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.
Photo courtesy of Joel Primack
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 Joel Primack by David Zierler on July 15, 2021,
Niels Bohr Library & Archives, American Institute of Physics,
College Park, MD USA,
www.aip.org/history-programs/niels-bohr-library/oral-histories/47296
For multiple citations, "AIP" is the preferred abbreviation for the location.
Interview with Joel Primack, Distinguished Professor of Physics Emeritus at the University of California, Santa Cruz. Primack discusses what he has been able to do in his free time since his retirement, including writing papers, giving lectures, hosting meetings at UC Santa Cruz, leading international collaborations, and supervising research. He sees the new data coming from the Vera Rubin Observatory and the Gaia Survey as exciting developments in the realm of astrophysics, and he is looking forward to adding to this data when we begin receiving images from the James Webb Space Telescope. Primack discusses his work with various simulations that he has utilized to understand what may be occurring within galaxies, and the growing importance of astrobiology in these simulations. He takes us back into his early years in Montana, where his passion for science began to develop, and how his high school education and internships led him to Princeton University for his undergraduate career. While at Princeton, Primack took classes from John Wheeler, worked at the Jet Propulsion Lab under Bill Pickering, and participated in the Students for a Democratic Society, where his interest in the combination of politics and science began to grow. Primack discusses how important the communication between politicians and scientists is, and he saw this need for improved communication early on. He started the Congressional Science and Technology Fellowship program as a preliminary way to work on the relationship between government and science. He then recounts his experiences at Harvard University and his eventual move to Santa Cruz, where he continued working on dark matter and dark energy, among other things. He remarks on his relationship and work with Nancy Abrams, including the courses they taught and the books they wrote together. He ends the interview talking about his family, his recovery from cancer, and the people he’s looking forward to working with in the future.
Ok, this is David Zierler, Oral Historian for the American Institute of Physics. It is July 15, 2021. I am delighted to be here with Professor Joel Primack. Joel, it's great to see you. Thank you so much for joining me today.
Well, thank you for inviting me.
To start, would you please tell me your title and institutional affiliation?
I'm a Distinguished Professor of Physics Emeritus at the University of California, Santa Cruz.
What does the title distinguished connote at UC Santa Cruz, and when were you named to that honor?
It was in 2007. The University of California has a micro evaluation system for faculty. There are six or eight steps of assistant professor and associate professor, and nine steps of full professor. At the lower levels, faculty evaluations are every two years. And the reviews, even at the senior professor level, are every three years or so. I don't know if people realize that this is how the University of California works. I think it's one of the reasons for UC’s success. The faculty at all the UC campuses are evaluated the same way. To get above step six, one must have a career that is considered internationally substantial, with letters of recommendation from people all over the world. There are no steps after nine, but one is still eligible for salary increases after positive evaluations. At UC the “distinguished” title is reserved for professors above step nine.
When did you go emeritus?
I retired from classroom teaching in 2014 and became an Emeritus Professor.
And in what ways have you remained connected to the department and Santa Cruz more generally since your retirement?
Well, it hardly changed anything, except that I no longer teach courses, I don’t serve on committees, and I don't go to faculty meetings unless I have some special reason to.
Retirement feels like a permanent sabbatical! I get to do what I love—research—and avoid most of what I don’t like. I’ve continued to write papers and give lectures, both for specialists and for the public. I’ve also recently served as chair of the American Physical Society’s Forum on Physics and Society, and as president of Sigma Xi, the scientific research honor society.
For many years, I've hosted a summer meeting at UCSC on galaxies, which started out as a meeting of my main collaborators and former graduate students. Then they started to bring their graduate students, their post-docs, and many other collaborators. Eventually this meeting became one of the main international conferences on galaxies, with a special emphasis on observation as well as theory. Unfortunately, we couldn’t hold the Santa Cruz Galaxy Workshop in summers of 2020 or 2021 because of COVID. I hope that we're going to be able to do it again next year.
I've continued to supervise research by graduate and undergraduate students. We also have a program at UC Santa Cruz called the Science Internship Program, SIP, that my astronomy colleague, Raja Guha Thakurta, started 11 years ago. In summer 2020, despite COVID forcing us to do everything remotely, we've had over 160 bright high school kids, including two working with my group, supervised by my grad student Clayton Strawn. In summer 2021, there are three SIP students working with us, accomplishing quite a lot despite having to work remotely.
In addition, I've continued to help lead international collaborations. We still have work to do on CANDELS, the biggest project in the history of the Hubble Space Telescope, and we've made major discoveries just recently. I'm also helping to lead the AGORA project, Assembling Galaxies of Resolved Anatomy, comparing simulations of galaxies starting from the same initial conditions by most of the leading codes. The simulations are calibrated at several steps to be as similar physically as possible. This is the only such study that is doing that. I’ve also worked on several new projects. So, I've continued to be quite active.
You've retained all the fun stuff is what it sounds like.
Exactly. The other thing I've continued to do is to write grant proposals, only some of which are successful. My success rate is definitely declining. I don't know if that's because I write in an old-fashioned way, or there's some ageism involved, or what. But I think the proposals are great. And the NASA Astrophysics Theory Program proposal I just submitted on July 1 is one of the best I've ever done. I'm hoping that one will be successful—my grad students need the support. I stopped asking for funding for myself years ago.
More broadly, in the world of observation, what are you most excited about right now?
First of all, I'll say that my general approach has been to go where the data is. To look ahead. For example, I was mainly a particle theorist from my graduate research in '66 to '70 and for the next decade or so, as a junior fellow at Harvard, '70 to '73, and then when I first came to UC Santa Cruz in 1973. But particle physics became boring by the mid-70s. There are only a few new things we've learned since then in particle physics, and mostly not from accelerators, except for the Higgs, but from neutrinos, which are mostly coming from space.
So, there was really a lack of new data to lead to new discoveries in particle physics. And at the same time, astrophysics became extremely exciting. Fundamental questions, the opportunity to propose fundamental answers, and huge amounts of data. And the amount of data that we're going to be getting in astrophysics in the next half-decade is going to dwarf anything we've ever seen before. The telescope that we're now calling the Vera Rubin Observatory, which is being finished in the Atacama Plateau in Chile, will be the largest-field large telescope we've ever had. It's an 8.4-meter telescope with a field with a diameter of more than 3 degrees. And it takes only half a minute to take images, and then it slews to the next 3 degrees and so on. It covers the entire southern sky every three nights in each waveband. All the telescopes in the world find a few thousand supernovas per year. The Legacy Survey of Space and Time (LSST) at the Vera Rubin Observatory will find about 10,000 new supernovas every night. It will have a database of something like 20 billion galaxies. These are staggering figures.
It also begs the question if we have the computational power to analyze all of this data.
The data is being sent by undersea cable to NCSA, the NSF Supercomputer Center at the University of Illinois at Urbana-Champaign, and as I recall there are copies sent to the supercomputer centers at Lawrence Berkeley Laboratory and Brookhaven National Laboratory. The 20 terabytes of data per night must be processed immediately because what they're expecting is to see a million transients every night, some of which will need rapid follow-up. And of course, all the big supercomputer centers go down occasionally. So that's why they have redundancy at other supercomputer centers.
The other area in astrophysics where we're getting a huge amount of new data is stars and planets in our own galaxy, including from the Gaia Survey. Gaia is this amazing spacecraft put up by the European Space Agency in 2013 that is measuring very accurate positions and velocities for two billion stars in the Milky Way. And starting with the second data release, you could see the sideways motion of the stars in the sky. For the bright stars, the accuracy is ten micro arc seconds. That's the size of a marble at the distance of the moon. Another new space observatory, eROSITA, is doing an all-sky survey of active galactic nuclei and x-rays from clusters of galaxies. It'll create the first catalogue of basically all the clusters of galaxies out to around red shift 1. That's about 100,000 clusters. We are looking forward to having crucial new data from James Webb Space Telescope launched later this year, the European Euclid satellite next year, and Nancy Roman Space Telescope a few years later. All of this new data is going to be absolutely fascinating. We'll learn huge amounts of new things.
Of course, all of this is happening in the world of classical computing. Do you see, at some future point, an obvious role for quantum computing in all of this?
I'm not qualified to say. I know some of the basics of the algorithms that will allow quantum computers to factor large numbers, and therefore, make current cybersecurity unworkable. But as to whether it'll allow us to process huge amounts of data that we're perfectly capable of processing with artificial intelligence and big machines today, I don't know. I wouldn't be surprised because, of course, it is inherently parallel computing. But I think the big challenge now is to use artificial intelligence. The amount of data that we need to deal with, and the need to discover new things that we aren't even sure what they're going to look like—these are perfect questions for artificial intelligence, deep learning, and similar things. My collaborators and I have been writing papers on that.
How does theory in astrophysics and cosmology keep up with all of this data? How are we going to make sure that the theorists are not overwhelmed with issues to consider?
Well, as usual, it's a combination of some fundamental theory, which is basically analytic calculations, plus a lot of computation including simulations of cosmic large- scale structure, and galaxies, stars and planets. And then careful comparison, typically now using artificial intelligence, between the calculations and the simulations on the one hand and the observations on the other. One of the things that my collaborators and I have been pioneering is an approach where we use the simulations to train the deep learning codes to recognize certain features. The advantage of the simulations is that we know exactly what's going on in them. We can take them apart, and we can see why they're doing various things.
So, we can tag the galaxies at individual time steps that we save from the simulations as having certain properties, that certain things have happened. And then, we can simply tell the deep learning code, "Ok, this is what we think this galaxy would look like as it would be seen by Hubble Space Telescope or James Webb Space Telescope." We make it the same resolution, same signal to noise, and so forth, in the same wavebands. Then we tell the machine, "Ok, learn how to recognize this feature of galaxies, and this other feature of galaxies, and this third feature of galaxies." And then, we test it by, of course, holding some examples back and seeing if it can correctly identify them, and it frequently succeeds.
And then, we apply the same code to the actual data, the images from Hubble Space Telescope or future telescopes. The danger is that we'll fool ourselves because the simulation may not be realistic. And the answer to that is to use more simulations. And do the kinds of tests that I told you about, this AGORA program, where we set up simulations by the main codes of exactly the same galaxy, same initial conditions, and as much as possible, the same astrophysics assumed and see how the codes differ as they simulate these galaxies. And incidentally, we're finding some things not so different, some other things extremely different. So, we have to be very careful about the ones that are extremely different. If the simulations are different, we can't really trust them unless we understand why that’s happening.
How prominent is astrobiology in all of these endeavors?
It's becoming more and more important. Of course, astrobiology is a subject where we only have one example, namely life on Earth, and we're desperate to have more. We don't know whether the kind of life we have on Earth, which of course all shares the same genetic code, is the only kind or just one of many. Here at UC Santa Cruz, we recently hired Natalie Batalha. She got her PhD at Santa Cruz, and later she went off to NASA Ames Research Center and led the Kepler program. And in joining the faculty at Santa Cruz, Natalie’s ambition is to create a major international center in astrobiology. I've been going to all the organizational meetings. And at one of the first meetings, one of my colleagues said, "You're a cosmologist. What are you doing here? Do you have anything to contribute?"
So, I said, "Yeah, I think there's something interesting." We had just discovered in 2017, because of the signal at LIGO and Virgo, the Italian LIGO-like detector of gravity waves, the gravity wave signal from a merging neutron star pair. Observations showed that this neutron-star merger had created a huge quantity of heavy elements by the so-called r-process or rapid process—elements like silver, gold, europium, thorium, and uranium. A total of about 5% of the mass of the sun in r-process elements was created in that one event. Something like ten times the mass of Earth just in gold. Now, whenever I see people wearing silver and gold jewelry, I think, "Ah, a neutron star merger made that."
The point I made is that these neutron-star mergers are extremely rare events, a few per million years in a galaxy the size of our Milky Way. Which means that you expect there won't be a very complete mixing of the output from these neutron star mergers, so that the amount of these radioactive elements should differ a lot from star to star and planetary system to planetary system, which turns out to be true. A factor of two up or down is quite common. Factor of three, we see fairly frequently. That's a big deal, because I knew that roughly half the heat of Earth is generated by radioactive decay. If you double or halve that, that could have a huge impact.
My UCSC colleague, Earth and Planetary Sciences Professor Francis Nimmo, who was at this meeting, said, "Oh, I can calculate that," because he'd worked out years ago the effects of radioactive heating on Earth. Francis reran his Earth models, changing the amount of radioactive heating. And if you double or triple it, you kill the dynamo that generates Earth’s magnetic field for at least hundreds of millions of years. And there's evidence that that didn't happen in the history of Earth. If instead you lower it significantly, you'll probably kill plate tectonics. And many astrobiologists think that both the magnetic field, protecting the atmosphere and the surface from radiation, and also plate tectonics, leading to recycling of things like carbon, are probably essential for the evolution of complex life.
We found that you can tell how much of these heavy elements there are by measuring the amount of europium in the star. Europium is an r-process element that’s fairly easy to measure in stellar spectra, and it’s been shown to track very well with thorium and uranium. So, measuring europium in stellar spectra will be an important clue to which planetary systems to look at for signatures of life. We published a paper about this in ApJ Letters November 2020, led by Nimmo.
We just submitted a NASA proposal to do much more detailed studies. Our paper considered an Earth-sized planet with basically the composition of Earth, except we just changed the amount of radioactive thorium and uranium, the two long-lived radioisotopes responsible for most of the radiogenic heat after the first billion years or so. But most planetary systems have “super earths,” planets with several times the mass of earth. Simple scaling tells us that if you have the same abundances of these radioactive elements as on Earth, those planets will be hellish lava worlds. Because the amount of heat goes as the volume, radius cubed, whereas the loss of heat goes as the surface area, radius squared. So, overall, it scales with the radius. And so, for the same composition with twice the radius, and you get a huge amount of heat.
For all of these new observational projects, are you more optimistic about fundamental advances in dark energy or dark matter?
I've no idea. Both of them are complete mysteries. I was, as much as anyone, the originator of the idea that dark matter is the lightest supersymmetric partner particle, which still seems like a good idea, since supersymmetry is still the best idea we have to go beyond the standard model. But as you no doubt know, we have no real evidence for supersymmetry. Alternatively, the dark matter could be axions, that's another candidate that has other reasons in particle physics to exist—although the axion could exist and not constitute a significant amount of the dark matter.
I'll be provocative. If the SSC had been built, and we saw supersymmetry, would we have dark matter solved at this point, if you had to put it all together?
Very likely. In supersymmetry, there's a natural candidate for the dark matter, which is the lightest supersymmetric partner particle. That's what we pointed out, Heinz Pagels and I, in 1982. The reason is there's this R-parity in supersymmetry. All the superpartners are R-negative, and R is a multiplicatively conserved quantum number. So, that means that if two of the superpartner particles come together, that's -1 squared. So, the pair can convert into ordinary particles—that's annihilation of dark matter with dark matter. But an individual dark matter particle, if it's the lightest superpartner, can't decay.
So, that gives you a natural explanation for why these things could be massive but not be able to decay. There are also many other possible dark matter candidates in supersymmetry depending on exactly how the supersymmetry breaks. I think if we had built the Superconducting Supercollider, we'd have known the answer a long time ago. It would've been 20 on 20 TeV. That was the plan. Instead of the measly little 6.5 on 6.5 TeV at the LHC. And if supersymmetry's right, we have strong reasons to think that that would've been enough energy to make supersymmetric particles.
The LHC is going to be turning on again with much higher intensity. And it's possible that we'll see evidence for supersymmetry or other new physics by just looking for rare events that were not previously discovered. They're also going to have cleverer schemes for detecting unusual events. Of course, the disappointment in the LHC is that we've really not discovered anything important since the confirmation of the Higgs. So, we're all hoping to get some interesting new physics out of the LHC, especially physics beyond the Standard Model.
Another big mystery question as it pertains to these new observational projects, what about the prospects of integrating gravity into the rest of the forces?
