Notice: We are in the process of migrating Oral History Interview metadata to this new version of our website.
During this migration, the following fields associated with interviews may be incomplete: Institutions, Additional Persons, and Subjects. Our Browse Subjects feature is also affected by this migration.
Please contact [email protected] with any feedback.
This transcript may not be quoted, reproduced or redistributed in whole or in part by any means except with the written permission of the American Institute of Physics.
This transcript is based on a tape-recorded interview deposited at the Center for History of Physics of the American Institute of Physics. The AIP's interviews have generally been transcribed from tape, edited by the interviewer for clarity, and then further edited by the interviewee. If this interview is important to you, you should consult earlier versions of the transcript or listen to the original tape. For many interviews, the AIP retains substantial files with further information about the interviewee and the interview itself. Please contact us for information about accessing these materials.
Please bear in mind that: 1) This material is a transcript of the spoken word rather than a literary product; 2) An interview must be read with the awareness that different people's memories about an event will often differ, and that memories can change with time for many reasons including subsequent experiences, interactions with others, and one's feelings about an event. Disclaimer: This transcript was scanned from a typescript, introducing occasional spelling errors. The original typescript is available.
In footnotes or endnotes please cite AIP interviews like this:
Interview of Sandra Faber by Alan Lightman on 1988 October 15,
Niels Bohr Library & Archives, American Institute of Physics,
College Park, MD USA,
For multiple citations, "AIP" is the preferred abbreviation for the location.
Interview covers Sandra Faber's childhood experiences; parental background; early reading; early preference for steady state model; relationship between questions and answers in science; confusion over being a woman and being a scientist; lack of female role models in science; education at Swarthmore and the influence of Sarah Lee Lippincott there; graduate work at Harvard; husband's job; graduate work at the Department of Terrestrial Magnetism; influence of Vera Rubin; early results of dark matter by Morton Roberts in the late 1960s; thesis work on photometric studies of elliptical galaxies; community's attitude toward excess mass in rotation curves in the late 1960s; motivation for work on the Faber-Jackson relationship between luminosity and velocity dispersion; motivation for work with the Seven Samurai (Burstein, Davies, Dressler, Faber, Lynden-Bell, Terlevich, and Wegner) on peculiar velocities; attitude of the community toward the Seven Samurai work on peculiar velocities; attitude toward the big bang assumption of homogeneity; attitudes toward the horizon problem, the inflationary universe model, missing matter, the flatness problem; discussion of what types of problems can be addressed in cosmology; attitude toward Center for Astrophysics (CfA) red shift surveys by de Lapparent, Margaret Geller, and John Huchra; importance of understanding how large-scale structure is formed; issues of gender in science and the experience of being a woman in science; the ideal design of the universe; the question of whether the universe has a point.
I wanted to start by asking you a little bit about your childhood. Do you remember any particularly influential experiences you had as a child?
I remember looking at the night sky a lot — just visually with star charts and a pair of binoculars.
What age was that?
Well, I guess with the charts, probably eight or nine. I had a big chart from the American Museum of Natural History. It was bulky; it was hard to use; it was something that I would roll up and roll up again.
Were you in Boston at that time?
No, I was in Cleveland. I was born in Boston, but I moved to Cleveland when I was three.
So, there was a night sky that you could see, away from the city?
It was a typical suburban night sky. I don't think it was anything to write home about. I remember spending evenings looking at the sky with my dad, who was interested. He was a civil engineer and was interested in science as a kid. And he always encouraged me. He found the binoculars for me. But I was the kind of kid who liked omnivorously almost all kinds of science; rock collections, fossils and I like leaves, I like plants, and I like biology. Almost everything. I didn't really focus until I got older.
Was your mother interested in science as well as your father?
Not at all. Well, let me qualify that. She was very interested in medicine, and she followed that. I think if she had been growing up today, she might have become a doctor. She renounced career interests and things like that in favor of family responsibilities. But all her life she did follow medicine, and she knew a good deal about it.
Did you build any things, any science projects as a child?
No. I'm very bad with my hands. I'm not an instrumentalist now. I'm a remarkable bull in a china shop, in any kind of laboratory. In fact, I really think twice before I pick up a screwdriver or do anything because I'm liable to do something really stupid. One of the most embarrassing failures of my childhood educational career revolved around science projects, which inevitably turned out badly. So, my talent isn't along those lines.
Do you remember any popular books in science that you read?
Yes. I read or tried to read popular books by [James] Jeans, written in the 1920s. My godfather, who was also an engineer, had collected some popular writings from the earlier part of the century. I should say by the way that there are almost two generations between me and my parents. I was born when my mother was 42 and my father was 45, and I was the only child of older parents. And so their friends were also older. Their education came from the beginning of this century rather than the middle of the century, as it would have had the age gap been more normal. So, that's by way of explaining why my godfather had in his collections books by James Jeans in the 1920s. So I read some of those. Stars and Their Courses is one I remember the best. I think, though, the most influential book for me from that period was Hoyle's famous book, which had a big influence on so many people. What was it?
Frontiers of Astronomy.
Frontiers of Astronomy, that's right, which I read when I was in high school. You're probably familiar with it. It dealt with steady state theory. And the aesthetic beauty of the steady state theory really appealed to me. I was quite convinced after reading his book that that had to be the proper explanation for the beginning of the universe.
Did he mention the alternative, the non-steady model?
So, among the two alternatives, you personally preferred the steady state?
Yes, I strongly preferred it because I thought aesthetically it was compelling. It definitely got rid of the question of the beginning by saying there wasn't one.
And that appealed to you?
That really appealed to me.
To remove the concern of the beginning?
Yes. It was my first exposure to something which has happened to me a lot of times since, and I feel sort of bad that I don't understand these things more deeply than I do. But it's a phenomenon that happens repeatedly in physics, astronomy, and cosmology. One poses what appears to be a perfectly reasonable looking question and then one finds that the answer is really not to answer the question as originally understood, but to show that the question somehow was flawed, that there was an implicit assumption that wasn't correct.
In this case, what would the question have been?
What is the beginning of the universe? Steady state answered it by saying that there wasn't a beginning, so we don't have to answer the question as posed. We replace it with yet another [view]. I no longer believe in steady state, but the demonstration that steady state was wrong was a lesson to me. It showed me that there's more to doing physics and astronomy than just aesthetics. But I still do find myself strongly influenced by aesthetic principles, as [do] most physicists, I think. My enthusiasm about inflation [the inflationary universe model] is an example of that.
