting to spend the final twenty years of my career as George Field did, a respected senior scientist doing a little work here and there. Had I gone back to astrophysics, that's what it would've ended up- at best.
Switching fields is not for everybody. It's not for the faint of heart. But when you do switch fields within science, you automatically bring to your new field a dowry. You bring a set of techniques and ways of thinking about things that people in your new field are not familiar with, that they don't even know about. With Ilya's CHAMP platform, the first thing John Hawkins and I did, faced with this picture full of noisy dots, was to say, "That's like an astronomical star field. What techniques do the astronomers use to locate stars in a CCD image and filter the noise out?" These would be things that I happened to know about because I was an astrophysicist, but no biologist would be expected to know about them.
Bill, in this ten-year period, you didn't keep up with the literature in astrophysics? So that you could make a smooth transition back?
I wasn't trying to do astrophysics while I was at Los Alamos. I was in management as a full-time job.
But you could have tried to at least keep up with the field.
I know people who tried to do that in similar cases, laboratory physicists, say, who became presidents of universities. Bob Dynes, who became Chancellor of Berkeley, is a good example. He was always proud of keeping his lab going. One day a week, he was just unavailable as chancellor of the university because he was in his lab. That was his decision to make, but I respectfully disagree with it. I think that these killer, high-level management jobs are not for part-timers. You have to be seen by the people working for you as being totally committed. You have to throw yourself into them. John Browne made pretty clear to me that if I wanted to come to Los Alamos as Deputy Director, it would be my full-time job.
So, when you left Harvard, you had already come to the conclusion that you were leaving astrophysics. The two were connected, as I'm hearing it.
There was the two-year period that I could've gone back to Harvard. That wasn't so long that I couldn't have caught up with astrophysics.
But even within that first two-year period, it sounds like you were not keeping up with the literature.
That's correct. But, maybe you've figured this out already, I have a high opinion of myself, whether deserved or not. I figured that I could do the Los Alamos thing, and then I would do something else. And maybe that something else would be going back to academia. It turned out to be that. Maybe that something else would be going into university administration. I didn't feel that I had to have an exact plan for what I did next.
Let me just interject on that point as well because even if you had a high opinion of yourself, you already cited the fact that you had not kept up with the literature, which you could have done. But the fact that you had not kept up with the literature, even within that early two-year period suggests, maybe subconsciously on your part, that by virtue of doing that, you were somehow not disqualifying yourself but self-consciously removing yourself from the kinds of interests that would have allowed a smooth transition back into the field. It's interesting. It strikes me as more dramatic than I had originally understood.
Wow, I need you as my psychoanalyst! I don't see anything dramatic about it. I didn't want to do the kind of thing Bob Dynes did. I was just going to jump feet-first into learning to be a good senior manager. And, as I later learned, there was a leadership component to that, which is quite different from management. Then, when that career stage was over, I looked around to decide what to do next. The real decision point was after Pete fired me, and I did have to do science. I mean, even at Los Alamos those three years, my job was not to sit around drawing a salary. My job was to do science research. Well, what research should I do? It was a conscious decision that my new field should be biology, that I should consider myself in training to do biology, and as early as possible trying to do things that were productive. That's when I turned my back on astrophysics. By the way, many of my best friends are still astrophysicists.
Did any of them encourage you to come back into the fold? You must have disappointed some people.
There's a more general question as to what the right role for scientists at the career stage of ages is fifty-plus. Some colleagues in that age range are able to contribute directly. But many, including some of the contributors, are also simply blocking the advance of younger scientists. It's simply a fact that the best science is done, on average and in many fields, by people under the age of fifty. In mathematics, they would say thirty. I didn't want to end up in a position where, behind my back, the junior scientists were saying, "Why doesn't he step aside?" Moving to biology, I could bring a dowry of computational expertise to justify my existence.
Move to University of Texas at Austin
How did Austin come about for you?
In kind of an indirect way. I had some connections to an institute at the University of Texas at Austin, the Institute for Computational Engineering and Sciences. I was on their outside visiting committee. So, I knew what went on there. It's now renamed the Oden Institute. It's unified by interest in large-scale computation and very broad as to the areas in which that large-scale computation is applied. It mostly came out of an engineering context, doing engineering models of things. But it also already had a bit of life science modeling, protein folding, solid state physics, other things like that. And, if you go back to early in my career, I knew UT Austin for its outstanding general relativity group at that time, with still some remains of that now.
It was probably Roy Schwitters who brought me to the attention of UT. Roy had been a professor with me at Harvard when we were both young, had then moved to Texas to become Director of the Superconducting Supercollider, which was, for all kinds of unfortunate reasons, never completed. He had ended up a professor at UT Austin in physics. But the triggering event had nothing to do with physics. UT was one of these places that once had a statistics department. Then statistics, as a separate academic field, went into eclipse. A lot of places, including Texas, merged their statistics departments with their math departments. We're probably talking about the 1970s or maybe the 1980s.
But now, here we were in the 2000s, and statistics, especially computational statistics with its connections to computer science and machine learning, were once again hot topics. They still are today of course. Mary Ann Rankin, the Dean of the School of Natural Sciences, decided that Texas should once again have an independent, free-standing statistics department, and that it should be very broadly based. It should have people in it who were not just PhDs in statistics but people from related fields. Maybe astrophysics, maybe physics, maybe biology, or something like that. And there I was, an astrophysicist doing biology and in the Statistical Sciences Group at Los Alamos, no less.
So, Mary Ann recruited me to come to Texas as one of the nucleating agents, that's one way to say it, of this new statistics department. Because I was being recruited by the dean, and for a purpose that didn't fit into any existing department, it was quite up in the air which existing department I should actually be in. As the negotiations proceeded, neither of us really thought of that until very close to the end. Mary Ann said, "Well, you should go meet your colleagues in the physics department." And I said, "Oh, I don't want to be in the physics department." I told her that I wanted to be in computer science, as actually quite close to my interests, and also in one of the several biology departments.
This even being possible was a credit to Mary Ann personally, also to the credit of the expansive Texan way of doing things, also always very welcoming. Harvard is stove-piped. If was being recruited at Harvard, it would be by a department. If I sort of breezed in and said, "Oh, I want a different department," they'd have looked at me like I was crazy. Texans, on the other hand, are open to wild, crazy ideas. Mary Ann said, "okay, why don't you go give a talk in the computer science department, and then give a talk in the integrative biology department," which was actually her home department, "and if they're happy, then I'm happy."
