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.
Please contact [email protected] with any feedback.
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 James Gunn by Alan Lightman on 1988 January 18, Niels Bohr Library & Archives, American Institute of Physics, College Park, MD USA, www.aip.org/history-programs/niels-bohr-library/oral-histories/34296
For multiple citations, "AIP" is the preferred abbreviation for the location.
Childhood experiences and working with father to fix things; building telescopes as a boy; early reading in astronomy and cosmology; early preference for steady state model; education at Rice and Caltech; thesis on correlation functions of galaxies; building an analog computer as a graduate student; reasons for building instruments; building a silicon-intensified target vidicon spectrograph; contributing to bringing CCDs to astronomy; importance of hands-on experience in observational astronomy; new questions in cosmology made relevant by the inflationary universe model; too many theoretical frameworks in cosmology today; attitudes toward the inflationary universe model, the flatness problem, the horizon problem, and the anthropic principle; attitude toward de Lapparent, Geller, and Huchra’s work on large-scale inhomogeneity’s; Cornell work on inhomogeneity’s; over interpretation of data; use of mental pictures in science; interaction of theory and observation in cosmology; the balance between observations and speculation in science; the ideal design for a universe; the question of whether the universe has a point; the conflict between big science and little science.
I wanted to start with your childhood and how you got interested in science. I’ve heard many stories about how you built things when you were younger. Could you tell me a little bit about that?
My dad was an exploration geophysicist for Gulf Oil. When I was a little kid, it was during the war. He was the chief of an exploration party, and you couldn’t get parts for anything. So anything that broke, he had to fix or make. And so we carried around with us — we stayed in one place six months at a time — we carried around with us this machine shop that was in a trailer. From the earliest time I remember anything, I was around power tools and my dad making things and fixing things. So it was a very natural thing for me to do, to grow up doing that kind of thing. I guess it was about when I was seven or so that I became interested in astronomy. I don’t quite remember what touched it off, but there was a little book that my dad bought me called The Stars for Sam. There were various science books for kids called The Ocean for Sam, and this, that, and the other...
The Stars for Sam? Was Sam a boy?
Sam was a boy.
Do you remember who wrote the book?
No, I have no idea. There was a whole series, as I say, and whether they’re still around or not, I don’t really have any idea. That really turned me on, and it just sort of took off from there. My dad had taken some astronomy courses as an undergraduate and still had his textbooks around — just descriptive kinds of stuff that a kid could easily read and more or less understand. It wasn’t but a year or two till I graduated to reading that kind of stuff about astronomy.
That’s before the age of ten, you think?
Yes, well before the age of ten. It was also a natural with all the other stuff that I build telescopes. So about that time I started building some telescopes and I built several by the time I was eleven or twelve. So it seemed a very natural thing to keep on doing because it was the thing I loved the best.
What size were these telescopes?
They started out being pretty small. You’d just buy lenses from what then used to be Edmund Salvage Company... little ones and mostly refractors. It wasn’t until I was in high school that I built my first reflector — four and a quarter inch. I guess I was in junior high, actually. And it wasn’t until the middle of high school that I started doing optics seriously. I made an eight inch mirror from scratch. That was the biggest telescope I built before I went pro.
I would say an eight-inch mirror from scratch at that age was ambitious.
It was pretty ambitious. It worked, though. It’s amazing, right? [Laughs]
You mentioned this book, Astronomy for Sam or something. Do you remember any other particular books, or any particular scientific ideas that impressed you?
The book later that really caught my fancy was Fred Hoyle’s book, Frontiers of Astronomy, and the notion that we’re all made out of the stuff that spewed out of stars and things like that. That kind of thing was not discussed in these old textbooks of my dad’s because they had no idea how the heavy elements were made. Hoyle is a very persuasive writer and that was a very cohesive view. The idea, to a high-school kid, that one could make a cohesive picture of the universe and why things were the way they are is a very exciting thought. It’s something that you don’t think about when you’re that young — that you can more or less sit down in your armchair, literally in Fred’s case, I think, [both laugh] and write down a theory of the world. It’s pretty powerful stuff, and it really caught my fancy.
You must have been at least susceptible to big thoughts. I mean at that age, as you say, most high school kids are just wondering how they’re going to get to steal their father’s car for the night, so you must have been at least occasionally thinking about big ideas.
Oh, I wasn’t capable of thinking big thoughts, but I was capable of thinking about big thoughts, anyway, at the time. [Laughs]
At this age, say ten to fifteen or sixteen years old, do you remember reading or hearing anything about cosmology? Did Hoyle talk about cosmology?
Yes, Hoyle talked about... The whole thing was, of course, couched in the steady-state idea and how things could come about in the steady state. I also read [George] Gamow’s books, and it was clear even then that there was some fundamental dichotomy in the world, although Gamow’s books were written before the steady state became current. Hoyle kind of soft-pedaled standard Friedmannian cosmology because he thought it was beneath contempt. But it was clear that there was some great dichotomy of thought at that point.
You were aware of that at that age?
I was aware of that and also somewhat aware of the kinds of tests that one might make to distinguish them. At that time the microwave background wasn’t known, so the understanding of things was really pretty primitive compared to... I’m sure it’s primitive now also, but [even] compared to now it was primitive.
Do you remember having a preference for any particular cosmological model at that time?
I liked the steady state, and I don’t know why. I guess just the idea of permanence. It’s something that’s hard to say. I have always had a philosophical predilection for universes that at least last forever. I don’t particularly care if they have been in existence for any amount of time or not. [Laughs] It’s probably one of these subtle biases that affect what one does, because I’ve come down fairly heavily on that side of the coin when it seemed to matter — I’m not sure it matters anymore.
The open universe?
