Oral History Transcript — Dr. Harvey Tananbaum
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.
See the catalog record for this interview and search for other interviews in our collection
Interview with Dr. Harvey Tananbaum
Harvey Tananbaum; June 24, 2002
ABSTRACT: A biographical interview; Tananbaum was director of the Smithsonian Astrophysical Observatory's Chandra X-Ray Observatory at the time of interview. Discusses his childhood and education including time at Yale and MIT; initial forays into X-ray astronomy; anecdotes about Riccardo Giacconi and launch of Uhuru Satellite in 1970; discovery of first black hole in Cygnus S-1 and confirmation of binary accretion model as source for x-rays. Transition of AS&E staff to Center for Astrophysics; building of High Energy Astrophysics division at CFA; Tananbaum's research interests; CHAMP project with Chandra. Discussion of building Chandra as well as comparison of Chandra project with Hubble telescope. The politics and finding for Chandra; its operation. HEAO-B (Einstein) mission covered including its precursors, detector technologies, and science contributions. Tananbaum's time on various NASA committees and the creation of a long-term strategy for space astronomy mentioning Chandra, Hubble, SIRTF, and NGST. Discussion of important topics in astronomy research including the merging of physics and astronomy in some areas. Committee contributions including decadal survey work. Thoughts on religion, family, values and personal meaning. Most significant changes in his career including personal rewards.
Session I | Session II
McCray:Not to sound like a therapist, but letís talk about your childhood. I know you were born in Buffalo, New York. Why donít we go from there? Tell me about your parents.
Tananbaum:Right, my childhood and my parents. Itís probably unremarkable in the sense that my dad was a salesman and my mom raised the kids. About the time I went off to college, she started working as a bookkeeper and office manager, and juggled working with raising my younger sister and brother. Iím the oldest of three children. I expect we were a classical middle class family in terms of social and economic; values were certainly put on family, education. We were Jewish and had a fairly strong tie to the Jewish community.
Tananbaum:No. Conservative. I think I was always a good student and I certainly was encouraged to value education and learning by my family, by my teachers. I went through the public schools in the City of Buffalo: P.S. 66 and then Bennett High School. Of course, education was probably a lot different. I donít think the teachers had unions yet. The fact of the matter is that they werenít particularly well-paid then just as they probably arenít particularly well-paid nowadays. Perhaps there was an even stronger commitment, dedication, recognition of the importance of teaching. It was a profession as opposed to perhaps a job. So I had the benefit of some superb teachers, and Iím sure some that werenít exceptional. But we were encouraged to think, and read ahead, and discuss, and debate. If you had questions that the teacher couldnít answer, there were cases where the teacher would go home—they didnít have the íNet—and I suspect the teacher went to the library or her reference text. In those days, we had both men and women teachers. I got just an excellent, excellent general education. Iím a stronger believer in public education. I believe itís very important that we have a mix for both social reasons and for economic opportunities, and hopefully to generate a somewhat literate mentality in the populace, our kids. I have two grown children and both children went through public schools in Framingham, Massachusetts where weíve lived since the late 1960s. They both went to Yale, so they both got a good education.
McCray:Your father, what kind of salesman was he?
Tananbaum:He actually sold what we would call home equipment which means beddings, linens, towels, pots and pans, clothes. He sold door to door in the city of Buffalo, working for a fairly substantial company, I guess you would call it. It wasnít his own business, but he would sell items mostly on credit, and he would collect a little bit of money each week, and then try to sell you more as your bill came down. This was before the large department stores and the easier access to credit.
McCray:Were your parents from Buffalo?
Tananbaum:No. My dad was born in New York City. My dad just died about two weeks ago. And he was born in the city of New York. He dropped out of high school, I think, his junior year. My mom was born and raised in the city of Buffalo and when she finished high school, went off to New York to work for a year or two and met my dad. They got married.
McCray:Did your parents move back then to Buffalo?
Tananbaum:Yes. I was born in Buffalo and six weeks after I was born, we moved into a two-family house with my aunt and uncle and their children. Lived in that house until I was about sixteen, seventeen years old. Then I went off to college and Iíve only been back to Buffalo really for summers and for family events, to visit family. Went to Yale also as an undergraduate and then MIT for graduate school; so Iíve been in the Boston area since 1964. I enjoyed it, liked it, decided to stay.
McCray:What was it like being a member of a Conservative Jewish family in Buffalo?
Tananbaum:It was fine. Buffalo had a pretty good size Jewish population. There were, I donít know, ten Temples and Synagogues from the Reformed, Conservative, and Orthodox combined. The high school I went to was a large city high school; there were six hundred students in my graduating class. We used to joke that about one-third of students were Jewish; about one-third were minorities; and about one-third were the rest which means probably Christian, white students. The point being thatís of some relevance, it certainly was a reasonably integrated school, although the Jewish and minority fractions were probably, substantially less than 1/3. There was a pretty decent sized Jewish population concentrated maybe more so in that high school than some of the others in the city. You certainly didnít feel as somewhat isolated being Jewish and I think that other minorities didnít feel isolated either there. With a school that size and with the makeup of the school, it was a pretty well integrated high school.