So, I ended up in those two departments. Two new tribes whose languages I've had to learn. I'm closer to the computer scientists professionally, but my wife and I socialize more with the biologists. Integrative biology has these fun people who actually know about animals and plants, as well as highly mathematical evolution theorists and ecologists, people that Arnie Levine wouldn't think are real biologists, but who I love talking to.
Interestingly, the whole reason I was being brought in—related to the statistics department kind of evaporated. There came a point where the nascent statistics department had to decide, did they want to be a department with this broad base of backgrounds—and there are statistics departments in the country that are like that, Duke, for example, or did they simply want to be excellent in the scholarly and mathematical field of statistics? Despite my being there, despite my being in on these discussions, in the end they went the second way. Texas did succeed in putting itself on the map with a good statistics department, and I still have some vestigial affiliation with it. But by and large, the people that they hire are people with PhDs in statistics who publish in statistics journals, which doesn't fit my profile at all.
Bill, I'm curious what relationship, if any, you've had with Steve Weinberg, either from Harvard or at Texas.
Friends from the time that we were both Harvard professors. Steve is an academic generation older than I am. Steve and Louise Weinberg are the founders of an organization in Austin that my wife and I are members of called the Tuesday Club. Members are drawn equally from the university and from leaders in the city of Austin, in government, business, or other fields. We get together once a month, hear an interesting talk, and sit at tables of eight or ten people, and get to know each other. In Austin it's something that the Weinbergs are known for.
Did you take on graduate students when you got to Texas?
Now, that must be sort of a delicate situation because you are a newbie insofar as being a graduate advisor as concerned in these fields. I wonder how you navigated that.
Well, I've taken graduate students not from the biology PhD program but from the applied math and computational science PhD program. I've always put my cards on the table, saying, "Come work with me if you want to end up more of a biologist than I am." I've insisted that my students spend at least half time across the street in the biology labs, working directly with biology graduate students and postdocs. The other half, they can be on my side of the street, staring at a computer display, or talking to me about algorithms, math, statistics, things like that."
So, you knew quite well that in order to become a graduate advisor at Texas in these new fields, you would need to take on a very unique kind of graduate student who had, baked into their own research agenda, a multidisciplinary plan. You must have recognized that from the outset.
My perspective was that as a young scientist I had had the fun of riding the wave of bringing large-scale computation and computational modeling into astrophysics. Large-scale at that time, at any rate. Now, in the 2010s, I saw going on around me the same kind of wave in biology. Why not ride this similar wave a second time, and teach students with a background in computation and mathematics also to ride it? I've been lucky twice, in quite different fields, helping to bring in larger scale or more sophisticated computational methods.
In what ways were those waves different, and in what ways were they the same?
Different in scale, obviously, because Moore's Law had advanced through many orders of magnitude. Different also because, when I was a graduate student in the early 1970s, it was even unusual for scientists to program a computer themselves. There were professional programmers and so on. Similar in the sense that both fields were welcoming this invasion by computer people. Far from finding the biologists resistant to this, I've found the biologists to welcome people with new computational and statistical techniques. In much less than one human generation, laboratory biology has advanced from mostly single test tube experiments to massively parallel experiments that require whole different approaches to analyzing data and visualizing results.
Now, by the time you had gotten to your second graduate student, had you sort of gotten the hang of it, and you thought this was something you'd be able to continue doing- turning out PhDs in computational biology?
Yes. My first two students, Jeff Hussmann and John Hawkins, both seem to be well-launched in their careers. They are biologists with professional research-level abilities in applied math, and computational techniques, and a lot more that goes beyond normal biology training. Biologists seem to think of them as biologists. So, in that sense, the model seems successful. I expected to take on further students. But, instead, my career took yet another unexpected direction. President Obama was elected.
President Obama's PCAST
And you were invited to join PCAST.
Yes, the President's Council of Advisors in Science and Technology. Initially I thought, Oh, this will be a feather in my cap. This will be a great line on my CV. And I'll still be able to continue doing computational biology at the same pace. But it quickly became clear that PCAST, if I really threw myself into it, would be something like a one-third to one-half time job. So, this became another conscious choice that I made between doing research and doing policy. Once again, research lost. I have a little bit of regret on the one side, but, as it turned out, I'm proud of what we accomplished in PCAST. And, of course, it was just so fascinating to see how the near-inner (not truly inner!) circles of the White House operated, especially with a president who was so knowledgeable about science and friendly to it. Within a year, Harold Varmus, one of the original PCAST co-chairs, left to go back into government at NIH, and the two remaining co-chairs decided to appoint two vice-chairs of PCAST, of whom I was one. Maxine Savitz, a retired executive from Honeywell, was the other. Then I really was committed.
So, what were the origins of PCAST for you? Did you just get a call out of the blue one day?
Yeah, exactly that. John Holdren was announced as the President's Science Advisor very early on—sooner in the Obama administration than in other administrations. Holdren, Eric Lander, and Harold Varmus had been on the Obama transition team, I think. They became the initial PCAST co-chairs. The story as I heard it was that the three of them sat around in Eric Lander's office in Cambridge and made lists of names on the blackboard, trying to satisfy all the constraints. You wanted both scientists and engineers. You wanted people from life science and physical sciences. You want a group reflecting diversity in several dimensions.
I didn't know any of those three personally. I had met John Holdren once or twice, and I had done some work for the MacArthur Foundation, whose board he was on. I knew of Eric Lander from his role in the human genome project. Harold, I probably only had read about as a Nobel Prize winner. Still, my name ended up on their blackboard. It probably helped that I had moved to Texas. They had no shortage of Harvard professors to choose from. Us people from the "fly-over" states were in shorter supply, in their possibly parochial view.
What was the request exactly for PCAST? Did you have the sense then that this would turn into a major activity?
I think nobody knew. In any new administration, everybody from top to bottom is just feeling their way, inventing or reinventing procedures as they come in. What gave PCAST influence was that John Holdren and President Obama quickly developed a close rapport. John Holdren didn't have Cabinet rank, so at Cabinet meetings, he wouldn't be sitting at the table; he would be sitting in the row of senior advisors behind the Cabinet officers. But when issues with a science component would come up, the President would turn to John. "Dr. Holdren, could you explain this to the Cabinet." John would have had no preparation or warning, but he was able to do well across a wide range of topics and was always able to give thoughtful advice. When a President's confidence in an advisor is this visible, everybody in the West Wing takes notice of it. Holdren and OSTP, the Office of Science and Technology Policy, were probably more influential than they had been under any previous president, and PCAST correspondingly so.