Let me ask you a little bit about your early professional career. Can you tell me a little bit about your undergraduate education at Rice?
Yes. I went to Rice for several reasons. I had several choices. The main reason was that my parents were not terrifically well-off financially, and at that time Rice had no tuition. I also got a quite nice scholarship for support. And doing as much research as one could do on the questions at the high school level; it seemed likely that I would get as good an education there as anywhere. I haven’t really regretted going there at all — it’s a very good school. Rice had no astronomy department, for which I have been eternally grateful since.
So you were forced to take physics?
I was forced to take physics, right. I took a double major in math and physics, and I think that was one of the wisest decisions I made — not to worry about astronomy until later. It wasn’t clear actually until the very last minute that I was actually going to go and do astronomy because I was taking physics fairly seriously. I applied to many physics graduate schools and also just sort of on an odd chance applied to Princeton, Chicago, and Caltech in astronomy and was admitted. And just in April — or whenever it is that you have to decide — of my senior year, decided: By God, this is what I really like; this is what I’m going to go do even though I had absolutely no idea what astronomical research was like. Zero. I didn’t know what professional astronomy was.
Well you certainly had been an amateur astronomer...
I had been an amateur forever, that’s right. I read lots and lots of books, but I still didn’t have a very clear idea about what a professional astronomer did, whereas I had a pretty good idea of what a professional physicist did because I had done... and also mathematicians. I had actually been more active in mathematical research than in physics. I had done quite a lot on numerical methods for solving differential equations and such and had a couple papers. But just at the last minute I decided to hell with it, and I never regretted that decision.
Once you decided to go into astronomy, which you said you did at the last minute, you entered graduate school at Caltech. Do you remember how or when it was that cosmology in particular attracted you?
From the very beginning. In fact, one of the main reasons I went to Caltech was that H. P. Robertson was there, and he was certainly one of the leading lights at the time. Now in the summer between the time I left Rice and came to Caltech, H. P. Robertson was killed in a car accident. With his passing, there was not even a relativist at Caltech, much less a cosmologist. So I think like many entering graduate students, I go to a place to do cosmology and most of them go away and do something else. I had no choice but to go away and do something else because there wasn’t anybody there even working in the field. That was something of a blow, but Caltech was a sufficiently exciting place with other things going on that I didn’t mind particularly. I proceeded to attempt to educate myself a little bit about relativity, and then my third year — maybe it was my second year, I don’t remember — Frank Estabrook from JPL came and started teaching a relativity course, and so I learned that.
With a lot of differential forms probably...
With a lot of differential forms, that’s right. Actually, it wasn’t my first introduction to differential forms. I took a differential geometry course my senior year at undergraduate school, which is the hardest course I’ve ever taken in my life. So I had already run into differential forms and was thoroughly scared of them. I didn’t get un-scared of them until Misner, Thorne and Wheeler came out, and I finally found out what they were.
Just back-tracking a little bit, you said you went to Caltech already with the idea of working with Robertson, who worked in relativity theory and cosmology. Do you remember how it was that you were initially attracted to cosmology?
I think again through reading Hoyle and Gamow. It was just the thing I wanted to do. It’s what every kid wants to do who thinks about it at all, you know, to understand the World with a capital W.
Well, you could think that you just wanted to go study how the sun works...
Yes, yes, yes, and there are people who do that. I don’t understand them. [Laughs] But actually during my senior year at Rice, Bob Kraft came around and gave a colloquium and by that time I knew I was going to Caltech. He stayed at Rice a couple of days, and I spent quite a number of hours with him just discussing this, that, and the other. So I felt I knew him fairly well. I then got involved with him almost immediately after going to Caltech on some things as far away from cosmology as you can imagine, namely studying the atmospheric abundances in F stars because that’s what he was doing. So my first year at Caltech, I got involved in observing. I went to the 60-inch [telescope] and got the spectra of F stars to do this with. So I got busy on other things almost immediately and didn’t really notice the lack until I had time to sit down and think about what it was that I was doing. So the loss was a disappointment, but not particularly a painful one. Caltech keeps you busy, as you know.
You said when Estabrook started teaching it, you took the relativity course. Did you then start finding more opportunities to do relativity and cosmology?
Yes, I did. I started thinking about cosmology in some detail then and doing little calculations. I got interested my second year talking to Guido Munch — which actually turned into a thesis about the statistical distribution of galaxies. He had done a similar thing for clouds with Chandrasekhar. It was some years before [James] Peebles’ [work on the clustering of galaxies]. That project in my third year turned into a thesis. I counted galaxies on plates and did correlation functions and stuff like that. I made models, tried to make evolutionary models without much success...
Of individual galaxies?
No, of the evolution of structure. And I did correlation functions. Actually for mathematical convenience, I did the correlation function expressed as a superposition of Gaussians. It’s one of the things that I regret in my career. I did a model with four Gaussian components and of course looking back on it now, those Gaussian components added up in such a way that they make a power law. But I didn’t recognize then at the time, right? [Both laugh] So I could have beaten that boat by a while, but it didn’t happen. Anyway, that’s water under the bridge.
Well, I guess one is free to choose one’s basis functions...
… one’s basis functions however one likes. But one should keep an eye open to what they look like at the end, right, when you’ve got it all done.
Did you have redshift data then?
No, no, no, this was just...
So this was two-dimensional...
This was long before you could even get redshifts for most galaxies like that. 3C295 still had the record and these were mostly objects fainter than that...
So you were doing theoretical calculations, too, and then you were projecting them?
Yes, that’s right. On the sky to compare with, right. The upshot was not very exciting. I got the form for the correlation function, which, as I say, I didn’t recognize. The depth of the material was not such that one could say very much about the many evolution models — which had cosmology and various assumptions about dynamics and things anyway in it. So there was not a lot to be done. It was a very exploratory kind of thing which didn’t really turn into anything.