McCray:One of the questions Iíll probably come back to later on is your current religious beliefs and the interface between religion and science. But is there anything, thinking back to your childhood, that stands out between those two areas?
Tananbaum:No. I think that it is fun and we certainly should spend a few minutes at some point over the time we talk this week figuring where many of us think science stops and religion, or philosophy, or something else must start. But as far as I can recall, I wasnít strongly influenced, say, by some aspect of a religious education to then want to pursue science. Itís possible as a result of science, Iím somewhat less religious in terms of separating the aspects of what the religion teaches versus what we can or canít do as scientific researchers. Itís a complicated interface.
McCray:Weíll come back to that then. Tell me about your interest in science as a teenager. Some astronomers have small telescopesÖ
Tananbaum:Yes. I thought a little bit and tried to connect the dots to ascertain how I became a scientist, but Iím not sure the dots actually connect in a sensible kind of way. I was always very good at math. My earliest recollections: math was very useful for figuring out batting averages, for example.
McCray:Were you a baseball fan?
Tananbaum:I was and I am. Although I think the Red Sox might finally do me in this year. [Note: Sox collapsed per usual, but did not ďdo me inĒ.] But in any case, you could figure out the percentages, and the standings, and the baseball batting averages. I would have fun doing calculations in my head, just fooling with numbers. I wasnít any particular prodigy with numbers, but I could do simple calculations; I could do them quickly and accurately. I always excelled at math. Through high school and even to a great extent through my undergraduate years, I really didnít know what field I might want to pursue from a professional point of view.
McCray:But you knew something in math and science?
Sciences probably more so than not. In high school, I got uniformly outstanding grades, but the math and the science werenít any better than the English, the history, the Latin. Latin was fun because there were rules and the concept of that language did resonate with people who like order and the mathematical background or interest in math certainly went hand in hand. I probably got the most encouragement; I had a wonderful math teacher. Her name was Theresa Podmele. She had a Ph.D. in math and she was teaching at the secondary school level and was called Miss Podmele, notwithstanding her Ph.D. She was my home room teacher and then I had her for math in tenth, eleventh, and twelfth grades. We were grouped, I guess what you call, homogeneous groupings so that the best math students were in a math class together.
You wouldnít necessarily be with exactly the same students in English or in history, but many of the same students were in some of the same classes. And she would push the class as a whole and then she would push some of us as individuals. So an example would be you might be assigned a homework problem and in addition to doing all of your homework, your job the next day would be to present the solution to a problem at the board for the whole class. I donít think there was any doubt that I routinely got the toughest problems she could find. She would have us tutoring some of the students who were having trouble and I think maybe so we wouldnít get a little too big for our britches in the sense that it wasnít easy for everybody to do the math. This is back in, well I graduated high school in 1960, so this was back in the late 1950s. We did do a year of calculus and advanced placement our last year of high school. So we got a very, very strong foundation. I knew that the math was a certain route where I could, coming into college, have some credits already. I had advanced standing in math, history, chemistry, and English, I guess.
McCray:When you were a teenager, what did your parents envision you growing up to be?
Tananbaum:Mom, of course, would say you ought to be a doctor. I think thatís given as a truism about Jewish mothers, but itís really not just true of Jewish mothers.
Tananbaum:And a doctor was certainly thought of as, I think, the highest of the professional possibilities back in those days. I think with the insurance industry and some of the ways in which medicine is dealt with these days, Iím not quite sure that it stands much above many other opportunities people have now. But in any case, I started out pre-med at Yale in a sense, although you didnít declare such a thing as pre-med.
McCray:Why did you pick Yale?
Because I didnít know any different. The story is actually— I donít know how much of this information is relevant other than the haphazard way in which choices get made even by supposedly intelligent individuals. We took an exam for National Merit Scholarships probably in our junior year of high school. Nowadays, at least for a while anyway, they use whatís called the pre-SATs, PSATs, as a qualifier for National Merit. But back then the National Merit had a separate exam and we either took it at the end of our junior year or the beginning of our senior year. I canít really recall which. We took this exam. We had to list our top three choices as universities. I was most likely going to graduate at the top of my class. I had already scored well in the 700s in my junior year taking SAT exams. I knew I was going to have an opportunity to go to a first-rate school; at least I thought I was. So as the three choices on that exam I wrote down in the following order: Princeton, Yale, and Harvard. And knowing very little about any of the schools, never having visited them, not having really researched them; just knowing that those were probably the three top Ivy League schools and perceived as such.