Inflation is something I want to ask you about later. Do you remember roughly what age you were when you read Hoyle's book? You said you might have been in high school.
Yes, I was a junior or senior, I'm sure. It was just about that time.
So that early you had been exposed to different models of cosmology.
Yes. And I was very interested in the nature of the universe, and I was very impressed by the people who even thought about it. Hoyle's book was a revelation to me on two levels. First of all, there was a theory there. But also, it was my first encounter with a group of people who were actively struggling with those issues. I was really inspired by their boldness and daring — how anybody could be rationally dealing with things which were so interesting; and so remote.
Approximately what year was this?
I graduated from high school in 1962. So this would have been somewhere between 1960 and 1962.
Did you know by the time you went to Swarthmore that you would want to be a scientist?
I wasn't sure that I could be anything, because I was quite confused at that time about how I could be a woman and at the same time pursue these questions which I thought were really interesting. I just was taking one day at a time. A very clear path was open to me at that time by education. One went to high school, then one went to college, then one could see going forward to graduate school. I could see that path fairly clearly. Beyond that, the future was a total mystery to me, as to whether or not I would have a job and work, whether or not I would get married and have children. I just had no idea what I would be able to do.
I was going to ask you about this a little bit later, but since you brought it up now, do you remember why it was that you felt confused about what the future would hold because you were a woman? Can you identify any influences that made you confused? Was it a message you got from your parents, or was it a broader message that it would be difficult for you as a woman? Had you gotten it from your teachers?
I think it was a broader, diffuse message. Just simply because at that time, say in 1960 or 1962, there weren't very many women scientists, and there were even fewer who had what might have been termed at that time normal family life. A woman scientist in the 1940s and 1950s, in general, was a single woman.
Did you know of some of those women?
No. I didn't know of any of them anyway. The only working women I saw whose lives interested me at all were school teachers in high school. But I didn't think that I wanted to be a school teacher in high school.
So you had no role models whatsoever?
I really had none that I can remember now anyway. I had none. The first role models that I began to see were women professors in college, and specifically there was a woman astronomer at Swarthmore, Sarah Lee Lippincott. She was on the one hand an inspiration. She was a woman who was professionally employed as a scientist and as a researcher too. She wasn't a faculty member. Her job in life was to do astronomy in the observatory. But on the other hand, I felt that I wanted to go beyond her position because she had never gotten a Ph.D., and that prevented her from becoming a faculty member or an astronomer at a larger, more forefront institution. I thought she was a wonderful person. But, I also felt that I wanted not to stop at a master’s degree as she had done.
Why was it important to you to get a Ph.D.? Why was there something more than just the research that was important for you to strive for?
I felt that if I had a master’s degree, I would be limited to a small institution. And therefore I would have limited access to tools. That's point number one. But I also felt very strongly that going on and doing a Ph.D. would add greatly to my education; a seasoning, actually. When I said that I was confused, it was really this diffuse message that I couldn't see too clearly people out there who had gone in the direction that I wanted to go. But I must say that as far as specific messages were concerned — with either my mother, my father, relatives, or teachers — I had fantastic encouragement from all of these people. That kind of encouragement was certainly responsible for my having gone along as far as I did. So, there was a conflict. My way of dealing with that was simply to take one step at a time.
You mentioned Sarah Lee Lippincott at Swarthmore. Tell me a little about the influence that she had on you.
She was not the only one but to address that question. One thing that really impressed me was the fact that she was very independent. She's the first woman I ever met who was in that situation. All of the women in my family were either married or divorced and living within a family situation. She lived by herself. And, on one occasion, she invited me over to her apartment and fixed dinner for me. She entertained me. It was very pleasant. I saw that she had made a life for herself. She had a place to live which she had created. She had hobbies. She was a photographer and had published a book of photographs of Philadelphia. She had a world for herself outside of astronomy, and she was doing this completely by herself. Later, she got married. She married Dave Garroway of the Today Show, and that was a very interesting match. I once visited them in their home together. It was lots of fun interacting with the two of them. But she was a very lively, vital person, clearly a very intelligent person, the first woman I had ever seen making it on her own.
This is a very hypothetical question. Do you think if she had been a male astronomer, who also had the same personal relationship with you, that you would have been as inspired to go into astronomy?
Probably. Because in fact she was not the only one at Swarthmore I interacted with. There were two astronomers at the observatory there that influenced me a great deal. The other one was Peter van de Kamp, and he obviously was a male. He was a fantastic and very positive influence on me as well. We had a very good relationship, a friendship. Somehow our two personalities clicked. I would say that was also the same with me and Sarah Lee Lippincott. Clearly, I didn't look at van de Kamp and say he was a remarkable person because he has a career and a family and a house — the sorts of reactions I had to Lippincott. But he certainly inspired in me a great love and interest in astronomy. The observatory was a very happy place. There were just a handful of students who were interested in astronomy, and it was a very small, congenial atmosphere, a very good place to get one's basic foundation.
When you studied astronomy at Swarthmore, did you also study cosmology?
Unfortunately, I didn't. This is one of the bad things about learning things in a small, liberal arts college. The astronomy that was done there was focused on nearby stars, as you know, and the astronomers there didn't really know much about the wider issues of cosmology. I didn't care too much about that at the time because I felt as though it was time for me to be learning fundamentals of a classical sort. And, even more than I studied astronomy there, I studied physics in the physics department. However, it would have been good in retrospect had I had some exposure at that time not only to classical cosmology, which I found easy to learn, but more importantly to particle physics, the physics of the early universe, which I find extremely difficult. Most of that I have learned completely on my own, since graduate school. I didn't get that in graduate school either. It's been hard for me to gather whatever understanding I have. I didn't have much of a foundation laid in my formal education. It's been a real handicap.
When you went to Harvard to do your graduate work, did you get an introduction to cosmology then?
Not really. I did take a course in general relativity, but it was a very classical course that just dealt with the equations and a few applications to the expanding universe and black holes and so on. But, aside from giving one some familiarity with what these equations look like, just from the standpoint of classical general relativity, it bore almost no resemblance to what we need now for cosmology. There was no particle physics.
