James E. Gunn

[Begin session 1]

De Swart:

For most people, I just, I looked up the publications, of course, and looked there, so I did, but I didn’t make like really extensive notes on that yet. So, it would be nice if we maybe had another opportunity. So, I just, I have looked at your publications, because you did a lot, of course, with Rich?

Gunn:

Right.

De Swart:

At the time.

Gunn:

Right.

De Swart:

But also with Ed, right? I think…

Gunn:

Not so much with Ed. Ed was, Ed was a student at Caltech while all this was going on.

De Swart:

Ah, yeah. Yeah. Yeah.

Gunn:

I think.

De Swart:

Oh, that’s really interesting. Yeah. Yeah.

Gunn:

And so, I worked with Ed. Although, I don’t think ever on anything like this subject. So, I was working with Rich and with Schramm a little bit, and Beatrice Tinsley quite a lot. She was a postdoc at Caltech while much of this was going on. Although, you know, this stretched over quite some time. And I was not directly involved in any of the sort of dark matter detection things. That is, you know, looking at rotation curves and looking at the dynamics. I did a fair amount of work on the dynamics of clusters, but Zwicky had already written that chapter, I think, you know, many, many years before. It was clear that the timescales were such that they could not be coming apart. It didn’t make any sense anyway. You know, they were made out of old galaxies with stars that were ten billion years old, and there they were and one could account for only five, ten percent of the mass that was necessary to keep them together. And, people were just, I think, unwilling to bite the bullet and say that there was something there that they didn’t understand.

De Swart:

So, to just get a general idea, so you started as an undergrad where?

Gunn:

Oh, I was an undergrad at Rice, in Houston, doing physics and math. So, not even, not even in astronomy at all.

De Swart:

And, when did you start to do it?

Gunn:

When I went to Caltech in ‘61.

De Swart:

Okay. Yeah.

Gunn:

And, then I graduated from Caltech in ‘66, came here for a couple of years, and then went back to Caltech in 1970 and stayed there for ten years, and then came back here. So, I bounced across the country a couple of times.

De Swart:

And so, and who did you do your graduate with?

Gunn:

I did my graduate work with a guy named Guido Münch, who was a quite eminent astronomer at the time but not connected with cosmology at all. He was an interstellar medium guy. But he was also the only person at Caltech in the astronomy world that was at all interested and knew anything about statistics. And so, I did my thesis on the correlation function of galaxies. And so, I basically worked with him because he was the only one who knew about the kind of mathematics I was working with, although it wasn’t… the extragalactic stuff wasn’t his field. And then, but during that time I was also very interested in cosmology. I did a bunch of papers on lensing, some of the very earliest ones on lensing, actually.

De Swart:

Oh, really?

Gunn:

Looking at sheer and magnification. And of course we thought, at that time, that the universe was baryon dominated. Didn’t know any different.

De Swart:

Yeah. Of course not.

Gunn:

And there were great controversies going on about whether Omega was Omega baryon, people thought, because they didn’t know about anything else. And a great deal of questions about whether Omega baryon was near one or small. And I argued for a long, long time that it really had to be small because just from mass-to-light ratio arguments, basically. I mean, you look at clusters. You look at things in the field. And you need a mass-to-light ratio a thousand to, to make Omega one, and which is no way that the average mass-to-light ratio was a thousand, unless there was a lot of unseen stuff around. And everybody thought unseen stuff must be baryonic because that’s the only kind of matter we knew. Right? So, it was sort of in the ‘70s.

De Swart:

And this was at Caltech?

Gunn:

This was at Caltech. This, and this was sort of while I was still a graduate student, and after I came back in the early ‘70s. And I can now not remember exactly when Vera published her first papers.

De Swart:

So, that’s in 1970, but most—but, it wasn’t, it wasn’t cited that much at the time? So, it…

Gunn:

At the time. That’s right. Although, there was sort of growing awareness of it. And, as we talked about before, there were at least a few indications, both from the radio, from 21 centimeter, from Mort Robertson people, and from the optical, from Horace Babcock’s work much, much earlier, that the rotation curve for M31 was flat as far as you could see it, and a hell of a lot further out than, you know, than you could see any light.

De Swart:

Yeah. So, did you know about rotation curves at all? Like, people doing measurements and properties of galaxies?

Gunn:

Oh, yeah. Yeah. Yeah. And the Burbidges were the foremost people doing this, you know, with their, with—and, the Brandt curve was a perfectly physically motivated thing, you know. You have light and it won’t fall off because it’s kept, going to be Keplerian in that large radii. And so, that, those ideas, you know, died very, very ungracefully. So, but in the back of everybody’s mind was this knowledge that they just didn’t want to quite come to grips with. We already knew that there were vast amounts of unseen matter in clusters, and that just had to be. There was no way around that. And so…

De Swart:

And whatever for?

Gunn:

Now, people were saying, “Well, okay, now we see it in galaxies. Right? Because, Rubin keeps, she’s got all these rotation curves and most of them are flat or rising.” And so, it just had to be dealt with. And then, I think Jerry and Peebles’ paper, which was what, 1972, ’73, ‘74, something like that, kind of systemized it and put it together. Although, I mean, there was nothing terribly new in that paper. They just pulled things and made a kind of systematic summary.

De Swart:

That’s what Jerry said. Yeah.

Gunn:

Yeah.

De Swart:

Yeah.

Gunn:

And then, I think people sort of accepted that from then on. I mean it, and of course we did not know yet whether the dark matter was still in some, you know, snowballs, or rocks, or something, you know, but had to figure out how they would form in the early universe. And the sort of realization that it was probably not baryonic matter was kind of growing at that same time, but slowly. And, you know, the primary piece of evidence early on about that was the primordial nucleosynthesis, which just didn’t work if there were that many baryons in the early universe, but seemed to work pretty well if the baryon density was very low. And Omega was small. And Omega-matter was small. Sorry. Yeah.

De Swart:

And about what time was this?

Gunn:

Oh, it was also in the ‘70s, but I don’t know exactly when. But by the sort of 1976 or ‘77 I think it was, I don’t think it’s proven. How do you prove anything in astrophysics, even today? [laughs] But, but it was pretty widely accepted. Even earlier it was very widely accepted that it existed, that dark matter existed. I think by the late ‘70s it was pretty well accepted that it was probably not baryonic.

De Swart:

Okay. So, and by which community do you mean “accepted”?

Gunn:

By, well…

De Swart:

So, are the…

Gunn:

Astronomers and physicists, you know, that community is, the boundaries are very, are very ill-defined.

De Swart:

But also, observationalists?

Gunn:

Oh, now that’s a harder question. I don’t know. You know, observers observe, and some of them sort of keep up with theoretical ideas, and others don’t. It’s much, it was much, much less integrated, the community, than it is today. Because…

De Swart:

Okay. That’s interesting.

Gunn:

By and large, I think we’ve talked about this as well, people worked in at least small groups, and there were observing specialists, and there are sort of interpretation specialists, and there are theorists, and there are people who work with software. And so, the communication across fields, across subfields, is much better than it used to be and people, there were still people that were fairly isolated. And so, I cannot tell you whether the average observer accepted this or not.

De Swart:

But did you get preprints from observers?

Gunn:

Oh, yeah. Yeah. Sure. Sure. There was an active preprint thing. So, we knew what was going on. And, it soon became apparent, you know, people would say, “Well, maybe Vera’s wrong,” but other people started looking at these things too with modern instruments. And technology made an enormous difference, because when the Burbidges were doing this they had photographic spectrographs. There was no way they could look at faint low-surface brightness outer parts. And so, they saw what they saw, and they extrapolated using the Brandt curve because it seemed like a reasonable thing to do. Right? And so, it wasn’t until image tubes, and later CCDs, and by this, by the time, by the ‘70s people were beginning to use CCDs, came along that this became feasible to do, you know, technologically and observationally. And then, it sort of became accepted and the, I think probably the landmark theoretical paper – maybe there were two of them, and I wrote one of them, but that would, later. I wouldn’t, I would never have done this by myself because I don’t do that kind of physics, but there was a paper by Steve Weinberg basically saying that the amount of dark matter that we needed was consistent with the simplest kind of WIMP you could imagine, that had the right mass, that had the right decay times, everything. It looked to be a sort of electro weak-scale particle. There was no reason to assume that these things didn’t exist, but we didn’t know it because it was too massive to have made it in any of the, in any machines of that day. And so, suddenly there was a kind of theoretical candidate for this stuff that was at least sensible, within the confines of—that was even before the Standard Model—but within the confines of what we knew. And, then I and a bunch of other people wrote a paper called, something called “Consequences of the Existence of a Stable Neutral Heavy Electron,” or something. I think Steve was even a coauthor on that. I was working with him quite a lot. It went through, basically, all of the stuff about the development of structure, all of these things. And so, it said that this, you know, this thing, Steve said this thing might well exist in approximately the right quantities. I went through the development of the perturbations and it showed that it would, in fact, help the galaxy formation problems, because it was cold and had a low, you know, a low… we were expected to have low-velocity dispersion. And so, it eased an enormous number of questions that were there at the beginning. One of the big questions was, “If the perturbations in the baryons were large enough to make galaxies that we see today, the microwave background would be much less smooth than it was observed.” Okay? But if you have dark matter, and have something heavy, then perturbations in it can grow for a long time even before recombination. Whereas the radiation pressure is keeping the baryons fairly uniform.

De Swart:

Oh, okay.

Gunn:

And that helps you by a factor of ten, and that’s what you need basically.

De Swart:

In the perturbations?

Gunn:

Yes. Right. Right. Right. So, that was another sort of circumstantial thing that pointed to some non-baryonic form for the, for the dark matter, that and the primordial nucleosynthesis, were I think the big thing.

De Swart:

So, this was beginning of the ‘80s?

Gunn:

Yeah.

De Swart:

Yeah.

Gunn:

Yeah.

