David Schramm

Lightman:

I wanted to start with your childhood and early experiences. I want to get some idea of influences that you may have had as a child, either from your parents or from other people or books that you read that got you interested in science.

Schramm:

I can't think of any specific thing that got me interested in science other than it just was a natural interest. My father was in real estate. My mother was a librarian. For some reason or another, I found things technical rather fascinating. I liked model airplanes and things like that. Initially, at about the age of ten, aeronautical engineering seemed interesting. Later, pure science began to catch my attention. I used to read the encyclopedia. That was my amusement when I was in about third or fourth grade. I remember the science sections were the things I found most fascinating — atomic energy, nuclear physics, things like that. I would do all my work in class very, very quickly and then had nothing else to do except [for] reading the encyclopedia.

Lightman:

Did you ever talk to your parents about science, or did they ever talk to you about science?

Schramm:

Not directly. They were very encouraging [of my] reading and studying. They were very supportive, but they had no specific science interests themselves. During high school, although I was always interested in science and involved in various science things, my main function was athletics. I played football. I won the state wrestling championship. I ran track, and put the shot. Athletics were my main activity in high school. I never studied. It seemed as if everybody I went to high school with went off to local state schools on athletic scholarships. I, without studying, got 800s on the SATs, so I went to MIT. There was a divergence at that point. But during high school, I was a standard mid-western boy going to public high school, playing football and wrestling, and all those kinds of things.

Lightman:

You said your father was in real estate and your mother was a librarian. Do you remember any hobbies that they had when you were growing up?

Schramm:

My father used to fly a small plane. Then, when he was having some health problems, he switched to flying model airplanes. I did that with him for a while, up until junior high school, at which time I was a juvenile delinquent for a while and then went into athletics. That took over my interests through high school.

Lightman:

Do you remember having any interest in astronomy at this age?

Schramm:

I always was interested in great questions: cosmological questions, theology, the theology-science interface. In fact, not so much on a scientific side but on the religious side, my family was very mixed up. My father was Jewish. My mother was Christian Science. Then my father converted from being Jewish to a branch of Christianity, a fundamental Christian group. My mother was Christian Science only because her mother was Christian Science. Then my grandmother died, and my mother shifted into a more deistic theology. So the religious debates in the family were fun, and I was always somewhat of a skeptic. Having seen all the different religions all at one time, I became skeptical of organized religion. [I] was looking to the more philosophical sides of things, including scientific things, to try to establish my own theology in some sense, a cosmology. So it was more not so much from the mechanistic side of science but rather the...

Lightman:

Philosophical.

Schramm:

Yes, philosophical side.

Lightman:

And you remember that as far back as high school.

Schramm:

Oh, earlier than that. In grade school I had an interest, because I remember the upheavals that were going on in the family, and I just had some initial interest in that. But certainly by high school, it was very strong, or even by junior high.

Lightman:

When you mentioned fundamental questions, can you mention one or two of them?

Schramm:

Things like everybody thinks of at one time or the other. I probably just dwelled on them longer. Questions like: "Why are we here?" "Where did we come from?" "Where are we going?" "Where did matter come from?" "Where did space come from?" The same questions I'm still asking. [Laughs]

Lightman:

With different tools. Let me ask you a few questions about your education. At MIT, did you start off as a physics major or as an engineering major? How did you begin?

Schramm:

[When] I started off, actually, I wasn't sure whether I was going to go into math or physics. I was very excited in high school about math, just because it was very easy for me. As I say, I got a perfect score on the SAT and all that sort of thing. I would go [though out the year] without ever missing a problem on any math assignment, without ever studying. Math just came very easily for me. So I was thinking that maybe that's the direction to go, but then the philosophical side of physics I always found [appealing]. The articles that appealed to me in the encyclopedia were frequently the things about physics, although history was also fun. So I wasn't sure whether I was going to go into physics or math. I really knew nothing about what it was to be a physicist or a mathematician. In the area I grew up in, a standard middle-class suburban neighborhood, there were no scientists. The closest to people doing science were the people who were engineers at McDonnell Aircraft, and they were on the engineering side — very applied. That was my only contact with [technical] people. The parents of some my friends might [have been] in that area. The most educated people in my family were two uncles who graduated from law school. In fact, how I ended up at MIT was due to their influence, because I knew nothing about any of these schools. The guidance counselors [in my] high school were not very sophisticated. So, it was my uncles, both who mentioned, "If you're interested in science you should go to MIT." In fact, in hindsight, I realize that I would have done just as well, if not better, had I gone to a place like Harvard or Stanford or Chicago, where there's an entire university rather than just one discipline. The science at Harvard is as strong as the science at MIT. But at that time, mainly because of what my uncles had said, I went to MIT.

Lightman:

When you were at MIT, do you remember any particularly influential teachers that you had?

Schramm:

Yes. MIT was a tremendous experience for me. Here I was coming from an environment where academics really were not what we did in high school. It was athletics and chasing girls. MIT was very different for me. The first thing that happened was I came there, as I said, not knowing whether I wanted to go into math or physics. All students [with an 800 on the math SAT] were put into one class. It must have been a class of over 50 students. So suddenly, instead of being unique, I was one of 50, and a lot of these people had had much more math and science than I had. It really showed in the math. There are two reasons I didn't go into math. One was just the simple [point that] it looked as if the competition was [hard]. They were all ahead of me. I was starting out from behind. It didn't seem like they [the other students] were as far ahead in the physics area. The other was the realization that math was more of a game. It was intellectual masturbation rather than playing with the real universe. Physics appealed to me because I was also doing math but applying it to something real, as opposed to just playing games with it. So that was my reason for leaning towards physics. I guess one other thing that may be of relevance is that between high school and going to MIT, I got married, which was not unusual for the area I came from. But it was very unusual for MIT. So I was married as an undergraduate and that would have made me a little more serious and focussed during that period.

Lightman:

Did you know anything about cosmology at that time?

Schramm:

I can remember [that] when I was asked by some fellow classmates about what areas in physics I was interested in, my standard response was cosmology.

Lightman:

Oh, really?

Schramm:

Without knowing exactly the details. Again, [this was] left over from discussions in high school of what was interesting for me. But I can remember one conversation with a friend of mine who was in biology — in fact, [now] a professor of biology at MIT. I can remember mentioning that [cosmology] was what interested me, but I don't remember the nitty-gritty of why.