And your appreciation of Obama's respect for science and scientists, that preceded your formal inclusion in PCAST. You understood this.
No, I don't think we knew. Obama had run on the platform of hope, not science. We twenty people on PCAST discovered his interest in science at our meetings with him. Over the course of eight years, we met with Obama something like twenty-five or thirty times. We were on his schedule even more than that, but crises came along. Anybody scheduled to meet with the President months in advance is going to be often disappointed.
There was an interesting learning process. Our first meetings with the President were very formal. They were in the State Dining Room around a big table. We each had our little pad of White House note paper and our White House pen. Souvenirs to take home.
Of course, you don't go into any meeting with the President just to wing it. In several hours of preparation beforehand, we had decided which important advances in science we should tell the president about. Maybe it was gene therapy, or graphene, or the discovery of the most distant galaxy. We planned to go around the table, and PCAST members would say, "Mr. President, I'd like to take a minute to tell you about an advance in gene therapy." That was the plan.
The President greeted us, walked around the table and shook hands with everyone. "Mr. President, we'd like to tell you about an advance in gene therapy." "Oh, I read about that in the New York Times. What else can you tell me about it?" We were sort of floored. We went on to the next topic. Same thing. Obama already knew about everything that a science groupie who read several national newspapers every day would know about. We quickly got away from thinking that our purpose was to educate him about New York Times-level science.
We started concentrating on writing reports with actionable recommendations. Many groups do this, National Research Council, JASONs, and so forth. But only we had the unique opportunity to present our recommendations on any subject directly to the President. And we were unique even among groups with access within the White House. I need to explain why. In a well-run White House, you can't have recommendations going directly to the President willy-nilly. You have a Chief of Staff organizing things, you have a system of staffing, and approvals. In his role as Presidential Science Advisor, if John Holdren wrote a memo to the President, it would get staffed first, which was a complicated process.
Bill, that's an insider-y term. Perhaps you could explain, what does it mean to get staffed?
Staffed means that John would write a draft memo to the President. It would go to the Chief of Staff. The Chief of Staff would send it out to many other people within the White House to comment on. They might not just comment, but in some cases, want to rewrite parts of the draft. Then, there would be a negotiation with John about the memo's final wording before it went to the President. And other people's comments would go with it, appended. If it did get to the President at all, that is. It has to be done that way; the President's bandwidth is too valuable. He can't be bombarded with a disorganized cacophony.
On the other hand, due to a quirk in the law, PCAST was different. Legally, we were a FACA committee, a committee chartered under the Federal Advisory Committee Act. Many federal agencies have FACA advisory committees, but we were the only one in the White House. By law, no government official can alter the findings of a FACA committee. So, at our meetings with the President, we could just leap right in, without any prior staffing or approval of our recommendations: "Mr. President, on this matter of such-and-such, we recommend that you do such-and-such."
We tried this in some of our early reports. We had some great recommendations. "Mr. President, by executive order, we think you should do A, B, and C." And Obama would say, "What does the Secretary of Energy think of that?" And we would go, "Well, we wanted to see what you thought of it first." "Oh, well, what does X think of it? What does Y think of it?" We would get what I thought of as the "Obama stare". I don't think the public ever saw this. Barack Obama was a charismatic guy and campaigned on hope. But he was also a constitutional lawyer who taught law students at Chicago. The Obama stare was chilling. He must've given it to first-year students when they didn't know the answer to a question. It was worse than being chewed out. It was just like, "Don't do this to me again, and don't waste my time."
We learned that we had to do our homework. We also learned that we had huge access and convening power, just by virtue of the fact that we could brief the President directly. For our later reports, we would get together the outlines of the report and a set of draft recommendations, and we would then shop them around town to the people who had a dog in the fight, the people likely to be named by Obama as "What does X, Y, or Z think of it?" We could say, "Mr. Secretary, we're thinking of recommending the following things to the President." "Oh, how are you going to do that?" "Well, we meet regularly with the President." "Oh…"
Now, when we met with the President, we could say, "We think you should do this. We've consulted the following people, and they concur. The following people have reservations, and here's what their reservations are." That was vastly more effective.
We also learned how to judge, with President Obama, how well we were doing. On a given topic, we would brief the President. He then might say, "Very interesting. Thank you very much." That was about a grade of a C- or a D. It meant, "You guys wasted your time. This isn't going to go anywhere." By the end of our eight years, he knew us pretty well, and we had gotten to know him a bit. We now met in the Roosevelt Room, right across from the Oval Office. We no longer got souvenirs. When we got his low grade, we could say, "What's wrong with it?" and he would spell it out. "There's this consideration, there's that consideration. Thank you very much. Next topic."
A higher grade would be, "That's pretty interesting. Go back to the Secretary of HHS and raise these additional issues. See what she thinks." That kind of thing. He would sketch a path forward and give us another chance with the issue. That was a B. There were usually some key, very senior White House staffers who attended our meetings with the President, either because that was their job, or else in self-defense. If our report hit a home run, as it were, Obama would point to one of his senior people in the real inner circle and say, "I like this. Follow up on it and see if we can make it happen." That was a grade of A.
Bill, I wonder if you can talk about, in contrast with the National Academies and the JASONs, the immediacy of that direct interaction with the President himself. In what ways was that satisfying, just in terms of cutting through all of the institutionalism and bureaucracy that might have been an enemy of good science policy that you may have experienced with your other affiliations and service work?
It was satisfying to find that the President was very pro-science. Of course, the people we were dealing a few levels down in the agencies were also pro-science, because they were the science people in those agencies, usually scientists themselves. But in-between, there were layers of traditional bureaucrats, often lawyers or government majors who got Cs in their science courses. They didn't trust scientists. They were Bismarck's "iron majors" standing against change. During our time, we got better at working the bureaucracy, more trusted by the White House senior advisors. But never completely.
Still, when it all did work, it was very beautiful, and we could accomplish things. Sometimes we barely left any of our own fingerprints. Effective bureaucrats in government know that the best way to get things done is to not try to have your name on it, or to take credit. Credit is another degree of freedom that you can use in negotiating. Sometimes there was going to be an event associated with the public release of one of our reports. John Holdren or Eric Lander would give a little talk about what we were recommending. These were considered White House events, even if the President himself wasn't there. When we were negotiating our draft reports with Cabinet-level people, we learned to ask them if they wanted to be on stage at our White House rollout. That made them more likely to spend time on our recommendations. We had to learn these personal touches.
