Notice: We are in the process of migrating Oral History Interview metadata to this new version of our website.
During this migration, the following fields associated with interviews may be incomplete: Institutions, Additional Persons, and Subjects. Our Browse Subjects feature is also affected by this migration.
We encourage researchers to utilize the full-text search on this page to navigate our oral histories or to use our catalog to locate oral history interviews by keyword.
Please contact [email protected] with any feedback.
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 Jim Peebles by Christopher Smeenk on 2002 April 5, Niels Bohr Library & Archives, American Institute of Physics, College Park, MD USA, www.aip.org/history-programs/niels-bohr-library/oral-histories/25507-2
For multiple citations, "AIP" is the preferred abbreviation for the location.
The interview focuses primarily on Peebles' many contributions to physical cosmology: research on nucleosynthesis in the early universe in 1965, the theoretical and observational study of large scale structure formation in the 70s, and the development of the cold dark matter model and numerical simulations of structure formation, to mention the most prominent topics. Peebles describes his interactions with colleagues and other influences that shaped his research, as well as describing his own research style and the style of physicists he admires (notabley that of his mentor, Bob Dicke). He discusses in detail his response to and assessment of various other topics in cosmology, MOND, various structure formation scenarios, and quantum gravity. Brief discussion of the institutional support and funding for cosmology and how that has changed over the course of Peebles' career. Peebles also describes several of the changes that have taken place in the practice of cosmology, from describing the introduction of numerical techniques and increased interaction in particle physics, to the increasing pace of research.
Editorial Note: Bracketed comments in the text labeled [JP: ...] were added by Jim Peebles on Feb. 19, 2003; other bracketed insertions were added by Chris Smeenk.
I wanted to start off this morning by asking you about something that we really didn't discuss much yesterday, which is teaching and what courses you've taught, how you see teaching as opposed to research, how you balance those.
Yes. I think they're a valuable complement. Certainly to me it has been valuable to have to think through the basics of physics in order to present them in a halfway coherent form for a course. That has led me to ideas in research. Even freshman physics leads to thoughts that lead to other thoughts that are stimulating. So I think for me it is valuable to have been at a teaching institute rather than, for example, a pure research place. I think that my research is valuable to my teaching. I think that the two complement each other and I'm able to present somewhat more stimulating lectures because of what's happening in research, so it's a good complement. We understand it's a very expensive operation. This is an amazingly inefficient organization but I think it does play an important role. I retired because I felt, in part, I felt I taught every course I want to teach enough times. It's stimulating to teach a new course. To teach a course three times in a row is, I think, about the maximum for me. On the second year - you know, the saying is that first year you learn how to teach the course, the second year you do it right, and the third year you're coasting and you had better move on to something else. I have taught a mixture of undergraduate and graduate courses, and found them both stimulating.
And the textbook on quantum mechanics that you published? Did that grow out of undergraduate or graduate courses?
That grew out of an undergraduate course that I taught quite a few times. I was not ever able to find a single textbook that really pleased me, so early on I started writing down on ditto master sheets - perhaps you have never seen them.
I have.
But they are rapidly becoming an antique. I wrote out the equa- tions with no words and each year I would add a little to the equations. Then I learned TeX and I noticed I could write out the equations more clearly. Then I noticed I could put in a few interpolated words. And so through the years the notes grew longer and longer and longer, until finally - it was very relevant that a young student got interested enough to go over the notes with care and ask me why did I do this and that and the other thing. And that led finally to a publishable-enough book. It was therefore pretty painless.
Who was this student who looked over the-?
It's embarrassing to admit ...[ laughs] I can't tell you. He went to the University of Texas. I haven't heard from him since. Since I can't put my hand on a copy of the book I can't even - I can. I saw him once at the University of Texas a few years after he graduated here. It's awfully handy to have the feedback of students each year, their complaints about the notes and that I could put into the next edition. David Riley. Yes. David got very impatient with me at times. "Look, I can't understand this at all." And that was valuable input. And I got a lot of that through the years on that course. I don't think the book is particularly good for everyone, because it's my own idiosyncratic approach to teaching. The quantum mechanics is entirely standard, but the approach is the one that I like and I don't think it's for everyone. It continues to sell. It's not a ... [best seller] but of course I'm not in this business to get rich.
Right. And were there any courses that you really enjoyed teach- ing? It sounds like you taught pretty much the span of the curriculum in the physics department?