You were not doing the observations yourself, I take it.
Yes, I was. So part of the time was getting the plates at Palomar. I got them. I counted them, which is something I never want to do again. Then I put the stuff in the computer, computed the correlation functions, and did comparisons with the theories.
So you were already doing both theory and observations at this point.
Observation, yes, right. Coming up from being an amateur and building instruments and such, I was already well into that at Caltech. I made several things while I was there. It never occurred to me to proceed in any other way. If you needed data, you go out and get it. I still think that’s the best thing to do.
Can you tell me one of the instruments that you built as a graduate student?
The major piece of equipment I built as a graduate student there was an analog computer, basically, that took data from a recording microdensitometer and ran it through the photographic transfer curve, which you could put in as an analog function. It came out with intensities, so you could scan photographic plates and output intensities or log intensities, however you wanted to set up the curve. That was done originally when I got there with an incredible machine in which you drew the curve on a graph in conducting ink. There was this little puck that followed the curve along the x-axis and output the y-axis. It was very erratic and very slow. The microdensitometer was quite fast. It would scan plates very fast, but this machine just couldn’t keep up. So I built an all-electronic thing in my spare time in my first two years there. I think in fact it’s still more or less in operation although there are much simpler ways to do it.
That’s fantastic. When you took the relativity course with Frank there, do you remember at this stage of your career — now you’re in graduate school — having any particular preference for one cosmological model versus another?
Oh, just the old preference that had been there for a long time, for things that went on forever. I was never convinced of the sort of neatness of compact [closed] models, both in space and time. It seemed to me rather a waste.
What about the question of homogeneity versus inhomogeneity — did you have any thoughts about that?
I had no real thoughts on the issue. Of course — people were already mumbling about the problem of a causal initial conditions and such things as that. That was a great mystery even then, although it didn’t bear much thinking about because there seemed to be absolutely nothing intelligent that one could say about it or do about it. So it was a thing that you talked about when you gave popular talks, and some place that science had to stop and something else had to happen. Which fortunately seems to have gone by the by... [Laughs]
Let me ask you a little bit about your instrument building. A lot of people say that some of the most important things that you’ve contributed to the field are not to be found in the journals but are to be found in some of the instruments that you’ve built. How do you go about deciding what instrument to build and how to design them?
I am nearly always driven by results I need. That’s been true in almost every case. I want to do something. A lot of the things have been done by the pressure of wanting to get the spectra of ever fainter galaxies at bigger redshifts and, by the by, finding those galaxies at big redshifts and being able to set them on the spectrograph. So most of the stuff that I’ve done in the last ten — no, more than that — fifteen years has been instrumentation aimed at that. When I first came back to Caltech in 1970, image tubes were just becoming off-the-shelf things that one could buy and think about seriously using. Bev Oke and I got involved almost immediately when I came back in this survey that still goes on for distant clusters. That was begun with Schmidt plates at the 48-inch [telescope], but it was clear even then that one needed to get better pictures with the 200-inch, and the way to do that was with image tubes. So the first things I built at Caltech were a couple of image-tube cameras. Then along about 1973 or 1974, we were doing spectra at that time with Bev’s multi-channel thing which uses lots of photomultipliers. 2-d electronic devices were just sort of becoming available — SIT videcons in particular. So the first really major project that I did in my professional career was building this SIT spectrograph for the...
What kind of spectrograph did you build?
SIT — it’s a silicon-intensified target videcon spectrograph. I spent a couple of years, essentially full-time, building that instrument. I didn’t do very much science during that time. I learned an enormous amount of electronics — analog electronics and digital electronics — my first real experience with digital electronics.
What years were these?
I can’t remember exactly, Alan, but I think sort of 1973-1974, 1974-1975, something like that. 1974-1975, I guess it was. That was also the time that mini-computers were becoming widely available, and so from the outset we designed this thing as a computer-controlled device. It was my first experience with computer control. So it was a very intensive, in an engineering sense, those two years. I did essentially all the design for the instrument and a lot of the electronic construction because there was this big laboratory called Astroelectronics Laboratory — may it rest in peace — whose purpose was to build electronics for the observatories. But in fact was not a very competent organization, and in particular it had no analog expertise at all and that was the hard part in doing this stuff. So I did all of that myself and in fact built most of the analog circuitry. Looking back on it, you could say it was two years gone, but I learned an awful lot, and it has come in very handy since.
Why would you say it was two years gone?
I didn’t do any science. So, as you say, there were almost no papers.
But you then used it...
Oh yes. Then I used it quite extensively. It was not as successful as it might have been because the detectors were not as good as they might have been. It did a number of things very well, but not the problem I particularly designed it to do.
That is to detect faint...
To detect very faint things, that’s right. Because it was limited by its own inherent kind of noise whose properties were not easily understood.
So did you then try to solve that noise problem in the next detector that you built?
Well, along about that time it became clear that CCDs were the wave of the future, and so it was not long after then that I got involved with [James] Westphal and the JPL CCD effort.
They were developed at Bell Labs, is that right?
Originally they were developed at Bell Labs, although by this time NASA had a big contract with Texas Instruments to develop this chip for Galileo. Later Galileo changed its mind and used an entirely different device, but that device became the space telescope wide-field camera CCD. And it was also about this time that Westphal and I got involved in the wide-field camera effort, so CCD’s were fairly easy to come by, and all the instruments I designed after that used CCD’s.
And that helped solve your noise problem?
That helped solve the noise problem, that’s right, because the [CCDs] really are essentially ideal devices. So the next thing I built was this thing called PFUEI that is a combination camera-spectrograph that would do almost everything I wanted it to do, except the field [of view] was too small. Then I designed Four-shooter to get around that.