De Swart:

So, what were you doing, like before you read Weinberg’s paper? What were you…

Gunn:

Uh, well I was still doing cosmology and worrying about mass-to-light ratios, and Omega, and things. I wrote some papers about the local group. Tinsley and Richard Gott and I wrote a paper sort of putting together all the arguments for Omega-matter being about .2, and Omega-baryon being even smaller than that. So, I was sort of working in this field but not, not directly with the question of, “What is the dark matter?” (

De Swart:

No.) Right. Right. Right. Yeah.

De Swart:

But there was a question going around, because there, I’m not sure but, so I, of course, scanned through your papers and the things you did, but I don’t have it with me. So, that’s a shame. Which was, I think Herbert Rood, which was already also a lot working on this virial discrepancy, at the time?

Gunn:

Yeah.

De Swart:

And you were… so, he was at Wesleyan?

Gunn:

He was at Wesleyan. That’s right. That’s right. That’s right. I think. Herb is in various places, and I don’t remember where he was at that time. I don’t know.

De Swart:

But did you know… so, so was this a, was this a thing, this virial?

Gunn:

Oh, yes. This was a thing. There were lots of people thinking about it. Right. Right. Right. So, you know, there was a virial discrepancy in groups in clusters. There was the growing virial discrepancy in galaxies. The one in clusters had been known for a long time. And, various people, including me, and Herb, and other people were just trying to put pieces of the puzzle together to figure out where it was. Scott Tremaine and I, people thought well maybe it was massive neutrinos. But that doesn’t work, because the phase density is too low and you can’t get it into, get them into galaxies. I think Scott Tremaine and I were the first people that realized that and we published a paper on that. So, trying to constrain, you know, what its physical properties, whatever it was, might be. And, nothing really seemed to work very well, until this idea of Weinberg came along, that it was WIMPs, what it could be WIMPs. And, uh…

De Swart:

Yeah. So, that’s interesting. Because, if you actually look at the citation pattern of people looking at virial discrepancy, missing mass problem, and things like that, you see a bump around ‘74. And, then it decreases again and then at the ‘80s, I think ‘81-‘82, you see exponential growth.

Gunn:

Growth. Yeah.

De Swart:

In the…

Gunn:

Right. Right. Right.

De Swart:

So, there was something already going on in the beginning?

Gunn:

Yeah. There was. And that happens often in astrophysics. I mean, somebody had an idea that’s clearly correct, but it’s just a little bit, you know, [laughs] too far into left field. And a little group goes and works on it and they finally convince everybody else, and then it takes off.

De Swart:

So, what was fashionable at the time you were doing the graduate? What was, like, in cosmology?

Gunn:

Well, we didn’t, we… I mean, all, basically what I was doing in graduate school all we knew was this problem that Zwicky had pointed out, forever ago, that there was this missing mass. We didn’t know that, you know, on what scale it was. We didn’t know anything about it. And the only place where it reared its ugly head, that we thought – I mean, because the stuff that Mort Roberts and Horace Babcock had done, the work was there but I don’t, I didn’t know about it and I think very few people did. They just didn’t notice. And I think both Mort and Babcock sort of said, “You know, well this is a kind of puzzle we don’t understand,” but they didn’t say that there’s dark matter out there.

De Swart:

No. No.

Gunn:

They couldn’t, you know, they just said, “I don’t understand the data. There’s something funny going on.” Right? And so, it was really only, I think, after Vera pointed out that these were, that this was an endemic thing that people went back and said, “You know, we knew this quite early.” [Laughs]

De Swart:

Yeah. Yeah. So, that’s ‘78, I think.

Gunn:

Yeah.

De Swart:

Right? That Vera, or even in the ‘80s, that you have these really extended far-extended radiation curves?

Gunn:

Right. Right. Right. Right.

De Swart:

Of the H1, of the optical?

Gunn:

But I mean we were already clearly in trouble with Vera’s first work. Right? And so, it either had to be wrong or there was something going on that was completely endemic that we didn’t understand. And I think people got that. Right? It wasn’t just a couple of measurements on one nearby galaxy that showed some, some funniness. So. You know, but these new ideas come in. They’re birthed very, their birth is very difficult. Right? [laughs] Right.

De Swart:

So, so, and at the time Ostriker, Peebles, and Yahil wrote this article, that was ‘74. So, you were, so were you here or still at Caltech?

Gunn:

I was still at Caltech. I was, between, at Caltech between ‘61 to ‘66, and then ‘70 to ‘80.

De Swart:

Okay.

Gunn:

Yeah.

De Swart:

So, do you remember, because I heard, for example, Rich Gott saying that he remembers this talk of Jim Peebles in Caltech when he, because he had this really small model, this simulation thing going on. Can you, can you remember that, when this paper, when it came out?

Gunn:

I’m sorry. Which paper are you talking about?

De Swart:

The paper with Ostriker and Yahil in which they show… so, you had two papers at the time. So, you had, in ‘73 you had the collaboration which looked at the halos?

Gunn:

Yes. Right.

De Swart:

So, they said that, “We need a halo to, for the stability argument.”

Gunn:

Right. Oh, yes. That was another point that I forgot, which was very important in this whole thing.

De Swart:

Yeah.

Gunn:

Right. Right. Right. Right.

De Swart:

But, so, can you remember at the time people coming to Caltech to give talks on this very subject of…

Gunn:

I’m sure there were, but I don’t remember.

De Swart:

Well, it just seems… well, why I ask is because it seems that in, at Caltech, because Zwicky was still…

Gunn:

Zwicky was still there. Yeah. Yeah.

De Swart:

He was still there? That it was kind of known that you were dealing with this virial discrepancy and that galaxies are, have a way too high velocity dispersion in clusters to make sense of the mass.

Gunn:

Right. Well, I think that was known sort of everywhere in the community. Zwicky was still alive and kicking, but Zwicky was such a difficult person, personally, that people didn’t necessarily pay much attention to what he said. He had a very strong tendency to be right and that wasn’t always popular either. [laughs] But, I don’t think that the, I don’t think that the virial discrepancy – I may be wrong about this – but I don’t think that the virial discrepancy was any, was any more the fabric of what was going on at Caltech than it was anywhere else. I think people just wanted to kind of ignore it because they didn’t understand it. And there didn’t seem to be any way to get a handle on it until the rotation curves came along. And then, you know, it was something every galaxy did, then Jill Knapp, my wife, and I worked on the rotation curve of the galaxy, showed that it was probably flat as well, and with the work on M31 and Vera’s work on other galaxies it, you know, people just had to accept that the stuff was there, not having any idea what it was, necessarily, but that it was there. [Laughs]

De Swart:

Did you know about the timing argument at the time?

Gunn:

Yes. Yes. Yes. I wrote an independent paper on the timing argument. [Laughs] Which showed how poorly I read the literature. [Laughs] Because, I wasn’t aware that it had been done before. But I wrote it and published it anyway. Right. Right.

De Swart:

So, yeah. So, maybe that’s why I thought about it, that I saw your name in that context. Actually, yeah. So, I think that’s, that’s why. Yeah. That’s interesting that it was more of a common, not-so-accepted, more on the edges of people doing cosmology.

Gunn:

Uhm-hmm.

De Swart:

Did you have a cosmology course during your…

Gunn:

No. No.

De Swart:

During your graduate?

Gunn:

I didn’t. As a sort of difficult story, I went to Caltech as a graduate student to work with Robertson, because I wanted to do cosmology. You know Robertson of Robertson-Walker. And he died the summer before I arrived at Caltech.

De Swart:

Oh, wow.

Gunn:

And, they didn’t even have anybody to teach relativity. So, I finally got a relativity course from a guy that they imported from JPL. But I, all the cosmology, no. I taught myself, because there wasn’t any, there weren’t any courses or anything.

De Swart:

Oh, that’s interesting. But there was before? So, there was…

Gunn:

Well, Robertson had taught, he had taught general relativity, which was very strongly sort of cosmologically-oriented, in the Physics Department at Caltech. And then, when I went back – so, I sort of taught myself and then when I went back to Caltech they had just hired Kip Thorne. And so, Kip and I taught, he was part of this wonderful book that he and Misner and Wheeler wrote, and I, I taught that course with him for several years. So…

De Swart:

So, you…

Gunn:

I learned a lot more when I was doing that, I’m quite sure. [Laughs] Right. Right. Right.

De Swart:

So, you were actually one of the first who was giving cosmology courses?

Gunn:

In Caltech. That’s right.

De Swart:

In Caltech?

Gunn:

That’s right. And, maybe the first, actually. Yeah.

De Swart:

Yeah. That’s really interesting.

Gunn:

Yeah. Yeah. Yeah.

De Swart:

Because it was coming up. Because you had this surge of books?

Gunn:

Yes. Of books.

De Swart:

Texts?

Gunn:

And interest in the subject, and papers in the literature, and people, you know, it was a subject that – it was a funny subject, because, you know, I don’t, there is an absolute seminal paper on the subject, written by Allan Sandage, and I know it’s in APJ 133, but I don’t remember what year that was. So, it was late ‘60s I – no, middle ‘60s. Anyway, and this paper had the amazing title, something like On the Ability of the 200-Inch Telescope to Distinguish Among World Models. [Laughs] And, what it was, a very long paper, which was a completely pedagogical paper about cosmology, and I think it was maybe one of the first that appeared in the APJ. But it went through the simple matter-dominated Robertson-Walker metric, the usual sort, sorts of models. I don’t think that it appeared in the astronomical literature before. He went through the Hubble diagram, number-count tests, all of these things that you use to, you know, say what cosmological model you’re in, and then he said, “Well, you know, with the current instrumentation the 200-inch can do this, that, and the other,” and came to a conclusion that just barely you might be able to tell. Right? [Laughs] And, Sandage’s favorite model at the time actually was Cnot=1, which is Omega = 2. And this is a model that expands, stops, and collapses again. And I don’t, where he, why he was in love with that, who knows. But, so, and that was the beginning, then, of, then many, many things came, came along, you know. He espoused the idea of the Hubble diagram and so he started working on brightest cluster galaxies, because he thought they were standard candles. Bev Oake and I came along later and followed that up with, with Bev’s – Bev Oake, Bev Oake I’m talking about – multichannel spectrometer. We were able to get red shifts at, at quite large distances. And, we knew, at the time, because we had – Sandage had done some work, which suggested that the evolution of stars was not important, in this. But it’s obvious, from the back of an envelope, that this is not the case. And so, we hired Beatrice Tinsley, because she was, she basically invented the subject of stellar population synthesis in galaxies, and the idea to try to use our data to fit population models and predict what the evolution of the stellar populations were in time, because it was clearly an important issue. And so, that was happening in the ‘70s, and then in the late ‘70s Ostriker and Tremaine pointed out that that was, the evolution of the stars was only half the puzzle, or maybe even less than half the puzzle, and the other half of the puzzle was the fact that it was clear that these big galaxies in the middle of clusters were eating other galaxies. And so, you had to worry about dynamical friction and their growth, and that made the problem basically impossible theoretically. So, I think that was the death of…

De Swart:

No more standard candles?