Well, the best idea is string theory, which, of course, is based on supersymmetry. I don't know of any other thing that looks like it would work. It's not something I've personally worked on, so I don't have any strong views on that. Of course, it's also possible that we'll get a clue as to how gravity works as we track down the dark energy. The dark energy question is, first of all, have we seen any evidence for change in the past? We have 5 to 10% accurate measurements of dark energy at low red shift in the nearby universe. As we get out to red shift 1, it's more like 50% uncertain, and as you go beyond that, it's even more uncertain. The reason is not just that we haven't done enough careful observational studies. It's that the dark energy, if it's a cosmological constant, is just something that's a property of space time. And as you go back in time, you get higher and higher density of matter, dark matter, and ordinary matter. And therefore, the relative contribution of dark energy is much less, so it becomes harder and harder to measure. But there are these massive programs underway, the Dark Energy Survey, which is concluding, Dark Energy Spectroscopic Instrument, DESI, which is now very much in progress, the European Euclid Satellite, James Webb Space Telescope, various others that are going to give us some major clues on that. And the other thing that we want to do is look in the nearby universe in different directions, in different regions to see if there's any evidence that the dark energy is not just a constant property of space time, which would be a cosmological constant. When we run big simulations, we just assume it's a cosmological constant. But it might not be. And if not, then that's going to tell us something about how gravity works as well. Because of course, the cosmological constant or anything like it is part of the gravitational equations.
A nomenclature question that has a historical bent to it, over the course of your career, there's astronomy, there's astrophysics, there's cosmology. Have you seen the boundaries of these fields shift over time? And where are they today?
I'll tell you a little story about that. I got my PhD at SLAC, and my advisor was Sid Drell, who was the deputy director of SLAC and head of the theory program. I finished in 1970. For several years after my PhD, I visited SLAC every summer, and when I came to UC Santa Cruz, I used to visit SLAC frequently. During the early 1970s, they had a significant program in theoretical astrophysics as part of their particle theory program. And several of the top people in the field got their PhDs or were post-docs at SLAC during that period. Then, the leadership of the SLAC theory group decided that they didn't want to have any more astrophysics, they wanted to concentrate on particle theory. Maybe budgetary considerations or whatever. And so, basically, it stopped being an interesting place for astrophysics.
Years later there was a change. Stanford University and SLAC created the Kavli Institute for Particle Astrophysics and Cosmology, KIPAC, which is currently led by my former graduate student, Risa Wechsler. And now, it's one of the major centers. And there have been some people from the theoretical physics program at SLAC and some of the experimentalists who moved over and spend at least part of their time now thinking about astrophysics.
At UC Santa Cruz we have SCIPP, the Santa Cruz Institute for Particle Physics, which was for ten years, until last December, led by Steve Ritz. Steve started as a particle experimentalist. But he was the chief scientist for GLAST, Gamma Ray Large Area Space Telescope, which was renamed Fermi Gamma Ray Space Telescope after it was successfully launched and went into operation. Steve continued as chief scientist for Fermi Gamma Ray Space Telescope for the first year after launch, until he became the director of SCIPP.
And now, Steve Ritz, for the last several years, has been the chief scientist for the camera for the Vera Rubin Observatory, the world's biggest camera. A room-sized instrument, 3.2 billion pixels. That's what makes it possible to take these more than three-degree images.
Many people associated with the Fermi Gamma Ray Space Telescope are in SCIPP. Also, people who are helping to lead the ground-based very high energy gamma ray project, VERITAS. These are basically people with particle physics backgrounds who are using particle physics techniques for astrophysics. And I think that's normal now. Because astrophysics is full of these interesting questions and the ability to get huge amounts of relevant data. And as I said, particle physics doesn't have those great virtues.
Now that we have talked about all kinds of issues in the present, the last one I'd like to ask before we go back and develop your personal narrative is one that we've all been dealing with, working remotely during the pandemic. What's been good about it, what's been bad about it, and what are you looking forward to as we hopefully are navigating a post-COVID future?
Well, what's been good about it is that I've certainly spent more time talking with people who are very far away than I did before. And of course, the other thing that we do is screen shares so that we can show each other calculations or results of studies of various kinds. So, in a normal day, I now have three or four or even five Zooms. It's rare that I don't have at least a few. And some people don't like that. But in my case, I think we've benefitted a lot. For example, my group meeting used to be held in a small conference room every week on campus. Now, collaborators are logging in from all over the world. And that makes it, I think, a lot more exciting.
So, that's the good. The bad? Well, I mentioned that I've hosted for many years the Santa Cruz Galaxy Workshop, usually the first week of August. And we had hoped we'd be able to do it this summer, but at least a third of the people come from outside the United States, and hardly anybody's able to get into the United States now. Or if they do, they have to quarantine. It's a big mess. Even getting visas is difficult.
And so, we've missed having it now for two years. One of the things we always did is to leave lots of time for individual discussions. We'd have a long coffee break in the morning, a long lunch break, a long coffee break in the afternoon, and also discussion periods after each group of talks. And those are extremely important opportunities to bump into people that you hadn't talked to before who will give you new ideas, become new collaborators, and that isn't happening nearly as much. Breakout rooms are not the same thing. So, that's the bad.
Well, let's go all the way back to the beginning. Let's start, first, with your parents. Tell me about them.
My father managed shoe stores. He got a big boost when I was 6 years old in 1951, and he became the district manager for Carl's Shoes for Montana, Eastern Washington, Idaho, and Wyoming. He opened stores, hired managers, and he used to visit the stores. And I would sometimes go with him. I grew up from age 6 to 13 in Butte, Montana, and I had an uncle, aunt, and two cousins in Billings, Montana. And the only rabbi in the state of Montana was then in Billings. I studied for my bar mitzvah by getting out of school early on Friday afternoon and taking the train across the state, the beautiful Northern Pacific. I would ride in the vista cruiser and do my homework. And then, I'd study with the rabbi and take the train back on Sunday.
I've been to that shul in Billings. I went to graduate school in Missoula for a period of time, and Billings was the only shul that offered high holiday services.
Exactly, yes. In fact, you're somewhat younger than me, so it may not have been the same rabbi. He was a wonderful man. And it was an exciting childhood because of all that travel. There was one Jewish summer camp in Red Lodge, up high in the hills looking over the northern entrance to Yellowstone Park. I learned to ride horses, usual camp kind of things. I was a cub scout, we did hikes, overnights, things like that. I was sort of a western kid, walking some distance to school, uphill both ways in huge snowdrifts. My mom and dad were not particularly intellectual, but they always encouraged me. We had a wonderful public library in Butte. And I used to check out a whole pile of books every week. First, kid books, but before very long, I got permission to use the adult library and check out biographies, history books, novels. But especially science books.
When did you start to get interested in science at a more advanced level?
Well, I was interested in science as a kid. I had one of those chemistry sets that came with huge numbers of chemicals where you could actually make dangerous things. Those were commonly available when I was a kid. And my dad let me have a space in the basement of his shoe store, where I also had an electric train set that he worked with me on and a photography studio. Also, Butte was, of course, a mining town. It was, then, a main source of copper in the world. One of our neighbors was a mining engineer, and he would give me beautiful rock samples. I would study them with an alcohol burner and things like that. I had lots of fun doing scientific-like things, even as a young kid. Also, Scientific American became its modern incarnation, I think, in '47 or '48, and I read every article in every issue from then until I guess when I left Butte. Then I subscribed to Scientific American, and I've continued to read it. Not every article in every issue anymore, but I try to keep up in many areas of science. So, my interest in science started very early.
Did you have a strong curriculum in math and science in high school?
No, although the chemistry teacher was good. At my high school, Gardena High in Los Angeles, we started using the Chem Study approach, one of two interesting new approaches to teaching chemistry. At the same time, PSSC physics came out, but the physics teacher didn't really know physics. His big achievement was being able to repair televisions, which usually meant just changing the old vacuum tubes. But the chemistry teacher hired me as his assistant, which gave me access to a nice chemistry lab, so I could do lots of experiments. And my job was often to set up demonstrations.
I had a lot of fun with that sort of thing. And as I told you, I'd earlier learned a lot of chemistry. But there were no AP courses. Gardena High was not considered any kind of special high school. Although, one of the things that was interesting about Gardena High was, it didn't have any majority population. It was maybe 30% white, 30% Japanese, 20% Hispanic, and 20% Black.
Between financial considerations, geography, your grades, what kinds of schools did you apply to?
Well, post-Sputnik, after '57, there were all these summer programs and other enrichment activities, including the PSSC and similar programs for high school education. I was admitted to a summer program for bright science and math high school students on the campus at UCLA. We high school kids stayed in the dorms. This was the summer after my junior year, 1961. And that was the first time I ever met other kids who were as smart as I was. We learned finite mathematics using the book by Kemeny, Snell, and Thompson. Group theory, Boolean algebra, statistics. And I found that fascinating, especially the logic of Boolean algebra. I read Turing's paper on the undecidability of the halting problem for Turing machines. I think I more or less understood it. And Alonzo Church's paper. I never did get through the original paper by Gödel on his amazing theorem that mathematics is, at best, incomplete.
This set me up for the summer job I had after high school, before I started at Princeton, which was designing the Air Force's logistics computer. They hired me to program a patchboard to go from Friden paper tape to IBM Hollerith punch cards, which took me about an hour. After that I got harder and harder assignments, and I ended up designing most of the central processing unit for this big Air Force computer. They had a room full of engineers, but they built my design. I was 16.
Now, I can tell you a funny story about college. I got in everywhere I applied except Harvard. And I'm sure the reason I didn't get in there was, there was this terrible interview by somebody who I think was antisemitic or something. Anyway, he absolutely was sure I shouldn't go to Harvard.
I had long thought I'd end up going to Caltech. But when I visited Caltech, I was a little surprised that it's really quite small, only a couple of city blocks. And my mom said, "Well, you should really take Princeton seriously." I had a National Merit Scholarship plus I was offered a Princeton scholarship that gave me a total free ride. My dad said, "Why don't you go out and talk to the admissions officer at Caltech?" So, we did. The admissions officer, a nice man, Dean Winchester Jones, asked me what I was interested in. And I basically said, "Well, I'm very interested in science and math. But I also read history and philosophy and lots of other subjects." And he said, "Well, in that case, you should go to Princeton." And then, he showed me his Princeton class ring. So, I went to Princeton, which I think was a marvelous move.
And was it physics from the beginning? You knew that's what you wanted to pursue?
Physics or math. As a freshman I took an honors introductory math course that was taught by James Milnor, who was then the editor of Annals of Mathematics. Brilliant mathematician. I remember the first examination, we were learning calculus from Courant’s textbook, and the examination was, "State and prove as many of the fundamental theorems of calculus as you can." In an hour or so. Which I could do, no problem.
But the real highlight of my first year at Princeton, 1962-63, was John Wheeler's honors physics course. At Caltech Richard Feynman had taught the first year of his two-year physics course for all the Caltech freshmen in 1961-62. Wheeler was convinced that his former graduate student was making a horrible mistake. He said that Feynman was “casting pearls before swine”—the swine being the mostly engineering undergraduates at Caltech who couldn't possibly understand what Feynman was trying to teach them. Feynman admits as much in the preface to his book.
Wheeler offered his honors course to only the students who tested the highest on a quiz for everybody who wanted to take it who had signed up for physics with calculus at Princeton. There were about 20 people who were allowed to start the course. And I think five of us finished—you could always transfer back to the regular course. All five of us who finished are professional physicists today. It was an incredible course. Wheeler taught in one year almost everything that Feynman taught in two years. Wheeler liked to quote what one of the students in the course told him, that trying to learn physics in this course is like trying to take a sip from a firehose.
But for some of us like me, it was fantastic. Wheeler really was an amazing teacher. He could draw these incredibly complicated diagrams in multiple colors of chalk at the same time he was speaking flawlessly. Every lecture was designed to be just the right length. He was a marvelous lecturer. And he always welcomed questions and never made you feel as if you were asking a dumb question. I remember feeling relieved when I asked a really dumb question that he gave a very smart answer to in one of the first classes. And then, I had numerous other interactions with Wheeler as an undergraduate, and with other faculty.
The following summer, the computer I had worked on in 1962 was already built and I couldn't get another job with that computer company. And so, my dad said, "Well, why don't you ask that nice admissions officer at Caltech if he has any suggestions for an interesting summer job?" So, I did. And Dean Jones said, "Oh, sure. My good friend, Bill Pickering," then the director of JPL, Jet Propulsion Lab, "will be glad to take you on as an intern."
And sure enough, he did. So, for four summers, I worked at Jet Propulsion Lab. That first summer, my first job was to edit the papers from the first fly-by of another planet, the Mariner 2 fly-by of Venus. I edited, for example, James Van Allen's paper on not detecting a magnetic field at Venus. I also designed a three-Tesla electromagnet and was amazed how hard it is to prevent the thing from collapsing on itself. And then, I learned to program in Fortran, and I was asked to program some calculations that some plasma physicists had done. In doing so, I discovered that there was a mathematical mistake in their paper. I became a coauthor of the revised paper, so my first published papers were in plasma physics. So, that was my first summer at JPL. JPL was a wonderful experience, again, thanks to Dean Jones. And then, two years later, on my birthday, I was in the control room as the first pictures came back from the second fly-by of another planet, the Mariner 4 fly-by of Mars.
As an undergraduate at Princeton, was your sense that black holes were something that people took seriously that early on?
No. Wheeler only invented that term somewhat after I graduated. I don't recall that he ever mentioned black holes.
I was fortunate to attend the lectures for the two books on general relativity that were the early masterpieces. One was Misner, Thorne, and Wheeler, Gravitation. I attended all Wheeler’s lectures in my junior year. Kip Thorne took the notes on ditto masters, and then he would hand them out to the students. If I didn't understand something in the lecture, I would read the ditto. If I didn't understand the ditto, I would go and talk to Kip. And if I still didn't understand it, I'd go and talk to Wheeler. So, that was a great experience. And then, in 1970-71as a junior fellow at Harvard, a post-doc, I attended the last series of lectures that Steve Weinberg gave at MIT before his book, Gravitation and Cosmology, came out.
I was probably the only person who ever attended in person both series of lectures. Very different style. Wheeler used modern mathematical techniques, differential forms. (I fortunately had taken a Princeton math course on this material by Norman Steenrod, who had invented some of this mathematics.) And the same equations were written in the usual tensor notation in all of Weinberg’s lectures. So, I got to learn both approaches.
I don't remember any discussion of the physics of black holes in either course. Of course, the standard equations, the standard solution for the gravitational field around a gravitating object, a star or black hole, we certainly worked those out. But the physics of black holes, the fact that they would have accretion disks that would be extremely bright, so there'd be a main source of X-rays and gamma rays, that they might also have jets, these were not things that I recall being discussed in those courses.
As an undergraduate, did you have a broader sense of the advances in particle theory that were happening in the early 1960s?
Oh, absolutely. Val Fitch was at Princeton, and Fitch and Cronin shared the Nobel Prize for the discovery of CP violation. Murph Goldberger and Sam Treiman were theorists in the Physics Department. Bram Pais and others were at the Institute for Advanced Study. I used to go out there, it was a short walk across the golf course, to attend seminars. Murray Gell-Mann had predicted the Omega minus, that there would be such a particle and what its mass would be. The “noble eightfold way” was the approach—that's the group theory associated with SU(3). Anyway, that discovery was made in 1964 while I was an undergraduate. It was very exciting because it looked like we were beginning to understand how the many particles being discovered all fit together. I was very much aware of it. Partly, I think, because a lot of this work was being done by people at Princeton.
By the end of your undergraduate, how well-formed was your identity in terms of the kinds of physics you wanted to pursue in graduate school? Were you still thinking about experimentation at all? Or was it theory all the way?