Yes, that's right. I think in the end I got very pleased with a course "Introductory Quantum Field Theory" for graduate students. You know that we teach two graduate quantum mechanics courses. One is for people who don't have much preparation in quantum mechanics. It's mainly a service course for people outside this department that covers about the contents of this book in two terms. Then for students who know quantum mechanics pretty well but need a little brush-up in fine details the same amount of material and more is presented in one term at breakneck pace, which I taught several times and enjoyed. But then it was the case that at the second term of that course - the high-speed one - was an introduction of quantum field theory, which gave you one term to do just the interesting things in quantum field theory, none of the heavy calculations. There wasn't time. But there was time to do interesting things, and I really liked that course.
What sort of things did you cover there?
Oh, for example, well, Hawking radiation. Turns out if you go to a de Sitter spacetime it's easy to write down the analog of Hawking's radi- ation. A couple of lectures and you've got it. It's really charming. Applied to a scalar field so you don't have all of the complexities of indices for a fermion field. But you could see all of the physics, easily explainable, in nice form. Then again, instead of going into quantum electrodynamics and all of the detail of renormalization groups and the rest - of course that's excellent physics but also pretty heavy - one could do fun games with what has come to be called- I'm trying to remember the name now. Box QED? A simple example, put an atom that is in an excited state and going to radiate near a conducting plane and you affect the radiation rate, both in direction and half life. Box quantization, yes.
That's not just putting periodic boundaries-?
Well, it is. It is. And you can make the atom do funny things, but by changing the boundary conditions by hand and putting a conducting plate or a cavity around it. And fairly easy to compute, and again charming to see the things that can happen, and the theory end can be observed in the laboratory. But then again interference with radiation is often found. It's sort of startling to think that photons that come from opposite side of a star can interfere with each other. It seems acausal, but they do. Easy to analyze. Elements of quantum field theory applied to solids. They are also easy to do, as far as superfluids. That's where I started. That's a course I really enjoyed. I'm not sure it's continuing. There is so much pressure on courses that are more immediately important to the student. That was a course more for students who weren't planning to go on to high-powered quantum field theory or superstring theory who wanted to have some insight into what goes on on that circuit. In fact it's a sort of ... You mentioned that the philosophy department here has someone interested in quantum field theory. It's at that sort of level: what does quantum field theory really do and how do you use it.
I wanted to ask you another I guess more personal question. I don't know if you hold any religious views, but if you do, how do those interact with your research work?
I don't. Actually, I guess the term I like to use is a convinced agnostic. I get offended by people who try to give me religious arguments. Why should I pay attention to these arguments? But I also get a little offended by people who tell me, "Of course, religion is bunk." How do you know? It's just an entirely different field of operation and actually I do like the words and music of some religions, so I have sat with pleasure through services - aside from the sermon. So no, I don't have any religious feelings at all. I often wondered what fraction of my colleagues are religious and how does that compare to the general population. What do you think? Have you noticed that there is any difference?
Just based on the interviews collected in Lightman and Brawer, very few people that they interviewed among the cosmologists professed any deep religious faith.
Huh. That's a good remark. It hadn't occurred to me to check there. I know a few respected colleagues who are religious, and what strikes me is I notice very little effect on-
Yes, if you look at the research work I don't think you would ever notice a difference-
That's right. That's right.
-that you would be able to say, "Well, here is clearly the research work of a religious person."
Yes, yes.
But I don't know in detail like what the statistics are and whether there is a difference between cosmology and other areas of physics, but-
It's certainly not what drives me, and I think most cosmologists are not driven specifically by religion.
Have you ever had, say, a strong response from people around you, maybe family members? Have you ever had the discussions about how cosmology relates to religion with people close to you?
Oh yes. My sister June's eldest son Craig is quite religious and is a little offended at my approach to the world. We have had discussions. He wants to save his immortal soul and he doesn't want to tarnish it by these bad thoughts. So it's fine. We get along well, but I try to avoid discussions of religion with him. I've given lots of public lectures, and on rare occasions have had people object to my comments on religious grounds, but very rare. It's just been a handful of times that I've had in particular public objections to my comments. I more commonly have had people approach after a public lecture with private concerns, but it's not been very frequent.
And are you the target of a lot of crank letters and postcards and that sort of thing?
Oh yes, yes, yes.
I would imagine there is a steady flow in?
Behind you, you see the big box? Well, it's full of questionable correspondence.
Oh, wow.
What I do is I set up such a box and when it gets filled to overflowing I toss it out and start again. I've always had this vague feeling that a student of sociology, or I'm not sure what, would be interested in looking through such correspondence and looking at what patterns emerge.
Yes. One interesting thing, I don't know if - there's a Gravity Research Foundation. I'm not sure-
Babson.
Yes. Have you submitted essays?
I did once, years and years ago.
It's interesting, because there's a selection of essays that are this type of essay that are always contributed to that. And in the early days it was mostly those kinds of essays.
That's right.