When did you develop Four-shooter? Was that the late 1970’s?
That was the late 1970’s. The design was finished and the construction was started when I left Caltech in 1980. It was on the telescope first — I can’t remember the date — I think it was September 1983 or September 1984. So that was also a fairly long time. But, fortunately, since that was happening there and I was here, it was a not an absolutely full-time effort. I was able to do some other things. It actually worked out extremely well.
You have a unique type of hands-on [ability] — when I say unique I mean there are very few young astronomers these days who are getting a hands-on experience both with building instruments and with actually looking out of the telescope. Do you think that that kind of experience has given you any special feeling for the objects that you study?
I think probably the contrary. The problem is that my efforts are so fragmented between doing various things that I probably don’t have time to understand the theoretical things as much as I would if I just did theory. One thing it does do — and there are two sides to this coin as well — is that it does give me a pretty firm handle on what and what not to believe of my own results and other people’s results because I feel that I know the limitations of instrumentation very well. I’ve noticed in the last few years — I guess one always gets negative as one grows older — that I believe less and less. [Laughs]
Of what other people are doing?
Of what other people are doing. Whether that’s healthy or not, I’m not entirely sure. Also, less and less of what I’m doing myself. [Laughs]
That’s a good answer.
But I think people who do pure theory or who do pure observations have a somewhat better chance of understanding all the facets of the thing they do better than I have been able to. It wasn’t so bad when I was younger and quicker, but I’m not so young and quick anymore. And I feel a real lack. I seem to do things more and more shallowly. That’s partly because — just as one ages, you get involved in more things, so I’m certainly doing more things. But I think also that the old gray matter isn’t working as well as it used to.
My impression, and you can correct me, but my impression is that there is a new generation of observational astronomers who have much less hands-on experience with a telescope and instruments than you have had.
So even though they’re doing purely observational astronomy — as you were distinguishing between doing a pure subject and doing lots of things — even though they’re doing pure observational astronomy, their experience is very different from what you’ve had.
Yes. I think that’s not very helpful. It’s fostered by the National Observatory System and such, in which you go to a telescope twice a year to use an instrument whose development you have not been involved in at all and that you learned about from a book. It’s almost impossible I think to get a real feel for what the limitations of that instrument are. I don’t know where the thing came from, but there’s a famous saying among instrumentalists that the problem with this instrument is that it always gives an answer. [Both laugh] So you take home a piece of data, but you don’t really have a very good idea of what it means, to what extent it’s to be trusted. I don’t know what to do about that because as universities become poorer and poorer and observational astronomy gets centered at fewer and fewer higher-and higher-powered places, there’s just not much way to educate people to do instrumental stuff. Not very many people are interested in the first place. In a way, it’s a return to the very bad old days because in the 1930s, observational astronomy was very much a kind of gentleman’s game. It was engineers who built telescopes. Even though the instruments were pretty simple, the astronomers didn’t really have very much to do with them. I think that had a deleterious effect and I think probably this new school is going to have a somewhat deleterious effect.
I’d like to ask your reactions to some of the discoveries of the last ten years. Let me start by asking you a general question, and maybe your answer will get to some of my specific questions. We said earlier that a lot has happened in cosmology in the last ten years. Have your broad views of cosmology changed in the last ten years as the result of your work or other people’s work?
Not so much as a result of my own work, but certainly as a result of other people’s work. I think that if you’re anywhere close to the game, you cannot have the same view of the universe now as you did even five years ago, because if anything like inflation [the inflationary universe model] is right, the universe is just a very, very different place than any of us visualized a few years ago. And the questions that were burning issues then are by and large not very interesting anymore in the global sense.
They’re made irrelevant?
They’re made irrelevant, and the kinds of questions that you ask are entirely different. It’s not completely clear to me that inflation is right, and there’s certainly no external evidence that it is. But it’s such a pretty idea and fits things together so well that it... it just enforces an entirely different kind of mental outlook on cosmology than before. I think it’s very important still to go on with the classical tests because that’s one test that you make. I mean if omega is not 1.0000, then something is screwed up, and we have to know that. We have to be able to test whether general relativity is the right theory of gravity, which is also something you can only test, I think, with global tests. But the reason for doing those tests in entirely different from what we thought the reason was a few years ago.
You said that no one doing cosmology or close to the game could keep the same views in the last five years. In terms of the change of viewpoint in you, has any kind of intellectual struggle gone on? Have you found any discomfort or hesitation or did you just sort of naturally glide from the old questions to the new questions?
No, because I don’t... a lot of people who work in cosmology I think have basically religious beliefs. I have had mild philosophical preferences but never very strongly-held views. It seems to me that the universe is the way it is and we’d better accept that. So it’s not been any problem; In fact, it’s been an enormous pleasure because all of these locked doors that were present a few years ago now seem to be unlocking. And even though I’m not working in the forefront of early universe theory, just the idea that these questions can be answered is enormously exciting, and all very positive. I wouldn’t have changed anything for the world in the last five years. The big problem, it seems to me, is that there are such a plethora of possible theoretical frameworks at the moment and no way of testing them, that the subject is sort of running open loop. That’s not very healthy from a purely scientific point of view. It’s very exciting, but from the point of view of trying to learn the ‘truth’, I think we’ve taken a large step backwards, and that’s just going to be the way it is, I think.
That’s an interesting way of putting it — that now we know how to talk about some of these questions, but the possibilities have multiplied enormously.
Yes. The problem is not anymore that there aren’t any answers out there.
There are too many.
There are far too many, right.