Lightman:

Had you read any popular books or magazine articles about cosmology?

Schramm:

Oh, yes, bits and pieces here and there.

Lightman:

You don't remember anything in particular?

Schramm:

Nothing that stood out. What really got me going in that direction, though, was actually a course by Phil Morrison at MIT. It was a course on galaxies, and Phil is a fantastic teacher. There's such a depth of knowledge there. I remember getting very excited about this very stimulating subject. What I mainly did, though, while I was at MIT, science-wise, was work in the nuclear physics lab. From my sophomore year on, I spent most of my time there in the nuclear physics lab. Either there or on the wrestling mat. I was also wrestling. Those were the two main interests that I had. [One was] playing around in this nuclear lab and discovering what science was all about. It's so different than what you learn in high school or what you read in the encyclopedia. You discover what scientists really are, what they do, that they are real people and have a good time doing science. It was an experimental lab, and we were doing real experiments. I published two papers [on nuclear physics] before I graduated. As a result, I knew a lot of nuclear physics by the time I graduated. But also, with Phil Morrison's course I got very excited about astronomy, and that's how I ended up going to Caltech for graduate school. I wanted to do nuclear physics, and I wanted to do something that was astrophysical because of Phil's course. When I mentioned that combination to people, they said, "Oh, you have to go work with Fowler." So that's how I ended up at Caltech. The actual thing of going to Caltech and working with Fowler didn't occur until my senior year, when it came time to apply. Up until that time, I, like most undergraduates, thought that my undergraduate institution was the place to be. The only place one would ever think about going to graduate school would be the same institution. As it came time to decide which graduate school to go to, [I remember] going in and having some discussions with various people at MIT about what I would do in graduate school and mentioning my interests. People were very honest and very nice — in particular, Icko Iben, who was there at that time. I remember going in and talking to Icko, and he said if that was my interest, I should go to Caltech. He said he would be happy to have me work with him at MIT, but that wasn't what I should do. I was very pleased to have good advice. I knew very little at that time about the relative differences, and Icko was a great help. But also a number of other people — Lou Grodzins, who is a nuclear physicist there, and Eric Cosman — were all very supportive of the Caltech idea. I thought they gave me good advice, and I ended up going to Caltech.

Lightman:

When you went to Caltech, did you take a course in cosmology, or in general relativity?

Schramm:

Well, I was in the physics department. I was coming from the physics direction nuclear physics. One of the things that I remember [about Icko] [was that] he also was a wrestler. So there was some interaction there that was very positive. When I came to Caltech, I was initially looking into the space physics program also, and I was taking some things in applied math. But then I settled in to working with Willie [William Fowler]. I took Willie's course, and Feynman's. I took quantum electrodynamics from Feynman, which was great. He'd come in and play bongos. I didn't take any astronomy. I sat in on one course on stellar evolution. There was no formal cosmology course during the time I was there. I took physics.

Lightman:

There was a general relativity course.

Schramm:

Yes, I took general relativity with Kip [Thorne].

Lightman:

He must have talked about cosmology.

Schramm:

Yes, in fact, [cosmology] was certainly in that course. I took that and [it] was very good. [When] I took it, Misner, Thorne, and Wheeler^[1] was still in the note form, and so we used notes. That was fun and I really enjoyed it. I remember that Kip had us do a paper, and I did my paper on big bang nucleosynthesis. In fact, it was shortly after [Robert] Wagoner, [William] Fowler, and [Fred] Hoyle.^[2] The first time I really looked into that with seriousness was for Kip's course.

Lightman:

When you had that cosmology section in his course and you were exposed to different possible models, do you remember having a preference for any particular kind of universe?

Schramm:

I was a student in the late 1960s, and it was shortly after the background radiation was discovered,^[3] and so there was already a leaning [towards the big bang model]. Certainly the big bang was what I preferred. I guess my prejudice at that time was [that] "closed models looked prettier. The idea of oscillating universes was sort of an initial prejudice. Kip's course also leaned in that direction. John Wheeler leans in that direction. Actually, I remember in the first or second year of graduate school thinking that that looked like a very nice model. [It was] my initial prejudice, which I don't have anymore.

Lightman:

Do you remember why you liked it?

Schramm:

Oh, Willie was, and still is to some extent, a steady state lover. The idea here was that at least you could avoid the problem of an origin to time if you would have many oscillations.

Lightman:

Was this [your own] feeling, or was it something you had read about?

Schramm:

Oh, more just my own...

Lightman:

Your own feeling. Because the famous Dicke, Wilkinson, and Roll paper^[4] in 1965 begins with the statement that an oscillating universe alleviates the necessity of specifying a beginning.

Schramm:

Actually I had not read that paper at that time.

Lightman:

But that was something that just appealed to you.

Schramm:

Right. [It] just appealed to me on some sort of aesthetic [ground]. [As I said], I certainly favored big bang over steady state, but by that time there was already the background radiation. Then there was this naive thing I felt about aesthetically liking the oscillating [model] at that time. But I hadn't read the paper. It was just sort of a bias. Then, as I said, in Kip's course I did the [paper on] big bang nucleosynthesis, and it was a way of coupling some interesting cosmology with my nuclear physics. My thesis was actually on nuclear chronology and dating the universe. Willie liked to have all of his entire students do an experiment, and so I did an experiment with Gerry Wasserburg, who's in geophysics. In fact, I ended up as a graduate student spending far more time with Gerry than with Willie, because experimental work takes longer than theoretical work. But in the course of doing work with Gerry, I also did some theoretical work with Gerry on nuclear chronology, as well as doing some of the measurements with a mass spectrometer and really getting into laboratory techniques and learning how hard it is to get real numbers. It made theory much more appealing. I guess one of the educational aspects of that is you realize that when you're doing an experiment you can only do one thing at a time. In theory you can do a lot of different projects, follow through a lot of interests. I had too many interests to...

Lightman:

To stay in experimental work?

Schramm:

Right. Gerry and I did some work^[5] on chronology which turned out to be very nice. We showed that the chronology equations could be separated. When you look at all the equations that apply, they all appear to be some big non-linear problem, which looks very difficult to solve. Previously, the only way to do it was [with] some sort of numerical technique.^[6] We were able to do mathematical tricks to separate the equations and [get a] simple solution. So one could get a physical feel of what was happening in chronology. I think it was [previously] too model-dependent, where you have to plug in a specific model. By doing the separation, we were able to then determine model-independent things. That was really just re-applying straightforward mathematics to these things [in a way that] people hadn't done before. That turned out to be powerful because then one got a feel for what was driving the chronology. Prior to then, we were really just plugging in numbers and seeing what came out. [This work] really [gave one] a much better feel. It was also brought on by some recent discoveries in mass spectrometry with regards to new chronometers. So those new chronometers all of a sudden gave a rebirth to nuclear chronology.