No, I was never thinking about being an experimentalist. I'm a klutz as far as experimental physics is concerned. When I was a kid in Montana, I got a Heath kit. It was supposed to be a shortwave radio receiver. And when I plugged it in, it blew up. I must've wired it wrong or something. One of my friends in high school, his father was the chief recording engineer for Capitol Records. He had a recording studio at home, and he hired me to solder connections—but he fired me after a few hours because my solders were too sloppy. So, I was always a theorist. Except for my early days of experimental chemistry and biology.
Did you take advantage of that initial plan to be more involved in the humanities at Princeton?
Oh, absolutely. I took courses in history, philosophy, and literature, especially the French enlightenment. I became good friends with existentialist philosopher Walter Kaufmann. I read all his books and even took his course on religion. We became personal friends, and I worked with him on projects, including a show of Immanuel Velikovsky’s sculptures at the Woodrow Wilson Society. At Princeton, when I was an undergraduate, almost every student went to an eating club in their junior and senior year. The eating clubs were rumored to be antisemitic, and I thought I would feel somewhat out of place in them. I was part of the agitation to create an alternative to the eating clubs. And at the beginning of my sophomore year, I was one of the first students in an alternative eating arrangement on campus—what we then called the Woodrow Wilson Society, which later became Woodrow Wilson College, and probably has now been renamed.
I became friendly with the first master of the Woodrow Wilson Society, Julian Jaynes. I got a budget to organize wine and cheese sessions, which faculty came to, to interact informally with students. And I also co-led, with Dan Kleinman, the “Film Cenacle”—an organization to show classic movies on campus. In those days, you rented the movies, they came as big 16mm reels, usually three of them. We used two projectors, and as soon as the first reel was finished, you'd start the second projector, and then you'd rewind the first reel, then put the third reel up on the first projector. We did that every few weeks, and that was the only way you could see a lot of these great classic films by Fellini, De Sica, Bergman and others. I think that counts as humanities.
I also used to go into New York almost every weekend for drama, opera, or dance. It was easy to do—one just took the “dinky” trolly from campus to Princeton Junction, and in an hour you were at Penn Station. The Princeton Music Department then was right across from Palmer Lab, which was the Physics Department building. You could write your name on a piece of paper and drop it in a box in the Music Department's office, and they would choose a few for tickets to operas at the old Met. And if I didn't get that, I would get a cheap ticket, $5 or so, for the fifth balcony, where you were so close to the stage that you got a better view of the prompter than you did of the interior of the stage. [laugh] And of course, these seats had big desks, because that's where students were following the score. But if your name got chosen by the music department, then you got orchestra seats. I still remember some of the operas I saw. In those days, the best way to have dates was to meet girls in New York. Because of course, Princeton was not coed.
What kind of advice did you get, if any, about any graduate programs to apply to or even specific professors who could be an advisor?
What happened was, I did my senior thesis with Gerald Edward Brown. He was one of the physicists who created modern nuclear physics, along with Ben Mottelson and Niels Bohr's son Aage Bohr. For my senior thesis work, I learned the field by reading the papers. No textbook existed. Many of the papers were published in a Danish journal in English. My senior thesis under Gerry Brown was to work out nuclear fission from the point of view of modern nuclear physics. The then-standard theory for nuclear fission was by Bohr and Wheeler, 1939, the so-called liquid drop model. In that picture, the nucleus is treated as an electrically charged fluid with properties like volume energy and surface tension. As the nucleus deforms into a prolate ellipsoidal shape, the protons repel each other; this will make it more and more aspherical, and finally it'll split apart. It's a quantum mechanical barrier penetration problem. You calculate the barrier penetration by integrating pdq, with q the deformation parameter as the nucleus deforms into this prolate ellipsoidal shape. The corresponding momentum p is proportional to the square root of the energy. Bohr and Wheeler had argued that the lowest energy trajectory would dominate. And then, they worked out what the half-life for spontaneous fission would be for isotopes of uranium and other heavy elements, and they got essentially the right answer.
In modern nuclear physics, the nucleons are pictured as inhabiting orbitals, despite the high density of the nucleus. As the nucleus deforms toward nuclear fission and becomes more prolate, the orbitals that fit nicely into a prolate ellipsoidal shape come down in energy, and the orbitals that are pancake-shaped go up in energy. Nucleons are paired with opposite spins, so there's no net spin and the pairs can quickly switch over from one orbital to another. It turns out that the fission barrier is much thinner than in the liquid drop model, but the corresponding momentum p is much higher. And when I worked out the spontaneous fission half-life in my 1966 Phys Rev Letter on this, my first Phys Rev Letter, I got it right.
Prof. John Wheeler was the person who evaluated my senior thesis. Which was critical of one of his most important pieces of work. I was told that my senior thesis defense was the longest on record. It lasted basically the entire day, until I convinced Wheeler that my thesis was basically right.
You were actually teaching him something.
Yes. And so, that was another one of my Wheeler stories. But let me tell you the extra part of this that makes it even more amusing. Wheeler was a right-winger. He and Eugene Wigner actually went to the roof of the Music Department building, which is right across from the Physics building, as I said. Some students had put up a big sign, "END THE WAR," and Wheeler and Wigner crossed out "END" and wrote "WIN." This, of course, was about the Vietnamese War. When it became clear that this thesis defense was going to go on, I said, "Can I take a break in the afternoon for a couple hours?" From 2:00 to 4:00, I think it was. And they said, "Oh, sure. We understand." Actually, Gerry Brown knew what my reason was, and Wheeler didn't. The reason was that I was one of the members of Students for a Democratic Society, and we had arranged to picket President Lyndon Johnson to oppose the Vietnam War. Johnson was coming to Princeton to dedicate the new Minoru Yamasaki building for what was then called the Woodrow Wilson School of Public and International Affairs—more recently “Woodrow Wilson” was dropped. We picketed. Then, I went back and finished convincing Wheeler I was basically right. I must have succeeded, since I was the valedictorian of my class at graduation.
To get back to that question about advice you may have received for graduate school…
Oh, yes. What happened was, while I was working on this project with Gerry Brown, he asked me where I would like to go to graduate school. And I said, "Well, I'm a Californian. I'd love to come back to California." And he said, "Great. I'll call Sid Drell. He's an old friend of mine." And he did. And after Gerry told him whatever he said about me, Sid said, "Ok, I'll be glad to take him on as a graduate student." So, it was all arranged. I do remember that I also applied to Princeton, and they turned me down because they said it's not good for an undergraduate to stay in the same place.
Another Wheeler story. After Gerry Brown’s conversation with Sid Drell and I'd already decided to go to Stanford, Wheeler asked me to come to his office one day. And then, he invited me back into his inner office, his private office, and there was Edward Teller. Teller wanted to convince me to go to UC Davis, and he promised me all kinds of scholarships and so forth to work at Lawrence Livermore Laboratory, where I would no doubt work on nuclear weapons or something. I politely turned him down and said I'd already committed to going to Stanford.
Now, connecting with Sid Drell, that firmly puts you on a path in particle theory. Were you thinking at all about astrophysics or cosmology at that point?
Well, certainly astrophysics. Hardly anybody took cosmology seriously in those days.
What about Jim Peebles? Did you have interaction with him? Or Bob Dicke?
I did. Yes, with both of them. With Bob Dicke, before I decided on a senior thesis with Gerry Brown, I actually went around and interviewed faculty. And Dicke wanted me to work on the sun being not spherical as a possible explanation for the fact that Mercury's orbit precesses, which, of course, general relativity nicely explained. But Dicke had this alternative to general relativity called the Brans-Dicke theory, where there's an extra scalar field. Dicke had also been the director of advanced laboratory required of all Princeton physics majors, and so I already knew him. Of course, he's a great scientist, but I thought that the project he'd proposed was not nearly as interesting as the nuclear physics project.
I went to all the Princeton physics colloquia, and I remember Jim Peebles lecturing about what must have been one of the first N-body simulations, about how a matter overdensity would collapse. Starting in the 1980s my students and I wrote key papers about that.
Gerry Brown told me that I was too smart for nuclear physics. He said, "The future is particle physics, and that's why you should go and work with Sid Drell." What ended up happening was, I wrote some interesting papers as a graduate student, which set me up for working as a postdoc on SU(2)xU(1), the electroweak theory, which became part of what we now call the standard model. The papers I wrote on that, the first ones were with Tom Appelquist, who was a postdoc at SLAC and then became an assistant professor at Harvard. So, when I went to Harvard as a junior fellow of the Society of Fellows, it was natural to work with Tom. And also, with my good friend as a graduate student, Helen Quinn.
I rescued Helen. Her husband, Dan Quinn, had come to Harvard to work on the Cambridge Electron Accelerator. He was an experimentalist. Helen had just had her first daughter, Elizabeth, and she was just a stay-at-home mom. I thought that was crazy. She's brilliant. So, I arranged for her to get a desk at the Harvard Physics Department. Helen, Tom Appelquist, and I wrote several papers. And very quickly, the Harvard physics department realized that they wanted to hire Helen. I also wrote a paper in '72 with Ben Lee and Sam Treiman, which was the first calculation of the mass of the charm quark. And when charm was discovered in November 1974, it turned out that we got it right. These are all new things that one could do with the electroweak theory. You could do finite calculations with weak interactions for the first time. So, that was a great time to be a theorist. You could make real progress.
When you got to Stanford, how much of your time did you spend in the department, and how much time did you spend at SLAC?
Virtually all my time at SLAC. And the reason was partly that the Physics Department had assigned me an office on campus with a chain smoker. I couldn't stand it. It was essentially impossible for me to share that office, so I never used it. I spent all my time out at SLAC. SLAC was also a much more interesting place.
And there was this totally nutty thing going on. I don't know how well-known this is. But there were two proposals that were submitted by people at Stanford for the next accelerator. The SLAC proposal was by Panofsky and Drell. There was an alternative proposal, the Mark 1 or something, by Bloch, Fairbanks, and others, which was not chosen. And when the Panofsky-Drell proposal was chosen, the senior people in the department had said, "Well, of course, we'll be happy to include yours under our rubric," or something like that. And Panofsky and Drell said, "Nothing doing. Your proposal lost. Ours won. We're independent." So, both of them were kicked out of the Physics Department. Sid and Pief had Stanford tenure, but when I was a student of Sid Drell's, I was treated by the Physics Department as if I was at some foreign institution.
Professor Dirk Walecka was my official advisor. Dirk thought that this was a very embarrassing position to be put in. He was as nice about it as could be. But it was crazy. What had happened was that everybody at SLAC was no longer welcome in the Stanford Physics Department, which I think was partly because the Physics Department was worried that they'd never get to hire new faculty if SLAC, which was then growing rapidly, was considered part of the Physics Department. But in any case, it made the relationship between the Physics Department and SLAC very complicated, especially for young people like me. I took the required courses on campus and did very well in them, but I did research at SLAC. My desk was in the seminar room of SLAC’s Central Laboratory building, along with desks of several other grad students. In those days Richard Feynman used to hang out there when he visited his sister, an astrophysicist at NASA Ames Research Center.
What was Sid working on when you first connected with him?
The first project that he asked me to work on was the Drell-Hearn sum-rule. Drell and Hearn had proven this rather fundamental relation, which says that if you consider photon scattering off of protons oriented one way versus the opposite way, you get a certain result. It's an integral equation, and you can predict the exact answer. And there was then a paper by Barton and Dombey, two British nuclear physicists, who claimed that the deuteron and other nuclei disobeyed the Drell-Hearn sum-rule. And Sid said, "It's obviously wrong. Find out what's wrong with it."
It turned out that they had not correctly calculated a composite system. Nuclear physicists, basically, had done it wrong. So, I wrote two fundamental papers with SLAC theorist Stan Brodsky on how you should calculate with a composite system. A key thing you must take into account in this case is that it is bound and has a unique spin. The deuteron has spin-1, and Barton and Dombey hadn't properly taken that into account. That was the first project I worked on. And it was good, I had some fun with that. One of the fun things was that to prove fundamental theorems like the Drell-Hearn sum-rule, you could make a sum-rule by using the Dirac equation. It's very similar to the Heisenberg paper that helped to start quantum mechanics, where he wrote a sum-rule for electron transitions in atoms based on the fundamental equation of quantum mechanics, .[x, px] = iħ. It turns out, you can do the same thing in the Dirac equation. And as far as I knew, no one had ever done that using the Dirac equation. So, that was an example of a fun thing that I learned how to do.
How much did you interact with Drell? Did he have time, given all of his responsibilities, to work with you as closely as you needed?
Not as closely as I would have liked. Sid was spending a day, or two, or three in Washington every week, pretty much the entire time I was his student. And we used to talk about that because I was also very interested in politics. And so, I would keep asking Sid, "How come this stupid Vietnam War is going on? You're meeting with the President frequently, serving on his Science Advisory Committee. Can't you tell him how dumb this war is?" And of course, Sid would gently say that he was working on lots of interesting things, and he wished he could tell me about them, but he couldn't. Sid had a big, booming voice. And so, his office had to have very thick soundproofing, because he would be having these classified conversations on the phone. We weren't supposed to be able to hear what was going on. And only much later did I realize that Sid, among other things, was the chairman of the Strategic Weapons Panel of the President's Science Advisory Committee, and he also was largely responsible for starting the US spy satellite program. But I only learned that at his retirement party.
Partly inspired by Sid Drell’s willingness to spend so much effort trying to help make better use of science and technology, I started something along with Joyce Kobayashi, who was Stanford student body president, and Bob Jaffe, who then went on to become a distinguished physicist at MIT and chairman of the faculty. Bob graduated from Princeton two years after me, and he also was valedictorian. Bob was also one of Sid Drell’s grad students. In 1969-70, my last year as a graduate student, we organized something called Stanford Workshops on Political and Social Issues, SWOPSI. That first year it was ten workshops, which were official courses at Stanford co-taught by graduate students and faculty. And each course was supposed to improve the world. That was our ambition. We wrote a proposal, and the Ford Foundation gave us $40,000, which was pretty big money in those days. So, we were a little independent of the Stanford administration, and the Ford Foundation was also, at that time, supporting a bunch of junior faculty at Stanford, and they suggested that we work with them. And so, a number of those faculty helped us, and also some of the senior faculty. A Stanford professor with the wonderful name of Lincoln Moses, who was the dean of the graduate school, was very sympathetic about what we were trying to do and was very helpful. And a number of these courses actually did change the world. The course that I and Bob Jaffe co-led with Frank von Hippel and Marty Perl–who subsequently won the Nobel Prize for discovering the tau lepton—was on how to get the government to do a better job of managing science and technology. One of the things we did was a questionnaire to members of Congress on various ways to do that. We convinced members of the House and the Senate from California, Jeffery Cohelan and Alan Cranston, to circulate our questionnaire, and they got very good responses from their colleagues in the House and Senate. The most popular idea was what was very shortly afterward created, the Office of Technology Assessment, a science advisory office for the Congress.
And the second most popular idea was to have scientists spend a year as interns on the staffs of members of Congress or Congressional committees. I made it a major goal to create that program in the Physical Society and also in AAAS. So, as a Harvard junior fellow I started the program of Congressional Science and Technology Fellowships, which still continues today and has led to more than 2,000 scientists working for a year on the staff of a senator, representative, or a Congressional committee. It subsequently led to the AAAS program on Science and Technology Policy that placed many more scientists in the Executive Branch.
One of the reasons I was deeply involved in these kinds of things was that for my second and third years as a graduate student, I was one of the two graduate student residents in the first coed dorm at Stanford. And of course, the students who would be in the first coed dorm were somewhat adventurous. So, I was aware of what was happening at Stanford. And part of it was, there were big demonstrations against Stanford's involvement in the Vietnam War. Among other things, that led to classified research being taken off campus and Stanford severing its relationship with SRI.