Gunn:

Of no more, yeah, no more standard candles, no more Hubble diagram.

De Swart:

Yeah.

Gunn:

In the process, we learned a whole hell of a lot about clusters, but nothing, but not anything about, about cosmology. [laughs] That’s the way it goes. But…

De Swart:

So, where was Sandage at the time?

Gunn:

He was, he… Sandage was always at Carnegie.

De Swart:

Carnegie?

Gunn:

In Pasadena, at Mount Wilson and Palomar. Just, although the divorce between Mount Wilson and Palomar was also happening about then. So, anyway. He was at Mount Wilson.

De Swart:

Sorry. The divorce?

Gunn:

Oh, this was Caltech and Carnegie jointly ran Mount Wilson and Palomar, and that’s a complicated political story, because the Carnegie Foundation had paid for Mount Wilson, all the telescopes on Mount Wilson, the establishment of the observatory in Pasadena, and then when George Ellery Hale was looking for money to build a yet-bigger telescope, Carnegie was unwilling. There was a lot of money in that. It seems like very little these days, but at the time it was an enormous amount of money, and Hale wanted to build a telescope more than he wanted to keep friendly relations with Carnegie. And so, he finally persuaded the Rockefeller Foundation to pay for the 200-inch. And there was no way that a Rockefeller thing and a Carnegie thing could exist under the same roof, because it was two rich people with, with their own piles of money. [Laughs] And, anyway, so that’s how the Caltech Astronomy Department got started, got started. So, they decided that Caltech should run… Rockefeller was going to pay for the 200-inch telescope and Caltech would establish, already had a very strong Physics Department, a very, very strong Physics Department, in Milliken and the like, and was going to start an Astronomy Department and run the 200-inch. But Hale was managed, did manage to get together a kind of marriage so that the observatories were run jointly by Caltech and Carnegie. And that went on for a very long time, and finally they decided, more or less amicably, to get divorced, because Carnegie was building the Los Campanas in the south. Caltech was much more interested in pushing high technology things, infrared and stuff like that, than Carnegie. And so, they decided to go their separate ways. And I’ve forgotten exactly when that happened, but about when I left Caltech. It was late ‘70s or the ‘80s, or something like that. That’s, but that’s another sort of astronomy sociology story.

De Swart:

But interesting, because most of the funding, where was it coming from?

Gunn:

Oh, I think already then it was coming from the government, mostly. The Carnegie people were pretty well supported by the Carnegie Foundation. And, Rockefeller built the 200-inch, but didn’t leave any, they left some kind of legacy for sort of maintenance and stuff, but nothing, nothing for science. That was something that Caltech was supposed to do. And government money was a great deal easier to get in those days. And so, the early, the early work, most of the early science on the 200-inch was actually supported by the Air Force. And that stopped at some point when somebody realized that the military had no business doing this. And then, after that, mostly the National Science Foundation. So. Yeah.

De Swart:

Because, was there any direct influence of the Cold War at that time, and in astronomy?

Gunn:

Well, that’s hard to say. I think the reason the military was so rich was because of the Cold War, and no one was paying very much attention to what they were doing. And so, for example, Jesse Greenstein, who was head of the department at Caltech, had a grant from the Air Force. And getting this grant basically involved writing one letter a year to the funding agencies and entertaining some captain when he arrived once a year. [laughs] Right? And, it was a, you know, these days I don’t think it would seem to be a big grant, but it was a big enough grant to support most of the science in the department. So, times were different.

De Swart:

Yeah. Yeah.

Gunn:

Very, very different. Yeah.

De Swart:

But, because it seems that before that high-energy physics was, of course, really interesting for the government itself, too.

Gunn:

Right. Because of the bomb.

De Swart:

Yeah.

Gunn:

Yes. Right. Right. Right.

De Swart:

So, I was thinking about the fact that you could seem to make a good point in maybe radio astronomy being a good government…

Gunn:

Oh, yes. Yes. Yes. Yes. And, and the NSF basically built the Owens Valley Observatory at Caltech, the radio observatory, and supported it really very well for a very long time. So, but money was a lot looser then. And I, actually the government is still supporting Owens Valley, but it’s a much tougher thing to, you know, to get now than it, than it was. So. But there [were] sort of interesting times and it’s fun to remember. But yeah. Yeah.

De Swart:

Thanks. So, maybe, is it okay if we have, like, another…

Gunn:

Oh, sure. Sure. Sure. Sure.

De Swart:

Maybe like another appointment?

Gunn:

Right. Right. Yeah. If you want to look at the stuff I wrote and then, we could have another talk later.

De Swart:

Yeah.

Gunn:

Right. Right.

De Swart:

So, so is it, what’s best for you? Is it best if I just come by again?

Gunn:

I think it’s probably the best thing to do, because, you know, I have this crazy schedule with phone confs that are ad hoc-scheduled, and things. So.

De Swart:

I’ll just annoy you.

Gunn:

Yeah. If, if, yeah. Just annoy me. [Laughs] Right. Fine. Yeah. Right.

De Swart:

Okay. Perfect. Thank you, so much.

Gunn:

Okay.

[End session 1]

[Begin Session 2]

De Swart:

Put this on, if you don’t mind?

Gunn:

Sure. No problem.

De Swart:

Yeah. And I, so we had already a conversation of like a half an hour, thirty minutes, and we spoke about some things, but there, I just wanted to go a little bit more into really the beginning of the ‘70s and you starting cosmology. And maybe we can start around the time you were doing your graduate. Because I think it was already some kind of a cosmologically-prone subject.

Gunn:

Yes. Yes. Yes. The thesis was about the two-point correlation function, actually, which, I think, as I remember, the basic ideas and the formalism, which is not really used anymore, but just a basic formalism, was written down by Nelson Limber. I think, oh I don’t know, it must have been in the very early ‘60s. And there is an equation called the Limber equation, which says how the 2D correlation function, sorry, how the 3D correlation function projects into two dimensions on the sky. And I don’t even remember whether he did this in the context of a relativistic cosmological model. The extension is trivial. So, I don’t think it matters.

De Swart:

But you did?

Gunn:

Oh, yeah. Yeah. Yeah. But, as I say, it was trivial. But I don’t remember whether he did it before I did it. That was, that was the point. So, I studied this by using plates from the 48-inch Schmidt. There had just been developed a new quite sensitive, for those days, quite sensitive emulsion called 3aj. And Kodak was making big plates in 3aj. And so, I took a lot of plates and counted galaxies on them. There were no measuring machines or anything, so you just put down a transparent grid on the plate. And I don’t remember what the cell size was, but something like a centimeter. And the plate scale was… I don’t remember. It was a couple, couple of three minutes of arc on a side. And so I counted galaxies, got a correlation function, certainly got a strong signal, as one would expect. But, Limber, in his original paper, suggested that the way to think about this was to think about it as a superposition of three-dimensional Gaussians. And so, I did that. And that makes the analysis very easy, because the integrals are separable. So, but I missed a very important point, and that was that the… so you, you get this function. You decompose it into Gaussian components, and there is a very trivial result that when you do that and you have a power law correlation function the power, the amplitude of the Gaussian scales as an inverse power of the size of the Gaussian you’re fitting. And I saw that and didn’t realize it. So, I just basically came out with this super [inaudible] Gaussian. And then, I think, two years later Jim Peebles did the same thing and pointed it out that it was just R^-1, R^-.8 on the sky. Right? And, of course, that’s what my data says, but I wasn’t clever enough to realize that it was a power law. [laughs] So, that sort of got me… well, no, actually I had gotten started already. I mean, I was very interested in cosmology, but then I, the next thing I did was to look into lensing in cosmological models. And there was an equation called the Sachs-Wolfe Equation, which Arty Wolfe did for his thesis with Ray Sachs, in which basically you take a ray from a distant object and you add up all of the gravitational influences from, from magnification that is matter in the beam, shear tidal effects on the beam and so on. And they wrote down a differential equation, which describes the evolution of the magnification and shear as a function of the affine parameter along the ray from the source to you. And I took that model, applied it to cosmological models, and pointed out a couple of things that I think had not been realized. One, of course, we thought the universe was very undominated. At the time, we thought all the matter was in galaxies. We thought all the matter was ordinary matter. And, if that’s true, and you think about it a bit, then a ray from a distant object to us is coming along a path which is biased. Because there aren’t any galaxies, any intervening galaxies, in that path, because you wouldn’t see the distant galaxy if the intervening galaxy were in the path. So, the path is emptier than the average path through the universe, and therefore the magnification is different and the red shift, well the flux red shift relation, the Hubble diagram, is different. And also, because there is shear from neighboring objects, the, they’re second-order effects because of… magnification does a first-order thing on the flux, and shear does a second-order thing.

De Swart:

And what implication would that have?

Gunn:

So, of course, in the universe as we know it now, the effects are tiny, because the dark matter is distributed everywhere.

De Swart:

Yeah. Okay. But at that moment?