Lightman:

Was osmium one of those [chronometers]?

Schramm:

No, that was later. Actually, the osmium idea was one that Don Clayton had had back in the mid-1960s,^[7] but it didn't really become a chronometer until some measurements were done in the 1970s. But iodine 129 and plutonium 244 both came on line just at this time. So the first calculations using all of these [elements], not just uranium and thorium but also plutonium and iodine, [came at this time]. That was what made it appear so complicated, and then that's what motivated us to do separation techniques and [to find] the moments. You really could determine what was happening, what was driving the system.

Lightman:

Let me skip ahead a few years up to the late 1970s and ask you a few questions about the merger of particle physics and cosmology. [Regarding your work on] limiting the types of neutrinos from helium abundance,^[8] do you remember what the motivating factors were?

Schramm:

Yes. That's an interesting thing. I think in many ways that [work] was the birth of this whole connection — particle physics and cosmology — because it was the first time when we were able to take something from cosmology and make a statement that was relevant to particle physics - as opposed to just trying to use it. [In order to explain] how I got into that and was even aware of the problem, you need to go back a little bit. In the early 1970s, there [were] two main areas I was working on. One was big bang nucleosynthesis, and establishing the relevance of deuterium,^[9] showing that deuterium couldn't be made in any other way but cosmologically. The other was gravitational collapse and supernova models.^[10] And that latter one is the only place in astrophysics where...

Lightman:

Where neutrinos...

Schramm:

... where neutrinos and particle physics came in. Because of my activity in that, I was following what was going on in particle physics, because that affected our input into these collapse calculations. In particular, when neutral currents were discovered in the lab, I was one of the first people to be involved in putting them in the stellar evolution codes and [considering] what the role of neutral currents [were] in calculating all the reactions and so on, and realizing that you produce all kinds of neutrinos in [stellar] collapse rather than just one kind. But that got me acquainted with the particle community on a firsthand basis, and particularly [with] people doing neutrino physics, and [with] the kinds of important questions people were asking in the particle world.

Lightman:

So those two ingredients came together.

Schramm:

Yes, so then those two ingredients came together on this. How the calculation actually ended up being done, though, was the result of Jim Gunn's visiting Chicago. We had Jim as a visiting professor for a few months. Jim asked me what I was doing, and we just started talking. I was talking about both the things I was doing, and in the course of the discussions with Jim, we outlined this argument. [It happened] just by talking to Jim, and [his] asking the kind of clever questions that Jim asked, and the [interaction] back and forth. So we sketched out this thing on — Jim's blackboard when he was visiting Chicago. Then I was starting to write it up, and I went to Aspen that summer, and Gary Steigman was there. I started talking to Gary. Gary and I had shared an office when we were both postdocs at Caltech. I was a graduate student and a post doc there, and we were both postdocs with Willie. We had never written a paper together. We shared an office, but we never did any science together. We were in different job areas. When I got to Aspen, I started talking to Gary about this thing that Jim and I were doing. And [I said that] I was writing it up. Gary had evidently done almost the same thing. So instead of writing two separate papers, we merged it together, and it became one paper. That was the summer of 1976, and it got published in 1977. So that's the origin of that paper. I think it was my interactions in the particle community for the supernova problem, the way we worded things that caused a break in how cosmology dealt with particle questions. If you look earlier, for example in [James] Peebles's book^[11] or at a paper^[12] by [V.F.] Shvartsman much earlier, [you see that] they were talking in terms of energy densities and were [commenting] that you could limit the energy densities of various kinds of particles, [but they were] never getting down to the real questions that a particle person asks.

Lightman:

[For example], what was the mass of the particle?

Schramm:

Yes, mass number, number of types, and the kind of real particle [physics]. Up until that time, most the people who were doing theoretical cosmology were relativists, [who] were thinking in terms of the stress energy tensors and the terms to put in. Whereas this approach of coming from the neutrino side [involved] asking the questions that a particle physicist is concerned with. He's concerned with how many flavors there are. [That] number [involves] formulating and parametizing things in a different way. It was that then that really made the connection [between particle physics and cosmology]. [Previously], cosmologists [were] talking in terms of something that a particle physicist couldn't care anything about. A term in the stress energy tensor is irrelevant unless you do a bit of algebra. But talking explicitly about a number that [particle physicists] were interested in [was different]. In fact, how many types of particles there were is one of the most important numbers for them. At the time [of our work], the particle community was thinking [that] every time you go to higher energy you get another new flavor. In fact, this was right after the November revolution, when charm had been discovered,^[13] and right around the same time as when the upsilon was found.^[14] So it looked like whenever you go up to higher energy, you find new flavors. The whole lore in the particle community was that this is just going to go on forever. At that time, we were the first to say, hold it. Cosmology wouldn't work that way. Our first limit, which is not as strong as it is now, was around 7 or so [types of neutrinos]. Gradually, we found it would be down to two to four. At the time, even saying that it was a small finite number as opposed to thousands was considered almost blasphemous. People were [saying] "Oh, we can just keep having families on up the line." Furthermore, somebody was making an argument from an astronomical thing about fundamental physics. That had never been done. The closest to it was probably when Fred Hoyle argued for an excited state of carbon.^[15] Then [experimentalists] went and measured and found it. But that was an excited level, and there are many, many levels. It's not something truly fundamental. Whereas here was something that was of central importance to' a real physics question as opposed to just one more level in the nucleus or one more level of an atom. We got a lot of flak from a lot of people, saying, "Oh, it's crazy to say something like that. You are talking about numbers of a few and astrophysicists are usually only right to a few orders of magnitude." Then, also, the trend at the time was toward more and more [families of particles]. Now, as time has gone on, people say, "Oh, of course, there's only a few neutrino [types]." It was certainly not the case at that [earlier] point.

Lightman:

Other things joining particle physics and cosmology happened around that time too. Do you think that for some reason the time was right?