Did you study cosmology any on your own at that time?
No, I didn't. At that time if you had asked me what I doing, I would have said I'm interested in galaxies. There was a lot to be done just in that field, at that time, having to do mostly with stellar and gas dynamics and so on. It was clear to me at that point that I was better suited to be an observer than a theoretician. And so I was pretty fully occupied mastering the techniques of observational work, data reduction, learning how to program a computer, which I had not done in college. So that was a new experience.
Do you remember any people who were particularly influential?
I had a strange career at Harvard, in the sense that I had by then gotten married. Andy and I had become acquainted at Swarthmore. My graduate career was impacted by his problem of serving in the Vietnam War. He was a member of that cohort of men who were permitted to have one year of deferment in graduate school but no more. This was a transitional policy. So he took one year of study in a master’s program of applied physics at Harvard. I'm a year older than he is. At that point I had completed two years of course work at Harvard, which, as in most places, is all you need. Then you go on to do your thesis. He then had to leave graduate school and either serve in Vietnam or move to Canada, which we did consider. Or he had to find a job which would give him a deferment. He had previously worked in underwater acoustics and now was able to find a job at the Naval Research Laboratory in Washington. I was not prepared to consider staying on at Boston and conducting a long-range marriage — which would have been possible to contemplate had we had more money to allow airfares and frequent visits. But we had very little money, so visits would be quite infrequent. That seemed very unpleasant. So since I completed two years of course work, I didn't really have to be in residence at Harvard. And they were quite flexible with me. They let me go away on a so-called travel and guidance program. But that was really quite a sham. Implicit in that program was the notion that you had found an advisor elsewhere who offered you an opportunity that Harvard couldn't provide. That was simply not true. My advisor was going to stay at Harvard, and I was going to physically be somewhere else. I wangled an office at the Naval Research Laboratory, amongst the astronomers there, who didn't do anything related to galaxies at all. They were doing things like measuring the temperature of Venus and studying water masers. So at least there was some astronomy, but nobody to talk to. After a year and a half there, I was invited to take residency at the Department of Terrestrial Magnetism, with Vera Rubin and Kent Ford, with whom I had worked one summer before. And I finished out my graduate thesis work there. That was a very good atmosphere because they were very interested in galaxies. Overnight I had found a very pleasant and productive niche. So that's what happened. That's a kind of preamble to your question of who was influential at Harvard.
Well, just during that period of time. I guess Vera Rubin was extremely influential?
She was more influential than anybody else. I like to call Vera my defacto thesis supervisor. My formal supervisor was John Danziger. He did a very fine job with me. He was a wonderful facilitator. Whenever I needed something from Harvard — my permission to go to Washington or travel money — he was right there with incredible support. On the other hand, he was a stellar astronomer and not too able to guide somebody who was doing a thesis on galaxies. Why was he my advisor? Well, he was the only optical observer at Harvard at the time.
Was Vera Rubin interested in cosmological applications at that time?
No. I think that is a fair question. She wasn't. But she was on the verge of thinking about things which were to have a profound cosmological application. I like to recall fondly one day when I was doing my thesis work there, she got a phone call from Morton Roberts, who was at the National Radio Astronomy Observatory. He said "Vera, I have some very strange results on rotation velocities in M31, and I'd like to come up and discuss them with you." He came up the very next week. I remember sitting around a table, and Mort Roberts was showing Vera this graph of rotation velocity versus distance [from the center of M31]. His new radio results went 50% farther out and showed the H I there going at exactly the same velocity as the inner stuff. Mort was clearly very excited about this. I said, "Well, so what? You know. Fifty per cent farther out is only 50% more mass. It's not a big change in mass." He said, "You don't understand. There's no light there." And I remember being profoundly unimpressed, I think partly because I was already convinced that there was something totally crazy about mass measurements in astronomy. My view then was that somehow these velocities didn't mean anything, that they were wrong, for some reason. I remember why this was. I had done a lot of wide reading on the structure of galaxies and the nature of groups of galaxies as it was understood at this time, looking for a thesis project. So this was the late 1960s. My thesis project was a photometric study of elliptical galaxies. But actually what isn't so widely known is that it was supposed to be a study of double galaxies. I was interested in whether or not the properties of doubles were more closely correlated than pairs picked at random. A crazy idea — it was a dumb idea, in light of hindsight. But it had led me to go through all the galaxy catalogue — by hand of course, because they weren't on tape at that time — and look for doubles. That brought me right up against the problem of when do you pick a double on the basis of proximity as opposed to velocity differences. I had found many doubles that had big velocity differences, and when I just calculated characteristic masses for them, [the masses] were impossible. Of course, I was aware that people were measuring masses of clusters and finding excess mass. So I attributed this velocity discrepancy in doubles to the same phenomenon — whatever it was. I knew that people argued about it. I knew it was a bottomless pit. Nobody had ever solved this problem. And I just didn't want to bother with it. I didn't want to think about it. It didn't seem to be a question that was ripe for solution. When Mort Roberts came and showed yet one more crazy velocity, my attitude was, "These velocities are never right. Don't bother me with that — we don't understand this. Why are you making such a big deal? It's just one more example of this problem." [Faber and Lightman laugh.]
Before I go past this period: Had you by the time you finished graduate school, on your own or through courses, seen mathematical expositions of the various types of cosmological models; for example, the open versus closed universes?
Yes, I had. In my general relativity course there had been a little bit of this.
Do you remember having a preference for any particular model?
Oh, even independent of mathematics, I was well aware that there were such things as open and closed universes, that this was [related to] q0, that Allan Sandage was very interested in measuring q0 from distant galaxies. I was familiar with those kinds of things.
Did you have a preference for any particular kind of model?
Ok. Let me go on a few years past your Ph.D. In your work with [Robert] Jackson on the velocity dispersion of elliptical galaxies, do you remember what your principal motivation was in deciding to do that work?