The guidance counselor whose name was Mr. Hill, walked over in the middle of the exam, looked over my shoulder, hmm-ed a few times and said, ďWell, if youíre not going to put Harvard first on your list, you probably ought to have Yale ahead of Princeton.Ē And since the test was done in pencil and that pencil had an eraser, I erased Princeton and put Yale first and Princeton second. As an interesting sidelight, the guidance counselor had also been one of my motherís teachers when she was a high school student. And knew who my mother was and even though it was a large school, there was a certain connection. I scored very well on the qualifying exam. I was named a semi-finalist, a finalist and awarded a Merit. That happened around November of our senior year and Yale sent me a letter. I think even back then normally, there was no such thing back then as early acceptances, and the normal time frame was probably April for acceptance letters coming out. Yale sent me a letter saying since I had been awarded a National Merit, it was their intention to accept me. It seemed fine. I had a few friends who were interested in going to Harvard and Princeton. One of the things that was often the case at a large city school, maybe only one person from your high school class was going to be accepted or at most two by the top Ivy schools. No one knew just what the criteria were, only that from year to year there would only be one or two. We can all figure that the factors were usually called geographic, but they werenít always just geographic. So I withdrew my applications from Princeton and Harvard at that point. I was perfectly happy to have the decision made that I would be going to Yale. It was a fine university. I did not know there were no women undergraduate students; that it was all male. I laugh in the sense that, of course, in some ways that affected the rest of my life and also indicates the limited knowledge that I had about the schools.
You might ask, people have, ďWhy wasnít Harvard just first on the list?Ē which is probably the most common experience. And I canít really give any kind of a rational reason other than I had chosen to do something a little bit different. Maybe I would have been accepted at Harvard and maybe I wouldnít have. We donít know. Most likely I would have, I would think. But in any case, the decision to go to Yale was just based on there being an opportunity to get a wonderful general education. The National Merit Scholarship was based on financial need and our income situation was such that I was eligible for the maximum, which was fifteen hundred dollars a year, which essentially paid tuition. Yale offered me a working job starting the first year in the dining hall and subsequently helping in the labs on different projects; the science labs. But basically, the education was paid for. I started out in what I call pre-med in the sense that, as I said, it wasnít a major. My declared intent at first was to major in math because I had more credits and could get my major done the fastest. But my freshman year, I took an advanced math and advanced English, organic chemistry, biology, and German. You had to pass a language exam. I think it involved a speaking requirement. For some reason what I had in Latin wasnít sufficient so I had a foreign language requirement as well. And of course having biology and organic chem, they were both lab courses. It was a huge load plus the advanced levels of the others. So I worked very hard, but organic chemistry was by far the greatest struggle.
I found that I had sort of been placed where I could take an advanced chemistry based on my one year of high school chemistry in a good school, and the SAT chemistry achievement exam. But I probably would have been better served taking first year chemistry in general. Everybody else in organic chemistry was an upper classman, and too many pre-med students too busy competing for getting the best possible grade so they in fact could go to medical school. And lab was challenging. The text, the class, the analytical breakdowns, and how do you synthesize this, and how do you analyze that; the problem solving was a blast. I mean, I was good at it and there was no problem getting an A. The lab was by far the greatest of my college struggles. The first experiment, I think, we had to calibrate a thermometer and then in the next experiment, I broke mine.
McCray:I have a couple of questions about Yale. But before we leave the teenage period, did Sputnik have any particular interest for you?
Tananbaum:Yes, yes. Certainly the beginning of the space age, it increased my awareness and interest in science; more so when we get to the graduate school phase and I had to pick what area of physics I was going to proceed with. The space program definitely influenced me and the opportunity to choose X-ray astronomy as a field. And so it certainly dates from Sputnik. I certainly paid attention. I was a Boy Scout and we did a lot of camping, summer camping and winter camping. I would have to check back. I donít actually believe I ever got the astronomy merit badge. I may have though. It wasnít a real difficult one.
McCray:Did you do the Project Moonwatch at all?
Tananbaum:No. But in any case, Sputnik was in 1957, so I would have been in the beginning of my sophomore year at high school.
McCray:How about the Cold War?
Tananbaum:War had a very significant influence on my life, but not the Cold War per se. My dad was wounded in World War II.
McCray:In the European orÖ?
Yes. He was in the European front and he was wounded in Germany. It was a wound that probably, in later wars, would have been stitched up at the front. Shrapnel exploded and tore his arm open pretty badly. He nearly bled to death and he was at various hospitals. First in Germany near the front, behind the front and then he was moved to England. I can remember in—it had to be sometime in 1945 or 1946 most likely—he was still hospitalized at the V. A. hospital in New York. There was a late night caucus (late night was probably all of nine oíclock at night between my mom and my uncle), and it was decided we would go to New York and visit him. I think I couldnít go into the V. A. hospital; I was probably four-years-old. I remember waving for him from outside his hospital room window. He had had health problems through the remainder of his life from complications of the War. Interestingly, he had a Veteranís pension and after he retired in his early sixties, he spent most of the next twenty years volunteering at the V. A. hospital in Buffalo.