In our committee of twenty, not everyone worked equally on every report. Still, the Obama PCAST issued more than thirty public reports in our eight years, briefing each to the President before it was made public. It really turned into a big enterprise, done with a tiny staff and tiny budget. It worked because all twenty of us volunteers were enthusiastic about it.
Specific PCAST Issues
As you were preparing to join PCAST, what were some of the major issues in science policy that you thought would naturally bubble to the top of the agenda, and then how well did those preconceived notions accord with what you actually found on the ground?
Yeah, that's a very good question. I would have to say that there was really only one PCAST report that fit my preconceptions of the sort of high-level, philosophical, policy wonk, et cetera. I played a big role in our 2012 report on the U.S. research enterprise, writ large. How was federal support structured for research universities? What was the interaction with industry? What was the role of the national labs? Where was science healthy, and where could it be improved? I think it was a good report. A number of its recommendations found their way, either directly or indirectly, into policy.
In truth most scientists would perceive most of our reports as being down in the weeds. Important, maybe, but in the weeds. My academic colleagues, during these eight years, would ask me what we were working on. I might tell them about the huge swaths of radio spectrum that were reserved exclusively for national defense use but really, using more advanced technology, could be shared broadly between defense and consumer use, for example, next-generation cell phones. That was a report we did in 2012. And my colleagues' eyes would glaze over. It sounded to them like an issue that could be resolved by some medium-level bureaucrat who would say, "Well, obviously. Just work it out." But, in fact, that report required extended negotiations between us, and the FCC, the White House office for radio spectrum, and the Defense Department. People at very high levels had to weigh in. Once you give away spectrum, you never get it back. Mark Gorenberg, our venture capitalist, was the PCAST member most responsible for this.
I give that one as an example. It sounds so prosaic, but if you asked an economist to estimate the present value, in dollars, of the future expansion of the U.S. economy made possible by this new technological idea of spectrum sharing, the answer might be a trillion dollars. By today, if you Google for it, you can find all kinds of stuff about spectrum sharing. When we wrote our report, the concept was almost unknown and to the extent known was violently opposed by the telecom companies. They lined up against it with platoons of lobbyists. But it happened. Now, they love it. 5G will probably come to depend on it.
We did a report in 2014 on big data and privacy. That was also interesting behind the scenes. The so-called "crypto wars" between civil libertarians and law enforcement aficionados go back to the 1990s. There was something then called the Clipper Chip, NSA's attempt to regulate all data encryption in the United States through a single hardware chip that would give them backdoor keys. Sounds like a crazy idea, but that same idea keeps coming up every five or ten years in different guises. As we speak, the current crypto war is over whether Apple should be required to make encryption keys available to law enforcement, requiring Apple to hold them all, versus whether they can, as now on iPhones, have the keys held only by the owner of the phone.
Around 2014, in response to the Edward Snowden case, there was a big push by law enforcement to severely regulate encryption and, by law, give the government backdoor keys to all communications. Obama was on the fence on the issue. Maybe surprisingly to some people, Obama was very pro-law enforcement. He was strongly influenced by John Brennan and John Podesta in these matters. We saw it directly. It really seemed that the President was going to tilt in the direction of regulating encryption. We wrote a report about big data and privacy. It was very carefully phrased, but its subtext was quite a firm stand against unfettered law enforcement access to encrypted data by technological means. In the end, I wouldn't say we won a stunning victory. The Obama administration didn't endorse our report. But neither did they proceed with the measures proposed by law enforcement. The matter just dropped. So that was good.
It's striking to hear you say that Obama was very pro-law enforcement, which is certainly not something that, in terms of his public persona, he talked about a lot. That was not a part of who he projected himself to be. And so, to get back to my question about what you found on the ground versus the kinds of things you assumed, I was really driving at the overarching science issues that Obama talked about with great passion publicly. I was expecting you to delve into, in PCAST, the kinds of big scientific endeavors that Obama strongly cared about. Climate change. International nuclear security. Healthcare. And I'm kind of struck that I'm not hearing these as the sort of top line agenda items that you were involved with from day one.
Yes, of course we were involved in those issues. But the digestible pieces weren't as big as those that you've called out. You couldn't do healthcare in one report. Out of our thirty reports, I bet ten of them are, directly or indirectly, about healthcare delivery or reform. About the benefits of a single-payer system, for example. We wrote several reports about data flows and the importance of reengineering healthcare so that, for example, every physician didn't repeat the same laboratory tests simply because they couldn't put their hands on the data of the previous tests (reason one), and because the payment system rewarded them for unnecessary tests (reason two).
Climate change. That's really an interesting one. John Holdren went in believing that this would be the administration to truly embrace action on climate change. A couple of members of PCAST, Mario Molina comes to mind, were hard-over that we should make very strong public statements about climate change. But Obama, in early meetings with us, told us not to waste a minute trying to convince him that climate change was real. He knew it was. And not to waste a minute trying to convince him that it was anthropogenic. He had seen all the evidence. He told us that he would love to act on climate change, but that he had to find politically tolerable actions. His instructions to PCAST were that we should take on issues that related to climate change: energy efficiency, future of energy production, clean air, renewables, end of the age of oil. Anything relating to climate was fine. But we should not put the President in an impossible position by calling for things that he knew politically he just couldn't get done. It wasn't until the last year of his term, at the end of 2015, that he felt politically able to put out a comprehensive plan on climate mitigation and adaptation. John Holdren's doing, of course.
So, the mandate was never a moonshot kind of mandate. He never went to you and said, "Figure out climate change."
It would be the components of climate change to deal with. That was PCAST's mandate.
Well, look, everybody hoped that once he was reelected in the second term, he could be more forceful about climate change, and, in the end, he did put forth a plan that you would call a moonshot. But the Democrats had lost the House in 2010, and the Senate in 2014, so many ambitious things that could have occurred in the second term didn't occur. Much that was hoped for proved to be impossible.
And I wonder, Bill, if this suggests that bold American leadership in international climate policy should have come sooner.
I think we're talking around a point that has been much written about, that there really were these two Obamas. There was Obama the charismatic figure, the campaigner, who brought a feeling of hope and of empowerment to many constituencies. That was real- there was nothing fake about it. But that wasn't how he governed. He governed like the constitutional-
Small C conservative. Small C conservative.
Yes. Or, as I keep saying, like a constitutional law professor from Chicago, methodical, organized, careful. We have to keep in mind that he was learning on the job—look at his history. You don't learn how to govern when you're a senator. The closest he came was way back in his career when he was a community organizer, organizing people to action. That was his previous experience in governing. As President he had to invent the rest. That's not unusual in many administrations. He had the advantage of being very, very smart. We on PCAST were always impressed with how smart he was. If only the Democrats had kept control of even one house of Congress, we might have seen very different results.