Let me ask you about your reactions to a couple of specific things: you mentioned the inflationary universe model, which I’d like to get back to in a minute. Maybe I should start with that, since you mentioned it already. Do you remember, when you first heard about the inflationary universe model, what your reaction was to it?
When I first heard about it, it was early inflation, or whatever it’s now called — ‘old inflation’ I guess.
Yes, old inflation.
I thought it was absolutely insane; it couldn’t possibly work because these bubbles couldn’t possibly ever get together. That [objection] turned out, of course, to be true. But that soon settled out. The idea was so pretty that I was immediately taken by it, but I realized there were big problems. Other people realized there were big problems, and now there are several ways around those problems.
Did you consider it to be highly speculative or did you think it had so much going for it...
I thought even at the time it had so much going for it that it was kind of natural. [John] Wheeler — was it Wheeler or Hoyle or maybe both — had kind of introduced this notion some years before that the universe was homogeneous and isotropic just because it was a greatly magnified view of some microscopic thing and that there was nothing magic in that. How you reconcile that with causality and Robertson-Walker [metrics] wasn’t at all clear. I think they all wanted some quantum geometrical effect for it rather than some phase transition. But that idea was already floating around, and I found that very attractive. But, of course, that was unanswerable because that pushed everything back to the quantum gravity era where you couldn’t understand anything.
Yes. I’d never heard that idea.
It was a kind of natural way, because you imagine that once the wave function of the universe was very small and causality was violated. After all it had to be, so maybe it could have been smooth then. And that was a kind of proto-inflationary idea which I already found very attractive but not scientifically addressable, whereas the inflation thing was...
Yes, a particular way of doing it. Let me ask you about the flatness problem which I guess got widely broadcast in the article by Peebles and [Robert] Dicke in their article in the Einstein Centenary book in 1979. Do you remember when you first heard that argument, — the flatness problem?
Well, I don’t know, Alan. It had been kicking around for many, many years before that, because I used to give popular talks and talked about it in the early 1970s, and it wasn’t my idea. So I think it’s true that that article really brought it to the forefront, but I don’t even remember where it came from. I think [for] anybody who really seriously looks at Friedmann, Robertson-Walter metrics, it’s a sort of obvious thing. I’m sure I read it somewhere and then went and looked at it. But I don’t remember where I got it from.
Well my impression, although I may be wrong, is that the problem is not as old as the causality problem.
The horizon problem.
The horizon problem, yes.
It isn’t exactly the same, it’s true.
It’s not the same and maybe comes after it in terms of being posed.
I think that’s true. I think that’s true.
When you first started thinking about [the flatness problem] in the early 1970’s and lecturing or however you addressed it, did you consider it to be a serious problem or did you think that it was just due to a random initial condition of the universe and not worth a physical explanation?
No, it seemed to be worth a physical explanation, but it’s sort of intimately tied with the causality problem. There were clearly things about the initial conditions that we didn’t understand and that were something else. I mean, the fact that it required such exquisite tuning was maybe no more bizarre than the causal nature in general of the initial conditions. So I guess I didn’t worry about it as a special problem and I didn’t attach a name to it. But it was certainly just one of the set of worrisome things about Friedmann cosmologies.
And you thought it was something that would require a physical explanation?
An explanation that somehow had to be locked back in the quantum era because that was the only way you could possibly explain it. So somehow the energy was zero, but you didn’t quite understand why and there must be some physics for that.
I know when I first heard the flatness problem, I didn’t think that it was a serious problem because I thought it was just an initial condition and the universe is as it is. My impression was that a lot of other people, astrophysicists, sort of reacted the same way to it, although some people took it very seriously. It was only when I looked at it — kept looking at and thinking about it that it bothered me more and more and more. For some reason, it didn’t bother me the same way that the horizon problem bothered me, but of course the two were...
...were very closely connected. One thing — looking back — that I don’t really understand was why I was not persuaded by these thoughts immediately that omega had to be one. Because it would be easy to say that the initial conditions have to be such that the energy is zero and that is somehow easier to swallow than to say that the energy is epsilon. But I was not persuaded of that argument, and now I cannot go back and reconstruct why that is so. I guess I was already thoroughly persuaded that the observations fairly convincingly told me that it wasn’t so or at least that the density was smaller than the [critical density] — which is something of which I’d become less and less convinced about as I grew older.
When you gave talks about [the flatness problem], did you describe it as a serious problem with the standard picture?
No, I think I just described it as a curiosity.
As a curiosity?
Yes, but not necessarily, not a serious problem because again the standard theory didn’t address the issue. The issue was somehow in the initial conditions, and the initial conditions were already a mystery. So it was one of the couple or three curiosities about the initial conditions that one wondered about but was not capable of addressing.
I know the problem really is only very dramatic [to me] if you can imagine that the universe could have been made with many different initial conditions, and the probability of this particular set of initial conditions is very small. So you have to be able to imagine lots of different possible initial conditions in order for this to be a dramatic problem. Did you think about it in that way?
Well, along in the late 1970s, the idea of the anthropic principle and such were current. I was never terribly convinced because ... I’ve never been convinced that the anthropic principle requires the universe to be as homogeneous and isotropic as it is. Among the manifold of initial conditions, maybe it restricts them, to be sure, but [it] does not restrict them to [the] nearly measure zero set that we see. And so I’d never taken it terribly seriously. So again it was just pushed back into the mystery of the initial conditions being what they were but we didn’t understand what. I thought about the many worlds [interpretation of quantum mechanics] ideas, but it’s clear even so that there have to be constraints. If that’s true, there almost have to be constraints on the initial conditions that we don’t understand — to account for the homogeneity and isotropy. So it was clear that that was not a very fruitful idea or didn’t seem at the time a very fruitful idea...