I think when he died, he had over seventeen thousand volunteer hours. So he wasnít bitter over how it had affected his life between the immediate wound and the complications that occurred subsequently. He had, certainly, physical ramifications and he had pain, persistent in various forms. So I certainly probably was affected more by the impact on him and therefore our family from World War II. Not to the extent of people that, of course, lost their fathers, their sons, their brothers and sisters. But I was affected by World War II more than the Cold War. From my perspective, it probably might have been a question of whether I was going to work in a science like astronomy, the field of X-ray astronomy in particular where things were pretty much open or in some classified area. There had been a few times in the past that some of the instrumentation that we were interested in using was classified or you needed particular access to the detectors and such. But one job offer I had upon getting my Ph.D. would have been to work on some classified work and I didnít rule it in or out. Ultimately, the opportunity to continue working with X-ray astronomy which had already been my thesis, was a more attractive option to me.
McCray:Did you read science fiction growing up?
Tananbaum:Yes. I read a lot of science fiction.
Tananbaum:Heinlein is the one I remember. I read a lot, of course, in the 1940s when I guess I learned to read, and in the 1950s. Television just was becoming in the middle or later part of the 1950s, when we all began to have more use of the TV. So reading was very important to us as children. It was a common thing to go to the library on a Friday afternoon and get four books for the week that was to come. Some were science fiction, some were sports, some were biographies.
McCray:Were there any particular popular science books that you remember?
Tananbaum:No, not really. Just that I liked to read. Certainly, the idea that there would be these other planets and different life. So it was very interesting.
McCray:It sounds like you started out with the standard pre-med and a lot of chemistry while at Yale. At some point, Iím assuming, you shifted to physics.
Yes. I think what happened was, I really wasnít very serious about going on to medical school. It was a possibility so I took the biology and chemistry the first year; I took physics the second year. I had all the science courses I might need for med school. I really enjoyed the English course I took the first year; it was English poetry from Chaucer through to T. S. Eliot. It was interesting reading it, and having to write essays, and interpreting what you had read. Itís not quite the same as solving a math problem or even an analyzing some organic compound. I took psychology; I took economics; I took symbolic logic as a philosophy course, but really it was more of a math course and it was an easy math course. I took comparative Western religion; I took twentieth century American history from John Blum; it was a wonderful course. A lot of the courses I took didnít have pre-requisites. Twentieth century American history wasnít a course that incoming students took; I probably took it my junior year.
But I had a great American history for a year-and-a-half in high school and it was easy at Yale. Miss Beach was my high school teacher for history. Iím not saying that the course was easy at Yale, but it was just easy for me to read, and do the papers, and the exams, and analyzing, and interpreting, and answering the questions. It was an enjoyable break from the physical sciences. It was a nice counterbalancing. I wouldnít trade the social science and humanities courses, the liberal arts education, which I obtained at Yale. I did pay a personal price when I decided to switch to physics in my junior year. I think the math courses that I was taking, I remember the one in differential equations in particular, and there is a wonderful construct called the Wronkian. All we did was prove that solutions to equations existed; we never solved the equations. I thought, ďGee, I donít care if a solution exists. The answer is what you want.Ē Not proving that you could solve it, but solving it. So I became quite, not so much disenchanted, but aware that probably pure mathematics wasnít a direction that I really wanted to go in. And so I became more and more interested in physics which was interesting because, of course, some of the equations that you couldnít solve analytically or prove a solution exists for were the things that then show up on the physics side. Itís actually the real world equations that are sometimes nasty.
So I ended up taking another physics course. In my senior year I took six semesters of physics so that I ended up with a double degree in math and in physics. I started to apply for graduate schools in physics. I had decided that physics was the direction I was going to go. Now actually, I think had I known I wanted to be a physicist and only that, maybe I wouldnít have gone to Yale. Yale didnít have a particularly strong physics department in those days. I had to take the GREs, probably at the end of the first semester of my senior year, and I didnít do particularly well which wasnít a shock in the sense that I didnít have a lot of basic physics yet. It certainly was a blow to the ego because I was actually in the top one percent; the top ten students in my class at Yale. I had outstanding grades and I had succeeded academically at everything Iíd ever done and I think my GRE in physics was in the mid-500s. So above average, but not great. I was also involved in a relationship with a woman who now has been my wife for thirty-eight years now.
McCray:So you met your wife while you were at Yale?