And then I don't know who ended up taking charge of the com- mittee but some-
I guess someone must have decided to get some experts on that selection committee. Nowadays it's very, very good people.
Oh yes, and very well known. I mean, the people winning are doing very good work.
Right. And of course it doesn't hurt that it's a fairly rich prize.
Right.
Do you know how much it is?
I don't know how much it is now. In the early days it was - even in the early days I think it was maybe five or ten thousand dollars, which was, in the fifties and sixties is quite a bit.
Big money.
Okay. I wanted to then shift and talk about - there were some recent observational results that we discussed briefly, but I wanted to focus on it in a bit more detail. First, the supernova IA results of Perlmutter and Reese, those groups.
Yes.
What's your initial assessment of those? Are you worried about possible systematic effects?
Yes. I mean, we all are, because the measurement necessarily is very indirect and there is a well-founded, well-accepted solution to this worry - get independent checks by other methods, which is happening. If we only had the supernova results I think most people would not be talking about lambda, the cosmological constant. Well, I mentioned to you, didn't I, that in a recent review this guy Bruno Leibundgut from Switzerland showed that you get a remarkably good fit in the old Milne model, which is simply Minkowski space time. No one would have paid attention to that in any case, but to draw the inference that lambda is present from that data alone is I think not "on." That's not to criticize in any way Pelrmutter and the other group.
Is it Reese? Is that right? Or is that-?
Reese is one of the leaders. The principal investigator - oh, why can't I remember his name? - they did magnificent work and I hope they continue. They certainly intend to with this SNAP satellite. But I think they would agree that we have to have an independent check before we buy it.
So it is the worry that there might be evolutionary effects that we just don't know how to control?
Yes.
Okay. And so related to that, do you think there's now enough independent support for this idea of dark energy?
Remember, we do have an independent check. It comes through the microwave background anisotropy and the position of that peak, which argues strongly for zero space curvature. Now again, one has to be a little careful because one can write down other models for structure formation but put the peak in other places for a given cosmological model. So far no one has been able to find any other of these models for structure formation that fits all of the measurements in detail, but no one has looked very hard, which is something that worries me. I can offer as an example a paper that I wrote with Sara Seager and Wayne Hu, both of whom at the time were over at the Institute for Advanced Study, in which we were able to contrive - and it is a somewhat contrived case - a model in which the universe has no dark energy, no lambda. It has the peak in the right place. The details I think we needn't bother going through. It is a cautionary illustration that things could happen in such a way as to mislead us on this point. We have to remember the universe is a complicated place. We jump on the simplest possible model of course, but the simplest model is not always the best.
So you would characterize dark energy as the simplest way for accounting for the supernova results?
Yes, and the anisotropy measurements.
Right.
But I would not say that it is yet definitively demonstrated ... It's not yet a compelling case. It's getting there. In another few years, with some more tests coming in, we might be there. I don't think we are there yet. In the draft paper I gave you, you'll see this described at length. [JP: And the WMAP results just about complete the demonstration that Lambda/dark energy exists.]
Are there some other exciting observational results in the last ten or fifteen years that have really shifted your views about the overall best fit model that we didn't discuss or that you want to discuss in more depth?
The big excitement we haven't mentioned is the observation of galaxies at high redshift. It is just so very impressive that one can see galaxies in considerable detail back at a redshift of three. You know, that's an expansion factor of four. The universe was a very different place then. The galaxies were very different too, but observable in remarkable detail and so there is now a rich fund of information about what galaxies were like as a function of time back to a time when the universe was considerably younger. The interpretation of these observations are still very much a matter for debate, heavy discussions, and conferences. I'm not sure that the definitive pattern has yet emerged, but it certainly has the potential to give us a convincing story for how galaxies formed. I for a long time had the belief that galaxies formed early because they're so dense, and the way to make them dense is to form them when the universe was dense. That's not indicated by this CDM model. It likes to form galaxies fairly recently and then to make them dense by having a lot of collapse — which seems to me to be, while certainly physically possible, the wrong first approach. So that has led to a considerable division of opinion between me and the community on when galaxies form, which is kind of fun. Certainly my old view that galaxies formed back at redshift 30 can't be right now, I think.
Why is that ruled out now?
Why is that ruled out? Because - well, you know I'm not sure it has been ruled out, but it gets embarrassing because if galaxies formed that early and made lots of ionizing radiation it would re-ionize the intergalactic medium too soon, and an ionized intergalactic medium scatters radiation and smooths out the microwave background radiation too much. So it had better not happen that soon. Also one sees a lot of evolution of galaxy properties - at least apparent evolution - from redshifts 3 down to redshift 1, indicating that they are still adding material, I think. In the next few years I think the observations are going to get rich enough that we'll come to an unequivocal interpretation, a clear interpretation of them and we'll know just how galaxies formed, at least in a broad sense. [JP: Another WMAP advance is the constraint on reionization. Reionization at redshift is convincingly ruled out; reionization at redshift somewhere around 15 seems to fit the observations.]