Schramm:

Yes. This particular thing was one of the [first] to start getting more communication between the fields. There were a few other things [in which] we started to put constraints the particle types. I did some stuff^[16] with Ben Lee just before he died. Lee and Weinberg put limits on neutrino types.^[17] Jim Gunn was involved with that as well. Rocky Kolb was a graduate student at Texas, and he did some work^[18] at the same time, starting to constrain particle properties from cosmology. Another thing was putting upper limits on masses. A little earlier, actually, Ram Cowsik had done some things on limits on neutrino masses,^[19] but at that time it didn't have much of an impact. Now Gary had also done some stuff on the antimatter-matter business.^[20] So there had been already some interaction [between particle physics and cosmology]. [There was also the work] on the maximum temperature — [R.] Hagedorn, [John] Bahcall, [Steve] Frautschi.^[21] None of [that work] really had a big impact, but by the time the 1970s came, the number of such [studies], was increasing. Also, the kinds of statements that were being made were more specific to the questions that particle physicists were interested in, rather than just vague numbers that really didn't affect people that much. Then, also, the big bang was really getting established by then. By the mid-1970s, Paul Richards' measurements^[22] had established that the [microwave background radiation] curve turned over. Big bang nucleosynthesis had come of age. It [not only] fit the helium [abundance], which is what people were doing back in the 1970s, but we were getting deuterium. That was powerful. [Calculations] were really fitting the rare isotopes, not just the 25% helium. Big bang nucleosynthesis was working very well. [To establish] those other isotopes took a lot of the work that we did in the 1970s. In fact, back in the 1960s, when the big bang nucleosynthesis work of Tayler and Hoyle^[23] first came out, and then [that of] Wagoner, Fowler and Hoyle^[24] and others, — although they calculated the [expected abundances] — the community (and the authors) ignored them because Willie had papers^[25] saying that you made all these other [elements and isotopes] in stars. It was only our later work that showed that you couldn't make them in stars and that the big bang was relevant. Not only was [it] relevant, but it was right on the money. I think a lot of that [work] really established that the big bang worked very, very well, and big bang said that it worked all the way back to [one] second [after the big bang]. If it worked all the way back to a second, that gave you confidence that you could extrapolate further back. Whereas, in the 1960s, we weren't even sure it was big bang or steady state. So it really took establishing that it was a big bang and that the big bang was working very well on the microphysics that we knew — nucleosynthesis at an [energy of an] MeV— and a time of [one] second. That enabled one to have confidence that we could extrapolate further. At the same time, in the particle community came GUTs [grand unified theories]. [Howard] Georgi and [Sheldon] Glashow were talking about SU(5),^[26] which is important at levels of 10¹⁵ GeV, way beyond what you could get in an accelerator. Where else do you have any experimental or observational consequences of GUTs? Well, it's the big bang. So you had a model that was finally saying that we seem to understand things back. To early times, and now you [also] had a motivation, coming from the GUTs side. So [we had the] Georgi and Glashow motivation of GUTs. And [we also had] the establishment of the big bang, [which] for these purposes was done with big bang nuc1eosynthesis more than [with] the three degree radiation — [big bang nucleosynthesis] arguing you knew about [the universe] at very early times, not just at 10⁵ years, when the three degree radiation [was emitted]. Then everything was in place. I think the things that really established the connection and made the growth go explosively were three key things all within about a year of each other. One was that GUTs enabled you to understand a long-standing problem in cosmology and that was the photon-to-baryon ratio. Again, [Andrei] Sakharov had done it earlier in 1967,^[27] but nobody paid any attention to it. When it was re-done by a half dozen different people at different locations,^[28] semi-independently and with the context of GUTs, you had a real model. Suddenly, there was success in answering a problem that had not been answered before — [a problem] in cosmology [with a solution] using particle physics. So that provided the other half of the relationship, to make it really symbiotic. Previously, we were saying that from cosmology we can tell you something about the number of particles, which is a very fundamental thing. Then the particle people were saying that we can tell you how you got the baryon-to-photon ratio, which was a fundamental unsolved problem in cosmology. So you really had information going both directions, rather than in the earlier decades, when astronomy was a parasite on physics, just using physics without giving anything back. I think that really cemented it. Then shortly thereafter were two other critical things. One was [that] with baryosynthesis, then you could have "inflation." Alan Guth recognized^[29] that the Higgs field from GUTs could drive inflation. But that had to occur after we understood baryosynthesis, because if you inflate a universe and you don't know how to make baryons, you've made it empty. So you had to know that you could make baryons at the end of inflation. Other people before that time, including my former student Demosthenes Kazanas,^[30] had talked about geometric models and de Sitter phases. And Richard Gott,^[31] among others, did later. But it took [the ability to] make the baryons at the end that made inflation a viable model- as well as Guth's answering several problems simultaneously. The next thing that occurred around the same time was curious. Two wrong experiments helped stimulate the whole thing of non-baryonic dark matter. These were the neutrino mass experiments. In fact, the year before those experiments were done, Steigman and I had written a little essay^[32] for the [Babson Foundation] gravity awards, talking about how the big bang nucleosynthesis numbers tell you that omega in baryons is small, and yet, for a variety of arguments, one probably thinks that omega is much bigger than that. We said okay, there probably is non-baryonic dark matter and the best bet is probably neutrinos. [That is] the least radical [hypothesis] is to just put a small mass on the neutrino. We looked at the numbers to estimate what it is and we said 10s of eV, and then shortly thereafter...

Lightman:

There was a Soviet experiment?

Schramm:

There was a Soviet experiment^[33] plus Reines^[34] claimed [to have measured a neutrino mass] also. The timing was great because those experiments came out right as the gravity award prize was about to be determined for these gravity essays. So we had a great press agent, and we won the prize for that. Both experiments are probably wrong, but the timing was [fortuitous]. Then, those measurements motivated a lot of people like [J.] Bond, [Joe] Silk, and [George] Efstathiou^[35]and a number of others to really follow through on neutrino masses as a possibility. There were lots and lots of papers at that point, talking about non-baryonic dark matter. We happened to be slightly ahead of it, but [we] hadn't really worked out all the galaxy dynamics because it was very speculative. Whereas afterwards somebody had something concrete to really do hard work on, rather than just wave [their] hands. That really established non-baryonic dark matter as a legitimate area to work on. Later trends went from hot dark matter to cold. And then also those experiments have ended up not being right. But that's somewhat irrelevant because you could still have a tau neutrino [with] a mass of 30 eV as the dark matter. Furthermore, the arguments of big bang nucleo-synthesis that Gary and I used in our essay are still the main arguments for why people discuss non-baryonic dark matter.