Because you didn’t think there was any hope with what you knew at the time of getting at those constraints?
At those constraints, right.
Or the physics behind them, which I guess Steven Hawking, has started to do in the last few years. So you recognized it as a serious problem but you felt like we didn’t have the tools to attack it.
We didn’t have the tools to do anything with it. I certainly didn’t have the tools to do anything with it, and it didn’t seem to me that anyone did.
So you just sort of quietly suffered with the problem.
Yes, suffered with the problem.
Let me ask you about another discovery in the last few years and that’s the work of [Valerie] de Lapparent and [John] Huchra and [Margaret] Geller on the large-scale structure. When you first heard about those results — or I guess they were extensions of similar results earlier by other people — do you remember how you reacted?
My reaction actually was that they had terribly over-interpreted their data. [Laughs] The data was very pretty, but at about the same time, and actually somewhat before, the Cornell group had published their 21-centimeter [radio] stuff on the Perseus-Pisces region, which didn’t look like the Huchra-Geller stuff at all. So here was another piece of the universe not very far away that did not have this property at all, so it seemed to me that...
They were imagining things.
…that they were imagining things. And I think it’s still quite possible. In the first place, the Cornell stuff was for a three-dimensional volume, right. Margaret’s and John’s stuff was for a two-dimensional slice, and you can easily be fooled by two-dimensional slices. And even if they were right, it was clear already that it wasn’t universal, because here was another piece of space that didn’t look like that. So I thought immediately that the idea that things were on these sort of regular bubbles had to be wrong and it was just statistical fluke that they happened on such a volume. I certainly thought Margaret’s notion that it said immediately that gravitational effects were [ruled] out was completely off the wall. Margaret usually doesn’t say off-the-wall things, but I thought this was [something] off the wall.
Before seeing these results, did you have any particular notions of your own about how homogeneous the large scale structure was?
Yes, because I had sort of believed the Peebles party line that the two-point correlation functions…
Died off [at distances larger than about 20 million light years]
Right. So one expected the universe to be pretty homogeneous on those scales, and I was very surprised. But I was already surprised by the Cornell results that that didn’t seem to be true. It didn’t take much thought to realize that the two-point correlation function was not a very good tool for exploring that kind of structure. And there we are. The question still remains: what scales are really relevant on large scales, and we don’t know.
Has your opinion about the de Lapparent-Huchra-Geller stuff changed since this initial reaction that you just described to me?
I still think they over-interpret the data a bit. It’s clear in their larger surveys that things are not quite as regular as they would have had you believe from the first ones. And still their region of space appears, I would say, qualitatively rather different from the Perseus-Pisces region, which says that there are differences from one place to another, which is another way of saying that there is really large-scale structure although it’s not a question of the density here being different, it’s a question of the kind of structure here being different. Whether that argues for some funny non-Gaussianness in the initial conditions, who knows? Until we have some better idea of what the spectrum [of density fluctuations] is initially, I think a lot of this can certainly be accounted for by a Gaussian [random] process that has rather more power at large scales than we’re currently capable of figuring out how it comes about. So I’m not persuaded that there’s any mystery. It looks like cold dark matter is in trouble, but it’s been looking for a long time like cold dark matter is in trouble. Qualitatively it’s okay, but whether it can give you the stuff on these really large scales, I just don’t know. The simulations are moot on that point, I think. I don’t think the correct simulations have been done.
What do you think they’re missing?
Oh, it’s just the problem with periodic boundary conditions.
Yes, it’s just the problem of doing a discrete N-body calculation.
Yes, right, right. You’ve got to have a bigger box and so you’ve got to have more particles, you’ve got to have a bigger computer, but preferably, you just need to be cleverer.
Yes. Is it true that your view about the large-scale structure has undergone a change?
Oh, yes, yes. It certainly has. Before that stuff, one had inklings that there were these structures on much larger scales than you liked to admit, but it was really that and the Haynes-Giovanelli stuff [see previous footnote] that put the nail in that. I think Haynes and Giovanelli should share as much credit actually as John and Margaret.
Yes. And your own thinking has changed on this issue. You said that earlier you had sort of gone along with Peebles...
Oh, yes. I mean it was clear. You’re confronted with the results. That’s what the universe looks like and you can’t ignore it.
One of the things that I’m interested in is how scientists use visual images or metaphors sometimes in their work. One of the most powerful metaphors in cosmology has been the expanding balloon metaphor that I think Eddington first introduced. Do you remember when you first heard about that metaphor?
Oh, in high school. Maybe it was in Hoyle’s book, maybe it was in one of Gamow’s books. One of Gamow’s books, I think. A very appealing idea. I knew a little bit about non-Euclidean geometry then in just sort of layman’s [language]. It was clearly a very powerful way to visualize things. I’ve used it thousands of times in talks and in courses.
Yes, I think I’ve even heard you use [the balloon analogy]. When I was at Caltech taking the relativity course, you gave a couple of lectures in cosmology. That would have been about 1971 or so, and I remember you used that. You also derived the Robertson-Walker metric in a very pretty way without any...
Without any relativity, yes.
…without any relativity, just symmetry principles and so forth. That’s a diversion, but do you find that you use visual images much in your work?
Oh, quite a lot. That also is something that happens more and more as you get older and are less able to figure out Christoffel symbols [laughs], differential forms, and such — you rely on mental pictures. But I always have, to a very large extent. That’s just the way I think. Unless I can make an image for something, I don’t really feel that I understand it. There are dangers in that because you can often make images that are not right. I think a lot of people manage to do that. It’s very powerful. I have never been a very high-powered technical person. I’ve never been able to push indices around and such very well, and it’s been a very powerful tool in the past, being able to visualize things geometrically. It’s one reason that I liked the Misner, Thorne and Wheeler book so much because it’s sort of Johnny Wheeler’s approach to the world and I find it enormously attractive. Also it’s Dick Feynman’s approach to the world. I’m less able to understand what he does, but it’s a very nice way to look at things.