But my attitude was, in addition to demonstrating and “sitting in” in buildings, Stanford students should be using their heads as well as their bodies. And that was the argument for these new SWOPSI courses. Another one of the new courses was led by Panofsky and graduate students in physics and Russian history, on nuclear security and arms control. About 150 students showed up and wanted to be in this course. Pief, I think, had expected 20. That helped lead to the long-running Stanford course on Arms Control and Disarmament and the creation of the Stanford Center for Nuclear Security and Arms Control (CISAC). Sid Drell was the first co-director of CISAC.
Did you ever think about becoming more involved in policy for a career?
Not really. When I was a junior fellow of the Harvard Society of Fellows from '70 to '73, I worked with Henry Kendall on nuclear reactor safety. I also wrote a book with Frank von Hippel that came out in 1974, Advice and Dissent: Scientists in the Political Arena. One of the chapters was on Henry Kendall, Dan Ford, and nuclear reactor safety. Another thing I did was to start the program of science and technology policy studies by the American Physical Society. The first of those was on nuclear reactor safety, and Freeman Dyson and I wrote the proposal for that (I also have some Freeman Dyson stories.) Several years later, Henry Kendall wanted me to come back, and I was offered a joint position as president of the Union of Concerned Scientists and a faculty member at MIT. And I basically said, "No, I'm really a scientist. I'll do this public policy stuff in my spare time. But I really don't want to make that my career."
What was the topic that you focused on for your thesis research?
It was just a bunch of papers I wrote as a graduate student. I mentioned earlier the papers that I wrote with Stan Brodsky. And the key question in two other papers was, why are the cross sections for elastic scattering of protons off protons so different from what we call deep inelastic scattering? Can we understand the elastic scattering? So, I worked it out with two different models. Basically, a pion exchange model for what gives rise to the nuclear force and a spin-1 particle exchange model, photon-like exchange. And the second one is actually the right one, except it's not the photon, it's the SU(3) gauge bosons–gluons, in other words. But I, at that time, didn't know how to do SU(3), non-Abelian gauge theory. This would've been '68, Yang and Mills had earlier written their paper on gauge theory, but no one had worked these things out yet. 1967 is when Weinberg's first paper on SU(2)xU(1) came out. But it was about leptons, not strong interaction theory. By doing those two papers, I was in a very good position for future work. And my second paper on that was with Appelquist. So, that was a very good starting point for us to do the early calculations on SU(2)xU(1).
It was several years after you left, but looking back, was there anything that foreshadowed the November Revolution during your time at SLAC?
Sure. I was aware that they were measuring the cross section for electron-positron annihilation, and that it was puzzlingly large. And I wrote a paper, just an internal document, with Helen Quinn, where we said that there might very well be resonances, and they should look for them by changing the e+e- energy just a little bit. And that's, of course, exactly what they did when they discovered the J-Psi. But it was really just a lucky accident. It turned out that they'd been measuring around 3 GeV. And 3 GeV seemed a little bit high. So, Roy Schwitters said one weekend, "Let's look around that 3 GeV. Let's change it a little bit." And it turned out that it's just a little bit above that, 3.1 GeV, and you get this enormous resonance.
I asked whether this internal document that Helen and I wrote had had any effect on that, and the answer was that the experimentalists had completely ignored it. And it was just the fact that they were just beginning to see the rise for that peak that made them come back and look at that. And of course, that was the same weekend that Sam Ting was visiting SLAC to serve, I think, on the shootout for some experimental program. And he had been denying that they'd been seeing anything at Brookhaven. But in fact, of course, they had. They just hadn't written the paper yet. And so, because I was known by people at SLAC to be involved in this, I got a phone call. "Come to SLAC on Monday." And I happened to be sitting next to Roy Schwitters in the audience. So, that's when I found out about the J-Psi.
After you defended your PhD thesis, what opportunities were available to you for post-docs? Or was Harvard wrapped up, and that was the main thing you were focusing on?
I think I got the offer from Harvard very early. Sid Drell had wanted to be a junior fellow himself. He very strongly encouraged me to do it. And the great thing about the Society of Fellows was, it was an opportunity to hang out with a bunch of very smart people in very different fields, and I've always been interested in what's beyond physics. So, it seemed perfect. And also, of course, I couldn't help thinking how ironic the one university that turned me down as an undergraduate invited me to come to their most prestigious post-doc. Anyway, I don't think I gave the subject much thought. I think as soon as I found out I was a junior fellow; I just took the opportunity.
And was that really the source of the attraction, being a junior fellow? Or was there a group or person that you really wanted to work with at Harvard?
Well, as I said, Tom Appelquist had just accepted an assistant professorship at Harvard, and he and I were good friends and enjoyed working together. And then, when I went there, I quickly managed to get Helen a position. And I interacted a lot with Shelly Glashow, but it was always the same story. Shelly would be the first one into the department, he'd wander up and down the halls, and then I would sometimes be the second one in. And he would have all these crazy ideas. And the job of anybody he encountered was to shoot down his craziest ideas and see if there was anything useful left. And I, unfortunately, tired of this fairly quickly. But other people who worked with Shelly didn't and benefitted tremendously because, of course, he was an amazing font of new ideas.
One of his popular ideas was charm—that there is fourth quark in addition to up, down, and strange. And I thought that naming it charm was trying to make something ugly look attractive. My intuition was it was wrong. But anyway, as I said, we worked out the mass of the charm quark, Sam Treiman, Ben Lee, and I, in the summer of '72, which I spent at Fermilab. And we showed that it had to be between 1 and 2 GeV. And my thought was, "If it's that low, it surely would've been discovered." I hadn't really appreciated that the phi boson, which is a strange anti-strange meson, was a good example of how you could have a charm anti-charm meson, which is what the J-Psi is, of course. I think we included a possible charm-anticharm meson in this little thing I wrote with Helen Quinn as one of the resonances that they might expect. So, it certainly wasn't a complete surprise to me at all that they discovered the J-Psi and that it had the energy it did.
What kind of interactions did you have with Howard Georgi?
There were two desks in one room and two desks in the adjoining room. Howard was not in the same office as me, but the adjacent office. He and I liked each other a lot, but very quickly, he became the one who shot down Shelly's ideas. And of course, some of those ideas were great, and they ended up writing some very good papers, including a very nice paper that he wrote with Helen.
Given your interest in the humanities going back to Princeton, you must've loved the junior fellow program and the exposure it gave you.
Totally. Well, not only that, but there's another thing that happened. There was always some very prominent physicist who was a senior fellow. My first two years, it was Ed Purcell. As I said earlier, I was very interested in trying to improve the way the US dealt with science and technology issues. And one of those was creating the Congressional Science and Technology Fellowship program. Another was starting the American Physical Society Forum on Physics and Society, and the APS program of summer studies on public policy issues. Ed Purcell liked all of these ideas, and he basically made it possible for me to pursue them. He was the president of the American Physical Society around that time. He introduced me to the relevant people, he got me put on relevant committees, so that was essential in my being able to actually promote these things that I did. So, that was another consequence of being in the Society of Fellows, my Ed Purcell connection.
What was your most significant physics work while you were at Harvard?
These papers on what we now call the Standard Model of Particle Physics.
Was your sense at that time that the standard model was getting closer and closer to completion?
Yes. So, first of all, it looked like gauge theories like SU(2)xU(1) were really going to work. For the first time, this allowed us to do calculations that gave finite answers. And a student who was working with Sidney Coleman, David Politzer, made major new progress. Sidney was on leave at the Institute for Advanced Study during, I think, '72. Politzer went down to the Institute to talk to Coleman about the latest calculation he'd done. He'd talked to me about it, so I knew what he was doing, and it was basically the idea that SU(3) as a non-Abelian gauge theory could lead to confinement. He told me that he had not only talked to Coleman, but he also talked to David Gross, who I guess was at the Institute, or maybe at Princeton, I don't remember. And I remember that David Politzer told me that Sidney Coleman had told him, "You better write this up really fast." Of course, at the same time, Gross and Wilczek wrote up the same thing. And I remember that Politzer told me that when he told Gross about this, Gross didn't say, "Oh, we've already done that." So, that led to the suspicion that Gross and Wilczek had really stolen the idea from Politzer. Of course, the subsequent history is that Gross and Wilczek did all kinds of brilliant things, and Politzer didn't. Anyway, all three of them shared the Nobel Prize for this work.
So, I was very much aware of SU(3) and what the implications were. And in fact, one of the things they'd discovered had already been discovered by Gerard 't Hooft, although the implications were not yet clear. I was at the conference where 't Hooft revealed this. It was the summer of '71, I think, in Cassis, a little coastal town near Marseilles. Tom Appelquist and I were there. We had already encountered 't Hooft earlier in his 1970 visit to Cambridge, where he explained how to renormalize gauge theories. It was at the Cassis meeting where 't Hooft said in his talk that the beta function was positive, which basically meant that you could get confinement. Anyway, yes, I was very much aware of that.
Did you consider a second post-doc? Were you ready for thinking about faculty positions at that point?
The amazing thing is that in '73, when my term at the Society of Fellows ended, I don't remember that I had to apply for anything. I had offers across the country. I had three offers in New York City alone. Columbia, Rockefeller, and NYU. I had two offers in Pittsburgh, Pitt and Carnegie Mellon. And I ignored all of them. I had fallen in love with the San Francisco Bay area, and I wanted to come back. And so, it just came down to Stanford versus Santa Cruz. And Sid Drell said, "You'd be crazy to go to Stanford. They're still nuts." And in fact, the faculty at Stanford who were trying to hire me said, "You better not do any of this politics stuff," like the Stanford Workshops on Political and Social Issues that I'd helped to start. So, I went to UC Santa Cruz. And it turned out it was another one of these incredibly lucky choices because Santa Cruz would increasingly become one of the great centers for astrophysics.
That had happened because basically, the astronomers associated with Lick Observatory all moved to Santa Cruz when the campus started. They had previously been living on Mount Hamilton, near the observatory. So, that was the nucleus of a great astronomy department. Also, the higher-ups had made the decision to have the astronomers and the physicists share the same building, and in fact, our offices were intermixed on the same floor. George Blumenthal's office was two doors down from mine. And George and I immediately became friends. We didn't work together right away, we invested together. For the first time, we were making money as assistant professors, and every week, we would go down to the Santa Cruz public library and study stocks to invest. And we did very well. We doubled our money several times, I think.
For example, we invested in a company called Molycorp, which was a main source of rare earths in the whole world at that time. Molybdenum Corporation of America. And the stock started to shoot up, as we expected it would, because people were finding good uses for these rare earths. And then, Molycorp was bought by a huge company, Amax, so you just got Amax stock. And we immediately sold it because nothing was going to happen with this huge company's stock. But that was an example of smart scientists investing. And then, a few years later, I started to work with George on cosmology because that was his expertise.
What was the initial connection that makes this such an important transition in your research agenda?
Actually, what happened was that toward the end of the 70s, I wrote a series of papers that were basically particle astrophysics. The most ambitious of them was sort of a technical tour de force. When I was a post-doc, two independent teams had figured out how to do quantum field theory at finite temperature, Louise Dolan and Roman Jackiw, and Steve Weinberg. Their methods looked quite different, but they just were different ways of doing the same thing. But no one had yet applied that to a real physical process. I was asked to referee a paper that calculated one Feynman diagram and drew some conclusions, which were completely wrong. You can't get physical results by calculating just one Feynman diagram. You have to calculate all the diagrams, otherwise what you have isn't gauge invariant, and so it doesn't make any sense.
No one had actually done a full calculation. And what we decided to calculate was neutron decay. And of course, you have to calculate, also, the reverse process, because at finite temperature, you have enough energy to go backwards as well as forwards. So, it's a bunch of Feynman diagrams that had to be calculated as a function of temperature, and no one had ever done such complicated calculations. A graduate student, Jean-Luc Cambier, and my then-post-doc Marc Sher and I did the calculation. We had two different people calculate each Feynman diagram, and we checked each other. This is now a permanent part of the standard calculation of what happens in the first few minutes after the Big Bang. Taking this into account slightly changes the rate at which neutrons decay.
Free neutrons, left to their own devices, decay with a half-life of about ten minutes. It's the neutron decay during that first few minutes after the Big Bang that determines how many neutrons are left to become locked up into mostly helium. Almost all the helium forms in the first few minutes—the helium that's made in stars by fusion is only a few percent extra. Many of the neutrons have decayed by the time the temperature is low enough that deuterium survives long enough that it can bind together to make helium nuclei. So, calculating that accurately is really important for astrophysics. At one second after the Big Bang the temperature is about one MeV, or about 10 to the 10 Kelvin. At that temperature we found that the neutron lifetime is, as I recall, about a fifth of what it is at temperature zero. But everything's still in equilibrium then because you get the reverse process just as efficient as the forward process, so what's determining the neutron lifetime is just the mass difference between the neutron and the proton. But as the universe expands and the temperature falls, the neutron lives longer. So, you have to calculate everything.
The thermodynamics of this was being worked out independently at the same time by people at Los Alamos, including Rocky Kolb. So, one of the things I did as part of this project was to visit Los Alamos and talk to the people, tell them what we were doing, find out what they were doing. We did the part of the calculation that was the particle physics part, and then you had to combine that with the thermodynamics part. Then the result was that there is an effect, and it's now part of all the standard codes for Big Bang nucleosynthesis. Anyway, that was an example of one of my early papers, which is particle physics, but applied to astrophysics.
Another particle astrophysics project was the paper I did with Pagels, which we wrote in the summer of '81, published in '82 as a Phys Rev Letter. That's the paper that pointed out that the lightest super partner is a natural candidate for the dark matter. We worked out a particular example, which was the gravitino. And we did it using this technique of sum-rules that I had figured out you could do with the Dirac equation, except this time, the equation that we used was the fundamental equation of supersymmetry. It turns out you can make a sum-rule out of that, so that was also part of the technical challenge of that calculation with Pagels.
When did you first meet Sandy Faber?
When I came to Santa Cruz in 1973. Sandy had come just one year earlier, at the same time as George Blumenthal. Later on, I was really impressed with the review article that Sandy and Jay Gallagher wrote for Annual Review of Astronomy and Astrophysics, ARAA. They basically summarized all of the papers, about 250 papers, on dark matter at that point. And the big question for astronomers is, is dark matter real? Is it really true that most of the mass on large scales is some invisible stuff? Dark, of course, means not luminous. It's a terrible name because people naively think it means it absorbs light, which is not the idea. But Faber and Gallagher wrote this wonderful review in 1979, and it turned out that of the 250 papers on dark matter, something like 200 were not convincing. There were all kinds of problems that they had not properly taken into account.
What were some of the zanier ideas at this early stage about what dark matter might be?
The question they considered was, is the evidence that on scales of galaxies and bigger, there really is something that's invisible that's providing most of the mass? The issue was not what it is. The issue was, observationally, is the evidence strong? And the kind of thing that people did wrong was, they would take the difference in velocity of, say, two galaxies or a few galaxies, and they would say, "Oh, high velocities. That means that there must be a lot more mass that's holding it together." And the problem is, there's backward and forward contamination. There might be another object that isn't part of that physical system, but you're including its velocity, even though it's not part of a bound system. That doesn't mean your calculation is wrong, but it means it's questionable. But of the roughly 250 papers, about 50 were simply not questionable. There was nothing wrong with them. So, their conclusion was that there must be a lot of dark matter.
In the summer of '81, Heinz Pagels and his wife Elaine had a little boy named Mark, who was born with a heart defect. They had just built a house in Aspen. Heinz had been one of the founders of the Aspen Center for Physics. But they couldn't live in Aspen because the little boy couldn't be at that altitude. So, my wife found them a wonderful house near UC Santa Cruz up in Felton, which they used every summer for six summers. And that first summer, Heinz and I worked together. We were old friends—Heinz had been Sid Drell's graduate student after being an undergraduate at Princeton. Heinz was also writing a popular book, The Cosmic Code. But in any case, that was when we had this idea of what the dark matter might be, the lightest super partner particle. And we found that its mass might be about a kilovolt. And that was our sort of natural value for the version of the gravitino that was then popular that we worked out, which we would now call warm dark matter.