Lightman:

Let me ask you how you've reacted to some discoveries in the last 10 years, both theoretical and observational. First, let me ask whether you remember when you first heard about the horizon problem? You don't have to tell me exactly.

Schramm:

Yes, I can sort of remember when I first became aware of it.

Lightman:

Was that in graduate school or was it later?

Schramm:

Yes, during graduate school, when I was at Caltech.

Lightman:

How did you react to it? Did it bother you a lot, or did you think it would be solved by some fairly straightforward means? Do you remember how you thought about it?

Schramm:

I can remember being bothered by it. I can remember thinking that it required some rather special initial conditions. It wasn't enough of a bother for me to say that I didn't believe the big bang as a model. But I certainly recognized that there was a problem, and it required setting the initial conditions [to be] very smooth. I had no idea how to do that.

Lightman:

Did you think that might be a solution to the problem, having very smooth initial conditions?

Schramm:

That's really all inflation does. It makes the initial conditions very smooth. It's just a mechanism to do that.

Lightman:

Well, right. Were any people talking about the universe beginning with the right conditions to...?

Schramm:

Yes, certainly when I first heard about the problem" that was the standard way of talking about the solution to it. It meant that the universe started out very, very smooth. When the metric was laid out, it was laid out smooth.

Lightman:

And how did you feel about that explanation?

Schramm:

I accepted that.

Lightman:

You accepted that.

Schramm:

It probably didn't bother me as much as it should have. But I recognized that you needed to fine tune things. My reaction was that sometime I'll have to think about that.

Lightman:

Did you get more bothered by the horizon problem as time went on, or did you just tuck it away in your mind as something that you knew was a problem but would come back to?

Schramm:

I can remember, in fact, in the 1970s when I was giving talks on cosmology, I would bring it up. It was something that I recognized was a problem with the big bang.

Lightman:

I see.

Schramm:

I can remember at various talks bringing it up and just mentioning that this requires something about the initial conditions. Usually, I would say these things more in semi-popular lectures rather than [in scientific talks], just mentioning that there are some problems and that we have to lay out the initial conditions very smoothly. I had no explanation for it, but I remember mentioning it.

Lightman:

Did the horizon problem take on any different significance for you after the inflationary universe model came out?

Schramm:

Yes. Then there was a mechanism to solve it, a mechanism to layout the initial conditions. Then certainly I put more focus on it. I would always mention it in talks, as opposed to just occasionally mentioning it, because there was a solution. You don't always show your warts, but you can always show your medals, and here was a nice successful... So I'd mention [the horizon problem] with great frequency at that point. In fact, the same thing could be said of the...

Lightman:

The flatness problem?

Schramm:

Yes, the flatness problem.

Lightman:

Yes, I'm going to ask you about that in just a minute. You're anticipating my questions. Before I get to the flatness problem, let me ask you about inflation. When you first heard about the inflationary universe model — many of us heard about it before the paper was published, of course — how did you react to it?

Schramm:

Actually, even before Guth inflated, I had talked to Henry Tye, who had been working with Alan on phase transitions in the early universe,^[36] trying to solve the monopole problem. In fact, Jim Fry and I had done some things on the monopole problem.^[37] So I was well aware of the monopole problem. I was aware of Tye and Guth's efforts at trying to do multiple phase transitions and so on. So I was quite up on what they were doing. Then I heard about Alan's inflation idea, and suddenly everything just fit together. It immediately hit me that he's found it. There's the solution, and it all fits together. I was an immediate convert, I guess. I had heard bits and pieces of it. I can even remember the initial problem that the phase transition didn't go to completion. It just seemed to me that it was such a beautiful solution to things, there was bound to be a solution found to that problem. That was my thought — that it was all going to fit together and that it had to work. There are still a lot of nagging problems with the fluctuations, but I remember right away, I was enthusiastic about it. [The inflationary universe model] was able to solve a number of things that had been in the back of my mind as problems with the big bang. All of a sudden, in one fell swoop, they went away. I was very pleased with that.

Lightman:

Why do you think that it [the inflationary universe model] has been so widely and so quickly accepted by the community?

Schramm:

I think it's pretty. With one idea you solve several problems at once. It really gave [a boost to] this interface of cosmology and particle physics, capturing some of the most exciting aspects of both. It set the initial conditions for the big bang, so it enabled all of the good stuff about the big bang to go there. You don't have to throwaway anything. You can remain a good, firm big bang lover. [Inflation] solves some of the problems, some of the things that used to be nagging embarrassments that you put under the rug. Now you didn't have to sweep them under the rug. So it enabled you as a big bang lover to be even more of a big bang lover. It also enabled you as a particle physicist to say, "Oh, particle physics can do something that's useful to somebody else. It can solve one of their problems." That was not occurring [before]. In fact, particle physics frequently had the problem of justifying its existence to the general public. Why does anybody care about how many quarks there are or something like that? Here is particle physics able to solve somebody else's problem. And [inflation] didn't solve [the problems] completely. So there was still work to be done. I think that was also an important aspect. If somebody solved something and there's nothing else to be done, you're not going to create an industry. But inflation did — and still does — have some problems, some loose ends. It's not all fixed. There is still room for people to do work. So lots of graduate students, lots of post-docs, and faculty people could then work on this problem and make contributions. It wasn't just a one-fell-swoop and it's solved. Not like Fermat's theorem.

Lightman:

I guess in mathematics what tends to happen is the method of solution opens up new branches of mathematics.

Schramm:

Yeah, right. With Fermat, also you might still continue to work on the problem because you might hope to find some cleaner or prettier proof.

Lightman:

You sound like you're pretty convinced that the inflationary model has a good chance of being right. I don't want to overstate it.