Gunn:

But at that time, if Omega in baryons… we didn’t know what Omega was… if Omega in baryons is near one, it’s a very big effect. It’s, you know, factors of twenty, thirty percent. So, that was potentially interesting, but, but we didn’t know about dark matter. [laughs] So, later when that, when we understood more like what the universe was like it turned out not to be important at all.

De Swart:

But when was this work you mentioned? Was it…

Gunn:

In 1965.

De Swart:

During your graduate?

Gunn:

Just after graduate school, actually, is when I did most of it, as I remember, ‘66. I don’t remember. There are a couple of papers. Anyway, I don’t think lensing is in the title because we weren’t quite thinking of it that way.

De Swart:

Yeah.

Gunn:

But something about the “Effects of Inhomogeneity in the Universe on the Hubble Diagram,” or something. I don’t remember what they’re called.

De Swart:

Uhm-hmm. And how, how common was it to do these kinds of theoretical papers?

Gunn:

Well…

De Swart:

On cosmology?

Gunn:

Theory was a lot more active in those days because there wasn’t any data. [laughs] So, it was possible to write arbitrarily complicated theoretical papers with no possibility that it would ever be checked. [laughs] This was probably one of them.

De Swart:

No, but cosmology itself, as a subject?

Gunn:

There was… but not very many people. There was quite a lot of very hairy esoteric theoretical work on different Bianchi types in cosmology that were anisotropic in various ways or that were topologically folded in various ways, and things, and these were really, you know, angels-dancing-on-the-heads-of-pins kinds of papers. But we didn’t know how homogeneous the universe was. We didn’t know how isotropic the… because we… the blackbody background hadn’t been discovered. And so, you know, the sky was the limit. You could write a paper about anything and nobody could say that, “This paper is nonsense, because we know the universe is different from this.” Right?

De Swart:

Uhm-hmm. But there was a community working on this?

Gunn:

Oh, yes. Yes. Yes. Yes. Yes. Right.

De Swart:

And they’re…

Gunn:

Mostly, mostly physicists, mostly relativists.

De Swart:

Ah, yes.

Gunn:

Who were working just, you know, in, in kind of theoretical relativity. Sachs-Wolfe was, when he, I think he had already sort of gotten interested in observation. Ray Sachs was a theorist.

De Swart:

Where was he?

Gunn:

Oh, where was this done? The University of Chicago, I want to tell you, but I’m not absolutely sure.

De Swart:

Well, I can check easily.

Gunn:

Anyway, it was Sachs-Wolfe was, was the paper. It was the Sachs-Wolfe equation. But there was very, very little work going on in observational cosmology, just because the data were so hard to get. You know, you had photographic plates that weren’t very sensitive. You couldn’t work very deep. You couldn’t get spectra, 3C295 at a red shift of .46 was, for years, the highest galaxy redshift that anybody knew, and it was a kind of fluke because it was a bright radio galaxy that had very strong emission. And so, Rudolph Minkowski got a spectrum with the 200-inch. I think it was an all-night exposure. One line and as he very boldly said, “Well, it’s, this looks like it has to be 3727.” And so, that got the redshift and everybody believed it, sort of, but, you know, that was all there was. But things really started happening about the time I graduated, about 1965, because the microwave background came, the quasars came, so we knew that there was stuff at there at very high redshift. Image tubes were beginning to be developed, so we still were stuck with photographic plates for detectors, but we now had photon amplifiers. And, although it was not until I built the first CCD spectrograph that the red, that that 3C295 redshift record was broken.

De Swart:

Oh, really?

Gunn:

And that was after I went back to Caltech. So, that must have been ‘72 or ‘73. We got a perfectly ordinary brightest cluster galaxy with a redshift of .61. And then, you know, the sky opened up then.

De Swart:

Yeah. Because of…

Gunn:

We broke one.

De Swart:

Because of the CCD?

Gunn:

Because of CCDs. That’s right. That’s right. Improve the detector efficiency by a factor of a hundred and you would be a fool if you don’t get science. Right?

De Swart:

Uhm-hmm.

Gunn:

So, so that was… and, when, as CCDs came along… and, I had started working on these very high-sensitivity plates and doing, trying to do cluster surveys because we were still under the impression that doing the Hubble diagram was the way to do cosmology and one could get Omega or “q-naught” [q0] as it was called then, because nobody called it Omega. And, as we talked about last time that turned out, because of various peculiar evolution, or not “peculiar”, but perfectly ordinary natural evolutionary effects that one couldn’t calculate just was not, it was just not the proper tool. But, I mean, I was very interested in clusters anyway, so I just kept on with that. And…

De Swart:

And who were the others at Caltech who were interested in that?

Gunn:

I did most of the work with a student, John Hessel, who did his thesis on brightest cluster galaxies. Another student, Don Schneider, and Bev Oke, who I worked with instrumentally the whole, the whole time, and really started with his, with his multichannel spectra, spectra… that was the other instrument that was very powerful at the time, and that was this scanner. But it wasn’t really a scanner because it didn’t scan. [laughs] It was just a spectrograph with a whole bunch of little multiple, or high-dispersion spread out the light spectrum over about, I don’t know, a little more than half a meter. And, with a whole bunch of photomultiplier tubes. There were thirty-two.

De Swart:

Oh, wow.

Gunn:

That looked at various pieces of the spectrum. And so, you could produce a sort of resolution, thirty or so, spectral energy distribution from this, and get pretty good redshifts. But we still, even with that machine we didn’t, we didn’t break the…

De Swart:

Schwarzschild?

Gunn:

Redshift record.

De Swart:

Uhm-hmm.

Gunn:

Until, until CCDs came along. So, I was studying clusters and doing a fair amount of theoretical work on cosmology, looking at the, whatever it’s called, anyway the collapse time of a local group, looking at… Beatrice Tinsley and Schramm, and Gott, and I were working on trying to sort of synthesize things from every direction to try to get a handle on the mass density. And…

De Swart:

Because that, was that the main problem that cosmologists worried about?

Gunn:

Well, it was certainly… I mean, we didn’t know about dark matter. Right? And it was coming along, and we sort of knew that we, we, if that became part of the, of the, of the kind of paradigm and people… it was kind of a relief because you just added up all the baryons you could see and it was Omega is .0-something. Right? And so, but the real question was whether… and… sorry. So, and if you took clusters, took mass-to-light ratios, took groups to mass-to-light ratios and so on, various people did this in various ways and we did this as carefully as we could at the time and came up with a number for Omega, which was .25 or something like that. Anyway, it’s very close to the… and, this was Omega-matter, of course, because we were adding up all the matter, dark and visible, that we could see in various ways. And that’s very, very close, of course, to what we believe today. It’s .27 or whatever.

De Swart:

That was the “Unbound Universe” article?

Gunn:

Yes. That’s right. That’s right.

De Swart:

Uhm-hmm.

Gunn:

That’s right. And then, there were lots and lots of arguments about where, if Omega was bigger than that the matter had to be distributed. It couldn’t, because you counted all the gravitational influences you could, and people thought it might be distributed between the galaxies. It didn’t make… people were beginning to do N-body stuff. It didn’t make any sense because when you do an N-body experiment and run it very long there isn’t very much matter left between the, between the big structures. [laughs] And people were working. Michael Strauss was one of the primary ones, and others, about looking at peculiar velocities in the universe. Peebles also worked on this. And those lines of evidence mostly suggested that Omega was large, it was near 1.5 or something. And exactly what went wrong with that I have never quite understood. There were funny selection effects involved in the data, and so on, and I think people are relatively happy with them, with the Omega matter that is derived in all these precision ways now with that. But, but somewhere the thread, the thread got lost. So, you know, whether it was Omega .2 or it was Omega 1 was the, the question.

De Swart:

And, when did the discourse change from, from thinking about the deceleration to thinking about the critical density?

Gunn:

It was very adiabatic, and I, I don’t know. Certainly, in the middle ‘60s people talked about q0. And I think by the early ‘70s people were talking about Omega. But I can’t point to exactly where it happened. It was just one of these kind of, you know, slowly creeping changes. [laughs]

De Swart:

Because what does the emphasis mean if you put the emphasis on Omega, what do you change?

Gunn:

Well, I mean, we thought at the time that the two were synonymous. Right? Because we didn’t know about dark energy. So, if you got the matter density you would have the deceleration, because Omega was 2q0.

De Swart:

Yeah. Yeah. Yeah.

Gunn:

And, it was, I don’t… I wrote a little paper with Beatrice, in… well, anyway, it’s called “An Accelerating Universe Question,” and it was a paper in Letters, in which we boldly suggested that the way around various problems—this was also, had to do with the age of the universe—was to reinvent Lambda. And it was also based on some observations and the observations were wrong. They were just… we were trying to work on faint galaxies, again with Oake’s multichannel spectrometer. And, when you work on a big telescope on an object that you can’t see because the, there was an integrating television guider on the instrument, which took a picture of the field. The spectrometer, the entrance aperture to the entrance plate on the spectrometer was a mirror that had a hole in it, two holes in it, one for the sky, and one for the object. And, you had to take a picture of the sky with the guider, move the telescope so that the object is in the object hole. You switch back and forth. And, if you can’t see the object, that’s very, very hard to do. [laughs]

De Swart:

Yeah.

Gunn:

So, you have to measure where stars are. You have a star and you figure out where the star, the object is with respect to the star, because these galaxies were too faint to show up on the integrating guider. If we had a CCD guider it would have been fine, but we didn’t, because we didn’t have CCDs. And, almost certainly… so, we got a set of fluxes for redshifts around, you know, .35, .4, something like that, which was as high as we knew, and they were faint. And, if the flux is low at a given redshift it means that the acceleration is small. And, in fact, if it’s low enough it means that the acceleration is negative. And so, we got, we got numbers which suggested that the universe was accelerating. And I think it was almost certainly because we weren’t getting the object in the hole properly. Because, later when we had CCD pictures and could do the photometry, it was clear that we were missing a lot of low light. So, you know, you say, “Well, that means Omega was high.” No, because, because of evolution and various things. Because you can’t do the Hubble diagram.