Bill, I wonder if you can talk a little bit about the breadth of PCAST. Talking about climate change, nuclear security, healthcare, these are all things, for example, that have a strong economic policy component to it. Were there opportunities to interact with scholars from other disciplines in or out of the administration?
Rick Levin, the economist who was president of Yale, was a member of the Obama PCAST. The PCAST membership was pretty broad. We had technology people like Eric Schmidt, then CEO of Google; Craig Mundie, a senior vice-president at Microsoft; David Shaw, who founded D.E. Shaw and Company and was a financial wizard before he went back to computer science at Columbia—to computational biology, as it happens. Those three were our billionaires. When we sometimes walked as a group for lunch at a nearby restaurant, one or another of our billionaires usually had a limousine creeping along at a discrete distance behind us. Our health experts were Christine Cassel, who was president of the U.S. internal medicine board; and Ed Penhoet, a public health dean. Our environmentalists were Mario Molina, Rosina Bierbaum, and Dan Schrag. Chris Chyba, from Princeton, had been on the National Security Council staff earlier in his career. He and I were supposed to hold down national security. We were frustrated because there were only very few national security issues that John Holdren thought appropriate for PCAST to engage in. Ernie Moniz was the energy guy, until he became Secretary of Energy and had to quit PCAST. Jim Gates, Chad Mirkin, and Ahmed Zewail were academic scientists. Ahmed and Mario were our Nobel Prize winners. John Holdren, of course, had to be an expert in everything as his full-time day job.
Bill, we talked earlier about how you balanced your responsibilities at Harvard as part of the NAS and the JASONs. It's a similar question with regard to PCAST and Texas. How did that work for you?
Well, there's an interesting bureaucratic answer. The Great State of Texas still considers itself in many ways to be a sovereign nation. Since I work for the University of Texas, I am a state employee. There's a provision in the Texas State Constitution that forbids any state employee from holding "any office or position of honor, trust or profit" with the U.S. Government. PCAST counted. Luckily there was an exception procedure, in my case a vote by the Regents of the University of Texas. It wasn't really as dramatic as it sounds. The Regents had a docket appendix, where hundreds of such items were all approved at the same time. But I got a formal certificate from the State of Texas allowing me to accept the PCAST appointment. I also got a formal document from the White House, a "commission" analogous to a military commission. It's very fancy, engraved, "Know Ye, that reposing special trust and confidence in the Integrity and Ability of William Henry Press, of Texas, I do appoint him…" This all made me feel like a real Texan. People who have been in government in various capacities have collections of these commissions.
Start of Trump Administration
We had one final meeting with the President after the 2016 election when Trump was elected. We met with him the first week in January, just three weeks before the end of his administration. You can guess that we were literally practically in tears. You can only imagine.
I've read that he was actually a consoler of people. He was the one that was bucking people up.
Exactly. He took it upon himself to give us a message: All things pass. It won't be as bad as you think. Unfortunately, it's been much worse than we thought. Those last few weeks were interesting, however. We received instructions from the White House Office of Personnel that we should send in our letters of resignation, to be acted on before the new President, Trump, took office. We weren't lawyers, of course; but we looked up the statute, and relevant executive orders, to see what happens to FACA committee members at the end of an administration. The rule was that FACA members would continue until replaced by the next administration. So, we asked the Office of Personnel what would happen if we just didn't resign? Would we automatically become members of President Trump's PCAST. It might be years before his administration even noticed that such a council existed. Maybe we could do some good in that time. We got back a very stern reply saying that President Obama had ordered all loose ends to be tied up, so that his administration would go out with style. If we didn't resign, we would be fired. Then, we would forever have to indicate in any application for a federal position that we had been fired from a previous federal position. So, we all did resign before the deadline.
What concerns might there have been for knowing what was going to become of science policy, or non-science policy, as it was, in the incoming Trump administration? Was there any concern that somebody's got to be in charge here. Somebody needs to take the reins because there's gradations of how bad this can get, and if someone stays on, maybe we can prevent it from getting as bad as our worst fears?
A small number of career federal employees on the professional staff of OSTP did stay on. And I'm sure they did their best. But it was a very rocky transition. The people in OSTP had prepared detailed briefing books describing their responsibilities, which were continuing issues, what policy things were still in the works that the Trump people could either continue or not. Voluminous briefing books. They could never find anyone to brief. I heard that they literally left them on their desks when they cleared out on January 20th. There was no uptake. A lot of people were trying to figure out in those early days, who could speak for the new administration on science policy, on technical matters generally. Was it Peter Thiel? Was it Jared Kushner? It was complete confusion. I wasn't there anymore, of course.
You now have a very unique perspective in science policy and service in this transition from the Obama to the Trump administrations in terms of the National Academy. Obviously, it's a very different perch than PCAST, but in what ways has being involved on any level in Washington during the years of the Trump administration are there obvious distinctions from his predecessor?
The differences are so obvious and so much a matter of newspaper headlines now, and soon history books, that I don't have much to say about that. I've stayed away from any association with the Trump administration. My wife has raised it as a hypothetical, "Bill, what would you do if the phone rang, and they asked you to take a job in the Trump administration, and as painful as that might be, as damaging to your reputation that might be, you knew you could do good? What would you do?"
I would have said politely no. There are times when you have to draw the line, when you can't buy the argument that it's better to do good on the inside. It's case-by-case. I joined JASON only a few years after the Vietnam era, and it was the same argument. Even though it might damage my reputation among scientists, I could do good on the inside, advising the Department of Defense. I bought that argument then, in that case. But the Trump administration is, in my opinion, too extreme. I would not associate myself with it.
It gets back to that question, who is there even to talk to? This wasn't just on January 21, 2016. You could probably also ask that question now.
There is a science advisor in the Trump administration, Kelvin Droegemeier. There is a PCAST; my former Harvard astronomy colleague Avi Loeb is a member. The Obama PCAST issued thirty public reports. The charter of the Trump PCAST forbids them to issue any reports. To my knowledge, they have never met with the President. It's been four years. I'm not a historian of American history, but I think it's clear, the Trump administration is the most anomalous administration in all American history, not in a good way. The only thing even comparable is Andrew Jackson. History has judged Jackson probably better than he deserves. I don't think Trump will get that benefit. But it's a joke in the TV broadcasting business that all live news clips end with, "…only time will tell."