Yes. I remember very well. My thesis was concerned with the spectral properties of these galaxies, and because I was limited to a small telescope, my measurements in the thesis work were rather coarse. I used intermediate band filters. When I came to Lick, at that time the image dissector scanner had just gone into operation. It was very clear now that for the first time people could obtain much higher resolution spectra of galaxies in a form in which you didn't have to laboriously micro-photometer the spectrum on a photographic plate and convert to intensity. In the IDS, we had a detector which was pretty linear. So the very first thing I wanted to do when I came to Lick was to take these spectra and see what they looked like in full detail. There had been many problems in interpretation of filter photometry, due to the fact that it was just so coarse. I didn't really know what spectral features I was measuring. My first work here was an attempt to find out what lines were actually in my band passes, and I was very interested in making radial measurements as a function of distance from the centers of these galaxies to find out how the spectral properties change as a function of distance. So I embarked on a program with the IDS which was purely centered on the composition, metallicity, and stellar population models.
So you were still following that thesis work?
I was, very much. However, when I looked at the spectra, any idiot could see that some of these things had broad lines and some had sharp lines. So I took a quick detour because there had been almost no work done on measuring velocity dispersions, and it seemed like this was something that had just fallen my way that I should follow up.
Did you have any idea that measuring these velocity dispersions would lead to some interesting relationship? Or were you motivated first just to measure them?
Well, as I was taking the data, I could not help but notice that the more luminous galaxies had broader lines. So, when we worked up the data quantitatively, it was with the intention of finding out how good that relationship was. Of course, it turned out to be a pretty good relationship. I think now the next question you might ask is, "At that time, did you have the faintest idea of how you might explain something like that?" I viewed the work we were doing there at that time as very similar to what you might have been doing at the beginning of the century when you plotted, purely empirically, luminosities of stars versus their surface temperatures — to make the first H-R diagram. I'm sure the people who did that knew that they had stumbled on something very fundamental. But for a long while people didn't really know why there was that relationship. They knew it was a key to explaining things. Likewise, at that time, I felt that the relationship we found was a key to explaining the origin of galaxies. But I had not the faintest clue at that time of any theory that might explain it.
To go back again, when you first decided that you had measured the velocity dispersions, you did that because you noticed that there were broad profiles as well as sharp profiles and that very little had been done on the broad profiles — on measuring these velocity dispersions?
Yes. I think at the time what I decided was that there had just been exactly one paper, the one paper by Minkowski. I was aware of that paper. While we were doing our work, Morton began to publish results. I think when we published our paper, those were the only two previous studies that had been done on that subject.
So this was an area that had not been explored, and you felt you were getting data in an area where there was no data?
Right. But I didn't have at the time a terrific sense of exploring wonderful, virgin new territory. It was just something that came my way, and it clearly looked like it was a fundamental property, so I should publish it.
Let me go on to some much later work that you've done in the last few years. Do you remember your own personal motivation in beginning the "Seven Samurai" project — the project to measure distances and velocities of a sample of elliptical galaxies?
Yes. In fact, you might like to read about that. I could provide you with my NSF request, which I didn't reread for this interview. It's possible that what I say here might differ some from what I wrote down at the time, but that is documented. I requested money from NSF to pursue that work. I seem to have a knack for doing things for the wrong reasons, and this was no exception. As a result of collaborating with Roger Davies and Roberto Terlevich and Dave Burstein previously, we thought we had discovered the scatter in the Faber-Jackson relationship was real. Is it okay to call it that?
There had been some controversy about that. "Different groups got different zero points from that relationship.
When you say the scatter was real, do you mean a departure from an exact relationship?
That's precisely right. But it was actually a little more profound than that. When Bob Jackson and I published our first relationship, we measured a particular zero point. That is to say, at a given luminosity our line predicted a certain level of velocity dispersion. [Paul] Schechter and a group of collaborators — [Wallace] Sargent, Boksenberg — observed another sample of galaxies and found the same relationship but a different position of the zero point.
This was 1976?
Probably, about that time. It was very close on to what we had done. And there were two interpretations of that. One was that there was a systematic difference — an error — in the actual sigma [velocity dispersion] measurements themselves. Since Schechter's group had done a better job with a better technique, most people believed that if there was an error, it was in "Bob Jackson's and my velocities, which were heavily weighted to eye estimates. So there was that problem. But an alternative interpretation, which seemed very unreasonable at the time, was to say that when you put both of these samples together, the total spread was real. Everybody's velocity dispersions were correct, and just by bad luck the two groups had picked two different samples of galaxies, one with systematically lower velocities than the other.
In general, you would expect that if that explanation were correct and you put the two samples together, the tightness of the relationship would diminish.
It did. If you took everybody's data at face value and just put it on the same page, you got a much broader relationship. And Jackson's and my data lay at the top, and these other data lay at the bottom.
So that would have less predictive power.
Exactly. Which nobody liked either.
I started by asking you about your original motivation for the "Seven Samurai" work.
Yes. That's right. I'm still on that subject. Now what Roger Davies and Roberto Terlevich and David [Burstein] and I showed was that if you took the same sample of galaxies and compared the magnesium measurement instead of sigma [the velocity dispersion], the galaxies lay in the same relationship on the page. A galaxy which lay high in one diagram lay high in the other diagram. And that suggested to us that everybody's sigma measurements were right, and that the scatter was real, that the relationship was unfortunately less predictive than we had thought. There wasn't any error in the sigma measurements. Everybody's sigmas were fine. And also there was a second parameter here.
Distinguishing one sample from the other.
Well, yes. Or somehow the second parameter is perhaps nothing more than a residual of any particular galaxy from the mean relationship. And furthermore — and this is where we went wrong — we thought there was a correlation with ellipticity in the sample, in the sense that the highly elliptical galaxies were on the bottom and the round ones were on the top. And it's true that in that sample there really was a correlation. Now, if that had been true, it would have been possible, in all probability, to say something about the true shapes of elliptical galaxies. That seemed to us to be a rather exciting prospect, just about then coming to the fore as a major question. We thought we would do a much larger sample — accumulate magnesium, sigma, luminosity measurements for a much broader sample — and in the end perhaps be able to tell you whether a round galaxy was round because it really was round or whether it was elongated and seen face on.
Everybody would love to have known the answer.