Steve Weinberg’s book, of course, is almost the opposite. I asked him this question, and he said that he doesn’t use visual images at all.
I believe that [Both laugh].
In fact what he says about elementary particles is that once you describe the symmetry properties of a particle with the various groups that it’s a representation of, then you know everything there is to know about the particle.
I believe that, too — I mean I believe that he thinks that.
It is very close to Bishop Berkley’s idea of dispensing with material reality altogether.
Yes right. I could not live in an immaterial world, [laughs]… unless I had a wall to climb.
We’re getting close to the end here. Do you ever try to visualize the Big Bang?
I used to, still do to some extent. I have all the naive wonders about where it happened and when it happened and still no matter how much you know, and know those are silly questions, they’re still sort of there.
What do you picture when you think about it or maybe you don’t picture it but...
Well, I doubtless picture it, but I don’t know whether I can describe what I picture. I don’t know, Alan. I don’t think I can answer that in any reasonable way.
But you have tried to picture it?
Oh yes, yes. One can think about the physical conditions and so on. I sort of make a movie — always going backwards. I never go forward in time, always going backwards.
Start with what you know, right?
With what you know, that’s right.
Let me ask you a little bit about theory versus observations. There’s certainly been a lot of controversy or conflicts between the two in cosmology in the last ten years. And your being both a theorist and an observer, it seems to me that you would be in a better position than most people to talk about how well the two have lived together in cosmology, let’s say in the last ten years. Do you have any comment on that?
It’s a very complicated topic. I think that the interaction has not been either very strong or very happy. There are two problems, one in the theorists’ camp and one in the observers’ camp. The theorists, especially in the last few years, have had this problem that the multiplicity of ideas has become so large that in some ways it has affected the seriousness of the field. I don’t know whether I can make that clearer, but there are so many ways that things could be that the fact that the observations appear to be at variance with one or the other of them is not...
I think I know what you mean.
The practitioners have so many things at their fingertips that the fact that you knock down one of them [theoretical idea] doesn’t make very much of an impression. In fact, it doesn’t make enough impression for them to abandon that one, much less any of the others, right? [Both laugh] The problem with the observations is, as always, that cosmological observations are always right at the hairy edge of the possible. Observers tend to over-interpret their observations. Theorists tend to over-interpret the observations even more.
Why do you think people tend to over-interpret?
I think it’s just ordinary human zealousness. You get a result, and you want to believe the result even though you know in your heart of hearts that it’s a little, maybe even a little more than a little, dicey. So there are a lot of things — I don’t want to jump on horses, but... A lot of the stuff about the large-scale structure, from the ‘observational’ point of view, I think, is not very well substantiated by the data. And yet it drives the theorists into feeding frenzy, right, because they have various things to say about it. So there’s a great deal of Brownian motion as a result in the field: theorists trying to explain the results of less than completely exemplary observations; observers going out getting observations to test various theoretical ideas and coming back with results that, one way or the other, are not necessarily believable. It’s something that will settle out, I’m sure. The catalog of the things... the correlation of what is really true about the universe and the set of notions that we think are true about the universe I think is not very high at the moment. We don’t understand the dynamics of galaxies very well. We don’t understand the velocities of galaxies very well. We certainly don’t understand lots and lots of things that the old-timers would worry a lot about, like the effects of obscuration on a galaxy, for instance. People now apply these cookbook things to their catalogs. It makes a lot of difference when you’re doing things like large-scale screening.
You think that’s because the big ideas have gotten almost within reach, and so we’re just reaching a little further, a little further
That’s right. People are reaching further without really carefully going over the ground. It’s forgivable, I think. It’s human. But I think that that ground is going to have to be tilled a little more before we can really believe a lot of the results that come. It’s in a way unfortunate because it must in some sense slow down real progress. It’s much better to build in a systematic way, but it’s very much more exciting to do what people are doing.
Once in a while some really unusual idea, like the inflationary universe model, will come out — which probably couldn’t have been gotten to by a very slow, linear...
I think that’s absolutely right. That’s absolutely right. We all know that progress comes in those two forms, and the important thing to do is to balance them in some way. I think that the galloping ahead now is the mode that we’re in, perhaps unfortunately. But you know it’s a very exciting thing to do, so people are going to do it. You can’t change the way people work.
Let me ask you to take a big step backwards and be as speculative as you’re willing to be and put a little of your scientific caution aside. If you could design the universe any way that you wanted to, how would you do it?
I actually can’t imagine a much prettier way to do it than the current notions of inflation and the idea that the universe is Infinite with a capital ‘I’ in a way that we couldn’t think of before — that there are regions of the universe far away which are very, very different from ours and that that reality will come crashing in on us in some exponential future, and that essentially every experiment has been tried somewhere in the universe, perhaps even to the extent of changing the physical laws because the symmetry may have been broken differently in different places That is just such an immensely appealing — it’s chaos — but it’s such an immensely appealing notion, that it should have been done in a way that lets everything.. happen the way it is without any forethought or any need for special this, that, or the other. It explains in a natural way the seemingly special things about our universe without imposing...
... any overall plan
... plan on the thing. I find it philosophically just enormously satisfying It worries a lot of people, I think, because since the horizon is the horizon, there are questions about this big universe that we can’t ask in principle. That would have bothered me perhaps even a few years ago but it doesn’t bother me now.
What’s changed in your thinking?