The Skip-Generation Nobel Club
Let's move on to this amazing club that you're a member of, being on either side of a Nobel Prize. Were you aware that there was such a club before you joined it?
No. I guess I should explain what this fictitious club is about. My colleagues Adam Riess, Brian Schmidt, and Saul Perlmutter won the Nobel Prize for the accelerating universe in 2011. I felt great about it, maybe even having helped it along by my work with Adam, when he was my student, on the technology leading to that amazing discovery.
And if you're going to be that humble, let's just cite the literature that states a little more clearly that this research had not been possible except for what had come before it and your contributions to that.
Well, I think it's always true in science that you can trace backward and find what the precursor elements were. But thank you. And then six years later, in 2017, Kip Thorne, my thesis advisor, shared with Rai Weiss and Barry Barish the Nobel Prize in Physics for LIGO. Again, something that I felt I had contributed to- more distantly, because my work on gravitational waves with Kip was by then more than forty years in the past.
However, Bill, in the world of theory, it's not such a long time in terms of how long it takes for theory to be recognized for Nobel-level work.
Maybe. Kip's was a wonderful prize because he was not a scientist who set out to win a Nobel. There are such people. Most of them fail, obviously, but a few succeed and have the satisfaction of having charted that course. Kip never set out to win a Nobel Prize. Kip knew from the Einstein equations that gravitational waves existed. He believed that they could be detected, and he devoted the large part of his career to that. In the end, it was an effort by literally thousands of people. It was great that he, a theorist in the collaboration, could share the prize.
It's been well-studied that the most effective way to increase your own chance of winning a Nobel Prize is to have a Nobel Prize winner as your thesis advisor. These prizes tend to run in academic families. Is that because there's too much of an old boys network, and if you get in early to a network with that kind of influence, it increases your chances? Or is it because there really is something about how to do science that can't be written about in textbooks, that has to be passed on as an apprenticeship from a great scientist to another scientist who can potentially become great? Or is it something about the elitism of institutions? Those are great questions for historians and sociologists of science. But until Kip won the Nobel Prize, I had not thought about the related question, how often does the Nobel Prize "skip a generation", as from my advisor, Kip, to my student, Adam- leaving me out.
In popular lore, we look at our biological grandchildren, and we see in them aspects of ourselves that skipped a generation, that we didn't see in our children. How did that work for academic grandchildren? It turned out I could write a computer program to go through all of Wikipedia and, by a little bit of natural language processing, I could ferret out people who were in my- I don't know whether to call it a happy or a sad, situation of being the skipped Nobel Prize generation. I found that there were about thirty of us in the last century. Edward Teller and Robert Oppenheimer were both on the list. Viki Weisskopf, who we've talked about a couple of times, was on the list. Sam Treiman, who was my colleague in JASON, and interestingly, Charlie Lauritsen, who was my first wife's, Margaret's, grandfather. My friend Rich Muller was also a member of this "club." Mathematically speaking, our club is much more exclusive than the better-known club of actual Nobel Laureates, because membership in our club requires a confluence of two Nobel Prizes, unfortunately neither to us.
I also became more interested in the longstanding question, how does winning a Nobel Prize change a person? There seem to be good Nobelists and bad Nobelists. The bad ones are the ones who decide that they now know everything about everything and start making inflammatory or scientifically wrong statements. William Shockley is everyone's best example of that. I'm happy to say that the Nobelists generationally above and below me are good Nobelists. Riess, Schmidt, and Perlmutter have done a number of things to recognize other members of the fairly large teams that they had worked with. Kip has been tireless in giving popular talks that promote science with young people.
Bill, how can we put this? Had you stayed in astrophysics, not gotten to Los Alamos, it's a pretty reasonable counterfactual, especially because the Nobel aura was all around you, that your subsequent work could have been recognized at that level. It's a feasible proposition. You can push back all you want on that, but it's clear in terms of who trained you and the trajectory that you were on, you could have been recognized at that level.
I'm not trying to be modest here. I made knowing choices early in my career that precluded me from ever doing the kind of work that the Nobel Prize recognizes. I didn't do that kind of science, and I didn't want to do that kind of science. So I think I'm a highly deserving recipient of the skip-a-generation club membership.
It's admirable to just follow your interests and not be concerned about, "Well, maybe I'll never get a Nobel Prize now."
I'm lucky that anyone even wants to interview me for an oral history! (In comedy scripts, you have to include laugh lines. That's supposed to be one.)
Research in Biology
But to get back to not being falsely modest, having a high opinion of yourself, why not try again in your second career, biology, to do work at that level? Why not?
You can't predict impact. You can work on something and do a piece of great work, and it will turn out to be a premature pre-discovery, as we already talked about. You'll get recognized only by the historians of science. You can work on something big for years and then get scooped by another lab. This also gets back to our discussion of Chandrasekhar versus Wheeler. The things that actually advance science are not exactly the things that get recognized by these big prizes. Despite your flattery—for which thanks—I'm not the kind of scientist who could do the big prize work. I've contributed a bunch of small, sometimes clever, bits across a broader than usual span of fields, not just in science, but in policy and science administration. I'm happy with that.
So, it's a good point now to get off of the modesty and what you haven't done. What have you done in biology that you're proud of? It is worth talking about?
Okay. Fair question. In biology, I'm not trying to go after big discoveries. I'm just not equipped for it. My goal is to seed new quantitative methods, methods involving larger scale computing, larger scale statistics, into the individual collaborations that I'm involved in. Helping to introduce the worm of applied mathematics into the apple of modern biology.
So, most of what I've done so far in biology is the same kind of thing that I did with Adam Riess, help develop new tools or technologies. My contribution to Ilya Finkelstein's CHAMP platform has that character. There's simply no way to tell whether this (or any other technology I'm collaborating on) will be the means of making a great discovery, the way Adam's was. I'll hope for Ilya to win a Nobel Prize, as statistically unlikely as that is.
But there are also other things that I just can't resist. One fun thing I've been doing with my former student John Hawkins, is developing an error-correcting code for large-scale storage of data in DNA. The idea of storing data in DNA has been around for a while. It's easy to make vast claims about it. You could store the entire internet in one cubic millimeter of DNA, for example. The real barrier now is that the economics just don't work out. It now costs about a tenth of a cent per bit to encode things in synthetic DNA, so to encode gigabytes, or terabytes, or exabytes is simply not feasible. But we like to point out that the cost of DNA storage now is the same as the cost per bit of the hard disk drive that was sold with the first Apple II computer. So, it's not beyond possibility that someday there'll be real applications for DNA storage. I think our work is definitely in the "premature pre-discovery" category, like gravitational lensing was.