I worked out the astrophysical implications with George Blumenthal, which led to a paper by Blumenthal, Pagels, and me. That was '82. And that was one of the main inspirations for Peebles's 1982 Nobel prize paper. You asked me about my interaction with Peebles as an undergraduate. I remember attending a seminar by Peebles, where he'd talked about one of the first N-body simulations, which he had done with a few hundred particles or so. So, I was definitely very aware that he was working on these things. And later on, of course, I got into that game myself. But anyway, in response to this paper that we wrote on warm dark matter with a mass of a kilovolt, Peebles said, "Well, why don't we consider the possibility that the mass is much larger than that?"
So, that's really cold dark matter. And what Peebles worked out in that paper was the consequences for the cosmic background radiation, which were only done in a very rough way. It was a very short little letter. But I think that's what won him the Nobel Prize. Meanwhile, Blumenthal and I were working out cosmic structure formation with cold dark matter, and in '84 I led the CDM paper with Faber, Blumenthal, and Rees—I actually wrote the paper—that became cold dark matter now applied to galaxies and structure formation.
I want to ask at this time, while we're in the chronology of the early 1980s, two other parallel advances that are happening in the field. The first is inflation. To what extent was this registering with what you were doing?
Absolutely, I was very interested. It's a natural thing from the point of view of particle physics because if you have a scalar field that's not at the minimum of its potential, then when you write down the energy momentum tensor, T mu nu, it contributes—there's a term of the form g mu nu times L, the Lagrangian. The Lagrangian is the kinetic energy minus the potential energy. If the potential energy is nonzero, if it's got some positive value, that's a negative number times g mu nu. That's a cosmological constant effect. But it's dynamical, it can change. This is standard field theory. Everybody who knew field theory knew this.
So, what inflation does is, it simply takes advantage of this. And once that scalar field becomes important cosmologically, there's a lot of energy associated with it, and that quickly drives it down to zero. So, that's inflation. And so, I wrote some of the early papers on inflation with George Blumenthal. We were paying a lot of attention to what Andrei Linde was doing. In particular, one of the younger Russian astrophysicists, Lev Kofman, worked with Linde and also with us. And we wrote several papers. One was a paper that was known in the trade as designer inflation. We considered the most general quartic potential—so lambda phi to the 4th plus something phi cubed, plus something phi squared, plus phi, plus a constant—and we worked out all the different possibilities. So yes, I was very much aware of inflation and worked on it.
Did you cross paths with Alan Guth at all?
Absolutely. As I mentioned, SLAC had this early period when it had some of the top people in the world in particle astro as graduate students or post-docs. And Alan was one of them. It was his third post-doc. He was very worried that he wasn't going to make it in the field. But one day, he had this brilliant idea that inflation was the way to solve the problem of overproduction of magnetic monopoles. In SU(3) or any grand unified theory, you expect that there would be stable magnetic monopoles, the 't Hooft-Polyakov monopole. And then, the question is, why don't we see them today? How could they all annihilate? 't Hooft had done a rough estimate of the monopole mass, and one of the first papers I wrote at Santa Cruz was with my first graduate student, Sander Bais. We calculated more precisely what the mass of the 't Hooft-Polyakov monopole is, and we showed how to calculate it to arbitrary accuracy. And so, the astrophysics problem was, these monopoles are so massive that even though they have magnetic charge, they wouldn't be able to attract each other strongly enough that they would annihilate. You need to get rid of them for cosmology to make sense. So, what you really want to do is prevent the energy from getting high enough to make the monopoles, or else inflate them away.
And so, that was the original reason that Guth invented inflation. And I was very much aware of these issues. And I was visiting SLAC all the time in those days, and I talked to Guth. He had just written a paper with a Chinese physicist—
Henry Tye.
—Henry Tye, who was at Cornell at the time. So, they, I think, were very much aware that there was this fundamental problem of getting rid of the monopoles. And I think it was largely because of that that Guth invented inflation. He realized that inflation could solve that by inflating the monopoles away. Inflation, of course, would decrease their number density once they formed. And I was totally aware of how important this could be, and I quickly got into the game.
The other one, depending on how you count this, however you want to name it, the superstring revolution, the second superstring revolution of 1984, when Ed Witten got involved. Did that register with you at all? Was that relevant to what you were working on?
Absolutely. I had learned about supersymmetry from reading papers and hearing people at Aspen talk about it, including lectures by Lochlainn O’Raifeartaigh. In summer ’81 Heinz Pagels and I were working through Witten’s papers, especially the one on dynamical supersymmetry breaking. That led to our paper that proposed that the lightest supersymmetric partner particle as the dark matter. I presented this and discussed our work with Ed Witten at the Banff meeting in summer 1982. I remember that Ed congratulated me for finding a dark matter particle more massive than neutrinos, a problem he had worked on unsuccessfully with Marc Davis and others.
On superstrings, one of the things I later worked on with my then-graduate student, Doug Hellinger, was calculating the rate of gravitational radiation from strings. The idea is that the superstrings will cross and turn into smaller things, and ultimately radiate away their energy as gravitational radiation. And the rate at which that happens, either for superstrings or other kinds of cosmic strings, was then not known. Graduate student David Bennett at SLAC was doing important work on that. But yes, I was very much aware of that, and we actually did some relevant work.
Talking about Witten, one of the crazy things that I learned about from Witten was Alice strings. These would be strings where, if they went past you, they would turn particles into anti-particles, which would have catastrophic effects.
Was your sense that you were very much part of a movement from particle theory in the 1970s to astrophysics and cosmology in the 1980s? Was your trajectory sort of suggestive of larger trends in the field?
Yes, very much.
What were some of the values of that, in having that particle theory background in looking at these new areas?
Well, I mentioned being able to do the calculation on finite temperature field theory for neutron decay, which is important in Big Bang nucleosynthesis. Also, there were other areas where I didn't personally make a contribution, but where people used the techniques of quantum field theory to solve problems in large scale structure. There was a French group. I remember I visited Saclay, and they were very interested in things I had done. The people that I talked to were particle theorists who got interested in astrophysics, and then used particle physics techniques to do large scale structure calculations.
I already mentioned that the Fermi Gamma Ray Space Telescope satellite was something that was brought into existence largely by particle physicists and became a very important source of new data. And the person who was credited with the design of that, Bill Atwood, is now an emeritus faculty member at UC Santa Cruz and SCIPP. The basic point I wanted to make in answer to your question was that gamma ray astronomy is an area where the techniques are particle physics because the energies involved are hundreds of GeV, TeV—particle physics technology. But the source of the energy is astrophysical, and what you're learning about is astrophysics. And you learn about two things. One, you learn about how photons can be excited to these extremely high energies. We've actually seen photons at 20 TeV. One TeV is the entire energy of four uranium-238’s mass converted into energy, and all put on one photon. That's one TeV. 20 TeV is 20 times that. So, somehow, a photon can have that much energy. And not only that, but these very high energy sources, we call them blazars, can turn up and down in a matter of minutes. Markarian 421 and 501 are nearby, 100 million lightyears away, the nearest blazars.
And so, those are jets pointed right at us from supermassive black holes. And they're putting out as much as energy in high energy radiation as the entire energy output of the Milky Way. And we've seen it change by a factor of two in ten minutes. It's like turning on a Milky Way in ten minutes. So, it's absolutely amazing stuff. And of course, what that tells you is that the region that's responsible is not more than ten light-minutes in size. So only black holes can do this kind of stuff. But anyway, the other thing is, this is extremely useful for measuring all the light in the universe. So, our best measurement of the so-called extragalactic background light comes from using these high energy gamma rays, and I had appreciated that very early. In the original proposal for GLAST, I said one of the things we could do with the Fermi gamma ray space telescope is to measure the extragalactic background light.
I supervised two PhD dissertations that really made major contributions toward this. One, by Alberto Dominguez, calculating extragalactic background light from galaxy observations and comparing with gamma ray attenuation, and the other, by Rudy Gilmore, by calculating the extragalactic background light from semi-analytic models of the entire history of galaxies. For example, consider a TeV gamma ray that hits a one eV photon. (One eV is light of wavelength one micron. It's a slightly longer wavelength than visible light, near infrared. Red-shifted starlight, if you like.) That's a million eV in the center of mass, square root of 10 to the 12. That's just enough energy to make an electron-positron pair. And so, that's what happens. These high energy gamma rays hit photons and are removed by being turned into electron-positron pairs. And that's the main thing that takes out the high energy gamma rays.
So, if you can measure the attenuation by comparing at different distances the amount of gamma radiation that you get from these blazars, preferably being able to calculate how much light they would've emitted at these high energies, you can measure the extragalactic background light. And we've done that. I pioneered that, and it's now become sort of an industry. So, that's another example of how high energy physics and astrophysics naturally work together. There's now this big commitment to build 100 of these ground-based gamma ray telescopes. They're being built both in the Canary Islands and in Chile, I think 20 in La Palma and 80 in the Atacama Desert.
In thinking about your work with Seckel and Sadoulet, I wonder if you can convey the level of optimism that the searches for dark matter would be complete, that dark matter would be understood in the short term circa mid-1980s.
Well, in our Annual Reviews article we basically considered the then best ideas, a WIMP, like the lightest supersymmetric partner particle, and we also considered axions, and our article summarized basically all the ideas then current. Hardly any new ideas have come out since then, except for different mass scales for what the dark matter might be. But I was very optimistic that supersymmetry would be seen at the Superconducting Supercollider and that the dark matter particles would be discovered by at least one, and preferably all three, of the methods that we discussed: direct detection, indirect detection, and production.
Direct detection is searched for in underground laboratories where the WIMP dark matter particle would hit a nucleus, and then we'd see an emission from the nucleus, a photon and usually electrons. And we emphasized that you'd have to see two different things. You can't just look for one to be sure that you're really seeing dark matter and not backgrounds like gamma rays. So, that's direct detection.
Indirect detection is, for example, annihilation. The dark matter should annihilate where it's dense, like in the centers of galaxies, and then you'll get characteristic radiation. Of course, we're seeing that radiation in the center of the Milky Way with the Fermi satellite. And as Dan Hooper likes to emphasize, it agrees perfectly with what we expected. There are other possible sources for the radiation, mainly millisecond pulsars, but the people who put that idea forward are now sort of backing off on it. So, it's possible that we're already seeing the indirect evidence.
And then, the last thing is making the dark matter particles. And to be really sure we've got the thing figured out, we want to see all three, and we want to see them all with basically the same characteristics. So, we want to have the same mass when we start making it at the Large Hadron Collider, for example. Although we have seen evidence for indirect detection, and some people would argue that there may be some evidence for that also from the alpha gamma ray spectrometer on the International Space Station, we have not seen direct detection, and the limits are getting tighter. The new measurements are going to improve by perhaps two orders of magnitude the sensitivity we've had up until now. The current state of the art is a metric ton of liquid xenon, the experiment at Gran Sasso and also an experiment that is not so well described in literature, so I don't know exactly what their status is, the Panda X experiment in China. But those are at the one-ton level. And then, there's LZ, which is going to be eight tons, in the Homestake gold mine. They're also souping up the Gran Sasso experiment, and I think the Chinese are also going bigger. And although it's only an increase of an order of magnitude in mass, the increase in the fiducial volume is two orders of magnitude. So, it should improve the sensitivity basically almost to the limit—because if you go much more sensitive, you've got a neutrino background. So, there's a real chance that when these turn on—which would've happened already, except for COVID—we may actually start to see something. Of course, the nightmare that we all have is that the dark matter will turn out to be none of the above, and it won't have any interaction with ordinary matter at all except gravitational. In which case, it's going to be very hard to discover what it is.
What's the path then? String theory, and only string theory?
I don't know. As I say, this is the nightmare.
Tell me about collaborating with Martin Rees.
Two important papers, cold dark matter and the other effects of a cosmological constant besides making the universe expand faster.
So, what happened about cold dark matter was, with George Blumenthal by the summer of '83 we had calculated the power spectrum. And we were working out the implications for structure formation for galaxies and so forth. The great thing about the cold dark matter is, you get this characteristic power spectrum. The idea is that everything comes inside the horizon at just about the same fluctuation amplitude. That's the scale invariant or Zeldovich spectrum of fluctuations. And that's what you naturally expect from inflation, something very close to that.
And then, you get only logarithmic growth in the amplitude of the fluctuations until the universe becomes matter-dominated, and then you get growth as fast as possible, which is linear with the scale factor once the universe is matter-dominated. And so, that gives you a characteristic shape of the fluctuation spectrum, but it also predicts when 10 to the 6 solar mass dark matter halos first form and make the lowest mass possible galaxies. Also 10 to the 9, 10 to the 12 halos—the theory basically says when these objects will start to form. And so, it makes all kinds of predictions about how structure forms, starting with small things and building up to bigger things. So, we had understood all of this certainly by summer of '83.
I met Martin in the spring of '83 at the Moriond Conference in the French Alps, where I was invited to come and talk about both astrophysics and particle physics, including the work with Pagels and Blumenthal, which was published in '82. But on my way there, I stopped in Geneva. And I remember sitting by the lake of Geneva, Lac Léman, and reading the papers, which I had printed out to take with me, that Rees had written. There were two papers, Rees and Ostriker, and White and Rees—Simon White's PhD dissertation. The first one was on, basically, gas cooling and understanding what the implications of that were for structure formation. The White and Rees paper was about what things would look like if galaxies formed in dark matter halos where most of the mass was not ordinary matter. And so, that was a very forward-looking paper, which turned out to be largely right. Anyway, I remember studying those papers, realizing that Martin was going to be there, and that this would be a chance to be sure I really understood those papers, discussing them with Martin. So, I made a point of reading them before I got to Moriond.
It turned out that Martin is a hunchback, and he didn't ski. And the way these ski conferences are organized, there's a little time for talks in the morning, and then in the late afternoon. In the middle of the day, everybody's out skiing. So, the first day I went and took lessons in downhill skiing. And I was a miserable failure. I kept falling, getting all wet and cold. So, I decided I was going to skip that. And so, every day, when everybody else was out skiing, Martin and I stayed behind, and that was when I started to work with Martin. And basically, I described how I thought it was going to work when you combined the calculations that we'd done of the cold dark matter spectrum with structure formation. And whatever I didn't get right, Martin straightened me out on.
Sandy Faber and George Blumenthal didn’t meet Martin until later. I was the one who basically linked Martin into this, and he certainly helped us make sure that we got it all right and pointed to lots of relevant references. Martin was scarily smart. As soon as you'd start to talk to Martin, he would immediately know where you were going, be three steps ahead very quickly, and ask brilliant questions. So, it was a joy and a challenge to work with Martin, but I could keep up with him more or less.
When did you start applying simulations to thinking about cosmological structure formation?
It was largely with another one of the young Russian astrophysicists, Anatoly Klypin. I told you that I worked with Lev Kofman. The last two astrophysicists who finished their PhDs with Zeldovich before he died were Kofman and Klypin. And I ended up being able to work with both of them. Sadly, Kofman died very young of cancer. But Klypin and I have continued to collaborate. We just wrote a paper, which came out a few months ago, on Early Dark Energy and what the observational implications of that would be. Early Dark Energy is my favorite way of solving the Hubble crisis. You measure the expansion rate of the universe nearby, and you get 73 kilometers per second per megaparsec, and if you use cosmic microwave background plus lambda CDM, you get 67. And the 67 is plus or minus 0.4, and the 73 is plus or minus 1. So that's a six-sigma discrepancy. Six-sigma discrepancies don't happen by chance. There's something going on.