Tananbaum:Well, actually we had gone to high school together and she had gone off to college in one direction and I had gone in another. We started seeing each other in the summers and then in our junior year, we decided that we were pretty serious and she transferred to Boston so we would be closer.
McCray:What is her name?
Tananbaum:Rona. We pretty much had zeroed in on liking Boston and graduate school in the Boston area. It is interesting, again, the basis on which decisions are made; not necessarily finding the best possible graduate school or schools or looking broadly. So I applied to MIT and Harvard for graduate school in physics and Brandeis as sort of a safety school. I wasnít accepted at Harvard which turns out not to have been a problem because I was accepted at MIT and it was my first choice. I donít know if that is still connected to the Harvard-Yale rivalry or just that I thought better of the opportunities to do graduate work at MIT, I think on average, people got their degree in four to five years at MIT and Harvard might have been six to seven years. I thought from a physics point of view that they certainly were competitive with one another. I think MIT must have had to look a bit beyond my GRE score. The first few weeks we had to take a qualifying exam. I proceeded to do not particularly well on that either. Again, I think anybody that had finished up as an undergraduate at MIT could zip through that qualifying exam. Some of the questions were things that were taught in courses their juniors and seniors routinely took. I probably had some of the basic physics that would have allowed me to answer some of the questions, but not quite couched such that I actually could do it.
McCray:How was physics taught at Yale? What types of coursesÖ?
Tananbaum:I think the courses were quite similar. One of the things was because I had to take them out of sequence and mostly at the same time, I didnít get all the honors courses in physics. There is an honors track and a non-honors track and I took, at least in the last year, some of the stuff wasnít in the honors track. My background in electricity and magnetism was pretty good; the mechanics was pretty weak. I took a course in electronics and lot of the lab work still was using vacuum tubes as opposed to transistors. So I think there was a case where—the teacher was a very nice individual—the course should have been updated.
McCray:Was there a lot of emphasis on quantum mechanics, relativity?
Tananbaum:Yes. I took a course in quantum and one in modern physics. I actually had a course that Alan Bromley taught. He was building a linear accelerator there. The course was some form of nuclear physics or modern physics; I donít remember the title. I did quite well. Probably the areas that I struggled with the qualifying exam were mechanics and probably even a little bit of the E & M. In any case, MIT sent me a letter saying that, ďWe think youíre going to have great difficulties succeeding here.Ē It was worded a little bit differently. ďYou should probably take your courses this year and graduate with a Masterís degree and think about doing something else professionally.Ē Iím not quite sure it was that negatively worded.
McCray:What was your reaction to this?
Tananbaum:Well again, having succeeded at just about anything I had tried, I didnít accept the conclusion. I mean, the premise that I hadnít done well in the exam and the GRE scores werenít particularly good was certainly valid. And I simply went in and said, ďWell, donít you think you ought to see how I do in the courses that Iím taking?Ē And they said, ďOh, yes. If you do well in the courses thereíll be no problems.Ē I got a B+ in quantum and I got Aís in everything else. So they sent me a letter, either at the end of one semester or the end of the year saying that I am reinstated in the Ph.D. program. And truthfully even at the end of the four years of the Ph.D. program, I didnít smash the generals. I probably passed with just a little bit to spare. And so I think you can draw whatever conclusions you choose. Some people out there, some of them might say, ďHeís still a lousy physicist, was and still is.Ē We donít really know, right? But in any case, I think the thread through some of that is that having jumped into physics rather late and having done it at Yale and then the non-honors kinds of classes, I was always a half-step behind. And some of that, of course, started from where I started late; and some of it then came by choosing to go to and thankfully being accepted at MIT, a top-rated place. So there was always a little bit of catch up going on. But anything I then took, each and every course, I excelled at. I had a very nice thesis project at MIT. In my first year, I had worked at the cyclotron and I wasnít really all that interested. It was where I was assigned in order to get my fellowship which paid tuition plus a small stipend. I was struggling, working very hard just with my courses. I was newlywed and so I had a family, I had a wife to go home to at night and weekends.
McCray:What was your wife studying?
Tananbaum:My wife majored in history and got a degree in education. She taught middle school. She now has a Ph.D. and she is a clinical psychologist with a substantial private practice. But while I was in graduate school, she made one hundred dollars a week teaching, and I made sixty dollars a week as a grad student. Tuition was paid. We felt we were wealthy. We did need both incomes, of course, to live.
McCray:A couple of questions about that. When you decided not to become a doctor, what was your parentsí reaction?
Tananbaum:My parents were always incredibly supportive. I think they realized, probably fairly early on, that I was an outstanding student, that I was going to do well in whatever I ended up choosing as a field. And I think they really didnít have much input into the choice of the colleges other than, ďYou must know more about it than we do. You should decide because thatís what youíre going to do for the next four years of your life and the rest of your life beyond that.Ē And so, they were very supportive. They didnít have a tremendous amount of information or expertise probably to offer other than to be very supportive.