De Swart:

How would you move the telescope? How did you do that to line up the hole?

Gunn:

What you do is you, you know accurately how far apart the holes are. So, this is just a TV camera looking at the sky. Right? And, you know that that is… I don’t remember what it was… forty arc seconds, I want to tell you. I think that was right, actually, forty arc seconds apart. So, you know what the scale is on this picture. And you take a photograph, because we didn’t have CCD images at the time, just a photograph, and you say, “The galaxy is here. There’s a bright star here.” So, you figure out, you know the scale on this picture now. You figure out where to put the bright star in the picture so that the galaxy is in the hole. And you move the telescope so that the star is in that… you just sort of mark a pixel on this. Put the star in that place and the galaxy should be in the hole. But, you know, the scale, it’s a TV camera, so there’s distortion. You don’t know the scale exactly. The star is probably pretty faint as well, because it’s, you know, it’s, the field is tiny. So, you pick something you can see, anything you can see. [laughs] It was a very kind of imprecise world. [laughs] So, anyway, I mean I think people ought to give me credit for, for, Beatrice and me credit for inventing the accelerating universe, but it was really on the basis of crummy data. So, I guess hard to, it’s hard to take credit. I mean, the supernova stuff later really nailed it and I’ve been, you know…

De Swart:

But you were well acquainted enough to put Lambda there and know what it meant?

Gunn:

Oh, yeah. Yeah. Yeah. Yeah.

De Swart:

Yeah.

Gunn:

I mean I had studied relativity a long time and knew the subject pretty well. So, so that was, that was that little flap. And I think that was, I don’t remember whether that was before or after the Gott, Gunn, Schramm, and Tinsley paper. I mean it was about the same time, but I don’t remember exactly where to put it, before or after.

De Swart:

And, so, and in terms of theoretical developments at the time, how important was, for example, Tinsley’s remark on the fact that you needed evolving galaxies to study?

Gunn:

Oh, I think, that was, I mean some people didn’t accept it, of course, because Sandage thought he had solved the problem with an entirely incorrect treatment. I think that that was accepted pretty, sort of immediately. Beatrice’s work, Beatrice was incredibly careful, and the work was pretty transparent. I mean, she wrote very well, and people read it and said, “Yeah. We need to worry about this.” Right? But the thing we didn’t know, the rate at which galaxies evolve, just from their starlight, depends critically on the rate at which stars come off the main sequence. Because it’s the light from the giants that dominates the light of the galaxies. And so, if you have a mass function that’s quite shallow so that, you know, today… and basically… so, how to think about this? Giants live a very short time. So, the light basically depends, to the first order, on the rate at which stars are leaving the main sequence and becoming giants. Right? So, what determines that rate? Well, that determines clearly how many stars there are at a given mass, because those, those are the stars, you know, right at say one solar mass, that are revolving off to become giants. Now, tomorrow lower-mass stars are revolving off to become giants, because those stars, one-solar-mass stars are already gone and you’re talking about .9-solar-mass stars. So, how the galaxy evolves depends on the ratio of the number of one-solar-mass stars to the number of .9-solar-mass stars. And so, if the mass function is shallow so that those numbers are comparable, then there are few giants later because the evolution time of the low-mass stars is lower, and so the luminosity of the galaxy drops very quickly in time. If the mass function is very steep, so that that difference is made up by the fact that there are a lot more .9-solar-mass stars, then the luminosity is, with time, is very shallow. So, it depends on this… and you can write locally. You can say that the mass function goes like M to some power. And, for various reasons, historical reasons, you do it as M to the 1 plus this power, 1 + X. And so, everything depends on what the value of “X” is. And it turns out that when X is 3, the luminosity is essentially flat. In the solar neighborhood, we don’t know the luminosity function very well, the mass function very well, but the suggestion is that it’s around 1. And so, one would expect elliptical galaxies therefore to decrease in luminosity very quickly, with time. But the rate at which they do so depends critically on this parameter of X.

De Swart:

And that was still unknown?

Gunn:

We don’t know. We don’t know. We had, well certainly we didn’t know it very, very accurately in the solar neighborhood. Who knows what star fraction is like? And all the people in galaxies can still [inaudible] that today, and we have no idea what the, what that, what the exponent is.

De Swart:

But how, and how much did this, did these developments, these kind of more uncertainties influence the whole field of cosmology, or people doing cosmology?

Gunn:

Well, again, the work on… I think that all of these things were suggesting more and more to people that observational cosmology, using galaxies, was really, very, very hard. And every, you know, everyday people would think of something new that, that people hadn’t thought of before. The thing that really killed it, though, was the Tremaine and Ostriker thing about dynamical friction, and cannibalism. Because it was clear that we couldn’t calculate that. [laughs] People still thought that maybe there was some way to get at the value of X by looking at dwarf spectral indicators. And, in fact, Charlie Conroy and, you know—Dutch—although he works in this country. It’ll come to me in a minute. Anyway, Beatrice suggested that the way to settle this was to measure a feature at, I think it’s at 9900, 9910 is the number that I remember, that is a band of, a molecular band feature of iron hydride, called the Wing-Ford Band, because they discovered it in dwarfs very, when, I think it was one of the first little bits of sort of near-infrared spectroscopy that was done. It’s quite strong in dwarfs. It’s absent in giants. And the reason is that it’s a monometallic molecule. So, it’s iron and hydrogen. And the binding energy is fairly low. And so, in a giant atmosphere, where the pressure is low, you don’t form this molecule. And, in a dwarf atmosphere at the same temperature, where the pressure is very high, the phase space pushes the molecules together just from this [inaudible] equation, and you get a lot of iron hydride. And so, Beatrice suggested measuring this feature in elliptical spectra and you should, at least with the power law model, you should be able to get the value of X. But the technology wasn’t up to it at the time and Van Dokkum and Conroy now have done this, suggesting that the mass function, at least in the middles of ellipticals, is very steep. That is to say, the evolution should be slow. But it’s only in the very central parts that they can measure this. And so, the whole, the whole question of what the mass function is in elliptical galaxies is still up in the air. Nobody thinks about it in terms of cosmological evolution anymore. But it’s an interesting question.

De Swart:

No, I was just, just wondering about if you felt you were doing something very new in cosmology at that time. For example, with these like-minded people like Beatrice Tinsley.

Gunn:

Well, we were certainly exploring territory that had not been explored before. In every… the, you know at the time, and I think that somebody wrote, there was a Scientific American article or something, not mine, that the, that the whole subject of cosmology was the quest for one number, and you could call it q0 or you could call it Omega. But, you know, if we knew that number we thought we knew everything. Then things got a little more complicated. [laughs] Right? Because dark matter was discovered so you not only had Omega-matter but you had Omega-baryon and you had Omega-dark, and it became clear that what was very important also, because now people were beginning to study structure, that the power spectrum, the initial power spectrum was very important. Zel’dovich had written down the scale and variant thing and it was very simple, and people were in love with it, but it wasn’t clear that it worked for a baryon-dominated universe. It clearly didn’t work for a baryon-dominated universe. And, so the problem just began to have a lot more dimensions, and now we know there are, what, seven or so parameters that describe things.

De Swart:

Uhm-hmm. And, as you said, you would pinpoint this explosion to after you graduated?

Gunn:

Oh, yes. Because, it was all, it was all, it’s all technology driven. Right? We, the microwave background came along. People argued endlessly when the microwave background was discovered. That, you know, we, this was another death knell for, for all baryon cosmology. You look at the structure you see today, you understand how the structures evolved because it’s basically just linear theory. You get to the epoch of recombination and you have 10^-3, 10^-3 perturbations. People didn’t see any perturbations at all. They certainly didn’t see perturbations at that level. And so, you know, something was terribly wrong, because we thought that structures grew gravitationally, but, and we thought we knew how, how they evolved. Dark matter saved that one, of course, as well. So, it, I think sort of in the middle ‘70s it became a really very active field, with lots and lots of people doing lots and lots of things, both observationally and theoretical. But the theory wasn’t fancy relativity anymore. It was growth of structure. It was all of these sort of quasi almost Newtonian things. Because, the background said, “You know, this is, this is just a plain old Bianchi type-one cosmology. It’s not, you know, doing any of these weird things that people…” So, all of that mathematics just was dropped in the wastebasket.

De Swart:

And this was the mathematics you were initially working on?

Gunn:

I was, I was never that fancy a relativist. But you know, but I did the lensing, and things, which was, was reasonably fancy relativity. And the lensing was also kind of forgotten, because in a dark matter cosmology it was just not important, and it wasn’t… that was a kind of funny thing, because we, somebody, Fritz Zwicky did it of course, he knew how massive clusters were. He didn’t have any clusters that were known that were high enough redshift to do strong lensing, but he knew that a cluster at a redshift of, you know, .25 or .4 or something, it was like the Coma cluster would do strong lensing. And he wrote it down and nobody paid any attention. And then, we started finding these clusters at high redshift and people, and quasars came along, of course. But the lensing just came out of left field, you know. It was just a completely serendipitous thing. So, 0957, you know, it was a double quasar. What the hell was it? [laughs] It was obviously a gravitational lens, but it took people quite a while to, for that to, to… and then, of course, then… who found the arcs? [Inaudible] at Kitt Peak took ultraviolet pictures of clusters, found these funny arc-like things. He… I mean, by then lensing was kind of in the air, and he knew what they were. And, then the subject, you know, just, just took off. And Tony Tyson and other people were working very hard on doing weak lensing, but the technology just wasn’t up to it. And so, when we first proposed Sloan we thought about, in a proposal, we thought about talking about weak lensing, but I said, “No.” Because it had been tried so often and was completely dominated by systematics, and so we didn’t propose that we would do weak lensing, when we did the original proposal. And, of course, when the data came along the data were better than anybody dreamed they would be, and it was one of the first sort of serendipitous results.

De Swart:

When was the first proposal?