One possible solution to that is that there's this brief episode of dark energy contributing about 10% of the total energy density of the universe for a brief period of about 5,000 years around 50,000 years after the Big Bang. What are the implications of that that you could actually observe? Because you can't directly see back to any time before the cosmic microwave background is emitted about 300,000 years after the Big Bang. So, it turns out that this Early Dark Energy theory, and any theory of this class, causes significant changes in the standard growth of structure from cold dark matter. Today, the predicted universe looks almost identical. But at red shift 1, you get 50% more clusters of galaxies than standard lambda CDM. At red shift 4, you get double the number of massive galaxies. And if you go back earlier, you get even bigger changes. And we figured out why this is true. We just published this article.
Now, as to my first paper with Klypin—first of all, you need to do simulations to work out structure formation because structure formation is inherently nonlinear. The first calculations are linear theory, but once you start to actually get structure forming, things collapsing, linear theory doesn't work. You must do simulations. And once you've done the simulations, you can then describe them analytically and put in the physics of galaxy formation. That's called semi-analytic models, and I also pioneered that with my former student Rachel Somerville, who's become one of the great leaders in that field. I first worked with Klypin almost 30 years ago. And we worked out first standard cold dark matter, lambda CDM, and also another version that looked promising, which is cold plus hot dark matter.
The story on that goes back to the late 1980s. My then-graduate student Jon Holtzman had worked with Sandy Faber to analyze the first images from Hubble Space Telescope. They wrote the code to analyze the images from the first camera on Hubble Space Telescope. And when the Challenger disaster occurred, the space shuttle blew up, I think the space telescope was supposed to be the second or third launch after that, but nothing was going to get launched for years until they figured out what had failed. So, Sandy Faber told Jon Holtzman to work with me.
Holtzman improved the code that George Blumenthal and I had used earlier, and he worked out different amounts of cosmological constant and cold dark matter, cold dark matter with warm dark matter, cold with hot dark matter, and so on. And we then compared to all the available data. This is before the cosmic background radiation fluctuations had been discovered, but there was data on galaxies, clusters, large scale structure, and so forth. We found that there were two theories that best fit the available data. And those were lambda CDM with omega matter of 0.3 and omega lambda of 0.7, close to the current values, and alternatively about 0.8 cold plus 0.2 hot dark matter. They were both good fits to the then-available observations. I reported this at a conference at UCLA in spring 1992. When the Cosmic Background Explorer (COBE) measurements of the fluctuations in the cosmic background radiation came out in April 1992, these same two CDM models were the best fits. Eric Gweiser and Joe Silk a few years later also found that compared to the same sort of data we had used, plus the cosmic microwave background data, cold plus hot was a good fit. It did, however, predict much later galaxy formation. Galaxy formation is much earlier with lambda CDM. So, as we started to see more and more evidence of early galaxy formation, that was one of the things that killed the cold plus hot story. What settled it was the discovery of the increased expansion rate of the universe.
But anyway, to work out these two theories, we had to do simulations. And Klypin had pioneered simulations in Moscow. He was at IKI, the Soviet Space Research Institute, where they had a bootleg IBM computer. They had somehow stolen the designs. But the Russians made their own transistors, which failed so frequently that the thing usually died after an hour. So, they had to constantly store outputs and then restart. But despite that, Klypin was already doing state-of-the-art structure formation calculations.
I went to Moscow in October of '88 to try to convince the USSR to stop putting up nuclear reactors in low earth orbit, another one of my science and policy things. And we did convince them. That's another long story.
Did the Chernobyl disaster help your cause?
That was much later. What helped our cause was that two of these nuclear reactor satellites had fallen back to Earth. The Russians had put them up to track the US Navy so it could be attacked in the event of war. These radar satellites, which the US called RORSATs, flew very low, 150 miles or something like that. Top of the atmosphere. So, they had little wings. They could be up between three months and a year depending on how active the sun was. The more active the sun, the more the atmosphere expands, and then these satellites can only stay up a short time. They had nuclear reactors on board, not plutonium power packs, 50 kilograms of high enriched uranium. They were fast reactors. That's what was powering them. And they had to fly so low because the Russians didn't have the electronic capabilities to do phased array radar on satellites, which is what we were doing. Instead, it was ordinary radar. 1 over R squared down, 1 over R squared up. So, 1 over R to the 4th, very strong motivation to be low and need to have a huge amount of power. Anyway, that was what the Russians were putting up. Two of them had come down. One in '78 over Northern Canada, one in '83 over the ocean. And I was working with the Federation of American Scientists. Frank von Hippel had been very deeply involved with Evgeny Velikhov, who was a leader of the Soviet Academy of Sciences, and other Russians trying to convince Gorbachev to help tamp down the arms race. I mentioned earlier that I worked with Frank when I was a graduate student. And there's a long story connected with that that I haven't told you in detail. We co-taught that first SWOPSI course. I continued to work with Frank. In the summer of '72 when I was at Fermilab, Frank was at Argonne, and we got together on weekends to work on our book, Advice and Dissent: Scientists in the Political Arena, which came out in '74.
Frank was the chairman of the Federation of American Scientists, and in the 1980s he was leading this effort to try to tamp down the arms race, especially working on the Russian side. And he asked me to get involved, and I said, "Ok, but I want to work on these RORSATs." Because the American plan was Star Wars. This is the Reagan era. And one of their best ideas, at least I thought at the time, was to have satellites with lasers to shoot down missiles powered by very large nuclear reactors. The first nuclear reactor in orbit was put up by the United States. But it was just a little experimental thing. The Russians put up something like 36 of them, almost all of them these RORSATs targeting the US Navy. Frank introduced me to Roald Sagdeev in fall of '87 in Washington. Sagdeev was the head of the Soviet Space Research Institute. And he was also Gorbachev's top civilian advisor on space.
Sagdeev agreed with that maybe the US and the Soviet Union could agree to stop putting these things up. And the reason the Russians would find that useful was, one, that would stop the United States from putting up these Star Wars satellites, and two, they could switch over to phased array radar, so they wouldn't need the RORSATs. And of course, the US didn't like the RORSATs. The main US official objective for creating anti-satellite weapons was to shoot down the RORSATs. So, if we could stop putting up nuclear reactors in low earth orbit, that would be a good thing from everybody's point of view. So, Sagdeev and I agreed with von Hippel that during Sagdeev’s next visit in May of '88, we would make a formal announcement that this would be what we'd propose that the Soviet Union and United States agree on. And presumably, Gorbachev would sign off on that. Or at least Sagdeev thought this was something that might work. We proposed this at a press conference at the press club in Washington. And it didn't get the back page of the New York Times, it got prominent coverage.
And the reason was that at about the same time, a British boys' school, which had been doing satellite spotting, announced that the latest RORSAT was misbehaving. And several foreign governments confirmed that they saw the same thing the boys' school did. And so, both the Washington Post and New York Times gave this prominent coverage and also mentioned our press conference. It was an incredible coincidence. That that would happen at the same time is amazing. But in any case, that gave us a lot of opportunities to put strong pressure on the Soviet Union to stop putting these things up. Two of them had come down, and the one that landed in Canada and had made a big mess. It contaminated an area 300 kilometers long by 100 kilometers wide with high radioactivity. Fortunately, it was very far north Canada. There are pictures of this, incidentally, in the Scientific American article that my US colleagues and I and the Russians co-wrote, which came out in '91.
Stopping the RORSATs was what took me to Moscow. I'd met a bunch of the Russian astrophysicists, including Zeldovich, the previous year at an international conference in Hungary. I had not met Klypin though. I first met Klypin at IKI. I was giving a talk at IKI about some of my latest work, which Andrei Linde translated. I would say a little bit, and then Andrei would translate. His English was excellent. I'd gotten to know Andrei in a previous conference, a weeklong International Astronomical Union cosmology conference in Balaton, Hungary, in 1987. And then, the following week, Alex Szalay, who later saved my life from pancreatic cancer, he organized this follow-on to the Balaton conference in Budapest with a dozen Russians and Americans.
I was fortunate to be one of the Americans involved in that. That's where I started to work with Lev Kofman. But Klypin was not there. Klypin’s office was on the 8th floor of IKI. It was October, and they hadn't yet turned the heat on. They waited until November, apparently. And so, everybody was drinking hot tea and wearing heavy jackets in their office that Klypin shared. And he and I, at that time, had both written papers about puzzling aspects of galaxy cluster-cluster correlations. And so, we had something to discuss, papers that we'd just written. And we really enjoyed talking to each other. That was the beginning of my collaboration with Klypin. He managed to get out of the Soviet Union a year or two later, as did Lev Kofman. And once he came West, we started working together. And it was the desire to work out the implications of these alternatives to cold dark matter, lambda CDM, and cold plus hot, which was our motivation for doing these early simulations.
What are steep central dark matter halo profiles, and what was the conflict that you were seeing between simulation and observation?
By steep central dark matter profiles, I guess you are referring to what’s now called the core-cusp controversy. My grad student Ricardo Flores and I had pointed out in a 1996 paper that dark matter halos have cuspy centers, which appeared to contradict the then available galaxy observations. This problem is now thought to be solved by including the effects of the ordinary “baryonic” matter. Well, at that stage, observation and theory were both quite primitive.
As to the structure of dark matter halos, it was later that was figured out. The most important papers on the radial structure of dark matter halos were, first of all, Navarro-Frenk-White, three papers from '96, '97, and '98, and they showed that the inner part of the dark matter halo is 1 over R squared, and the outer part is 1 over R cubed. So, logarithmically, it starts like that and then becomes steeper. That's the radial dark matter distribution. And the way they did this was by doing re-simulations of individual halos. And that had the advantage that they achieved high resolution. It had the disadvantage that they never actually saw the time evolution of a bunch of different halos. Klypin did much higher resolution simulations, and I worked with my then-graduate students on this, especially James Bullock and Risa Wechsler. We discovered how the radial structure of dark matter halos evolved, and how it depended on the local environment.
My group also worked out in detail what the 3D shapes of the dark matter halos were. There had been a series of papers by a bunch of different people that looked like they totally disagreed with each other. But it turned out that they'd measured halo shapes differently. The only way to straighten this out was to do the same calculation that every different group had done but one of the same halos. And we were able to do bigger calculations than people had done before and store data at many output times. So, instead of having to re-simulate a bunch of halos, we actually simulated a whole region of the universe, which is now standard, but we did some of the first such simulations. My student Brandon Allgood was the first author of one of these papers, and Ricardo Flores was the first author of another of these dissertation research papers with me.
The Allgood paper was published in 2006. And what we found was that the dark matter halos start out quite prolate, and then as time goes on, they become rounder. But they're still rather prolate. The long to short axis ratio of the dark matter halo of a galaxy like the Milky Way today is nearly a factor of two. And in our paper, we showed how these halos evolve—and why. Most halos form along long dark matter filaments, and much of their dark matter falls in along the filaments. That results in anisotropic velocity dispersion, with higher random velocities along the filament direction. It’s that, not rotation, which supports these prolate dark matter halos.
And then, the next thing on the shape of dark matter halos story, this turned out to be extremely important as we started to understand the shapes of galaxies. I and probably most astrophysicists expected that galaxies would start out as disks. Basically, for the same reason that proto planetary systems start out as disks. Newton could not explain why all the solar system planet orbits are in a plane, which we call the ecliptic, and they're all going the same way around the sun. And it's also the same way that the sun rotates. Newton was asked about this, and he answered, "This is evidence of God's handiwork."
Laplace, in the 1780s, wrote a book on the origin of the universe, actually the origin of the solar system, and he argued that this was basically just angular momentum conservation. You have a gas cloud that's got some rotation, it would shrink as it radiates away its energy, and it would become a disk because it must preserve its angular momentum. And the center would become the star, and the disk would become planets because of gravitational attractions. And that's basically the right picture for the origin of planetary systems. And now, we see this with radio telescopes, the ALMA array especially.
And so, why wouldn't it work the same way with galaxies? The dark matter would form a halo. The ordinary matter could continue to lose energy because the atoms would collide, they would get excited, they'd radiate away energy. As they lose energy, they'd fall in closer and closer to the center, and then finally, you'd get the same story. They'd make a disk.
And why wouldn't it work the same way for galaxies as for planetary systems? When we actually looked at the data, we found that galaxies start prolate. This was first convincingly shown in a CANDELS paper led by Arjen van der Wel in 2014. The Hubble images make it obvious that the galaxies start out prolate. They start out pickle shaped, not disks at all. You often see these forming galaxies looking like this. How do we know that that's not just an edge-on view of a disk galaxy? And the answer is, you're going to see galaxies in every orientation randomly. When a disk galaxy is seen face-on, it will appear nearly round. So, if you don't see some nearly equal short and long axes, some very small short axis to long axis ratio, and everything in between, then the galaxies can't be disks. They must be prolate. We recently did a more detailed analysis of the final CANDELS complete data set. Sandy Faber and I worked on this a couple years ago with a brilliant Chinese student, Haowen Zhang. We reached similar conclusions, but in more detail.
When we first realized that the Hubble images proved that galaxies start out pickle shaped, I have to admit that we had already done simulations which showed exactly that. But I never thought to visualize the simulations to see what the galaxies looked like until we got the Hubble observations. Then, as soon as we visualized the simulations, we saw the same thing. The galaxies start out prolate.
The reason is that the dark matter is almost all the mass, and it's controlling, even in the center, the shape of the forming galaxy. And only after a lot of ordinary matter falls in does the ordinary matter become most of the matter even in the center of the forming galaxy. And as soon as the ordinary matter is the majority in the center, our simulations showed that you start to form disks. That turns out to fit the data quite well. So, that's a problem that we've now solved.
There's another very big problem though, which is that most star-forming galaxies have these giant clumps. The biggest clumps of stars in the Milky Way are globular clusters, and they top out at 10 to the 6 solar masses. And the same is true with most other nearby galaxies.
But these giant clumps are 10 to the 7, 10 to the 8, sometimes even 10 to the 9 solar masses. So, 10, 100, or 1,000 times more massive than the most massive clumps of stars that we see in nearby galaxies. And we make them in our simulations of these early galaxies—but the simulations that make them produce too many stars. If we turn up the feedback until we get the right number of stars, the right mass of stars, then we kill the clumps before they become massive enough. And the main competitor to our group, the FIRE group, led by Phil Hopkins at Caltech, has the same problem. They don't make the clumps. But the clumps are a ubiquitous phenomenon observationally. I just submitted a proposal where we are going to try to figure out what's going on. But the prolate galaxies turn out to be a consequence of the prolate dark matter halos. And the key paper on that is still the Allgood et al. 2006 paper.
Tell me about Ari Maller's thesis research.
Ari's one of my smartest graduate students ever, but Ari was not as ambitious as some of my other brilliant students, like Rachel Somerville, James Bullock, and Risa Wechsler. But Ari loves New York, and he seems quite happy at New York Poly. And he also has the advantage now that there's the Center for Computational Astrophysics, the CCA, which is part of the Simons Foundation’s Flatiron Institute. I understand that Ari's been spending a fair amount of his time there. Rachel Somerville is the leader of the CCA galaxy group. I think Ari is continuing to do interesting work. Anyway, the subject of his thesis was to understand that when you have two galaxies that lie more or less along the same direction in space, there are two things that you can do.
One is, you can look for gravitational lensing of the light from the background galaxy by the gravitational field of the foreground galaxy. And the other thing you can do is to use the light of the background galaxy to study the internal structures in what we call the circumgalactic medium surrounding the foreground galaxy. And the idea of his thesis work was to do both of those, to look both at the gravitational lensing phenomenon, but especially at the structure of the dark matter halo, the gas that would be absorbing light from the background galaxy. And what we ended up doing was creating a semi-analytic model of dark matter halos and their gas content. That subject that was then on people's minds—and there's now much more data thanks especially to the COS instrument, the Cosmic Origin Spectrograph on Hubble Space Telescope, which wasn't available at the time. We're talking about 2002. Ari finished his thesis research with me somewhere around 2002. So, at the same time as James Bullock and Risa Wechsler, and just a little after Rachel Somerville, an amazing group of graduate students.