So you said your views differ from the community.
Well, there is a lot of acceptance of the idea that many galaxies formed at redshift less than 1. There is some evidence of it. In particular one sees obvious cases where galaxies have merged, which is a process of formation of course. A famous example that maybe you've seen pictures of is a nearby elliptical galaxy called Centaurus A. And running across a spiral galaxy that evidently is just plowed into it. Here is an example of galaxy building. There are a few famous examples of this galaxy building, but what impresses me is that wherever you see these discussions of galaxy building at low redshift you see the same old images. In other words, the building process certainly does happen. No question about that. But the rate looks to me awfully small.
Right. So it could just be a random collision.
It's the occasional random collision. It's the tail end of the building process in my mind. Then again, well if we want examples, you know our galaxy is surrounded by gas clouds that have the familiar name high-velocity clouds, atomic hydrogen clouds that are coming toward us. Some people think that this is an example of galaxy building, that these are clouds at considerable distance that are falling into us to build up the massive galaxy. My bet is that that's exactly wrong, because if our galaxy were building by clouds falling into it, we should be able to see the same effect around other galaxies, and we should see clouds around them - observable in 21-cm emission. And you don't. They are very rare. So again a heavy debate between those who accept the notion that our galaxy is still growing at a pretty substantial rate through accretion of gas clouds and those who doubt it. I shouldn't say I stand alone on this. There are a set of issues, and on each case there are certainly two sides. Quite a few people doubt the idea that the Milky Way is going through accretion of gas clouds. And quite a few people doubt that the elliptical galaxies could have formed at redshifts less than 1. But I think the community view - communities do establish views, and they are not always rational-
Right. In one of your reviews you mention the element of social construction you see.
Yes, there is. This is by no means a bad thing. It's the way people operate. They tend to reinforce each other's opinions. A good example is the Einstein-de Sitter universe. People held onto it much longer than I think they should have. And then I think they swang over to the dark energy universe much more firmly than was justified by the evidence.
So you think that now the conventional wisdom is, say, omega matter .3, maybe lower, and then dark energy making up the-?
Right. The rest. That is now the community view, and I think the community is more firmly in favor of it than is supported by the evidence, just as five years ago perhaps the community was more firmly in favor of a Einstein-de Sitter universe. There's no complaint in that. It's just an observation of I think the way communities operate. [JP: and now WMAP adds considerable evidence to Omega matter close to 0.3 with most of the rest in dark energy!]
There is one observational result you mentioned in one of your review articles which I didn't really understand. It was Bachall's work about how you can look at galaxy clusters and get some cosmological tests out of those.
Yes. Yes. For example, one can look at the motions of the gas and stars in the outer parts of our galaxy and from their motions try to estimate the amount of mass needed to provide the gravity to account for those motions. Now you notice that if you look at motions in the inner part of the galaxy you are quite insensitive to the mass in the outer part of the galaxy. Remember Einstein or Newton's iron, sphere theorem: Inside an iron sphere you see no gravity. And so when you look at the motion of the inner parts of the galaxy you don't see that outer halo. It means that as you look to greater and greater distance you may expect to see more and more of the mass of the galaxy. And indeed, as you increase the distance at which you make these observations of motions of stars and gas you do see an increase in mass. But then Neta Bachall and others emphasize that as you keep increasing the length at which you make these measurements you find that the mass stops growing and you reach an apparent plateau. This is part of the biasing concept we mentioned yesterday. For under the biasing picture much of the mass of the universe is distributed in the outermost parts of the concentrations of galaxies and will be seen as a contribution to the relative acceleration of matter only when you get to quite large scales. That effect that you might have looked for is not there. It was one of the reasons why people eventually have given up the notion of strong biasing.
And are there any other observational, important results that we haven't really discussed, that had an impact on your thinking?
I mentioned to you a point that I'm very excited about these days. First the general point that general relativity theory is wonderfully successful, but nonetheless we're making a long extrapolation, so we should be reconsidering the tests of general relativity on the scales of cosmology. And I had been collecting these tests. They are summarized in the review that I gave you. And I'm surprised at how demanding these tests are becoming, and so far how successful general relativity has been. It's quite impressive. And in particular I'm here scribbling away on this test of the inverse square law but on large scales, and it's just startling. The constraint is not that dramatic; this is not a precision experiment of the sort that we have on terrestrial scales. I think we will be able to say that the inverse square law, 1/R² for long range gravitational interactions can be much different from 1/R(²±.03). It's impressive that you can= put such constraints on what's happening at such enormous scales.