Gunn:

Early ‘90s. We, we wrote, and that was a kind of funny thing. We wrote a proposal, I think it was ‘91, we wrote a proposal to the NSF. And, the NSF did not have a program for large proposals at the time. So, they were all just completely ad hoc. So, if you wanted to build a 100-meter telescope you would write this proposal to the NSF. It would disappear into this black hole and they would eventually tell you, “Well, we’re going to think about this, but it will be a couple of years before we say anything.” So, this wasn’t nearly that big, because we terribly underestimated how much Sloan would cost. So, we sent them the proposal and they said, “Well, this looks nice. We will send it around to people and we’ll let you know in a couple of years.” And then, we discovered that the Sloan Foundation had some money that they were forced by law to get rid of because foundations can’t accumulate money. We had this proposal. We sent it to the Sloan Foundation. Two months later we got a check. And so, that’s how Sloan got started and they have been incredibly generous supporters ever since. And, you know, they’ve gotten a fair amount of good publicity out of it, as well. But, so, but they’ve just been very, very good. Where were we? [laughs] I’ve forgotten how I got…

De Swart:

So, yeah. And so, let’s get…

Gunn:

How I got on this thing. Oh, lensing. Lensing. So, you know, the first sort of early data came out in, whenever it was, 2001, the EDR, and people immediately realized that the, we, mostly because of the scanning technique that we were using, that the systematics were really quite well controlled. And so, the first lensing results began to come out.

De Swart:

So, if we go back now to your time at Caltech, right, which was still after you’re a graduate you were still at Caltech?

Gunn:

Yeah. I graduated in ‘65. I came, then came… I worked at JPL for a year, a couple of years, a couple of years, came here in ‘67. Went back to Caltech. Came here in ‘68, went back to Caltech in ‘70, and was there from ‘70 to ‘80, and then, came back here, came back here in ‘80. Yeah.

De Swart:

So, do you, in Caltech then you were working on clusters of galaxies, and you had this pretty famous article with Rich Gott …

Gunn:

Right.

De Swart:

…in ‘72, about the inflow of matter, you already mentioned. So, I was wondering how much you were aware of this mass discrepancy problem when you were looking at this?

Gunn:

Of the mass? Well, I think by the time that paper was written I think that everybody knew there was dark matter. I, I honestly don’t remember, in that paper, what we said about dark matter, if anything. And it was just dynamics. We knew how massive the cluster was and certainly we knew there was dark matter in clusters, because of Zwicky, you know.

De Swart:

Because of Zwicky you knew that?

Gunn:

There was, there was just no question about that.

De Swart:

Okay. Because all the people at Caltech will be aware of that and what Zwicky…

Gunn:

I think so. I think so. Yeah. Yeah. I mean, there weren’t very many of us who were thinking about cosmology. So, when you say, “all the people at Caltech,” it’s not very many. [laughs]

De Swart:

But were there other institutions in the U.S. who…

Gunn:

Oh, yeah. Yeah. Yeah. I mean, there was a lot of work, a lot of places where the relativity was going on, Chicago, and NYU, and Harvard, and Cambridge. The, so a lot of this sort of synthesis work, the kind of stuff that, that Beatrice, and I, and Schramm… Schramm was at Chicago. Gott was at Caltech, then. He was a postdoc. So, a lot of people were trying to sort of synthesize things, because the data were so sparse that it’s clear you sort of had to grab at every straw you could get hold of to, you know, to get ahead.

De Swart:

To synthesize what?

Gunn:

Hmm?

De Swart:

What specifically to synthesize?

Gunn:

Oh, just, you know, what we tried to do in the “Unbound Universe” paper was to find all the evidence we could find from every possible walk. And, you have to be very careful when you’re doing that, because it’s easy to select things that support your point of view, of course, and to ignore things that don’t. Beatrice was very, very good at this. [laughs] She was, she knew what was right and what, you know, what was reasonable, and what was not. But I think we were pretty fair-minded about the way we went about it. And a lot of people were doing it. At the same time, the technology was advancing and so we were getting real data and it was just, as I say, the field just exploded. Because there were physicists coming in. There were not so many old-time astronomers, but a fair number of people who were interested in, in galaxy research and observational cosmology. And the whole thing about galaxies, and mergers, and the formation of clusters, and the formation of structure, and all of this, was sort of going on all at the same time. It was, it was really a, it was a beautiful golden era. People were… a lot was happening. Communication was pretty good. There were really very good meetings that people from sort of all walks, x-rays, the beginnings of gamma rays, radio astronomy, the back of, the microwave background, sort of, you know, putting things, trying to put things together and to make a picture. But we still didn’t know what the dark matter was. We didn’t know that the universe, we didn’t know about dark energy. So, there were still a fair number of mysteries. We did know what Omega was and we knew that if dark matter probably, that you couldn’t make Omega one… okay. Let me back up a little bit. So, we… because there was, there’s another, it’s another big factor in here. As soon as the COBE results came out we knew a couple of things that we didn’t know before. We knew that the spectrum was very accurately blackbody and there was a lot of work going on based on, on the space experiments that failed in one interesting way or another. This suggested that it might not be. I remember there was, was it an AS, no it was an APS meeting at, in… was it Baltimore or D.C.? It was D.C. I think, in which the lead talk was the COBE talk. And I now don’t remember which COBE team member it was that showed this slide of the spectrum. And there was drawn 3D blackbody, dots like this, and somebody from the audience asked, “Well, what about the error bars?” “Well, the error bars, sir, are smaller than the dots.” [laughs] And, at that point everybody in the audience got up and clapped. Right? That problem was solved. It really was blackbody spectra.

De Swart:

Wow.

Gunn:

And they put limits on the end, on the fluctuations of, I don’t remember, 10^-3, something like that. It was really quite good. But they didn’t see any fluctuations. Right?

De Swart:

Yeah.

Gunn:

They wouldn’t have, because it was, you know, a factor of almost 10³ smaller. But, that told us right away that the universe was homogeneous and isotropic. And so…

De Swart:

Yeah, but that was somewhat later, right?

Gunn:

Hmm?

De Swart:

COBE, when was this?

Gunn:

When was COBE? Uhm…

De Swart:

That’s a great deal later?

Gunn:

It was later, yes. We knew already, from the discovery of the blackbody background, and various experiments, that it was very isotropic. And, but we, and that the spectrum, at least at low frequencies, was quite accurately blackbody, but it was the high-frequency stuff that was the problem. COBE must, it must have been after I came back. So, it must have been… god, I don’t remember.

De Swart:

Alright.

Gunn:

’80-ish. But, but I, I don’t know. So, the universe was very, very accurately isotropic, and that seemed causally unlikely. And I don’t remember when, and I actually don’t remember who, who invented inflation.

De Swart:

Well, in ‘81 you had the famous paper of Alan Guth.

Gunn:

Yes. It was Alan. Right. And so, I guess the real problem with Omega started then. The problem started earlier because the universe was very accurately isotropic. The standard relativistic models all become causally disconnected at early times and there’s no way that you can apply causal initial conditions and make an isotropic universe. The horizon problem, or the blackness problem, which are the same thing, basically. And so, this was sort of hanging on everybody’s, in everybody’s mind. It was clear that there had to be fluctuations, smaller, if there was dark matter than if the universe was baryon dominated. But we still, it was possible that the dark matter in clusters was baryonic in some form that we didn’t know, you know snowballs, or dust, or anything. Nobody could figure out how to make it, but it could have been. But then the blackbody… sorry. Yeah, the blackbody nucleosynthesis, we knew about the microwave background. Everything was suggesting that the Omega and baryons was a very small number. And the work on mass to light, and so on, suggested that even with the dark matter Omega was small. And, nobody could make sense out of an Omega, out of a small Omega universe that was not, and there, it was therefore not spatially flat, that just made any sense. I mean, it was, it, whether it was actually better to have an acausal flat universe or an acausal non-flat-universe was not clear. But it was, it was just confusion in the, in the ranks. Didn’t know. Inflation, the idea, it was just an idea that Alan had, of course, and he worked through and most of what I think in those original models was just now known not to be correct. But anyway, the idea was there. And it was so appealing because it fixed the flatness problem. It fixed the horizon problem. It fixed all kinds of sort of mathematical relativity things about cosmology, but it demanded that Omega be one, demanded that the spatial sections be flat, and it was that, then you were in problem because if you counted up all the matter and you could only get Omega as a quarter, or .2 or something. And so, that was really waiting, you know, for Lambda to come along and save that. Right?

De Swart:

And if we then go back to the earlier searches for the Omega, in which you said that you ended up at a .2 or something, after you did these cluster works with Richard Gott you went into this problem, like the problem of the mass of the universe?

Gunn:

Oh, I did lots and lots of things. I worked with Griffin on globular clusters. I did the cluster, the galaxy, cluster of galaxy survey. Schmidt, Schneider, and I were working on quasars with the scanning instrument.

De Swart:

Oh, yes. Right.

Gunn:

That Gott developed. Beatrice and I, until she died, continued to work on stellar population things. I did a big catalog of stars for stellar populations with the Oake multichannel spectrograph, with Linda Stryker and that, amazingly enough, is still used. I don’t remember. But I was doing lots of things. Not, not, just that one.

De Swart:

Yeah. It’s just because I came across the Texas symposium in 2000, sorry, in 1974, where you were giving a talk on the mean mass density of the universe.

Gunn:

Uhm-hmm.

De Swart:

Discussing some results. But I…

Gunn:

I think that’s basically the Gott, Gunn, Schramm, and Tinsley stuff.

De Swart:

Ah, okay.

Gunn:

Because it was about the same time. Yeah.

De Swart:

Uhm-hmm. Uhm-hmm.

Gunn:

Yeah. Yeah.

De Swart:

Ah. Okay. Because in this one I think you do mention the work of Ostriker and Peebles, which I’m not sure you do any unbound… maybe you do, by the way.

Gunn:

Yeah. I don’t… by ‘74… I forget when that came along, but about then. Right. Right. Right. Right. Right.

De Swart:

Yeah.