McCray:Also, Iím curious about the interaction between physicists and astronomers at Yale. I realize you were undergraduate there, so your perception of it may have been different. But did you get any sense as an undergraduate physics student what the physicists thought of the astronomy department?
No, not too much. They were separate departments at Yale. At MIT there isnít a separate astronomy department and so thatís, of course, quite relevant to the rest of my graduate education. But at Yale I had a roommate who took a course in the astronomy department. I think it literally was in celestial mechanics, which of course is one of the traditional ways in which astronomers either say theyíre doing physics or the physicists say theyíre not doing anything thatís particularly interesting or difficult. But I think, and Iím not really completely up to date, I believe that in recent years Yale has made a greater attempt to increase the interaction between physics and astronomy; seeing the advantages from doing that. But I think back then, the fields were quite separate and there are classical statements that physicists think what they do is hard science: hard both in terms of the degree of difficulty and in terms of the science itself. Astronomy, you donít control the experiment generally, you control what you look at and you get to conclusions sometimes by looking either at different wavelengths or at multiple numbers of objects, and you build your picture by looking at different systems at different stages. Every so often something happens.
If you want to study a solar flare, for example, the sun is going to flare and it is going to be more active at certain points, and you can trace the time history, or the spectral evolution, or the spatial properties of a flare. So while you actually donít control the parameters of the experiment, you certainly could trace an event choosing when and how to look, and observationally do something thatís probably not all that different from smashing a bunch of particles together in an accelerator and analyzing the debris. But in any case, I am aware that there have been instances wherein physics people donít think as much of the astronomers. It probably was true that fifty years ago there were a few astronomers who were outstanding scientists and then there were a group of people who charted the path of a comet in the sky or something more like that, but nowadays, there are many more similarities between astronomers and physicists.
McCray:So astrophysics didnít really seem to be a big part of the Yale curriculum at that time?
Tananbaum:No, I donít think it was. I donít think I took any courses that were related to astrophysics during the time I was there.
McCray:So when you got to MIT, how did you make the transition then from physics into astronomy? How did that happen?
Tananbaum:First of all, MIT, of course, was just physics and still is just physics. But astrophysics is a significant area at MIT. The first year, I had this assignment to help at the cyclotron and really, I didnít show a lot of interest. There were safety rules that said nights and weekends there had to be two people in the building if there was a run going on. So they would look for someone to shanghai into coming in to the help the grad students who were thesis level doing their experiments. And I learned not to answer my phone if it rang after four or five oíclock in the afternoon if I had other things I preferred to do. From a research perspective, I didnít put much in and I surely didnít get much back out.
McCray:So particle physics wasnít ever something thatÖ?
That I was all that excited about? I could have, in principle, could have gotten excited about it or been more interested. I canít say why I was or I wasnít. I soon became aware of x-ray astronomy at MIT—because this would have been 1964, the Fall, when I started there—the field of X-ray astronomy was two years old. I was certainly aware that these X-ray sources had been discovered and people didnít quite know what to make of them. They were very mysterious as to how they could generate so much energy in the form of X-rays. George Clark was the senior professor at MIT - I donít think he was forty yet - but he was a full professor and he was doing some balloon experiments. American Science and Engineering, where Giacconi and Gursky were headquartered, was just around the corner. Itís now where the MIT health center is, on Carleton Street. It was right in the middle of the campus. I went and I talked with George later in my first year of graduate school about the possibility of joining his group in my second year and starting to do some work in X-ray astronomy.
George had several graduate students already. Even then, you needed outside funding to support the students who worked on your projects and he was fully committed. He suggested I speak with Jim Overbeck, who was a former student of his, who was either a lecturer or some kind of a junior faculty position. As young as George was, Jim was even younger, just a few years beyond his Ph.D. Jim had decided to build his own gondola, which is a balloon payload. The nice thing in those days was that rockets were where most of the discoveries were still being made in X-ray astronomy. But a graduate student could work on a balloon payload where probably the whole experiment cost one hundred thousand dollars. I may be off a little, but the experiments were such that you could get your hands into the middle of it. You could learn how to, if you were interested, how to weld the frame together or how to design and build the electronics to read out the scintillators. We did buy scintillators and phototubes. We didnít make every single thing ourselves. But there were decisions as to what kinds of detector to use and how youíre going to read it out, the processing and storing of the data, the commands up to the payload during the flight, the pointing system and how to get it to point stably while it was hanging under the balloon. Jim Overbeck was a real gadgeteer. He really was very, very talented with his hands (and mind), both mechanically and electronically. I think he enjoyed building stuff more than even doing the observing. So he took me on. Again, it is interesting in that Iím not particularly skilled in working with my hands. My wife hangs the pictures in our house.