But that makes me want to tell you about a couple of other things in my category of "premature pre-discoveries", papers I've published that, I thought, deserved more attention than they got. (You did tell me not to be modest, right?)
I wrote a paper some years back on how to do clinical trials by a vastly more efficient statistical method than the methods that are currently used. These days, during the COVID pandemic, we read a lot about the "gold standard" of randomized controlled clinical trials.
Double-blind, randomized, that kind of thing?
Yes, randomized, double-blind, two-arm clinical trials. Two-arm means half the participants get the drug, the other half get a placebo. There's nothing actually wrong with that protocol, statistically. It's easy to understand. It's easy to explain. But it's vastly inefficient. There's a topic in statistics called, believe it or not, bandit statistics, named fancifully after the "one-armed bandit" slot machines in Las Vegas. This subject explores the question, how should you integrate incoming data, each pull of the arm of the slot machine, into your decision about whether to make a bet or not? Clinical trials today don't do that. They collect all the data until they reach an end point, and then they evaluate the endpoint statistically. My paper in PNAS lays out an efficient statistical way to use the results of individual cases as they occur in a clinical trial to reach a conclusion about the efficacy of a therapy much sooner and much more efficiently.
I could wave my hands and say, "Many lives would be saved if the world adopted my way of doing clinical trials." Unfortunately, life is complicated. My paper is somewhat abstract, so more work would be needed to put into clinical practice. But here's the real problem: I talked to people who work professionally on statistical methods for clinical trials, and who try to bring about much more modest reforms in the way they are done. They convinced me that anything as different as my proposal was hopeless. The field was so conservative, so regulated by well-meaning traditionalists, that it would never be possible to make a leap of the kind that my paper proposed. Interestingly, when I gave a talk on this in Israel, I got a much warmer reception, because it's a small country, and perhaps thus more nimble. But even so, nothing came of it.
I'll give another example of a paper by me, also published in PNAS. This one might actually have had an effect, but it is hard to tell. In the early 2010s, TSA, the people who inspect you at airports, were very much into risk profiling. They would pull people out of the line for secondary screening based on a set of government-defined secret characteristics, which, to external observers, seemed very highly correlated with: how dark was your skin, and did you look like an Arab? It really amounted to racial or ethnic profiling of a bad kind, although TSA always denied this. They claimed to have a set of objective criteria that had been shown statistically to indicate that you're more likely to be a terrorist. Good luck with that!
I thought that this deserved some mathematical analysis. My paper showed that if you had such a set of criteria, even if they were statistically valid, and if you applied them even the smallest bit unevenly or inexactly, or if the criteria themselves were the slightest bit wrong statistically, then you would actually do better just to pick people for secondary screening randomly. The reason is roughly that everybody you pick out for secondary screening lets some corresponding number of people pass through without being screened, because the secondary screening is a scarce resource. You can only do a certain number of them per hour at any airport. So, you could do the math (as people say) and discover that random is almost always better than profiling.
My paper generated some interest, and I wrote a couple of popular columns about it in various places. It got a little bit of press. A couple of years later, TSA embraced random selection for secondary screening. Was that because someone, somewhere, who was in a policy position, read my paper? Or was it the general backlash against what was obviously racial profiling? You never know how these things come about.
Treasurer of National Academy of Sciences
Let's talk about the National Academy, your work as the elected Treasurer. Was this a way for you to sort of stay in Washington or part of Washington circles?
It's a part-time job, so it doesn't really take me out of Texas except for a handful of meetings a year. As we discussed, I was involved with the National Academy over the years in various ways. The minor embarrassment of my father being a former president of the Academy faded after he retired, and decades went by. But I became Treasurer of the NAS really by accident. Jerry Ostriker was stepping down as treasurer, and I was on the Nominating Committee that selected candidates for his replacement. The rules were that the Nominating Committee had to find two candidates to run in a contested election. We had found one, a highly qualified person, a much-honored academic researcher, former CEO of a significant corporation. He was just the perfect Treasurer. But we couldn't find a second candidate willing to run.
It was afternoon, and everybody was looking at their watches to leave for their planes out of Reagan National Airport. I finally said, "I'm going to step out of the room, and then you can decide whether you want me to be the second nominee. It doesn't matter to me one way or the other, because I'm certain to lose." But, unexpectedly, the membership of the National Academy elected me Treasurer. At that point, I decided that, regardless of how it had come about, I had to do a good job, so I set about learning mathematical finance and investment theory, starting with Bill Sharpe's famous book (I had once gotten to know him at a conference), and the book Active Portfolio Management by my former student Ron Kahn, the one who became a Wall Street rocket scientist. I'm now in my second four-year term as Treasurer.
It's a lot of fun, but also a serious job. NAS has a full-time Chief Financial Officer who, with a staff, thankfully, does the day-to-day stuff; but it falls to the Treasurer and Council (of which I'm a member) to make strategic decisions. The Academy has an endowment of about a billion dollars. Our Investments Committee includes a Nobelist in economics, several Academy members with investment bank experience, and Jim Simons, the billionaire who (many would say) invented the hedge fund. Jim, who was elected to the Academy for his work on Chern-Simons theory in pure mathematics, is an engaging person and a great philanthropist. Thanks to my getting to know him in this connection, I've become a board member of the Simons Foundation. That involves travel to New York City as well as Washington. Pre-COVID, I felt like a regular on the Acela train. Jim and Marilyn Simons are doing great things, and I like being involved.
Bill, at this point, let's bring things up to the present. We're in the middle of the COVID-19 crisis. What does your day-to-day look like? I know we're all on Zoom, everything is over video conference these days. But substantively, what are you working on, and who are you working with?
Well, right now, I'm physically not in Austin, but in Los Alamos, where we still have our house from when we lived here full-time. We've been here waiting out the big spike in coronavirus cases in Texas. We may drive back to Austin in a couple of weeks.
There's been a group of us former members of the Obama PCAST, that subgroup who most participated in writing the health-related reports, and we've written a series of reports on the pandemic with the kind of advice that we would've given the White House if we were still in office. We've done six reports so far, on things like the national stockpile, testing, data, and modeling. Just as when we were in PCAST, we've tried to give concrete, actionable recommendations. The group includes John Holdren, Eric Lander, Harold Varmus, and others, it's a distinguished bunch. We have no official standing. We post our reports on the web and send them directly to people in the Trump administration, in the Biden campaign, journalists, Congressional staff, "opinion leaders", and so on. We've had some good uptake.