Gunn:

I didn’t quite realize how bad it was, I think. [laughs]

De Swart:

No. No. But, I was wondering at how much this, this impressed you at the time, because, and if you knew of these rotation curves, because which was kind of a different sport. Right?

Gunn:

Right.

De Swart:

Making rotation curves of galaxies.

Gunn:

Well, I certainly knew about Vera’s work. And so, I don’t really remember what I thought. I would like to think that I, you know, sort of understood that there had to be dark matter that, that it’s sort of where we think it is now. I certainly understood for clusters. And so, but I don’t remember, exactly.

De Swart:

But what was the problem at the time? Was it that you were, you wanted to be, to have an Omega = 1 universe?

Gunn:

No. No. No. No. I was, I was never, I was never a… Omega = 1. [laughs] In fact, at the time I accused my colleagues of Omega = 1 being a religious point of view. I was pretty open, I think, to the… I didn’t know what Omega was. I mean, Omega = 1 has these nice properties that it, you know, it’s a zero-energy solution and all of these things, nice mathematical properties. Didn’t know about inflation, so it wasn’t, you know, absolutely necessary. But people just, I think people wanted Omega = 1 on aesthetic grounds then, but that’s almost the same as religion. What I wanted to do was find out what it was. And so, you know, and every, every attempt I made to get at the number gave me numbers like .2. And cosmology was a great inexact science and so, you know, I could be making an error of a factor of five, but I could do lots and lots of approaches, and I always got .2 that were completely independent. So, no, I didn’t believe that Omega = 1.

De Swart:

And how were these approaches conflicting, the religious approach and…

Gunn:

There wasn’t any data, see, except people just thought Omega should be one. And then there, as time moved on and people started doing these, these… oh, come on… peculiar measurement, peculiar velocity measurements nearby and started getting large values of Omega then it was clear that something was wrong, because that was also sort of counting the mass. But it was counting all the mass that people… well, so you take the clusters, you take mass-to-light ratios, and you say, “Well, you get Omega as .2,” or .25, or whatever you get. Anyway, a number much, much smaller than one. “So, how can we reconcile this with a religious or aesthetic view that Omega should be one?” Well, what you have to do is take… the important thing to know is that the things you’re measuring, clusters, and galaxies, and things, occupy a vanishingly small fraction of the volume. So, you can put a lot of stuff out there, in between the galaxies, in between the clusters, that you don’t see and make Omega as one. Now, if you go and do the N-body experiments you discover that that doesn’t work. Right? It’s because maybe of some high-pressure thing. Maybe, you know, who knows? But it can’t, if it was just there and noninteracting and dust, as the cosmologists like, it would fall into galaxies and clusters. Right? It wouldn’t be out there. So, so that was the sort of tension at the time. If Omega is one, this stuff has to be more uniformly distributed than anything we know in the universe. And, of course, it turned out to be exactly the case, because of dark energy. Right? And not dark matter. But that was…

De Swart:

So, how was it not dark matter?

Gunn:

Hmm?

De Swart:

How was, how was this not dark matter?

Gunn:

Well, again, if you, if you do the, the structure, you know, the N-body experiments it just doesn’t happen. You don’t get this, this, this uniform low-density stuff that all, gravity just pulls it in to, to structures you find. Right?

De Swart:

Right.

Gunn:

And you actually measure the correct Omega matter by doing these mass to light ratio experiments. And, but, you know, the numerical experiments weren’t very advanced at the time and people thought there might… and we didn’t know what the dark matter was, you know. Maybe, it couldn’t all be high-pressure or it couldn’t fall into galaxies. This was this phase space argument that Scott Tremaine and I made. But there might be two kinds. Right? There might be… suppose that the dark matter was neutrinos and the tau neutrino was a lot heavier than the electron neutrino. It’s still a problem because the numbers have to… that doesn’t work out, because the numbers have to be the same, and you, the electron neutrino could have a low enough phase density that it could be spread everywhere but it wouldn’t contribute anything to the mass. Anyway, but, but people thought about sort of exotic kinds of elementary particle kinds of solutions to this using different kinds of dark matter. And it could still could be right, except that we know that the dark energy fixes the problem. So, yeah.

De Swart:

But, and if you say, “the dark matter problem,” what exactly do you refer to?

Gunn:

What, just what is it? What is it?

De Swart:

Yeah. But…

Gunn:

You know it’s… I think, even then we knew its dynamical properties damn near as well as we know now.

De Swart:

And which, which dynamical properties were there?

Gunn:

That it’s, that it, it’s cold. That is, it does, in the universe, in the large, it does not have a high-velocity dispersion, because it couldn’t, if it did it couldn’t participate in the formation of the structures we see. It’s crucial that it participate in the structures we see or you screw up the microwave background. And you also see dark matter in galaxies. There are rotation curves say that they’re there. And so, you know, it can be, it doesn’t have to be dead cold but it has to be pretty cold. So, it has to be cold dark matter. And, because if you heat it up enough for it to be even warm it, it doesn’t work if there’s only one species for it.

De Swart:

Yeah. But what I meant is more when you were working on this mass density problem with Tinsley and Gott, in the beginning of the ‘70s, the dark mass problem you would also say these dynamical properties we would kind of understand already?

Gunn:

I think so. I think so. Yeah.

De Swart:

Okay.

Gunn:

Yeah. Yeah. Yeah. Yeah.

De Swart:

And, how important was the rotation curve you mentioned from Vera Rubin?

Gunn:

Well, I think that it sort of tied things down. Because, with Fritz’s work you knew that this stuff was there in clusters, but you didn’t know where it was. I mean, you knew that it was roughly coextensive with the galaxies. Right? But the clusters have a velocity dispersion of 1,000 km a second. And so, what this has to do with galaxies wasn’t clear. But what Vera did, when she showed the rotation curve, she showed… and her rotation curves didn’t go far enough to get anything like the mass-to-light ratios you observe in clusters of galaxies, but she did show that there was stuff out there, which was gravitating, which we did not see. And furthermore, that when she last saw it the mass was going linearly with radius. And so, the sort of leap that this stuff was dark, was associated with galaxies, therefore when the galaxies, when the cluster formed, you know, galaxies fell into the cluster, because I think that’s, that was also something that sort of came out of the, out of the paper with Richard, with Gott, that we… I don’t think people have thought very much about sort of how clusters got assembled. And so, that simple spherical model was a kind of a paradigm that then people used for a long time. I mean it’s certainly not very accurate, but it’s qualitatively correct. So, when galaxies fall into a cluster they bring this dark matter with them and from the, from the mass-to-light ratios in clusters you can sort of say, “Well, you know, how much dark matter is there in a galaxy in order that when they fall in they make the clusters like the clusters are?” And that factor is about ten. You know, ten-ish. And, then that sort of reconciles the, the Omega-baryon that you get from blackbody nucleosynthesis. The Omega is .2 that you get from, from dynamics and all these other things in the universe, but you still can’t get to Omega as one.

De Swart:

Hmm. But, by the way, “baryon” was not something which was…

Gunn:

“Baryon” was not a word that people used in those days. That’s right. Because it was not suggested. I don’t remember when. I mean people, particle physicists just talked about baryons and that was not the issue. But you didn’t use the phrase “baryonic matter.” And that was another one of these adiabatic things, I think. There was some work, and actually there was a paper. I think one of my favorite papers that I ever wrote was a paper, the, one of the first papers about WIMPs in astrophysics.

De Swart:

With Gott?

Gunn:

With, no, with Schramm and Lurch, and Steigman.

De Swart:

Yes.

Gunn:

I think it was called…

De Swart:

It’s ‘77.

Gunn:

“On the Existence of… On Some Effects and the Existence of a Heavy Neutral…”

De Swart:

Lepton.

Gunn:

Lepton.” Right. Right. And I don’t remember whether Weinberg was on the paper or not, but it was definitely, I had…

De Swart:

I don’t so.

Gunn:

I talked with Steven. Steven had written this paper suggesting that there could be something heavy like this. And, that if it was, if it had ordinary weak interactions that you get the density about right. If the particle weighs a few hundred GeV, and has sort of ordinary weak interactions, then these things are made thermally. In the early universe they mostly annihilate. But the, and all of this happens when the temperature, when the temperature is, you know, several Ge, or a GeV, about. And if they interact weakly enough, if they, if the annihilation cross-section is small enough then there will be some left. And, that the matter density you get from this is at least in the ballpark of what you need to do the dark matter.

De Swart:

To do, and “to do the dark matter,” you mean the discrepancy or to get Omega 1?

Gunn:

No. To get, to get Omega as a .2.

De Swart:

Yeah. Yeah. Yeah. Yeah.

Gunn:

Right. Right. So, and the… well, I say Omega is .2, but the, you know, this was just a sort of back-of-the-envelope calculation. [laughs] And, and the exact cross-section things depend on… so you could make Omega as one just as easily as not. It wasn’t Omega is 1,000 and it wasn’t Omega is .001. But certainly, somewhere around one.

De Swart:

Was it the Lee and Weinberg paper?

Gunn:

Yeah. It was the Lee and Weinberg paper.

De Swart:

Okay.

Gunn:

Right. Right. Right. Right. I think there were several papers, and I don’t, sort of following this theme, and I don’t remember which one actually suggested this.

De Swart:

Okay. Yeah.

Gunn:

And I don’t actually know why Steve wasn’t on that paper that I was first author on, because I spent a long time talking with him about this and making sure that I understood the mechanisms and so on. Maybe he didn’t want to be. I don’t remember.

De Swart:

Uhm-hmm. And, you said something…

Gunn:

I don’t remember whether we first used the phrase “non-baryonic dark matter” in that paper. I don’t remember. But anyway, it was about that time when people started seriously thinking that these could be some other kind of new elementary particle. That was when people started using that phrase.

De Swart:

Where did the particle physics come in? Because, particle physics, you weren’t doing particle physics. Right?

Gunn:

No. No. I mean; I was interested in it entirely from the astrophysical point of view. And, you know, because this paper that Steve said, “Well, maybe this is what the dark matter is.” Right? Just purely from a particle physicist’s point of view, but he didn’t do anything about growth of structure, or any of this. So, I think that that was mostly in the paper of mine and others.