Electronics, I was pretty reasonable at. Circuit design is almost analytic and something you could do a little bit by trial and error. If youíre not getting a square enough pulse, you can change the value of the capacitor or the resistance on the input or the output and watch and see if you can square your pulse up. You put it on an oscilloscope, you can see if youíre timing your signals in sync, the way it needs to be, or do you need to put a time delay, or the like. So it was sort of an interesting thing. I was in Course 8, in Building 6 at MIT. The office had been, Charlie Townesí, where he had done a lot of the early work on masers and lasers, and he had just left to go to Berkeley.
Tananbaum:So Jim inherited that lab in the second floor of the physics building. The first job he gave me was to rewire. There was a big table put in the middle. I was supposed to rewire the lab. I did understand that there was a two-prong and a three-prong outlet, and the third prong was to safely ground things. But whatever I did, which I donít remember of course exactly, if you made the mistake in the next three or four years—maybe even in the next twenty or thirty years—if you simultaneously touched the outer part of the bench strip and the shell of the oscilloscope, you always got a good jolt. I had clearly not wired it correctly. I was over there a few years ago, and I wandered through, and it has been reconverted to office space. Paul Schecter, who is an astronomer, is in that office. I asked him if he ever had any problem with electrical shocks. I guess when they took the lab table and the benches out, they fixed the wiring, if not before. So in any case, it was an interesting mix. Jim was patient with me. I did learn to solder and became adept enough at that. I did work on circuitry. He handled more of the mechanical structures. We built this payload.
McCray:Where were these launched from?
Tananbaum:Palestine, Texas. Alan Womack was also a graduate student who came on a few months after I did. Jim and he were interested in building a germanium detector and working at still higher energies and looking for lines possibly. So Alan and I went out to Texas. It was the Fall of 1966, midway through my graduate career. We flew the payload with Jim. Thereís where I learned to wait for the weather. Launching the balloons is an interesting art all of itself. The balloon is, a little bubble inflated and held by a trailer, and then there is a long layout of the balloon uninflated on the ground. When the winds are calm and there is just enough wind, the bubble sort of rises straight up and when it is over the payload, the payload is released and it is supposed to go up instead of down. Usually it does, but plenty of the balloons leaked in those days, and sometimes the payload had to be cut loose, and bounced off the runway. If the weather forecast was for the least amount of wind above, say, five miles an hour at sunset, which is when the winds would be the calmest, the weather forecast would be enough to scrub any attempt to even launch that day. We were there for three weeks and there are three places to eat. We ordered all four things on the menu at each place. We were ready to leave. We would say to the launch and weather crew, ďWell why donít we set the payload up and maybe the winds will be calmer than theyíre predicting.Ē Because, of course, it was imprecise. No, this was the way they did things. At a certain point, Jim didnít want me to go to the weather briefings because I was getting so impatient that I might antagonize the people who had to give us the go-ahead to launch.
McCray:Who funded this?
Tananbaum:It was funded by NASA. Jim had a grant from NASA. Maybe he had a couple of grants because the germanium detector work that he was doing with Alan Womack probably was funded on a separate basis.
McCray:And for the detectors, were you working with commercial firms to get parts?
Tananbaum:Yes. We had a sodium iodide detector. I should remember the name of the company (Harshaw, I think). It was headquartered in Cleveland and made the detectors. But in any case, it was fairly standard with a little glass window so you could see the light flashes with photomultipliers that also were commercial tubes. And of course, you had to build things so they would work at the low pressure. The balloons flew to an altitude of approximately 135,000 feet in those days, so you had to worry about arcing and coronal breakdown, and plotted things around the high voltage to prevent breakdown. We flew the balloon about four time Ė once in September 1966, twice in May 1967 and once in June 1967. We got some arcing and breakdown in the middle of one of the flights, but also lots of great data. What was interesting is that Jim was more interested in the next generation detector, the germanium detector he was going to build. And he said, ďWhy donít you analyze the data from these last three balloon flights and see if we have anything interesting.Ē
McCray:That became your thesis?
Tananbaum:Yes, it did. Although it wasnít officially my thesis yet because we didnít know if there were any results.
McCray:How was the data analysis done?
Oh, goodness. For starters, we flew a tape recorder to record the data. It was a supermarket tape recorder, the same kind that they used to play music in the supermarket or used to in those days. Jim, I donít know exactly what he did because that was an area he took charge of. But he could get this kind of a tape recorder, and it held a lot of bits, and it didnít cost very much. He could take the front head off or something, so that it would not need sound to record, but it would run electronically. He modified the front end of it and it worked fine. When we got back, we had to take the tape out of the tape recorder and eventually read it into some kind of a reader machine. And he told me to go and build a turntable to do that. And I didnít have a clue other than it had to be some kind of a smooth, flat disc with a hole in the middle so you could put it on a spindle. And I remember working with one of the engineers who occasionally would take mercy or pity on a grad student. You know, what materials should we use, how to drill a hole right in the middle of it, and getting the darn thing to not wobble so that when we played it back it didnít wobble.