Schramm:

Well, the way I would word it is I think the inflation paradigm is right — the idea that we went through some de Sitter phase very early. It could very well be that the whole mechanism and the details that we're currently talking about are wrong. But my feeling is that the inflation paradigm [is right], that there was an early de Sitter phase. Richard Gott argued on geometric grounds that you needed something like that, at a comparable time and even before Alan was inflating. It's just [that] Alan Guth gave us a very nice mechanism to do it. But I'm not sure that Alan's mechanism is the right one. It runs into problems. Maybe as we learn about the "theory of everything" or some super-gravity or string theory, we'll find that inflation occurs in some different way. Rocky Kolb,^[38] for example, has talked about [the idea that] maybe out of ten dimensions, three inflate, back at Planck scales or something like that. That may be the way it occurs, which is quite different [from] the Higgs field driving inflation, [which is] what Alan was initially talking about. Now Andre Linde has shown^[39] that any scalar field will inflate. You don't need any sort of special thing. It's very easy to get inflation. So we've gone from where we didn't know about inflation to now, where all you have to do is propose a scalar field at early times and you get inflation. So it might not be Guth's mechanism in detail, but just the idea that something from particle physics or something from the early universe drives a rapid expansion phase, a de Sitter phase. That is the key. The details of that expansion I don't know and I don't buy. In fact, what I can say is I'm pretty sure that everything we're working with right now is not the way it happens, that we're leaving out some important ingredient. Right now, it's a little too contrived. You have to fine tune the parameters too much to avoid the fluctuation problem. In fact, I remember when we first started focusing on the fluctuation problem, Alan and I were giving lectures together I on cosmology. We went through a series of lectures. We were a team going from London, to Moriond, and then to Oxford, at various conferences. It was in 1983, right after new inflation came out. I remember asking him about the problems of fluctuations — that was before they were investigated.

Lightman:

Now we've got to explain.

Schramm:

Now we've got to explain [the fluctuations] because [inflation] wipes everything out. In fact, the worry I had was you were going to wipe them out and how were you going to have any fluctuations at all? And, of course, when Alan and others looked at it in detail, they discovered the opposite problem. [Because] of quantum fluctuations, you actually end up making fluctuations too big. But I remember talking to Alan about it and motivating him, I think, to look at the problem. And he got the opposite conclusion of what I was worried about.

Lightman:

Let me ask you about the flatness problem, which we started to talk about. Can I ask you when you heard about it the first time, if you can remember?

Schramm:

I guess it was a Dicke and Peebles argument.^[40] I'm not sure whether it [was in]a talk that I heard Bob Dicke give or somebody referring to a talk that Dicke gave.

Lightman:

Do you know approximately when that was?

Schramm:

Mid to late 1970s.

Lightman:

Mid to late 1970s?

Schramm:

Yes. Before Guth, but after graduate school, sometime in the 1970s. In fact, in various vague ways, I can remember hearing it maybe even earlier. I remember somebody — I forget who it was — went through a similar argument on the blackboard one time. We were talking about the big bang and mentioning this one problem. In fact, it was back in 1974 when [Richard] Gott, [James] Gunn, [Beatrice] Tinsley, and I did a paper^[41] on the open universe. I can't remember who it was, but I can remember when we talked about everything fitting remarkably well at an omega of 0.1. In fact, for baryons everything still fits remarkably well for an omega of 0.1. I can remember somebody — I can't remember who it was — going through what is now known as the flatness argument, saying, "Well you realize this requires some very special fine tuning." It might have even been. Jim

Lightman:

Yes, I think he was talking about it at that time.

Schramm:

Yes, it might have even been Jim.

Lightman:

Do you remember how you reacted to [the flatness problem] when you heard about it — whether you considered it to be a serious problem or not?

Schramm:

I think again it was [a problem that] I assumed just required that we had some very special fine tuning and very special initial conditions.

Lightman:

Was that an acceptable solution to you?

Schramm:

It was a problem like the horizon thing. And you could do [solve] it that way. You could say, "In the beginning, we had this condition." But it was again always something nagging. It would be nice if there was some better solution. But, at that time, I certainly was not bothered about having an open universe. In fact, I had switched from liking a closed universe, which I had mentioned in graduate school...

Lightman:

As a philosophical...

Schramm:

As a philosophical thing, to thinking an open universe is fine and that the data all points to it.

Lightman:

Let me ask you about that, then. When you found that the data pointed towards an open universe, did that bother you at all, given your earlier philosophical prejudices?

Schramm:

Yes, it took me a while. It took me a little while on that. I didn't like it, but it just seemed to all fit together so well. We got this beautiful little triangle and everything from the different arguments [fit together well for an omega of 0.1. The thing I'd been working on, the deuterium, and also the age argument were two of the three lines of the triangle. Then Jim [Gunn] and Richard [Gott]^[42] were doing the bird's-eye view of the universe with the dynamics of halos, and so on, and came up with the density from dynamics being about the same as the deuterium density. And then Beatrice [Tinsley] was the glue, in that she put it all together in a nice way. I can remember when that happened. That was just one of those things when everything fits together, and I realized that was much more powerful than my prejudices. So that convinced me for a while. But I was certainly at that point very well aware that the deuterium argument was only on the baryons. We put a caveat in the paper that it's true that omega is about equal to 0.1 unless you have something else, some non-baryonic matter. We also mentioned black holes, primordial black holes. But if you had some other kind of dark matter — we didn't call it dark matter at that time, but just some matter other than baryonic matter — then the deuterium limit, which was what was really forcing us to the lower bound, was not the only value [not a strong constraint]. But at least when it was all fitting, I can remember being convinced in that direction. Now I've gone back the other way [laughs]. Not completely, though.

Lightman:

Do you remember how you reacted to the results^[43] of [Valerie] de Lapperant, [Margaret] Geller, and [John] Huchra on the large scale structure, when you first heard those results?

Schramm:

Yes. There were a couple different reactions. One was that I had already thought from stuff like what [Robert] Kirshner had done^[44] that there were voids out there. And I was also still [asking if] we could somehow save hot dark matter. Hot dark matter predicted larger structures and so on, whereas cold dark matter — [in] the models that people had done up to that time — didn't. People have since massaged the models so that they give more structure. So I thought it was very nice. My first response when I saw the Geller, Huchra, de Lapperant stuff was that the voids are real and maybe we lean more back toward hot dark matter. We had more reasons to try to save hot dark matter, which had run into difficulties because of the galaxy formation problem. And so I viewed this as very nice in that regard — as one additional piece of support in favor of the large scale structure which had already been indicated by the Kirshner, [Guss] Oemler, and [Paul] Schechter work, the KOS void. So they just lent [support] in that direction. Plus, at Chicago at the same time, Rich Kron with David Koo, in their pencil beam approach, had seen shells or indications again that there were [large] structures.^[45] I had seen their data even though they hadn't published it then — but they had also seen with their pencil beams much deeper than the Harvard survey. So I didn't view the Harvard thing as unique, but just one more piece of evidence. They seemed to have better press agents than everybody else, and they've gotten more publicity on it, but it just seemed to me it was all part of the overall package. It all fit together very well.