Right. So you are talking the scales of galaxies?
And out to the Hubble length.
Oh, out to the Hubble length. Okay.
Yeah. So these are-
[One of the most impressive] things about physical science is how broad is its reach with such few basic assumptions of the laws of physics. To think that the inverse square law that's so familiar here on Earth can be extrapolated all the way to the Hubble length without failure or a dramatic failure is so impressive.
Yes. So you've emphasized that you would be interested in seeing more observational work testing general relativity cosmological scales.
Yes, and it will happen. It is happening.
Are there any other areas of observational work that you think have been neglected or that you would like to see being focused on?
Of course I can echo a complaint that many have in extragalactic astronomy and cosmology that the small-scale structure of the universe is understudied these days. It's again only human nature to want to go after the big problems and the front line research, and in astronomy that often means that one is neglecting areas of research that could have been done decades ago but people haven't gotten around to it. There is still work on the structure of our galaxies and on the relative motions of the nearby galaxies. What are things like on the scale of individual galaxies? Not as much work on this subject as there could be, simply because it lacks the romance of looking at a galaxy at a redshift of 3, I think it's fair to say. This is again no complaint. This work will get gone, more slowly I think than would be rational, but it is being done. I'm desperately trying to think of an example. Well, we mentioned the notion of galaxies merging to build bigger galaxies. In some of these high redshift studies of clusters of galaxies people have noticed there is a curious number of pairs of galaxies — two very close together — it looks sort of like a dumbbell — and have said, "Look. Here are galaxies that are in the process of merging." And that may be true, although it's not the place where you expect to find merging. In a cluster the galaxies are moving fast, and they tend to just rip through each other. But never mind. There is I think undoubtedly a tendency for galaxies in high redshift clusters to appear in pairs. And furthermore, naive estimates of how long it will take for them to spiral together are quite short. But I was reminded of an observation made by an old friend, Herb Rood, and another acquaintance, Mitch Strubell decades ago that in nearby clusters of galaxies there's a curious tendency for galaxies to appear in pairs like dumbbells. So I managed to get some of the old photographic images — that are now available on the Web. That's amazing. And yes, there they are, little tight binaries. They look a lot like what we see in these images from high redshift. So I asked a few people, "Do you believe these binaries are there and real?" You would check by doing some statistics. Of course some galaxies will appear to be close just because they are seen in projection. The check is easy. Look at the frequency distribution of apparent separations, angular separations, and see whether you see a sharp spike in the distribution. It hasn't ever been done. It would be dull as dishwater to do it, but it really ought to be done, and then it ought to be done as a function of redshift to see if there is any tendency for this effect to become stronger with increasing redshift. No one I have talked to has expressed an interest in doing this. In fact they say, "Why should I waste my time on such a silly thing?" But it will be done eventually. It should be done. Oh, another example. I remember being very hot and bothered about this. In the cold dark matter model there is what seems to be a pretty reliable prediction that the dark matter that spreads into the outskirts of the galaxy goes into the center and piles up into a sort of cusp-like distribution. Perhaps you've heard this discussed. What is the name? It's the core problem, the dark core problem. Galaxies tend to have compact dark cores of matter not inconsistent with what you see in something like the Milky Way. But there are these galaxies that happen to have very little visible matter, so- called low-surface brightness galaxies, in which you can much more clearly trace the apparent distribution of the dark matter because there's not much luminous matter to get in the way. And there is a heavy debate going on about whether or not there is this predicted concentration of dark matter toward the very center of one of these low-surface brightness galaxies. I was just at the conference I mentioned in France, and there was a little bit of debate between me and Simon White over whether or not this is a problem. And I was unable to assure him that people have mapped the motions of the little bit of hydrogen that is observable in these galaxies in enough detail to test whether these cusps are really there. So here is a simple prediction. A simple set of observations match in all the detail you can ever master, a motion of gas in the center of this to see if it's swirling around a concentration of dark matter. I got in touch with a young guy who I mentioned yesterday in connection with MOND, Stacy McGaugh at the University of Maryland. He has been doing some work on this. He's feeling very frustrated about how his work is being played down. He's feeling a little marginalized, which is a very bad thing because he's doing good work. This is good astronomy. It's tedious astronomy, and not a lot of people are doing it. So I, talking to Julianne Dalcanton, who is at Seattle, University of Washington, she and I, she showed me an image of a very nearby low-surface brightness galaxy. It's speckled with these little concentrations of stars, each of which can be observed and its redshift measured to see whether or not these guys are swirling around a dark matter core, and I was saying to her, "Why aren't you measuring all these?" She said, "Oh, it's so tedious. We'll get around to it sometime."