Everybody would love to have known the answer. Now there was a great deal of skepticism, but our original correlation of ellipticity was statistically valid. The tests that we did said that it was significant at the 99% level. And I still believe those tests. Still, in the light of a subsequent sample, we know that there's no correlation there, and that was just a piece of bad luck. Many statistical things turn out badly. That's an example. So, if you read this NSF proposal, there's a great deal of energy devoted to explaining why there might be more than one parameter and why the second parameter might be related to galaxy shape and therefore how people should go out and collect bigger samples to study elliptical galaxies. And somewhere in there, in like two sentences, it says that after studying all these properties maybe we can develop a better distance indicator. Maybe we can somehow tune-up the original Faber-Jackson relationship and collapse the scatter once again. And if we can do that, then we propose to study deviations from the Hubble flow. But the whole thrust was on the galaxies; it wasn't on the Hubble flow.
Did you need a more accurate Faber-Jackson relation in order to measure the deviations from the Hubble flow?
I didn't really think about it quantitatively. I would say now, with the benefit of hindsight, we probably could have seen something with the old Faber-Jackson relationship, but there would have been severe problems. We collapsed the scatter by about a factor of 2 with a new distance indicator. A factor of 2 has a profound effect on one's interpretation of these diagrams. So it was a good thing.
Let me ask you this question. You said that in your proposal the goal of measuring deviations from the Hubble flow was just a footnote.
Two lines. When the work got started going and you started getting into the work, did you expect to find significant deviations from the Hubble flow, or were you sort of surprised in that result?
We were completely surprised. We were shocked, elated, and didn't believe it. One day we believed it, the next day we didn't believe it. In the meantime, [Mark] Aaronson and his group had published their 1982 analysis of the Virgo infall into the local super cluster. So we were very interested in trying to find out whether we could confirm their measurement of Virgo infall.
Was that infall relative to the microwave background radiation?
Well, looking back on their first paper, they really didn't emphasize the fact that the local group was moving at 600 kilometers per second with respect to the cosmic microwave background.
So, they were just measuring relative velocities?
They were just measuring relative velocities. Now any thinking person — they probably realized this themselves — would have realized that their whole system was translating rather rapidly with respect to the cosmic microwave background. But I must say that I wasn't really a thinking person. In fact, it's amazing that as I look back on it, I was wading through all the icky parts of data reduction, just keeping my eyes focused on day-to-day activities, not thinking about larger implications. I knew those two facts. I knew that the local group was moving at 600 and I knew that the Aaronson analysis had shown nothing very interesting in the local group coordinate frame. That meant that everybody [all the galaxies] must have been moving rather rapidly [relative to the mean rest frame of the universe]. But it took me a long time to put those two things together.
Do you think that you didn't see it because you weren't looking for it?
One of my faults as a researcher is that I do not think all the time. I spend large amounts of time just dealing with trivia — going observing, doing all the myriad trivial things that you have to do to produce data. And then, suddenly when I get the data, I think. I have a very productive period of thinking, and then months can go by after that [in which] I hardly think at all. So, I was in my non-thinking mode here. [Faber and Lightman laugh.] I was just wading through this big project to get all this stuff done, and that's all I was doing. Now, there were a couple of things though that began to point in this direction. One of them was a statement by Donald Lynden-Bell, who contributed a great deal to this project. He said: "I think we ought to conduct all of our analysis in the microwave background frame.” Up to this point we had never really said in what frame we were working. He said, "the natural frame is the microwave frame, so why don't we do that.” We were standing on the street corner in Hanover, New Hampshire. And right then I could see: "Hey, that's going to put 600 kilometers a second on all of our velocities… That's interesting." I got very excited and said, "Donald, that's a brilliant idea. Why hasn't anybody done that before? Obviously that's what we need to do." That was point number one. After we adopted that as the fundamental frame, the person who really first saw in our data that there was an area that was doing something totally bizarre was Dave Burstein. Dave was always out in front of the rest of us. He was our "Czar of Data." He kept all of these big data files. He would punch in the numbers and run a quick plot. The rest of us were still cleaning up the loose ends. Meanwhile, Dave was out there as our advanced scout. And at our next meeting, in Pasadena, lots of us were doing deadly, dull, trivial little things and getting programs to work while Dave was making these plots.
Looking at all the data.
"Look at this," he said. "There's a whole region in Centaurus, and it’s moving at 1000 kilometers a second." Another tendency of our group was that nobody wanted to think about things until it was the right time. If I was doing some little operation that was dull and boring but had to be finished, usually I'd say, "Well I'll think about that later. Don't bother me with that now because sometimes it didn't turn out to be right.” [I] knew it might go away in two days, and why should I waste time thinking about it until we were really ready to think about it? But he was right about that one. He kept bugging us about it. As the weeks and the months went by, that region got more and more interesting. It wouldn't go away.
You said you spent some days believing this and some days not believing, and so forth. Did you reach a point, before you published, where all of you believed [in the bulk deviation from Hubble flow]?
Oh yes. Before we published anything, we all believed that the individual peculiar velocities attached to every galaxy were correct, modulo random errors of measurement. We had really convinced ourselves that we didn't have any systematic problems with the data, and this pack of arrows [galaxy velocities] in the Southern hemisphere at high velocity was a real phenomenon. Where we did have a disagreement was over our first paper, on which Alan Dressler was the first author. That was the first analysis, a very simple one of bulk motion, in which we claimed that when we averaged together all the galaxies in a fairly large sphere (6000 kilometers a second), we got a bulk motion of over 600 [kilometers a second]. Now Alan had a great deal of difficulty with that and he thought that the bulk model was not a good description of the data. He pointed to the fact that when you just looked at our plot, what you really saw was a large number of points moving very rapidly on one side and weak, if any, net bulk flow on the other side. Dave Burstein and I were the primary people who sat down with Alan and did some bulk flow motions in different subsets of the data. We showed Alan that within the errors, they were all consistent with the bulk flow, not only in the direction of Centaurus but also in the direction away from it. I think all of us now believe that was a mistake — the publication of that bulk flow paper. Alan was much brighter on that subject than the rest of us. The real significant feature in our data was the high motions in the direction of Hydra-Centaurus, coupled with lesser motions in the vicinity of Virgo. There was only a very ambiguous signature of bulk flow on the opposite side of the Local Group away from Hydra-Centaurus. In fact, I would say at the moment that we still don't know from any of the data sets what really is going on in that direction. The aims of a project that I'm doing with a student here and also with Alan are to look at that region of the sky and try to figure out what is going on.