I've been able to get finished several papers that I had been working on. I volunteered to write a 5,000-word biographical memoir on Freeman Dyson for the series that the National Academy of Science publishes, memoirs of its deceased members. Ann Finkbeiner, the science journalist who wrote a book on JASON, is my collaborator on that.
Well, one of the remarkable things about COVID is that scientists from all kinds of backgrounds are throwing their expertise at this. A big theme of our talk has been your willingness to take yourself out of the comfort zone that you self-consciously were so aware of where you were as this wunderkind at Caltech, for example. And so at these decision-points in your life, leaving Harvard, going to Los Alamos, entering in biology, how self-conscious were you that you wanted to pull yourself out of that very comfortable track that you might have seen yourself on as you were embarking on this career in physics from Caltech?
Reflection on Cross-Disciplinary Science
Well, yes. I love to do different things. I'm a scientific boulevardier strolling down the incredibly interesting avenues of science of technology, of policy, and so on. What I wrestle with is the fine line between being broad and inter- or cross- or trans-disciplinary (good) and being a dilettante (bad). I pose that question to my social conscience: Am I actually contributing? If I'm cross-fertilizing different fields, if I'm injecting ideas into these scientific establishments as I stroll by them on the boulevard—I really love this metaphor—then I feel good. But I don't want to fool myself. I've known promising scientists whose ability to be broad turned them into dilettantes, flitting from problem to problem without digging in enough to make a contribution that anyone would recognize. In a few cases, such people find other productive niches, science popularization or journalism, for example.
Science is fun, and it is full of really interesting, smart people. The chance to interact with these people across a wide range of fields is just something I've never been able to turn down. It might be a shameful thing for me to admit, but that's been more motivating to me than any desire to stay laser-focused on pursuing a single deep discovery.
To the extent that you do separate out your achievements and your contributions, in the policy realm, within the scientific realm, and many different disciplines, what are those contributions that have brought you the most personal satisfaction?
My satisfaction is more from the process, I think. From seeing a problem, maybe a tiny, little problem not recognized by its established field, and being able to dive into that problem, and if it's amenable in a mathematical or computational sense, solve it, and make a contribution. Sure, recognition is always fun. I would never say that I don't love it when I do a piece of work that gets recognized. But it's not the recognition that drives me to go on and look for the next little problem. It's the fun of doing it, especially when that fun is not just working alone, but, as I said, interacting with smart, creative people.
Going all the way back to when I was a graduate student, and we would eat lunch with Dick Feynman, the conversational topic one day was: "If you were going to be shipwrecked on a desert island all alone, what physics books would you want with you?" Some people thought it would be Morse and Feshbach's two-volume set on mathematical physics; or, of course, the Feynman lectures in physics; and so on. I shocked everybody by saying that, if I was shipwrecked on a desert island, I wouldn't do physics at all. There'd be no one interesting to talk to. There'd be no one to share ideas with. Feynman scowled at me. He was absolutely clear that if he had the means to do theoretical physics on a desert island, that's how he would occupy his time.
Bill, your emphasis on both process and the pleasure of interacting with people, I think that's a great summation for the ease with which you slid into policy and advisory work coming from a science background. That's probably the key to success in combining those fields.
I'm curious if you think that as science becomes less stove-piped, if there is a greater focus on multi-disciplinary or interdisciplinary, your approach might be a vanguard, a way for people to stop thinking of themselves particularly within the bounds of academic disciplines and to start thinking about where they find their interests and where they might pursue them, and limitations about what a textbook is named be damned.
People talk about interdisciplinary science. That's been a buzzword for a generation. Some so-called interdisciplinary science is not actually interdisciplinary—it's just in a field that hasn't gotten its own name and stove-pipe yet. But some interdisciplinary science actually is, in that it requires a collaboration of people who individually have deep subject area knowledge in fields to come together, and more than one such deep subject area is required to get to the solution. In those cases, yes, there's a role for people, for us less-deep people, as the glue, helping to give structure and direction to the interdisciplinary effort.
How should you train new generations of such glue-people? I've certainly puzzled over that. Prospective students come to me and say, "You've worked in all these fields, and you're so interdisciplinary, and I want to be like you. I don't want to work on a narrow thesis problem. I want to work on everything, like you."
But that's not how you approached it. That's a retrospective.
Exactly. I say to those students, in a kindly way, "Please don't let the door hit you in the rear as you leave." If they're a bit more serious, I try to talk them out of their ambition. I don't think my career is a good template for anything. I think it's been a lucky journey, but not a template.
It's a unique career.
But if they still insist, my advice is: You have to start out by mastering a deep discipline. You have to do that for a while. If you haven't ever done that, you'll never understand the importance of depth in other disciplines, ones you don't know anything about. You'll be someone who believes that with an afternoon of reading, you can master any field—and that's just never true. Science advances because there are deeply focused people.
Okay, so you master an area of science. Only then do you look around and say, "Do I have either the courage or the foolishness to drop this and dig into something else different?" And if you do, try it. And if you succeed in it once, try it every five years. But not every three months. Malcolm Gladwell took a lot of flak for putting forth the glib idea that it takes 10,000 hours of "deliberate practice" to master a field. But I think he's approximately right, that's five years of full-time work. So, spend five years in a field, then try to move. And if you do it once, you'll do it again. If you don't succeed, you can go back to your old field.
Last question's a simple one. Based on everything we've said up to this point, what's next for you? What else do you want to accomplish?
Great question. I like being in an interdisciplinary institute at UT Austin, and I like collaborating with biologists right across the street in the bio labs. So one possible path is more of that. Maybe take on additional graduate students, maybe not. My gig as NAS Treasurer still has a ways to run.
I doubt that I'll go back into government advising with the same intensity that I did for the Obama PCAST. That was once in a lifetime. If not advising the government, then advising the advisors, some of whom are younger proteges of mine. I'm also enjoying my role as a board member of the Simons Foundation. These days, it's worth thinking hard about the role of philanthropy in support of science. We are in an era of huge wealth creation, with great fortunes made by tech entrepreneurs who are interested in science. What is philanthropy's unique role? How can it partner effectively with government, with the private sector, and so on? What can it do that would otherwise be left undone? Okay, but the real answer to your question is a constant during my career: Something interesting will come along and, hopefully, I'll have the ability to grab it.
Well, I hope so, too. Let's leave it at that.