So there is part of this which is motivational - right? - that these would be tedious observational projects.
Yes.
But like the example of McGaugh, is it also partially that this type of work isn't necessarily rewarded?
Here is an example where I think he is not rewarded as much as he should be. JP: But he has just been promoted to tenure at the U of Maryland.] By asking iconoclastic questions he's perhaps putting his reward a bit in jeopardy. I don't know how serious that problem is. I haven't experienced it, but then I grew up in a different age - and also I think I was much more conservative. Yes, I was. I was conservative. I think there is a danger that the person, particularly the young person, who asks unpopular questions will not be adequately rewarded. That's a concern. He is working with front line astronomers. Perhaps you know the name Vera Rubin.
Yes.
So she is working with him on this, and she is I think a considerable shield, just because she is so stable - you know, a rock solid person. There is no question. That is very beneficial to him I'm sure. So there are two examples here I guess. One is, yes, there are observational programs that I wish were more strongly supported, and I've given you examples. In some cases the support I guess is not present because the work is tedious; in some cases maybe because you are asking unpopular questions. Does asking an unpopular question put you in hazard? Well, perhaps in a way, because people are saying, "Why are you spending your effort and these very short funding dollars on a crazy question?" And that question has to be asked. "Why are you looking for binary galaxies that appear as dumbbells? We know they don't last any length of time. They must be rare."
There is the flip side of that question which is the tendency of hot research topics to draw a really disproportionate amount of effort, and I'm wondering if you see any trends in cosmology where you think there's been a lot of effort devoted to something which maybe is more speculative or hasn't necessarily paid off that effort.
Well, I would put this in two parts. Undeniably there is a strong impulse to go after the popular and rewarding problems, but I don't think that's wasted effort. People go after these problems because they are rewarding. The observations of high redshift galaxies are very difficult but so rewarding that they command enormous amounts of telescope time, computer resources and people's time. Whether or not it's a disproportionate amount is a judgment call I think you could argue either way. Without a doubt people race after the popular problem, no question about that. But for a rational reason: the good rewards. I'm not thinking of the rewards of support, monetary support, but the rewards of exciting new results. You go after the hot result. The somewhat, slightly counterproductive aspect of that is that I think the most popular areas get overcrowded. But that's a choice that each astronomer has to make, "Do I want to work in an area that is crowded and in which I will get results but always have to worry about the other people breathing down my neck, or do I go off into a less popular field?" But the odds are that the feedback in new science will be less. On the other hand the competition will be less.
I also wanted to ask you about a few more general issues in cosmology, and one is just the status of the cosmological constant. Since you have recently written this review article ...
Yes. I think I'd give slightly more than 50 percent odds that it's there in the form of dark energy or some analog. I wouldn't give 99 percent confidence. [JP: After WMAP, I go to 90 percent!]
So would you put your money on a true cosmological constant or something that is just the vacuum energy of a scalar field that's-?
Well, I much prefer the latter. That's only emotional. This cosmological constant has such an absurd value that I'm hoping it's not there.
It's got to be very, very close to zero.
It's ridiculous. Whereas a scalar field naturally would pass through this value on its way to zero.
And then the question is why are we around at precisely the time where it's just starting to-
Yes, yes. Bharat Rhatra and I had some heavy debates on whether or not to get into these anthropic arguments. If you read far enough into that draft you will see that we reduced it all to one footnote.
Well, that actually leads into my next question. Is there a proper role for anthropic reasoning in cosmology? ... I've seen people give talks where they call it the "A principle" or say, "I'm not going to mention the A word because it has a bad name."
Yes, yes. And yet, people who I deeply respect emphasize this principle very strongly.
Dicke for example-
Bob Dicke originated it, although I'm not sure how he would react to the present use of the argument. Well, we see Steven Weinberg, Martin Rees, Alex Vilenkin at Tufts, very capable people paying considerable attention to this principle. And I think they are right. I personally don't get into it, but that's a matter of taste. And I mean the fundamental point is really kind of elementary. We don't see people living in Antarctica. There is a very good reason. And we don't see people wandering on the surface of Mars, and there's a very good reason. We see them here for a very good reason, at least in part. It's an hospitable place. So that version of the anthropic principle, the universe had better not be only a million years old, is self-evident. Whether we can turn this into science of the sort we're familiar with is the open question.
I'm wondering what you think of a lot of the arguments involving, say, assigning a probability distribution over an ensemble of universes and then using anthropic reasoning in that context. Are you fairly skeptical of those?
I think it's good science, but I'm very skeptical of it. The trouble is - as you have noticed, I am a pretty pragmatic person. I would like first to be sure that this ensemble of universes exists.