Let me ask you a little bit about the reception by the community when you began talking about the results. How did the community react to this work? Maybe it's not fair to generalize everybody in the community, but give me a flavor of the range of reactions.
Well, they ranged from extremely uncritical and accepting [to the other extreme]. There was a spate of papers, as you probably know, which took our first bulk flow model at face value — a whole sphere of radius 6000 kilometers a second moving with such and such a flow — and said this can't possibly be consistent with cold dark matter, which was the paradigm cosmological theory at the time and still is. Then, there were people like Jim Gunn, who looked at our data and said "I really see these arrows on this side [Hydra-Centaurus], but over here I really don't see very much." That was echoing what Alan had already said. Then, [there were] people like Nick Kaiser, who I think played a very positive role in pointing out the fact that there's a certain window function for any analysis like this. There's lots of statistical weight near the Local Group; it falls off far away because the errors get larger. We knew this on one side of our brain. We knew, for example, that when we did solutions within a certain radius we got a certain bulk flow motion; when we added galaxies beyond that, we got hardly any difference. So we knew that the galaxies at the edge of our volume weren't having much statistical weight on the results. But somehow we weren't very good at telling anybody else in the community what the implications of this were. We just wrote our paper citing 6000 kilometers a second. It didn't really occur to us that we had to be cautionary. If people were going to make a comparison with theory, they had to take this into account. We didn't emphasize it in our conclusion. And then there was yet another group of people, mostly Mark Aaronson and his group, who said that we weren't showing anything new with the Aaronson data. Because that was one of the things we were also doing at that time that was good. We didn't just analyze our own data, thanks mainly to Dave, who is voracious with data. We really made an attempt to put lots of different data sets together. We began to analyze the Aaronson spirals that way. And we discovered this very predictable thing, namely, that the Aaronson spirals were moving rather rapidly, and Mark thinks he of course already knew that. Well, he probably did. But this was an example of a paper which, for the benefit of the reader, hadn't really pointed out a very salient and important factor. Then, in addition, Aaronson and company had just published their second paper on clusters, in which they showed that eleven distant clusters weren't moving. Those clusters were nominally inside our volume. So we had to face whether or not we thought the clusters were moving. I think early on we thought their clusters were really moving, too, along with our sample, but because of an unfavorable distribution on the sky, that motion really couldn't be seen that clearly. I think now we're much less certain because our new Great Attractor model says the velocities fall off in the direction of their clusters. So it's possible that the Aaronson clusters really are stationary and that Aaronson et.al. were right about them.
Let me ask you a broader question. Has this work raised any concerns for you about the validity of some of the assumptions of the big bang or the interpretation of the cosmic microwave radiation? This is one possibility that you mentioned in one of your papers.
How do you feel about that now?
Well, I think we mentioned that purely for the sake of intellectual honesty. I personally don't think that's the right approach. I tend to think the cosmic microwave background really is the right rest frame [of the universe].
So then do your results have implications for the degree of homogeneity in the universe? If there were huge bulk flows, wouldn't that call into question the assumptions of large-scale homogeneity?
Not necessarily. The challenge is to find a theory in which one can simultaneously match the scale of velocity flows we observe here with the other major constraint, [which is] what goes on in the cosmic microwave background fluctuations. I think that we don't know yet whether or not that's possible. I tend to think that it is possible. For example, there is a student at Santa Cruz, John Holtzman, who has calculated motions and fluctuations for a hybrid theory, which combines cold dark matter with a fairly light neutrino. That looks like a theory which on paper can match any constraint. So I would make two statements at the moment. First of all, even if you take our results here at face value and interpret them naively, it's not at all clear that there aren't theories that are consistent with both streaming motions and any other observational constraints like the microwave background. Secondly, I think that what we are learning is that comparing these streaming motions with theory is a very subtle business. We haven't seen the end of how best to do that yet.
So your faith in the big bang model has not been shaken?
It's not shaken. No, I don't really think it's going to be that bad. It's going to be an interesting constraint; but I don't think it's going to be earth shattering.
Let me ask you, do you remember when you first heard about the horizon problem?
No. It would probably have been a few years after graduate school.
When you did hear about it, do you remember whether you considered it a serious problem?
Yes, I did. I believed that it was telling us about a very fundamental problem in our way of looking at the initial conditions, and our whole concept of the initial conditions as being classical in expanding universe models.
Did you think that the resolution of the horizon problem might have something to do with the initial conditions?
Did your view of the horizon problem change any after the inflationary universe model?
Well, I heard about inflation and its ability to solve the horizon problem almost in the same breath. So I would say yes; instantly. When I first heard about inflation, one of its great virtues was the fact that it may be able to solve [the horizon problem].
You mentioned earlier that you like the inflationary universe. I can't remember exactly how you put it. Why do you think the inflationary universe model has caught on so widely?
Because it solves these three or four classic problems of conventional cosmology. The horizon problem is one, the flatness problem is another. Those really speak to me. One conventionally mentions the monopole problem, but, not being a particle physicist, I have a dimmer appreciation of how serious a problem that is. But they do say it's quite serous.
Do you think that the inflationary universe model has a good chance of being correct, or do you think it's too speculative?
I don't know any more what people mean when they say "the inflationary model.” I'm kind of thinking that inflation is a process — which is likely to be important, but exactly at what phase in the evolution of the universe, I can't say. And I don't have an opinion on whether there are multiple inflationary periods. I think inflation is a general concept that we need to use.
How do you reconcile the amount of matter needed for inflation — omega equal to one — with the observed value of omega? Does it bother you that those two values are in conflict?
It bothers me intensely. The only way out of it, I think, is that there are baryons in the voids.
You make it [the dark matter] with baryons?
No, I mean that there is both dark matter and baryons in the voids, in the normal proportion of about ten to one. I like to think that the universe was well mixed in its early stages. We know that only a few percent of the baryons have shown up in galaxies. My guess is that the other baryons — perhaps 5 times as many baryons — are out there not in galaxies, probably in voids.
I'm trying to find out exactly how you reconcile these things, or confront this. Are you sufficiently impressed by the inflationary universe model that you are willing to take on faith that there are these baryons or missing mass in the void, even though [this matter] has not been currently observed?
Well, I never believe things perhaps as strongly as other people believe them. So I would just like to say parenthetically that if you made me bet and put money on something today, that's where I'd put it. But my world would not fall apart [if it turned out otherwise].