And what was then, I think, most on people's minds, as far as the gas in and around dark matter halos, was what were called Damped Lyman-alpha Systems. So, that's what we concentrated on. And the person who invented this terminology of Damp Lyman-alpha Systems and pioneered the field is Art Wolfe. Wolfe had earlier done really important calculations on gravity and things like that. I think he was at Berkeley, then he moved to UC San Diego, and one of the points that he had made was that the total amount of mass in the hydrogen clouds in these dark matter halos that you could measure with the background quasars, you look at the absorption of Lyman-alpha, the n=1 to n=2 transition in neutral hydrogen, and you can measure the amount of hydrogen.
You look at the shape of the absorption spectrum, and that lets you determine what the total mass of hydrogen is per square centimeter along your line of sight. And Art Wolfe correctly argued that the total amount of hydrogen that we could see at red shift 3 in these Lyman-alpha clouds was equal to the mass of all the stars in disk galaxies today. So, his picture was that all these clouds of neutral hydrogen in these big dark matter halos had somehow collapsed to make the disk galaxies of today. That's not what we now think. But it was certainly a plausible idea at the time. Art and his most brilliant student, Jason Xavier Prochaska, who likes everybody to call him X—he's my faculty colleague here at Santa Cruz—measured various features of these hydrogen clouds.
Associated with the hydrogen clouds, there are also what are called metal-line systems. So, in addition to the neutral hydrogen, there'd be little clouds of magnesium and things like that. And you could measure how fast they were going. You'd measure some over here, some over there, so you'd get an idea of how rapidly things are moving in the dark matter halo. They had all kinds of statistics on this. And the question was, can you put this together into a unified picture? And that was a key project of Ari's dissertation work with me.
One of the first of the gravitational lens systems that we looked at was a foreground galaxy that had made a double image of the background galaxy, the background quasar. There are several other cases where there are quads, you get four different images of the background quasar. The one we analyzed was a double.
An interesting question having to do with disk galaxies is just how much mass there is in the disk. That was uncertain both for the Milky Way and other galaxies. It's still somewhat controversial, but it's now been cleaned up considerably. I can tell you more about that, but it's not my work. What we were able to show was, from the amount of gravitational lensing, we got a strong upper limit on how much mass there was in the disk of the foreground galaxy. And this disproved, at least for this one galaxy, what a number of people had claimed about how much mass there was in the stars. Subsequent analysis with the quads and other methods have confirmed what we found. Especially the work on this by Matt Bershady at University of Wisconsin, which I think has been very influential.
Broadly conceived in all of this work on galaxy imaging, what were some of the key advances in instrumentation that were relevant for your research?
As far as imaging of galaxies, there's no question, Hubble Space Telescope has had an enormous impact.
You're talking about after the repair, I assume.
Yes, of course. The repair was basically to put in a new camera, basically corrective lenses. What happened was, Sandy Faber and Jon Holtzman arranged to go to Space Telescope Science Institute right after Hubble was launched to be able to take advantage of Jon's work writing the software to get all the images, and for the first time to be able to see these clear images of what galaxies really look like. And of course, what we saw instead was that there was a horrible optics problem with the mirror. There were two teams that independently tried to figure out what was going on. One got access to the previously classified data at PerkinElmer, where the mirror had been incorrectly built. The other was that Sandy Faber and Jon Holtzman analyzed the images, and they figured out what was wrong. And fortunately, the two teams got the same results, and so it became clear what you needed to do with adaptive optics to fix it. So, basically, after the fix, Hubble was just about as good as it was originally planned to be.
After that, new instruments were installed on Hubble. The second camera, the one with the corrective optics, is called Wide Field Camera 2. What's now on Hubble replacing that one is Wide Field Camera 3. And what came in between those was the Advanced Camera for Surveys, ACS. So, those are now the two major cameras on Hubble, ACS and WFC3. And then, the third truly important instrument is COS, the Cosmic Origin Spectrograph. And all of them have been totally revolutionary.
There are several small regions of the sky where there has been very deep imaging, first with Wide Field Camera 2, then with Advanced Camera for Surveys, and now with Wide Field Camera 3. Wide Field Camera 3 has ultraviolet capability, which has been somewhat disappointing. But mostly, it gave us infrared capability. Its longest wavelength is 1.6 microns, what we call H band. And that lets you see the blue light that stars emit out to red shift 3 and the red light out to red shift not quite 2. That's important because with Advanced Camera for Surveys and the original Wide Field Cameras 1 and 2, we only saw optical light, basically, the light our eyes see. And the problem is that the most important source of the short wavelengths is very young massive stars. When you make a bunch of stars, you always make some O stars, the most massive stars, which radiate a lot of their light in UV. UV emitted at red shift 2 or 3 converts to optical when you see it because of red shift. So, all you could see before Wide Field Camera 3 was basically UV from galaxies at red shift 2 and above. You couldn't see the light of ordinary stars at these redshifts. And so, we didn't know what the galaxies really looked like. We only saw what the regions of star formation looked like, which are usually little clumps.
Once we had Wide Field Camera 3, which was installed in the last Hubble service mission in 2009 along with COS, that totally revolutionized our imaging. One of the first things we learned, as I told you, is that galaxies start out prolate, they don't start as disks. There had been a competition to oversee Hubble once Wide Field Camera 3 was installed, and CANDELS was the survey that came out of that. Sandy Faber had submitted a proposal, there had been another proposal submitted by people at Space Telescope Science Institute, and the two were combined to make CANDELS. And I was fortunate to be the main theorist working with Sandy.
Another CANDELS discovery was that galaxies appeared suddenly to contract in size. We call it compaction. That's not what's really happening, but if you look at the size of galaxies as a function of their stellar masses, there's a sudden drop right around a stellar mass of 10 to the 9.5 or 10 to the 10 times the mass of the sun. The reason is that there are these in-falls of gas that go right into the galaxy centers, and that's a lot smaller than the disks that were forming or the prolate structures that were forming before that. So, we called that compaction, a word that's starting to catch on. And so, one of the things we did was, I was mentioning earlier, we trained a deep learning code using our simulations. We identified galaxies as pre-compaction, compaction, and post-compaction just from simulations, not looking at the images.
But we also made realistic mock images, taking into account stellar evolution and dust effects, and we told the computer, "These are pre-compaction, these are compaction, these are post-compaction." And the code can do the same analysis on the real images from Hubble. We basically took the entire output of Hubble imaging, applied artificial intelligence, and we saw the same pattern, and we saw the transitions at the same stellar masses as in our simulations. Our paper on that was published in 2018.
The word partner can take on many meanings in the context of this question, but I'd like to ask specifically about your work with Nancy Abrams on the book projects.
Nancy and I have had a wonderful marriage. Another one of extremely lucky things in my career was meeting Nancy, which only happened because I was doing this science and public policy stuff. The way we first met was, I was in Washington, I had started the Congressional Science Fellow program, and I'd written this book with Frank von Hippel, which came out in '74. Some people on Senator Ted Kennedy's staff, and apparently Kennedy himself, read our book. And they wanted to pass some new legislation that would improve the way Congress dealt with science and technology, the way the United States dealt with science and technology.
And so, Kennedy asked Frank and me to come and talk with him in Washington, and what came out of this was something that we called Science for Citizens. Frank and I organized two sets of hearings, and just as Kennedy had predicted, the thing passed. He told us right at the beginning, once we had cooked up the scheme, who was going to be in favor of it, who was going to be against it, how he was going to make a deal with this person or that person. The stories you read about him being a brilliant legislator—it was amazing. He really saw the whole picture. It worked out just the way he said it would.
But in any case, I was in Washington for one of these hearings, and Kennedy had a Congressional Science Fellow who was on leave from Avco-Everett Research Lab. His boss Arthur Kantrowitz was on the President's Science Advisory Committee. They were pushing an idea called the science court. And so, they invited me to come and critique their idea at one of their meetings, a subcommittee of President Gerald Ford's Science Advisory Committee. And I basically tore it apart. Nancy was there because she had been working for the Ford Foundation, where she'd been leading their program on environmental mediation. Kantrowitz and his colleagues wanted to get money from the Ford Foundation to have a trial run of the science court.
Nancy told me at the end of this meeting that she thought the only one who had said anything sensible was me, that she totally agreed with my critique, and that she thought the science court was a terrible idea. She said, "The purpose of courts is to resolve disputes without violence by following standard procedures. It is not to find the truth. If it happens to find the truth, that's fine. But that's not the purpose of courts. And to imagine that a court-type process would be a good way to find the truth is to completely misunderstand the whole point."
Nancy and I kept getting together in connection with this, and we really liked each other. And before too long, we got married. Nancy's background is history and philosophy of science as well as law. Her undergraduate degree at the University of Chicago was history and philosophy of science. I'd heard about Ernst Mach. Nancy had actually read his book. I have her copy, all marked up, where she studied this stuff. So, she knew a fair amount of physics. But of course, she was especially interested in public policy issues. We bonded over that, but then as I was more involved in these discoveries in cosmology and astrophysics, she was always very interested, and she had enough background in science to understand what was going on. And she would come with me to many international conferences, and meetings, and summer schools.
And in addition to everything else, Nancy's also a musician. After she finished at the University of Chicago, she spent three months in Israel on a kibbutz and got really bored. And then, she went to Paris, where she studied mime with Marcel Marceau's teacher. And she taught English to French people. She spoke French. Then, she went to Italy and learned Italian. And she was recruited into an Italian cabaret troupe. And she and the other members of the cabaret troupe traveled all over Italy for over a year, singing mostly in Italian, but also she would translate some of their songs into English or French, so they could appeal to the foreign tourists. So, she's a professional musician. Later on, Nancy wrote many songs about science and policy, and also about other things.
Nancy’s science songs became a big treat at conferences. I would be invited to speak at a conference, and Nancy would be invited to sing and perhaps write a new song. Dennis Overbye's Lonely Hearts of the Cosmos ends by quoting all the lyrics of a song that she wrote for a conference in Kona, Hawaii for the groundbreaking of the Hubble Space Telescope, a very important conference in 1986, to the tune of Yellow Submarine. "We all live in an expanding universe, expanding universe, expanding universe…." Nancy's going to be doing a concert of some of her science songs at the March meeting of the American Physical Society in Chicago next year. She was supposed to do it two years ago, but that conference, of course, became virtual. It was supposed to happen this year, but the March and April conferences, again, were virtual.
How did writing these books with Nancy sharpen your ideas specifically as a scientist?
Well, Wheeler used to say that if you really understand what you're doing, you should be able to explain it to your grandmother. So, having to understand things well enough so that I could explain them so Nancy, who was a critical judge, could understand them certainly sharpened my understanding and my ability to present them. But more than that, because Nancy was so interested in what I was doing, she pushed me, I think, more toward interesting subjects.
So, for example, after this early work with Sander Bais where we calculated the 't Hooft-Polyakov monopole mass, we did very detailed papers on monopoles in larger numbers of dimensions and things like that. I also worked with Richard Brandt on hypothetical particles with both electric and magnetic charges. Very esoteric stuff. And Nancy was really bored with this. And then, I started to work on cosmology, and she thought that was much more exciting. And so, frankly, that made it more interesting for me, too. She definitely pushed me toward working more on cosmology, which in the long run had turned out to be a really smart thing to do. But it was partly because Nancy found it more interesting that I did it.
And then, of course, we started to give these talks that we would do together. Well, actually, what we started to do was give a course. We gave a course called Cosmology and Culture. It started as a little seminar course at Crown College, the college at UCSC that I've been affiliated with since I came. And the students loved it. And we made the mistake of expanding it into a very large course, 150 students or so, which was totally unwieldy. Because we had wanted to have the students write essays and things like that. And there's no way that you can read that many essays, even with TAs to help. In following years, we gave the course with a requirement that students had to write a little essay on why they wanted to take the course. We didn't pay any attention to why they wanted to take it, we just paid attention to whether they could write. Because we didn't want to read essays by a bunch of kids who couldn't write. So that's how we narrowed down the students. And we used to have, I think, 30 or 40 students. We taught this over a decade, almost every year, and we got better and better at explaining cosmology and the whole modern picture, including some physics, as we tried it out on students again and again.
For example, we started with the common view that the fact that we're made of stuff that's created in stars and stellar explosions, basically the dregs of the universe, and that we're made of the stuff that is the lowest fraction of matter in the universe. We're not made of dark matter, not even mostly hydrogen and helium, although, of course, there is some hydrogen in our bodies. We're made of oxygen, carbon, nitrogen, and things like that, which, at most, represents 100th of 1% of cosmic density. So, we're very much a cosmic afterthought. At least, that's one way of thinking about it. But the other way of thinking about it is that we're made of the rarest stuff in the universe, and that makes us pretty special. And intelligent life may also be extremely rare in the universe.
So, presenting it that way is much more effective at communicating. Because if you present it the other way, that tends to turn students off. This idea of seeing us as being cosmically special or central became the theme of our book. The title of our first book, The View from the Center of the Universe, was not our idea. I don't remember what our working title was, but it was nothing like that. The publishers liked that title. In fact, they liked "The Center of the Universe," and then we added "The View From." And I remember what the publishers told us: "The title of a book is a marketing decision, and what do authors know about marketing?" But the theme of the book was an outgrowth of this idea that I was just saying—that we're not cosmic dregs and an afterthought. You can also think of us as being very special.
And so, the way the book is organized, after a couple of chapters on the history of cosmologies starting in the Middle East, the next several chapters are looking at different aspects of the universe. Size, mass, what it's made of, things like that. And then, in every case, seeing that we're central or special. So, for example, on the size scales from the smallest size that physics knows how to deal with, the Planck length, to the size of the entire visible universe, is about 60 orders of magnitude. And our human size is close to the center of that, logarithmically.
So, that very much came out of our experience in trying to teach these concepts to a bunch of undergraduates, almost all of whom were not science majors. Then, our second book, The New Universe and the Human Future, was basically because people at Yale read our first book and invited us to give the 2009 Terry Lectures. Our joint lectures were the first time in a century of Terry Lectures that two people gave the Terry Lectures. I think we were also the first ones to show lots of images and videos. Yale University Press publishes all these art books, and they treated our book basically as an art book. And it's a beautiful book, with nearly a hundred color illustrations and a linked website with many videos, although unfortunately it never sold very well.
What have been some of the values that you've seen from a policy perspective in this broader science communication and the need to bring in that broader public on all of these endeavors that are happening right now?
I was asked to run for president of Sigma Xi, and I thought, "If I do, what could I usefully do in this area?" And what I decided I would emphasize was teaching scientists how to communicate better to the public. I was already worried before the Trump Administration that scientists weren't getting across to the public key ideas like climate change and how dangerous it is. My first forays into science and policy were during the Nixon years. And, when he was reelected in '72, Nixon abolished the Presidential Science Advisory Committee. So that encouraged me to get the American Physical Society to do our own studies, rather than waiting for the government to finance studies, either through the National Academy or PSAC.
The Congressional Science and Technology Fellowship program was, to some extent, an outgrowth of reaction against Nixon, the Vietnam War, and all that. But Frank von Hippel and I thought that the situation under Nixon, the supersonic transport SST, the anti-ballistic missile system, and so on, which were scientifically bad ideas, was as bad as it could get. Actually, it got much worse under Reagan, wasn't so great under George W. Bush, and of course, Trump is so much worse yet. So, I thought that educating people better is really important, and part of the reason it's not getting done is that we scientists don't do a very good job of communicating.
Most scientists are very reluctant to talk to the public. And if they are invited to do so, they usually hedge, the way we scientists usually will do. We're very careful about making claims and so forth. Which is fine when you're talking to scientists, but it misleads the public because they think we don't know what we're talking about. And then, the other problem is, a lot of people see largely old white men like me as typical scientists. And we do not represent a demographic that most people relate to very well. So, I also thought it was really important to broaden the scientific communication to young people, including graduate students, post-docs, and junior faculty, as well as to diversify it in every other way. So, that was my big goal in my presidential run in Sigma Xi.