Which is kind of hard.
There it is. So I support this science as good and interesting, I sit through talks, but I have never felt moved to join in that line of research.
This is related to a comment you made yesterday, that at some point you were interested in quantizing gravity. I don't know how serious that interest was at the time, but have you followed research in quantum gravity?
The interest was certainly real. I wrote a long paper which never got published, for good reason. Interesting considerations I still think, but very much less fruitful than the present line from superstring theory, a subject I have not attempted to follow. I'm aware of, you know, the basic arguments and I'm impressed with conventional GR, general relativity theory follows in a pretty direct way from these considerations. And I'm certainly hoping that this can be turned into a real physical science with solid predictions that can be tested in these observations. That hasn't happened yet. There are other approaches to quantum gravity physics, quantum cosmology, as Jim Hartle follows, for example. It's exciting, it's important, and they are addressing a real problem. But it's not something in which I can offer any informed comments.
I also wanted to ask another type of question, a very general question about the relationship between physics and astronomy. Historically these are two very distinctive fields with distinctive traditions and we're really at the point where there's a lot more interaction between the two fields. And I'm wondering how you see the two traditions interacting or how people from the two different fields interact.
I certainly agree with your comments. There are very different traditions. But present day astronomy I think of as a branch of physics, along with particle physics, biophysics, nuclear physics, condensed matter physics, astrophysics. We are very lucky that historically there has been an astronomy department in this university, so just next door there is a building full of what I would call physicists. It certainly makes our presence in physical science much more strong to have that whole department over there that is really a branch of physics. That is in no sense demeaning. It doesn't mean that they have advanced either. [JP: I have no idea what I was trying to say in the previous two sentences.] It's just that it's branch of physical science. Great. What will happen in the future? I guess I will be a little surprised if we keep a department of astronomy. It does seem a little irrational. I think many universities are merging their physics and astronomy departments. I don't know what's happening at Penn [State]. Perhaps in some universities there never was a separate department of astronomy.
Right. At Penn State there actually is, and I don't know if there are any plans to merge.
Right.
The University of Pittsburgh on the other hand has just one department.
Yes. You can see why it makes sense to have a separate department of molecular biology. In fact I was astounded to hear that their graduate students don't have to take any physics beyond freshman year. But then they have so much rich wetwear to think about and all of the phenomenology. I guess I would compare molecular biology to where astronomy was in the 1900s, in the 1800s, where there was not a tight connection to physics, to developing physics. There were connections of course, but not tight.
And so do you see astronomers becoming more and more like physicists in terms of their general outlook? Because it seems when I talk to astronomers or just talking to physicists they still have very distinctive approaches.
I guess so.
Maybe - sometimes that breaks down.
Well, of course, the approach is conditioned by the subject. If you talk to a particle experimentalist you'll get another worldview entirely. But their goals, their methods of analysis of results, the methods of presentation, the interpretations I think are very familiar. So, no, aside from the adaptation to the conditions of what you are trying to do observationally or experimentally I don't see much difference between the astronomers I talk to and the physicists I talk to. When I talk to a superstring theorist I'm talking a very different language from when I'm talking to a condensed matter physicist or to an astronomer.
Right.
But we are under the same big tent, whereas I'm not sure a molecular biologist is quite under that tent.
Right. ... I would think the area of early universe cosmology or very early universe cosmology is much more similar to talking to the superstring theorists.
It certainly is.
And do you see that as developing more towards connecting up with fundamental physics or more towards a richer phenomenology in the future?
I don't know. I hope it's both, of course, and I don't know which is going to advance more rapidly. It's very hard to judge. It will depend on how lucky we are on both sides whether people can make the theoretical breakthroughs that will make more convincing connections, whether phenomenology will show up that will give us new leads to what's actually happening experimentally. Hard to make the call.
So, looking back over your career, which has spanned a really amazing change in cosmology.
Yes.
Just the changes in the field, the changes in our understanding of the universe. What do you think have been the most dramatic changes, either in terms of how cosmologists work, the practice of the field, what people do when they're doing research, or in terms of changes in ideas?
I think both have changed dramatically, but I am impressed with the thought of how similar the ideas we work with today are to what you can find in Tolman's book. I think that if Tolman could be brought back and set before us and we explained the state of present research in cosmology, much of it would be familiar to him — the thermal radiation and of course the cosmological tests. The tests that we're now so excited about were in large part invented in the 1930s and described in his textbook. The subject hasn't changed all that much in some respects, and where it has advanced it has advanced in directions so far that are fairly natural extensions. So to my mind — and remember we had general relativity theory all that way back.
Right. And people had the Freidmann-Lemaitre models.