You wouldn't put a lot of money on it.
Yes, right. [Faber and Lightman laugh.] But I was one of those who voted for omega equals one at the dark matter conference. Were you there?
I wasn't at the conference.
That was quite amusing. Scott Tremaine took a poll in the audience, and he asked people to vote in three categories. This was at Princeton.
Yes. That was about two or three years ago?
Yes. He asked the audience to vote [on the value of omega] — those between 0 and .9999; .9999 to 1.0001; and over that. Almost nobody voted for omega much larger than one. I think the rest of the audience split about 50/50.
Yes. It was interesting.
Do you remember when you first heard about the flatness problem? You said a few years after graduate school for the horizon problem.
Yes, about the same time.
Did you also consider [the flatness problem] to be a serious problem?
Yes. I did. In fact, I mentioned that at a talk that I gave at Swarthmore, a Sigma Xi lecture. But I didn't consider myself a cosmologist at that time. I considered myself to be someone who was interested in the structure of the galaxies. And I knew that they were related. But I didn't have much familiarity or any real contact with the bigger questions of cosmology.
Did you have any opinions about how the flatness problem might be resolved?
Yes, I've actually had an opinion all along on this. And I still hold this opinion.
I'm glad I asked.
Again, this is something I wish I knew more about. But at some dim level, I perceive a relationship between doing cosmology and trying to answer the questions of cosmology, and Gödel’s theorem. It just seems to me that trying to explain the universe viewed as a closed system, like a closed mathematical system, is incomplete. I don't think it will ever be possible on logical grounds to explain everything about the universe in a completely self-consistent way. So, to arrive at the properties of the present universe, I think it will always require dragging in some kind of information outside this closed system. Now that extra information could take many forms. I think of myself now in terms of the ancient Greeks. What they saw as their universe was the sun, and the planets, and the earth. And they might have said, "Well, why is the earth the way it is, or why is the sun the way it is?" We now know that the right answer to that question is to say that the question doesn't make any sense. The earth is the way it is because we happen to be on it. This is the old anthropic way of answering cosmological questions. I really think that we will probably find that our universe is the way it is to some extent just because we're in it. Implicit in that is the notion that there are many different kinds of universes. One of the questions I am interested in, although I have no idea of how to address it, is: What is the total possible range of parameters you could think of for something that you might call a universe? What can be different? Probably things like omega can be different. Can the mass of a proton be different? Can particles be different? Can space and time be different? I'm interested in those questions, but I frankly have no idea of how to address them.
Of course, that way of thinking about the flatness problem [requires] a meta-framework, sort of implicit in quantum gravity.
Do you remember when you first heard about the results of de Lapparent, Geller, and Huchra on the large scale structures, or the "results of Haynes and Giovanelli? Did those results surprise you?
No, they didn't surprise me. They didn't surprise me because I had gradually begun to become more familiar with our local territory, just from looking at the spiral samples. And gradually it has been dawning on me over the years that the galaxy distribution is very, very inhomogeneous. When I looked at their picture I could see voids that were comparable to the voids that I already knew existed around here. Now the other interesting facet of their picture, of course, is the thin walls. I am worried about that because they're making their map in velocity space, and we were simultaneously showing that regions can have big peculiar velocities.
So you don't really know whether the velocities correspond to distance.
I don't really know. On the other hand, you could imagine velocity fields that are such large scale that they merely distort the picture but don't fuzz it out. I don't think that their results are necessarily inconsistent with what we're finding, but the interpretation of the location and sizes of their structures might need some refinement. At the same time I think most people believe that their first slice was more striking, as far as the sharpness of its edges, than any subsequent slices that have been made. So we probably don't have a fair sample yet of how sharp these boundaries are.
Yes. I would assume from your earlier comment — that the large scale streaming motions are not really inconsistent with the big bang picture — that you would say the same about the Geller-de Lapparant-Huchra work that those in homogeneities do not necessarily conflict with the homogeneity in the microwave background.
I believe that to be the case. But I think in the case of their work, it has not yet been demonstrated. I know that in the case of our work there are calculations that show that is true.
That you could have a model that will satisfy both constraints.
Besides these examples, have any other developments in cosmology in the last 10 years or so really changed your thinking about cosmology in major ways?
What are the examples that you've given?
Just the things that I've brought up — the inflationary universe, model, and the large-scale structure, and your own work.
Actually, I think you haven't mentioned the one that's been most crucial to me. That's this whole picture of the growth of gravitational instabilities. That's implicit in what you've said, but if I could just expand on that briefly. I used to read papers on galaxy formation in the 1960s and 1970s. None of them made any sense to me whatsoever, because the authors — maybe they had it in their [own] minds — didn't succeed in putting in the paper a picture that I could make in my mind of what was happening. I think very geometrically, and I can't evaluate a model unless I can generate a mental movie of what's happening and what's moving where. None of those early pictures of galaxy formation models were able to make me generate my movie. And this has all changed for me now with [James] Peebles' work, which however is not very visual. But people have built on his concepts. The [Simon] White and [Martin] Rees paper was a profound influence on me — galaxy formation due to the collapse from dark matter haloes; then the N-body simulations, which actually showed us how things were going. That I think has been very influential for me in developing intuition [about these things.]
Would you say that has really changed your thinking?
I can't say that it has changed my thinking because my thinking back then didn't exist. It's given me the crucial conceptual peg to hang my hat on — the whole scenario of fluctuations coming out of the big bang, being modified according to some transfer function, and then after recombination growing into the structures we see today. I don't bet much money on too many things, but I'll bet you that basic picture is correct.
Would you say that that's one of the outstanding problems or areas of cosmology to be filled in?
I would say the concept is a tool now to be used — the way you might use the concept of nuclear burning as a tool now to understand the sequence and evolution of stars. We're in kind of the same situation.
Let me come back to the issue of gender in science. Would you say that your experience in science has been any different because you're a woman?