De Swart:

So, that was the era that particle physics got involved with astrophysics?

Gunn:

Yes. That’s right. That’s right. That’s right.

De Swart:

So, what were the, the other connections? So…

Gunn:

Not, well, not really very much then. It sort of grew. Alan Guth was a particle physicist and, you know, he was studying scalar fields and realized that this rather magical thing happened, could happen. And, Weinberg, of course, was already interested in cosmology, because I think that was shortly after The First Three Minutes book came out. And, also maybe his cosmology, first cosmology book. So, he was certainly aware of how little we knew and he was a pretty bright fellow. [laughs] So, he was interested in this stuff but was not particularly interested in sort of the nitty-gritty astrophysics. He was, you know, more interested in the kind of particle physics point of view. But I think it was shortly, it was around that time that people suggested that there, that the dark matter might be axions. And so certainly WIMPs and axions were the only… and maybe heavy neutrinos. I don’t remember when heavy neutrinos came along. Because there, there’s just a, just a sort of slight variation on the WIMP thing.

De Swart:

So, maybe in a different context, how would you describe the status of the, of the dark matter problem and its, and its development?

Gunn:

Well, I think, you know, in about 1980, let’s say, so sort of after, after, well, after inflation.

De Swart:

Okay.

Gunn:

After the idea of inflation came along.

De Swart:

But maybe…

Gunn:

Maybe after, after COBE.

De Swart:

And, if we start earlier, what was this, was it a problem?

Gunn:

Well, I’m not sure what you mean by the “dark matter problem.” You know, people knew, in the back of their minds, as soon as Fritz wrote this down about clusters that we had a problem. That, you know, that there was a lot of matter that we—that was gravitating that we didn’t see. Vera said, “Yes. Yes. Yes. This is right. It’s associated with galaxies.” And so, the question that’s still there is, “What the hell is it?” We don’t know. The other question was, “How much of it is there?” And, again, I think almost immediately, and I may be over, I may be over-optimistic, we understood its dynamical properties pretty well, and we knew where it was. We didn’t know what it was. We didn’t know exactly how much there was, except that, again, we knew how much there was in clusters of galaxies, and from groups, and we therefore knew sort of by, I think, pretty straightforward induction, how much there was in galaxies. And that the dark matter associated with structures we saw was enough to make Omega matter about .2 or about .25. And so, just to go back again, you know, if you want Omega to be one, for some reason, then you have to find some way to distribute the stuff out in the, in the otherwise unpopulated parts of the universe. Nobody knew how to do that. The simulations were getting better and better. And so, it was really looking like Omega was .25 or thereabouts. Then, okay, another important thing. Then Boomerang flew. Because the story now depends… and I don’t remember when Boomerang flew… it was before COBE flew. But it must have been, it must have been after 1980. So, it must have been, oh no, ‘70s maybe. I don’t know. Anyway, look up balloon experiments.

De Swart:

Yeah. I can.

Gunn:

You know. You know.

De Swart:

Yeah.

Gunn:

So, the new thing that that… so, until then we didn’t know what the fluctuations were. Right? We knew they should be there. We knew that the, that a dark matter dominated universe with Omega is .25 or, point, or one. Didn’t matter. Was, gave fluctuations that were consistent with the limits at the present, at that time, for the, for the observed fluctuations in the background. We had no idea what the spectrum of fluctuations was. So, when Boomerang flew one of the things that they really came down on was that the spectrum did not fit our expectations unless Omega was one. So, that was the, I had forgotten that in the earlier conversations. So, now we have a big problem, because we see the universe today. We know roughly how it has evolved from gravitational collapse. We know what those perturbations had to look like at the epoch of recombination. But the geometry is such that you don’t… and then you observe the spectrum at the Boomerang, observe the spectrum at the epoch of recombination. And the two do not agree unless Omega is one. So, whenever that was that Boomerang flew, late ‘70s maybe. It was a couple of years before COBE. So, there was something really wrong and there were many, many things that people thought might be wrong. You know, it was some funniness that we didn’t understand in, in the growth of structure, or almost sort of had to be that I think. Or there was something funny with the blackbody background because we, Boomerang measured fluctuations but we still, I think at that point, did not know that the background was accurately blackbody. So, there was a lot of flailing in the, in the community. People were writing crazy things and thinking crazy things. And the problem sort of got nailed with COBE, because COBE said, “Yes, it’s blackbody, and the spectrum and the Boomerang spectrum is correct. And so, Omega has to be one.” And so, now we have a big problem, because you add up all the matter, you get Omega as .25. You look at the blackbody background, Omega is one. So, what is the, we now know what the resolution is, but that had to wait until the supernovae came along to fix.

De Swart:

Yeah. So, what I just was thinking about is the problem of missing matter. You keep talking about the dark matter, but in the beginning, it wasn’t…

Gunn:

It was missing matter.

De Swart:

Yeah. Yeah.

Gunn:

We didn’t know what it was. Right.

De Swart:

So, what it meant for the developing field of cosmology in the beginning of the ‘70s. So, how central was this problem.

Gunn:

Well, I don’t think the field was very organized. So, I don’t know whether I can answer that question. In a way, I’ve sort of, I’ve given you my take on it, which may or may not be, be terribly accurate. There was a lot going on and there was no, I don’t think there was any consensus in the field. There was sort of an opinion for every worker that was doing various things, and it wasn’t until the data started getting better, until we really knew what the blackbody background was, that then people began to agree on any kind of synthesis of stuff. So…

De Swart:

When, for example, were the first undergrads or graduates you trained in cosmology? When was that?

Gunn:

The… let me think. When I was at Caltech, I had both theoretical and observational postdocs, graduate students. And so, I had people, Hessel and Schneider, working on observations of clusters. Hessel also, and I, worked with Bill Press and they developed the Press-Schechter formalism, which is a very nice sort of… it’s better than qualitative but isn’t quite quantitative, description of the growth of structure. I had a couple of students doing relativity, but not in cosmology.

De Swart:

So, it’s more than I was thinking. When did the cosmology you were doing, turn into the stuff that students started learning from lectures?

Gunn:

Well, I think all through the ‘70s I gave the extragalactic/astronomy course.

De Swart:

Okay. In Caltech?

Gunn:

And so, you know, sort of as soon as this stuff started being developed and I was sure enough of it, I started teaching it at Caltech. And I think that this was going on in parallel at other places, the University of Chicago, in particular, because Schramm was very, was doing this stuff and largely with us, and he was teaching and having students.

De Swart:

And did you use your own notes on that? Did you use a book?

Gunn:

Oh, no. I’ve never used a cosmology book. I just use my notes. I don’t think there are any good books, actually.

De Swart:

Still?

Gunn:

Weinberg’s book is too formal. And lately there are some but, you know, for a long time there just wasn’t a suitable text. But I think that was, I think that the sort of graduate training program was going on kind of in parallel with this, and it tends to happen in astronomy because you get behind immediately if you don’t keep up, of course. [laughs]

De Swart:

That would be interesting to see when were the first things, for example.

Gunn:

Nice to go back and look at those lecture notes, but I think I no longer have them. [laughs] So, I couldn’t.

De Swart:

It would have been. So, yeah. So, let me think. So, maybe one more question. So, how did what you were doing contrast with what earlier observational cosmologists were doing? (Sigh) Did you feel like you were doing?…

Gunn:

Well, no. I was doing new stuff and I, there wasn’t very much earlier observational cosmology is the truth of the matter. I mean, again, Sandage’s paper in APJ133, on, you know, the “Ability of the 200-Inch, etc. to Distinguish World Models” was basically the observational cosmology paper.

De Swart:

Oh, wow.

Gunn:

And, you know, Burbidge and folks working on rotation curves, they were working on galaxies but they, nobody tried to put anything in the cosmological context. And there was a famous meeting that Beatrice ran at Yale in, it was in the ‘70s. It had to have been in the ‘70s. She died in ‘82. So, it must have been around ‘76. So, you should look and see if you can find the proceedings from that. That was the first meeting. She was very, she had a sort of omnivorous mind, and she was very interested in understanding not only galaxies, per se, but how galaxies interacted, how they sort of fit in the cosmological framework. There weren’t N-body cosmological models by then, but she was thinking about these things. And she designed the program for this, for this conference, to concentrate on those things. And I think that was the first conference at which people really began seriously to think about galaxies as inhabitants of the universe instead of, you know, individual things to study. So, you know, the ideas of mergers, the ideas of interactions, the ideas of all of these funny things that we see in clusters, with winds blowing past galaxies, and all of this stuff. I think the beginnings of that was that, was that meeting. And, then that fit immediately sort of naturally into this sort of growing cosmological awareness that people were having. So, it was all kind of an organic thing that was the course. Yeah.

De Swart:

I was just wondering if there were really sort of some hostile, maybe…

Gunn:

No. No. No. I don’t think so. I mean, you know, the old guard dies away, not very gracefully, usually. [laughs] And so, there were a lot of people who were saying this was nonsense. But it didn’t really matter. [laughs]

De Swart:

Yeah.

Gunn:

I have a talk I have to go do.

De Swart:

Yeah. So. Thank you, so much.

Gunn:

You’re welcome.

De Swart:

Yeah. I’ll actually be leaving again on Saturday. So, I think I, I’ve had some really [inaudible] conversations and this was also… because, I would say that you were one of the first in these new physical cosmology…

Gunn:

I think so. There weren’t very many people when I started. Yeah. Yeah. I don’t know whether I, I don’t think people followed me, it was just that it was, it was time.

De Swart:

Yeah.

Gunn:

It was time for this to happen.

De Swart:

Yeah.

Gunn:

Right. Right.

De Swart:

Yeah. So, that’s really interesting to see how this problem of dark matter actually traces its development of, in cosmology.

Gunn:

Right.

De Swart:

So, thank you very much.

Gunn:

Okay.

[End session 2]