Also, I was recently in my second term as chairman of the APS Forum on Physics in Society. I had helped to create the Forum on Physics in Society in 1972, and I coined this name “Forum.” Bill Havens, who was still running the American Physical Society in the early 70s, said the Physical Society had divisions and regional groups, and this didn't count for either of those. So, I said, "Oh, let's just call it a Forum."
Another of my arguments for the Congressional Science and Technology Fellow program, the Forum, and the APS studies on public policy issues was that these would give physicists and other scientists a much bigger perspective and change their career ambitions and success. And Havens thought that was a very good argument because we were already overproducing PhDs. I've had something like 40 graduate students where I was either the only or one of the mentors. And that's obviously vastly overproducing. But then, many of my recent students have become data scientists. At least eight of my former PhD students are now playing that sort of role. Which I think is fine, if they like it, they're making lots of money and presumably having lots of fun, maybe even doing useful things.
But I have to say, my efforts at Sigma Xi were quite unsuccessful. I tried to get them to raise money, to allow me to raise money on their behalf but with their cooperation, because ultimately, I'm not going to be the grantee. And part of the problem is that Sigma Xi was recovering from a financial disaster, and they were very conservative about what they were willing to do. And they didn't have the staff. They still don't have much professional development staff—in other words, professional fundraisers. They have one now, who basically is raising money from members, but not from foundations. I had good connections with foundations that I thought would be productive, but they never let me pursue them. Fortunately, other scientific organizations, AAAS, the American Physical Society, are definitely moving in that direction. And Sigma Xi is starting to do so now, although I can't claim to have had much to do with that.
A technical question to sort of summarize your work over the past 10 or 15 years. Do you see observation and simulation in cosmology as a two-way street? In other words, simulating the universe enhances our observations of the universe and vice versa? Or do you tend to think of these as separate paths of inquiry?
Very much the first. And as I told you, one of the things that we've pioneered is to use simulations to train artificial intelligence codes, which then can analyze the observations. And that's a new way to connect the observations and the simulations. I told you about this example, Marc Huertas-Company, me, Dekel, and others. The 2018 paper that I described to you, where we trained the deep learning code to recognize three stages of galaxy evolution—the mostly prolate stage, the compaction stage, and then the post-compaction stage, which is typically a spiral galaxy. The papers we just recently published using deep learning were on the giant clump story. We trained the deep learning code to measure the masses and ages of the clumps.
Again, I don't think there's any way we could've done that without using the simulations as the training set and then applying it to the observations. And, of course, to the extent that there's a disagreement, then it's a challenge to us theorists to try to figure out why things aren't working, what we must change. And I think right now, the biggest thing that we're doing wrong is what's known in the trade as feedback. So, when stars form, a certain fraction of them are very massive stars that become supernovae, and that can blow the gas out of the region where stars are forming and prevent further star formation. But it can also enhance star formation if the expanding shell concentrates the ordinary matter, by shockwaves, and so on.
It's a very tricky thing because the phenomena that we're talking about, star formation, supernova explosions, the shockwaves, and so forth, occur on very small scales. Scales like an astronomical unit, the size of the earth's orbit, or the size of proto planetary disks or, things like that, whereas when you want to simulate a galaxy, you have to simulate its dark matter halo and even the region somewhat beyond that because things are going to be coming in. So, you need to have a simulation of at least megaparsecs in scale, whereas the physics that we're talking about is occurring on scales much smaller than a parsec, three light years. And so, what we do in the large-scale simulations when we simulate galaxies is, we have what's called subgrid models for things like feedback. And they're not well-constrained. And every code does them somewhat differently. So, one of the things that we're doing in the AGORA project, where we're comparing simulations by the leading codes of the same galaxy, is trying to understand the uncertainties in feedback, and then also looking at things like these massive clumps as a feedback test.
Another source of feedback comes from the giant black holes in the centers of galaxies, the so-called active galactic nuclei. As matter falls into them, a huge amount of energy is radiated. And that is probably what's responsible for turning galaxies into what are called quiescent galaxies, stopping star formation. And that, again, is not something that's very well-understood. And so, we're also working on that.
So, what I'm telling you is that simulations can help us interpret observations. For example, I told you that we're learning how to measure the masses of the clumps. The problem with the clumps is that they're mostly smaller in size than a kilo parsec, 1,000 parsecs, about 3,000 lightyears. Most clumps are smaller than that. But Hubble Space Telescope's resolution at red shifts of one and above is about a kiloparsec. So, in other words, we're not resolving the clumps. So, if we try to measure the masses of the clumps, we're including the light from a lot of the rest of the nearby galaxy. So, how do you correct for that? Well, in our simulations, we have much better resolution. Our resolution in our current simulations about 20 parsecs. So much, much better than a kilo parsec. What we do is, we figure out what a clump’s mass is by measuring much is actually bound together into these clumps in the simulations. And we teach the computer, "This is how much this clump really weighs. This is what it's going to look like. You figure out how to correct for this." And you give the machine a hundred-thousand examples, and the codes are really good. You tell them, "This is what the truth is, this is what you're going to see. Figure out how to connect them." That's what they're very good at. I call it “face recognition for galaxies.”
On the one hand, we're helping the codes interpret the data, measure things that we can't measure directly. Fortunately, James Webb Space Telescope has three times better resolution than Hubble at the same wavelengths. So, we're going to get a test right away of how well we've done this when James Webb starts giving us better images of the same galaxies.
And on the other hand, when we do these projects, we find out that our simulations are not adequate. What we find is that our simulations and others' simulations that correctly reproduce the clumps are overproducing the stars. And if you cut back on the stars by upping the feedback, but still doing the feedback the usual way, just turning a knob to make it a little stronger, then you kill the clumps. So, there's clearly something wrong.
That's one example of how it really is a two-way street. We're using the simulations to help interpret the observations, but we're using the observations to tell us what we're not doing right in the simulations.
And, incidentally, let me make a contrast with cold dark matter in the early days. When you're just doing large scale structure, it's really pretty simple. You can learn a lot from linear theory, which is simple calculations, analytic calculations, and then the next generation of simulations, these large scale structure simulations where you just use dark matter—my group has a series that we call the Bolshoi simulations from the Russian word for big in the sense of great, and also another series called MultiDark because that was the name of the European grant that funded some of that. Those simulations are comparatively simple to run and to compare with observations. And so, a lot can be done without this two-way street thing. But galaxies are extremely complicated.
[End 210715_0642_D]
[Begin 210715_0643_D]
Because there's all this feedback. It's not just that the stars form, it's that they then modify the background in which they're forming. And especially once you get these supermassive black holes, they also have a huge impact on the galaxy. The way most science works is that there's an interplay between theory and experiment, or theory and observation. And that's very much going to be true as we try to understand both astrophysics of galaxies and the components of galaxies.
And also, astrobiology. For example, I think it's likely that life formed on Mars. Mars, in its first billion years, should've been quite hospitable to life. It had lots of liquid water, it had an atmosphere, had a magnetic field. So, since on earth, we know that primitive microorganisms formed at least by 3.7 billion years ago, the stromatolite fossils that we have, and they're already pretty complicated. So, some life formed before that, and that's very close to the end of the cosmic bombardment, the Late Great Bombardment. So, there's every reason to think that some life formed on Mars, and it's entirely possible that there's still life below the surface. Mars has almost no atmosphere, the surface is too cold and gets too much cosmic radiation.
On earth, there's supposed to be about as much living mass below the surface as on and above the surface. We keep finding life deeper and deeper, more than a mile deep. And this underground life, it looks like, is largely powered by radioactive decay. Radioactivity splits molecules, and then combining those atoms and molecules is a source of energy. So, there are more and more papers indicating that that may be the way life could continue on Mars.
The big question with life, one of the big questions, is, is it based on the same genetic code or something completely different? And the only way to find out is to have at least another example of life. And Mars or the icey moons of Jupiter and Saturn may be other possible places because, of course, under the ice is liquid water. And possibly, life could've gotten started there. Further away, we won't be able to encounter it directly and will have to look for indirect evidence of life like atmospheres. So, that won't be able to tell us whether it's based on the same genetic code.
But in any case, we desperately need more data on astrobiology. There's every reason to think that we're going to explore the solar system, and maybe we'll be lucky enough to encounter some other life.
So, in anything complicated, you can't expect that science is going to come up with the right story just by deep thought. Gravity was simple. It just took Newton deep thought and a couple of pieces of information. And Einstein, of course, came up with general relativity almost entirely just based on deep thought, with almost no observational data, although he was aware of Simon Newcomb's calculation of the perihelion precession of Mercury, that it was 43 seconds of arc per century after correcting for the effects of the other planets. And Einstein supposedly had heart palpitations for two days when he got 43 seconds of arc per century with general relativity. Because that meant the theory's right. But that was one of these cases where it was basically, essentially, pure thought. And cold dark matter was pretty close to pure thought. But it's only in the beginning of a science that the simple, low-hanging fruit is like that. Once you get into the real complexity, you don't expect theory to come up with the right answers right away.
Well, we began our talk discussing your current interests, so now that we've worked right up to the present, for the last part of our interview, I'd like to ask one broadly retrospective question about your career, and then we'll end looking to the future. So, of all of the things that you've worked on, with all of your collaborators and all of the areas in physics in which you've been involved, what stands out in your memory as simply being the most enjoyable, however you might want to define that? Being a scientific eureka moment, the pleasure of a real intellectual partnership, the pleasure of learning something that was so difficult for you before. What stands out in your mind?
The 1984 paper on cold dark matter. And especially, when the cosmic microwave background fluctuations were discovered, and they were what the theory had predicted, basically. The same amplitude. And as more and more data came in, and we were able to do the more detailed calculations, the same power spectrum, etc. But I remember when we first did the cold dark matter paper, we were basically walking on air for a couple weeks. Because when we compared with the then-available observations, everything worked.
The then-available observations were what was called the CfA Survey, by the Harvard Smithsonian Center for Astrophysics. They had done the first "large scale structure survey." They had something like 2,000 galaxies in both the northern and southern hemispheres. And the classic paper that analyzed that was Davis and Peebles, 1983. And they had one thing right, and they had one thing wrong in their paper. What they got right was the two-point correlation function of the galaxies. What they got wrong was the mass of matter in the universe. They had it too low. And that was because of a mistake that they made. I discovered that in the first paper I wrote with Rachel Somerville, where I had always suspected there was something wrong. We tried to reproduce this other result in the Davis-Peebles '83 paper, and we could not do it. Even working with the same data they had worked with.
So, I arranged for Rachel Somerville to go talk to Marc Davis at Berkeley, and she told me that he asked all the same questions I had asked. "Did you try this? Did you try that?" And she'd already tried everything he'd asked about, and she told him about the results in detail. So, he finally went and found the original computer code he'd used and a computer that could actually read it. And he started to read through it, and Rachel said his face turned completely white, and he asked her to leave the room. And he called her back in a while later, and he said, "I found the mistake." And Rachel, Marc, and I ended up writing a paper together where we showed how to do it right.
In our 1984 CDM paper we had the CfA data, and that was the first large scale distribution of galaxies, the sizes of voids, the clumping of galaxies, the number of galaxy groups and clusters, and we compared this with what we predicted from our semi-analytic models, where we had the large scale structure in linear theory, and then we worked out, assuming just simple spherical collapse, what this would predict for galaxies of different masses, and for groups of galaxies, and so on. And for the first time, we actually had data that we could compare it to from the CfA1 Survey. And everything fit. And it fit especially well if we assumed that Omega matter was quite low. We worked out Omega matter of 0.2 and 1, and we said 0.2 is a much better fit to the data. And we now think that the correct Omega matter is 0.3.
I remember Jim Gunn visited just as we were finishing this paper. Gunn is, like Sandy Faber, one of these amazing triple threat astronomers—observer, instrument builder, and theorist. Sandy and Jim, I think, are the only people that have done this, where Sandy is, of course, a great observer, having made major discoveries in observational astronomy, she's done important work in theory, including this paper with me, and she's built one of the most important instruments, the DEIMOS Spectrograph, an extremely difficult thing. Fantastic success. Jim Gunn, of course, hand-soldered the detector for the Sloan Digital Sky Survey, and he also, in his basement, built Wide Field Camera 1 for Hubble Space Telescope. That, of course, got replaced by Wide Field Camera 2 with the corrective optics.
Gunn visited us, and I remember our joy in telling him about what we were doing in the cold dark matter paper. This would've been late '83 or early '84. The paper was submitted in '84 and published later in '84 in Nature. But we were just amazed that it looked like we really had the right idea. And of course, we did, basically.
Last question, looking to the future. In light of your miraculous recovery from pancreatic cancer and the amazing things that immunotherapy can now achieve, it's probably a very personal question, but I wonder if you've ever thought about the lease on life that you've been afforded and how however many extra years you've been blessed with might allow you to understand things that wouldn't have been possible had the diagnosis gone a different way.
I don't know how to answer that. I've continued to be active, and I always try to work on what I think is the most exciting and interesting thing to work on. So, I've continued to do that, and I've continued to have good graduate students to work with and wonderful colleagues like Sandy Faber, Avishai Dekel, Anatoly Klypin, and several of my former grad students. And so, I'm just incredibly lucky in all these respects. And of course, also, Nancy and I continue to have a wonderful relationship, and I must mention, also, our amazing daughter, Samara, who uses her stage name Samara Bay. Samara is one of the top dialect coaches in Hollywood. But for years now, she's also been coaching people to give talks for other purposes.
For a couple of years, Samara worked with the Alan Alda group training scientists to give better presentations, and her work on that was part of my inspiration for my efforts that I mentioned to you earlier with Sigma Xi. But in the last couple of years, she's especially worked with Move On helping female candidates for office, including a number of people running for the House of Representatives. She was their go-to person for that. And now, she's been training other people to do that. She did a podcast with something like 25 episodes called Permission to Speak on iHeartRadio, where she interviewed a lot of interesting people, including Michelle Obama's speechwriter, Sarah Hurwitz. And Samara got a big advance, more than half a million dollars, from Penguin Random House for her book on how to use your voice to get what you want, but especially aimed at women and other people whose voices are usually denigrated. Every woman politician is criticized as being shrill. Because the female sound of power doesn't fit with most people's expectations. Samara has made it her goal to change that. She's amazing. If anybody can do it, she can.
We also have a wonderful 6-year-old grandson. One of the positive aspects of this immunotherapy treatment that I was in is that I had to be in Los Angeles for eight days every month. And the hospital where I was treated is only a 20-minute drive from Samara's house, which my wife and I bought for Samara and her husband by trading a house we owned in Santa Cruz. So, they're renting it from us. It’s a house with an extra bedroom and bathroom for us grandparents. So, that gave us a chance to spend a lot of time with Samara, her husband, and our grandson. And so, we had a chance to sort of see Samara much more in action than we otherwise would've. During the height of COVID for about a year, we didn't drive down together, I just drove down by myself, and I stayed in an Airbnb in case I had been exposed to COVID at the hospital. And I just got together with Samara and her family for walks in the woods, that kind of thing. And then, starting last April, we started to stay with them again after everybody, except for our grandson, was vaccinated.
So, I've really been incredibly fortunate in so many different ways, including, of course, recovering from pancreatic cancer. I've sort of asked myself, "Is there some big project I'd like to take on?" And I haven't really thought of one. I've been sort of writing up a little timeline and essay on what I call The Luckiest Astrophysicist. Because I've been so fortunate in so many different ways. I'm not sure who this is going to be useful for. I asked our literary agent some time ago if there was any market for a book like that. And he said, "No."
But you're aware of how fortunate you are, and that's key to everything.
Yes, absolutely. And throughout my life in so many different ways. It's really been quite amazing.
Joel, it's been so fun spending this time with you. I'm so glad we were able to connect, and I'd like to thank you so much for doing this. I really appreciate it.
Well, thank you. It's a wonderful opportunity.