Yes. I would say it's probably been somewhat different. On the most mundane level, I've had less time to do science than most of my male colleagues. That was really a factor when my children were young. I came to Santa Cruz in 1972. I really didn't publish a paper aside from my thesis work until Bob Jackson and I published the work on velocities. That's a gamble; three years or so. People wouldn't survive very well with that kind of hiatus now. I felt acutely aware of it at the time. I felt very badly I wasn't getting any work done. It was hard. I had my first child nine months after I came here. So that's been a definite handicap. I think, though, I've had a lot of opportunities. I've been appointed to lots of committees, national committees, and other things because somebody wanted representation from women. Committees have played a very big role in my education. You get to meet people that way, and you review all kinds of scientific issues. I think astronomy should ideally be one of the most synthetic subjects. It's hard to do effective astronomy if you're too specialized. Ideally, astronomers should have a very global view. You don't know if the solar neutrino problem is going to have something to do with the big bang. Who would have guessed fifteen years ago that that might be a very intimate connection? So you like to stay as broad and as flexible as possible. Moving on the national scene and serving on committees gives you a way of keeping in touch. I think therefore I've had an advantage.
Do you want to say anything about how the experiences of men and women in science, in general, might differ? Stepping back from your own situation, do you have any opinions about that?
Yes. I think women are under an additional handicap in that they are typically younger than their husbands. Their husbands therefore are always a little bit more advanced and ready to make career decisions when the women have not quite reached the optimum point. So in the normal course of things, I think women, always being a couple of years behind, are more forced to make their career decisions conform to what's important for their husbands. Also, because the husbands are always a bit in advance, their earning power is greater. So even if a couple wanted to be totally fair to both parties, they'd still look at the fact that, say, the woman is still in school and the husband is now going to earn $25,000 a year for his first job. That earning differential is a very big factor. So, it's really a difficult question. One would like to be able to play the game of accepting offers, moving around, moving up. That's great if you can drag somebody else along with you, but for women it often doesn't work.
Let me end with a couple of philosophical questions. If you could have designed the universe any way that you wanted to, how would you have done it?
Gosh, has anybody given an interesting answer to that?
Oh, I've had a lot of interesting answers.
I don't know, I have to beg the question because, again, I echo the theme that troublesome answers to questions usually mean the question is flawed in some way. I'm sorry, but this strikes me as a flawed question.
Well, you don't have to answer if you don't want to.
The reason [the question is flawed], I think, is that it is trying to put human values on a question which is intrinsically totally unrelated to human values. And that's why I'm having difficulty with it.
Okay. That leads naturally to the last question that I have. There's a place in Steven Weinberg's book The First Three Minutes where he says that the more the universe is comprehensible, the more it also seems pointless. Have you ever thought, yourself, whether the universe has a point or not?
Yes, I've thought a great deal about it. And my answer here relates to what we were saying before about why is the universe the way it is. Let's use the same analogy. Why is the earth the way it is? To me, the questions are very, very similar. I don't believe the earth was created for people. It was a planet created by natural processes and as part of the further continuation of those natural processes, life and intelligent life appeared. In exactly the same way, I think the universe was created out of some natural process, and our appearance in it was a totally natural result of physical laws in our particular portion of it — or what we call our universe. Implicit in the question, I think, is that there's some motive power that has a purpose beyond human existence. I don't believe in that. So, I guess ultimately I agree with Weinberg that it's completely pointless from a human perspective.
That's a good place to end the interview.
 J. Jeans, Astronomy and Cosmogony (Cambridge, 1928); The Universe Around Us (New York: McMillan, 1929); The Stars and Their Courses (Cambridge, 1954)
 F. Hoyle, Frontiers in Astronomy (Heinemann: London, 1955)
 S.M. Faber and R.E. Jackson, "Velocity Dispersions and Mass-to-Light Ratios for Elliptical Galaxies, " The Astrophysical Journal, vol. 204, pg. 668 (1976)
 S.M. Faber, “Variations in Spectral-Energy Distributions and Absorption-Line Strengths Among Elliptical Galaxies,” The Astrophysical Journal, vol. 179, pg. 731 (1973); “Ten-Color Intermediate-Band Photometry of Stars,” Astronomy and Astrophysics Supplement, vol. 10, pg. 201 (1973)
 R. Minowski, in Problems of Extragalactic Research (IAU Symposium No. 15), ed. G. McVittie (New York: Macmillan, 1962)
 D.C. Morton and T.X. Thuan, The Astrophysical Journals, vol. 180, pg. 705 (1973)
 A series of papers have been published on this project, beginning with A. Dressler, D. Lynden-Bell, S. Faber, D. Burstein, R. Davies, G. Wegner, and R. Terlevich, “Spectroscopy and Photometry of Elliptical Galaxies. I. New Distance Estimator,” The Astrophysical Journal, vol. 313, pg. 42 (1987)
 See Ref. 3
 W.L.W. Sargent, P.L. Schechter, A. Boksenberg, and K. Shortridge, “Velocity Dispersions for 13 Galaxies,” The Astrophysical Journal, vol. 212 pg. 326 (1977)
 M. Aaronson, J. Huchra, J. Mould, P.L. Schechter, and R.B. Tully, The Astrophysical Journal, vol. 258, pg. 64 (1982)
 A. Dressler, S.M. Faber, D. Burstein, R.L. Davies, D. Lynden-Bell, R.J. Terlevich, and G. Wegner, “Spectroscopy and Photometry of Elliptical Galaxies: A Large-Scale Streaming Motion in the Local Universe,” The Astrophysical Journal Letters, vol. 313, L37 (1987)
 M. Aaronson, G.J. Bothum, H. Mould, J. Huchra, R.A. Schommer, and M.R. Cornell, The Astrophysical Journal, vol. 302, pg. 536 (1986)
 See Ref. 11
 A. Guth, “Inflationary Universe: A possible solution to the horizon and flatness problems,” Physical Review D, vol. 23, pg. 347 (1981)
 V. deLapparent, M.J. Geller, and J.P. Huchra, “A Slice of the Universe,” Astrophysical Journal Letters, vol. 302, pg. L1 (1986)
 H.P. Haynes and R. Giovanelli, “A 21 Centimeter Survey of the Perseus-Pisces Supercluster, I. The Declination Zone +27.5 to 33.5 degrees,” Astrophysical Journal, vol. 90, og. 2445 (1985)
 S.D.M. White and M.J. Rees, Monthly Notices of the Royal Astronomical Society, vol. 183, pg. 341 (1978)
 S. Weinberg, The First Three Minutes (Basic Books: New York, 197), pg. 154