Lightman:

At this point, had you already accepted an inhomogeneous universe?

Schramm:

Well, that there's large scale structure. Let's see. It was also right around the time that Alex [Szalay] and I had been doing some stuff^[46] with Neta Bahcall and Ray Soneira's data.^[47] At just about the same time, we had done this thing where we showed that the cluster-cluster correlation function — in the way Neta had analyzed it — is much, much greater than the galaxy-galaxy correlation-function. Which at first seems ridiculous because it's looking like larger scales have...

Lightman:

More power.

Schramm:

We found that you could analyze it in a scale-free way, if you put in dimensionless units, and then you found that it was just all a fractal, which would imply very large scale fluctuations, very large scale structure. It was a fractal structure. It looks homogeneous, but isn't. That sort of thing occurs in phase transitions. So we had come to that [conclusion]. We had written that paper over the summer, or a little before Geller and Huchra. [So] my personal response [to de Lapparent, Geller, and Huchra] was "Great! This is all further support that there is some sort of structure out there on very large scales." I was very pleased by it. I had already been looking in that direction, so I liked it. All this fit [my little] piece of evidence. The other [amazing] thing — and I mention this again when I give talks — was that the de Lapparent, Geller, Huchra work could in principle have been done earlier. You know, it [the capability] was really built up when Mark Davis dedicated a telescope at Harvard and the Smithsonian to do [redshift surveys]. The technology to do that was really available long before. The thing that was so amazing was why hadn't the astronomical community recognized this structure many years earlier? The reason is the sociology of the astronomical community: the way telescope time was awarded and the way people do projects of looking for the weird object of the month. If you have only a couple nights a year on a telescope, you're going to get this couple nights a year by promising that you're going to publish some earth-shaking paper, not that you're just going to have all the positions [of] all the objects. So the net result is that the whole astronomical community had really missed the forest for the trees. They just ignored the background. They just looked at weird objects. They looked at the clusters. They sure knew where Virgo was and how far away it is — well, to within factor of 2. But they just hadn't filled in all the other ones. And yet, from a scientific point of view, that's the obvious thing to do. You should know what your background is before you go looking at the weird things. But because of the curiosities of the sociology... everybody wants to have an object named after them, so they try to find the strangest object, rather than trying to find what the average is. And to get telescope time, you go try to find the weirdest thing rather than the average. So they'd missed this basic structure.

Lightman:

That's an interesting explanation. Let me switch gears a little bit. One of the things that I'm interested in is how scientists use visual imagery and metaphors in their work. To what extent are these important in thinking and research? Do you use visual images very much in your work? In cosmology, the classic metaphor is the expanding balloon.

Schramm:

Actually, I don't use that. I use a loaf of raisin bread.

Lightman:

Okay. [Laughs] I apologize.

Schramm:

One of the things about the balloon is a bias towards the closed model.

Lightman:

Yes, of course. That's one problem with it.

Schramm:

That's one of the reasons why I started using a loaf of raisin bread, because it's a fiat model. It also has fiat, three-dimensional geometry, and the problem I found in popular discussions with the balloon is even though...

Lightman:

You say that it's just the surface, but people ask what's in the middle.

Schramm:

With a loaf of raisin bread, you don't have that problem. It's really three-dimensional. Then all you have to do is move the walls of the pan out to infinity. I think that's conceptually easier than going from two to three dimensions.

Lightman:

How about in your actual thinking?

Schramm:

In my actual thinking, yes, I think more in terms of the expanding raisin bread. It's what I visualize more than a balloon.

Lightman:

Do you use visual images much in your work?

Schramm:

Yes. In fact, my way of thinking is really a lot of visual images. I also find I do my best work interacting [with others], rather than sitting alone. I tend to work better in a collaborative mode rather than just sitting at a desk and calculating. I used to [work alone] when I was a post-doc doing computing work and [similar things], but lately -– for example, the thing with Jim Gunn and Gary — it has been more interactive. I find I work best when I'm talking with somebody else and bouncing ideas [off each other]. In fact; I discover my own ideas get formulated in discussion, to some degree, rather than something I hold before I go into a discussion. But then in a discussion, particularly if people question me, I think about it and formulate an answer. [That] is when I realize what I'm really thinking. I will frequently not have realized that before I'm questioned and I find out how I really think about an issue.

Lightman:

Have you ever tried to visualize the big bang or do you just not think about it?

Schramm:

Oh no, I visualize it. I have found that my visualization is different than the popular conception. I visualize it as a huge, very dense fireball everywhere. I don't visualize the pea blowing up. I think it's because of the way I visualize the horizon problem and also my paper in 1974 about the open universe. And although now I believe omega is most probably, the visualization I have is still always an open surface rather than a closed surface. I guess a lot of my visual formulation is from the early 1970s, when I was thinking in of terms open models. Open models work best if you don't visualize a pea blowing up, but visualize an infinite density.

Lightman:

I wanted to ask you about how theory and observation have worked together in cosmology in the last decade or so. Do you think that they have worked well and complemented each other or do you think they've gone off in separate directions? How do you see this relationship?

Schramm:

I think what's caused the explosion in cosmology has been two-fold. In fact, I just wrote this in a report. One is the cosmology-particle [physics] interface which we've talked about. It's brought in a whole fresh supply of people, smart people who hadn't previously looked at some of the cosmological problems and weren't brought up with the prejudices that others [had]. So they can come in with new ideas. The other [factor] has been the explosion of new kinds of observational information. That's ranged from the 3 degree radiation and the constraints on the anisotropy, which has forced us to go to more exotic models [than] we might have otherwise done. The things we were talking about, like the large scale structures of Geller, Huchra, and de Lapperant; Koo and Kron; Kirshner, or Temler, and Schechter. All of them have changed our way of thinking or supported it. As I was saying, I had some prejudices already in thinking that way, but it's nice to have your prejudices supported with these things. [Also, there is the work] on the number of neutrinos. If that had just been some number that was going to sit up in archives and nobody's going to ever check it, it's irrelevant. The thing that's made that so important is that people can do an experiment and are doing experiments to check it in the laboratory and really see whether it's right. For a purely theoretical number, you don't have the drive that you do when somebody can check it and do observations or experiments on it. In fact, I think one of the biggest problems with inflation is it’s going to be very hard to' prove whether it's right, to really do an experimental check. Probably [two] things are going to matter most on it. One is you can get omega down to be 1 to within plus or minus epsilon. The other is learning about Higgs fields and scalar fields. If you can ever find them in accelerators, then you can learn something about them. But really, that's the difficulty with inflation. It's a pretty idea, but you don't have that kind of nice experimental interface. Another thing [that] I think has been driving a lot of the dark matter stuff lately is that it helps that people are going out and trying to search for it. Again, it's no longer just intellectual masturbation. It's really getting down and doing that interface with experiments. So [this recent work] has changed cosmology from a subject which, up the last decade or so, was just redoing Hubble's observation over and over again — trying to find two numbers, Ho [Hubble's constant] and qo [the deceleration parameter] and not making much progress because of all the systematic errors. Now there are all sorts of different kinds of measurements that you can do, which have opened up in all different wave-lengths as we heard today — the infra-red, the X-ray background. All these different things have importance.

Lightman:

What do you think are the outstanding problems in cosmology today?

Schramm:

I think certainly the whole question of galaxy formation, dark matter, large scale structure — which I think are all the same problem. It's just the blind man looking at the elephant. Which parts do you focus on? But it's really all one problem. It's the major one, because it's the confrontation of the traditional astronomical observations [with the] interface of all this early universe stuff, with what we traditionally have called cosmology looking at the objects in space. I think that's really the prime problem that a lot of people are focusing on, and there's progress being made. There are ingredients coming in from particle physics, from traditional astrophysical theory, from particle cosmology, and from the observational ends. And the observational ends are in all wavelengths. The 3 degree background, the infrared, and the optical observations of the galaxies all are contributing. I think that's a real dynamic interface right now. And even the dark matter searches — underground searches, accelerator searches, different kinds of things. It all comes together on that problem. As a result, there [are] a lot of articles, and the in-vogue opinions on what is the favorite candidate keep shifting from year to year because they're constantly having new input. I would say that would be the number one problem. I think the number one theoretical problem is: what is the vacuum? It's worded in different ways by different people, but when we talk about the Higgs field, when we talk about the scalar fields that cause inflation, what you're really asking is, "What is a vacuum?" The Higgs field [that] causes phase transitions is caused by some sort of vacuum expectation value. The cosmological constant — why is it zero or why is it so small? Which is again, what is a vacuum? In all of these theories — in fact, even in inflation — the final value you set [the vacuum] to is totally arbitrary, and we choose zero because we think it's small. But there's no reason why there can't be some other arbitrary number. The superstring theories attempt to go at it, but there's really no good answer to [the questions]: what is the vacuum and why is the cosmological constant so small. That I think is the theoretical problem: the phase transitions and the nature of the vacuum, which ties into why [there are] three versus ten dimensions and so on. That, I think, is the big theoretical problem, and the phenomenological problem is how to make galaxies. So I think there are two major problems.

Lightman:

Let me finish by taking a big step backwards and asking you something much more speculative, maybe asking you to put some of your natural scientific caution aside. If you could design the universe any way that you wanted to, how would you do it?

Schramm:

I guess if I was going to design the universe... We actually have a pretty good one. The one that we come up with, the single one-shot big bang, omega equals one [universe] is a very pretty one. Flat space, one-shot [and] you create a universe. You have a creation and it's all very smooth and nice and very pretty. We clearly have some problems still of how to make galaxies and everything. It would be nice to [make] them in some sort of direct phase transition, which may be what it is. It's a very pretty universe. I like it. In fact, people used to ask me that question, and the answer I used to give was [that] the prettiest universe was the steady state universe. This may have been a prejudice of working with Willie Fowler and Fred Hoyle and so on. I used to think that was a very pretty one because of infinite time, and everything's always the same. Everything always is, always was, and always will be. [That's a] philosophical way of viewing it. But I guess recently I've come to think that maybe this nice one-shot big bang — but a flat one right at the boundary — is a very pretty one. Everything is nice and clean. I like the idea [that] you start out with one grand unified force, a theory of everything, and the rest of the universe is a consequence of that. It takes only one ad hoc assumption — your initial physics – and then everything else follows. I like to minimize the number of ad hoc assumptions. And in fact, if you can somehow argue that that law, which some people are searching for and haven't found yet, is the only law that's consistent, then the ultimate statement you're making is that mathematical consistency is the only law, which is very nice. I recently said this in a popular article I'm writing. [That] is really what they're saying when they talk about superstring theories. The argument was that there's only one [theory] that's anomaly-free, that doesn't have infinities. So then the ultimate thing they are saying is that mathematical consistency is the only law that you require. Then everything else follows as a natural consequence of that. I think that's very good. So you just have to say: in the beginning, there was mathematical consistency. Everything else follows, including us. [Laughs]

Lightman:

Let me ask you one last question. There's a place in Weinberg's The First Three Minutes where Weinberg says that the more that the universe becomes comprehensible, the more it also seems pointless.^[48]

Schramm:

Yes, I've read that.

Lightman:

You noticed that.

Schramm:

Yes. I take some issue with it. If we take what I was just saying about mathematical consistency, in some ways that seems to me very beautiful. It's a very nice thing. It's saying that that's the underlying cause of it all. Now you might say, well to what extent then does that affect your daily life and everything, and conclude that it is pointless. That's probably what Steve was referring to, that all these other things are just irrelevant consequences of this initial law or whatever it was. I like to think that it's not that they're irrelevant, but they're a beautiful demonstration of it, that all the intricacies of life and all the things that can happen develop from one simple thing. It's a wonderful demonstration of it, of the beauty of nature. It's like cellular automata. Depending on what formula you chose in the beginning, you can end up with all sorts of different patterns. In some sense, that's what the universe is. If we take this logical law as our cellular automata pattern, then we end up with a whole complexity of things and all this kind of funny stuff that's happening is lots of fun and yields good things, bad things. We end up with a delicate interplay. I think it's more a demonstration of how beautiful it all is and how wonderful. It makes me feel good, and it's what we're searching for. We're searching for the underlying beauty and it's this underlying beauty that can yield richness. So I'll take the same set of facts that Weinberg did and...

Lightman:

Wouldn't describe it the same way.

Schramm:

I would describe it exactly the opposite way.