People had the Freiedmann-Lemaitre models all that way back. It is startling to think how little the basic concepts have changed. Now of course there have been dramatic additions to early universe physics and the dramatic addition of our understanding of structure formation. These are important. The analyses of galaxy evolution are now a rich subject; then practically nonexistent. So there have been great extensions in concepts, but I think the revolution - it's not been a revolution I would say so much as evolution in a fairly linear growth of concepts. The conditions of research have dramatically changed, of course. I stressed yesterday that when I was a beginner in this subject I could get to know almost every significant contributor to the field — there were just a handful. The subject was moving very slowly. It was physical science, but it was not very rich. Now it is an exceedingly rich and very active subject just in terms of the amount of research going on. There has I think been a dramatic change in quantity - and in quality too. One could write very simple papers in those days. Considerations were simple because they hadn't been thoroughly explored. Now there are very, very sophisticated analyses called for and being done.
So what do you think is different, say, in the training? Let's say you had a grad student walk into your office in 1970 versus now. What would you say the grad student now needs to have in the toolbox of methods and theories that they are familiar with in order to be doing research in cosmology?
Hah. Hah. I'm not sure. Since then numeral simulations have become a very active and important and rich subject, and the ability to do numeral simulations is something you can pick up yourself but you'll be very inefficient at it. That is I suppose an important part of the toolbox, to understand numerics. But then and now I would say the first thing you better understand is basic physics. I consider this a branch of physical science, as we have discussed, and that means you really have to understand what's going on in atomic physics, nuclear physics, subatomic physics, relativity physics. But these considerations are much the same now as they were in 1970.
I have just been struck, looking at — I don't know if you have, you probably do, Peacock's recent book, Cosmological Physics?
I know it but don't have it on my shelf.
Yes. I've just been struck by how much quantum field theory and other-
Well, that's a good example. Quantum field theory was not needed in 1970. Now it is if you want to at all understand concepts about the early universe. So I wasn't thinking about early universe physics when I was answering your last question. When I consider it I realize, yes, there are a lot of new concepts that go into inflation that one really ought to understand if one is going to take part in researching cosmology. It's a little ironic of course that we don't know if inflation is the right picture for the early universe, and therefore whether the tools of quantum field theory are relevant. I think the odds are good that they are, because we do believe quantum mechanics is an excellent approximation. We do believe, and we certainly have pretty good evidence that GR is good too. The two don't fit, but in limiting cases they can be accommodated. It's easy to do quantum field theory in curved spacetime that is classical. And that certainly sounds a lot like the early universe.
Right. Yes, it would be hard to imagine a limiting energy that would be low enough that you could apply something other than field theory.
Yes. So yes, that is an addition. A very solid one. In conferences that I go to I don't hear a lot about quantum field theory these days. I hear a lot about brane worlds, an attempt to add extra dimensions to make interesting cosmological models that more naturally treat dark energy and more naturally finesse a singularity and in a simple-minded solution. Many of these arguments strictly use classical physics. They don't even put in the quantum corrections. But no, I shouldn't have said I don't hear much about field theory. Indeed people often first use a classical treatment of brane worlds and then attempt to do a little bit of quantum field theory within the brane world context.
So by classical treatment you just mean lowest order [in perturbation theory], you don't add any — ?
These are classical fields, no quantum mechanics. Yes, I mean classical relativistic [field theory], but without quantum treatment of the extra dimensions, for example, or of space time fluctuations. But no, no, I shouldn't have said I don't hear quantum field theory mentioned much. I do hear it. On the other hand such a large part of cosmology happens in redshifts, let's say, less than 1020. It's an enormous, rich subject in which I don't see much use for quantum field theory. Some aspects of it are used to treat the large scale evolution of density fluctuations. The methods of quantum field theory are well adapted to perturbation theory and general relativity. And that is a rich subject. That's good science.
So let me ask you a slightly more speculative question. If you could have a cosmologist say for fifty years in the future, you're imagining resurrecting Tolman. If we could bring someone back.
Yes.
What are some of the questions that you would want to ask them about what the future holds?
Well of course I would want to find out what we have ... First question, what is the dark matter? Is it really a simple particle, say the lightest supersymmetric particle, or is it something more interesting? Is its physics really as simple as that of a point-like particle? Then next, have you convinced yourself that dark energy is there, and if so, what is it? And have you found some connection between the two? I would be charmed to learn what is known about how galaxies form at high redshifts, what is the sequence of events by which galaxies formed. Galaxies have remarkably rich systematics in them. Lots of regular properties that indicate to me they aren't just trash heaps. They are formed through very systematic processes, and have nice physical regularities that I don't grasp, and I suspect people fifty years from now will understand well, I hope.