Oh, just this notion. That was unthinkable. Well, not unthinkable, it just hadn’t been thought of a few years ago and one was under this immensely self-centered notion that the rest of what — we knew couldn’t see all the universe and my own notion was that the universe was infinite and I had the temerity to think that the rest of the universe was just like the part we see because the part we see is homogeneous and isotropic, right? And that’s just absolutely crazy. Its counter to every kind of Copernican development that has come in the world and this is kind of the ultimate Copernican ideas that not only are we of no conceivable consequence, but even our universe is of no conceivable consequence. That’s very pretty.
Your answer to this is closely related to another question that I was going to ask you, and you’ve about answered it. I was going to mention that somewhere, I think in Steve Weinberg’s book on gravitation and cosmology, is the astounding line — astounding just because it’s out of character with the rest of the book — that the more that we learn about the universe, the less it seems that the universe has a purpose. Have you ever thought about this question of whether the universe has a purpose? Maybe you’ve just answered the question already.
That’s a hard question, and I’m not a religious person — have never been since I was a kid. I toy with the idea that the only conceivable purpose that I’ve ever seen the universe might have had is the development of things that can understand it. I find that a very attractive idea, but I can stand back and realize that I find it an attractive idea because I’m a human being, and I like to think that we’re important, right’ So I’ve never been persuaded of its correctness. It’s not, I think, at variance actually with these new ideas, because intelligence can take probably as many forms as the universe can take. So I think that one could entertain this notion even in a chaotic universe of the sort that is envisaged today, but certainly one is not compelled to. I’m not quite sure what my own feelings on the subject are.
But you mentioned just a few minutes ago that if you could design the universe any way that you wanted, you would find the present universe — which can live happily without any grand plan — to be very appealing.
I find that to be very appealing, that’s right. But there’s the set of meta-questions. The universe exists, and you have to ask: “Well, why?” It’s the same question as before, the universe is just different than the way you looked at it a while ago. And the answer is on a sort of exponentially grander scale maybe, but it need not be a different answer. Perhaps if I thought about it more carefully, it would change my views on the subject, but I don’t ... To the extent that I have thought of it, it hasn’t changed my views on the subject, but I don’t know exactly what my views are. So...
That’s about everything that I wanted to ask you. If there’s anything that you want to elaborate on or any additional points that.
Nothing really comes to the top of my head. I’m very worried... I’ve only said a very little bit about the field, but I’m very worried about it, actually. I’m very worried about research in general. These are not scientific remarks; they’re sort of political remarks. There has always been this question — the conflict between little science and big science. I think that our lifestyle — the way we’ve been doing things — is very, very much in danger, and I see almost no way around it. I think the collapse of NASA is kind of a paradigm for what’s happening to science in this country in general. I worry about it, and I think that people should be more worried about it than they are. I haven’t any idea what to do. I’m not quite sure — I mean it’s fairly clear that the individual can’t do very much, but I think we’re in serious trouble. It’s especially sort of sad in this field because things are so exciting now but... I can see the possibility of the field really grinding to a halt. The whole idea of success-orientation of projects, I think, has now caught up even the ground-based [observations]. People have these grandiose ideas to build telescopes whose technology is completely untested. They go out and they scrape up enough dollars to do it, but not enough dollars to have any contingency. I can see several disasters on the horizon in the field in general. So I think we’re very healthy at the moment, but I see that we might not enjoy that health for very long.
When you mention this in relation to the big science-little science issue, do you mean that the projects are costing so much money now that we’ve been forced to not have any backups?
To not have any backups. And furthermore, when they get into trouble, they have to be so expensive that the agencies have to reach in their pocket the only place they have left, namely the support for the little guys, and that’s where it’s going to come from. It’s happened in NASA, it’s going to happen in the NSF. Now when they thought they had a bigger budget, they were going to make these centers. Everything is becoming concentrated and big. Freedom to do the kind of research that has led to all this stuff is going away, so that’s a very negative note.
 F. Hoyle, Frontiers of Astronomy (Heinemann: London, 1955)
 George Gamow, One, two, three…infinity (Viking: New York, 1947); The Creation of the Universe (Viking: New York, 1952)
 C.W. Misner, K.S. Thorne, and J.A. Wheeler, Gravitation (W.H. Freeman, San Francisco, 1973)
 A. Guth, “Inflationary Universe: A possible solution to the horizon and flatness problems,” Physical Review D, vol. 23, pg. 347 (1981)
 R.H. Dicke and P.J.E. Peebles, “The Big Bang Cosmology-Enigmas and Nostrums,” in General Relativity: An Einstein Centenary Survey, ed. S.W. Hawking and W. Israel (Cambridge University Press, 1979); the flatness problem was actually stated earlier in R.H. Dicke, Gravitation and the Universe, The Jayne Lectures for 1969 (American Philosophical Society, 1969), pg. 62
 stated qualitatively in R.H. Dicke, Gravitation and the Universe, the Jayne Lectures for 1969 (American Philosophical Society, 1969), pg. 61; stated quantitatively in S. Weinberg, Gravitation and Cosmology (John Wiley: New York, 1972), pgs. 525-526; the problem was probably under stood well before these references
 V. de Lapparent, M.J. Geller, and J.P. Huchra, “A Slice of the Universe,” Astrophysical Journal Letters, vol. 302, pg. L1 (1986)
 H.P. Haynes and R. Giovanelli, “A 21 Centimeter Survey of the Perseus-Pisces Supercluster, I. The Declination Zone +27.5 to 33.5 degrees,” Astrophysical Journal, vol. 90, pg. 2445 (1985)
 S. Weinberg, Gravitation and Cosmology (John Wiley: New York, 1972)
 Editor’s Note: This statement actually appears in S. Weinberg, The First Three Minutes (Basic Books: New York, 1977), pg. 154