Of course, the different tracks Ė I think it had eight tracks on it Ė would have different information on the different tracks, the different bits were on the different tracks. So that was a major undertaking again because itís in a mechanical area. Once we were able to play this thing back (maybe by taking it to a supermarket and playing the tape then) and get it into digitized form, I actually then got the assignment to learn how to write small computer programs. We had, I think, the Lab for Nuclear Science had an IBM 7090 as a primary computer, ran off of punchcards, and plots were made on something called a calcomp plotter. I remember writing plotting programs and submitting this stack of cards that, of course, were a couple of feet deep. Then you would go the next day hoping you got your output back. Some days you got it the next day; sometimes it was two days. The first ten times you submitted a job, of course, it didnít compile or work properly. You would get it back with some kind of an error message. You would fix that mispunched card. Resubmit it. Then you had the next problem. And so, it was again, a sort of a painful, slow experience because I was learning to program and I was learning to analyze the data. But that was a very useful part of my education. One of the strengths that I feel that I have had, postgraduate school, is the ability to look at data carefully, see how it was analyzed, to ask questions of the people writing the code.
McCray:Iíd like to come back to that at a later point because that ties into some other questions about how people currently do astronomy with all the tools for understanding the data.
Tananbaum:So in any case, what happened then with those data, the program finally did work. I was able to plot the light curve. And one of the sources we looked at was Cygnus X-1 and it changed its intensity over a few months between two of our observations by a factor of between 2 and 2Ĺ. It was still a pretty open debate, a pretty wild debate. We didnít know what was powering these X-ray sources. There had been some reports in the literature for Cygnus X-1 and maybe a couple of other sources that the results from different rocket flights gave different values. But none of the detectors was precisely calibrated. So the group at AS&E in particular hadnít seen any variability in their own data. And since the variability came from their chief competitor—Friedmanís group at NRL—they tended to say, ďOh, they must have made a mistake.Ē So the people at AS&E didnít believe that sources varied.
McCray:Were you interacting with people much at AS&E at this point in time?
No. I didnít have any interactions at all with the people at AS&E. They were still building and flying rockets. We were doing the harder X-rays; ďharderĒ in terms of higher energy and fewer of them which I guess made it a bit harder. But the nice thing about a balloon is the balloon flights lasted eight, ten, twelve hours whereas rocket flights lasted five minutes. We did have fewer photons because we were nearly a decade higher in energy. And the falloff, whether they were thermal or power law, you were down a factor of several or even an order of magnitude or more in terms of the photon fluxes. But payloads could be heavier and they could stay up longer. And so we published a paper, Jim Overbeck and myself, saying Cygnus X-1 increased by a factor of 2.3 + 0.3 from September 1966 to May 1967 and then decreased by a factor of 2.6 + 0.3 in the one month from May to June. We were confident the results were real because it was done with the same detector in the last three flights. That was by design in the sense that we flew the same detector. We didnít know the source was going to vary, of course. So you get into the things that astronomers do which is the same as physicists do, in the sense that we had a stable detector.
It was a scintillator as opposed to a gas-filled proportional counter, so you couldnít get into arguments over the pressure in the detector, or was the efficiency somehow different, and so on, and so forth. We also, over the course of the last three flights that we did, saw a variability in Sco X-1; similar factors of a few, but brighter in June 1967 than in May 1967, the opposite direction of the change observed for Cyg X-1. I think it was certainly an interesting result. It was a significant contribution to the debate and probably the development of the field. As we go forward, weíll talk a little bit about how that influenced some of what we did with Uhuru when I was at AS&E.
McCray:How did astronomers react to some of the data that you were getting with this variability?
Tananbaum:Well, astronomers. Itís funny. Until 1970—
McCray:With the Uhuru?
Yes. In 1971, we started reporting results which is a whole interesting experience. The community to which we belonged—AS&E, the group at MIT—was the physics community not the astronomy community. Most of the people had degrees in physics. I was, certainly, at MIT getting my degree in physics. I passed the physics general exam when they didnít just ask me about astrophysics Ė maybe one question was on astrophysics.. I had nuclear physics, solid-state physics, and quantum mechanics, and E & M, and so on and so forth. Giacconiís paper in 1962 was published in Phys Rev Letters, the discovery paper. [Note: In late 2002, Riccardo Giacconi received (Ĺ) the 2002 Nobel Prize in Physics for his pioneering work in x-ray astronomy). Some of the papers by the mid-1960s were routinely being published in the Astrophysical Journal. But by and large, often as not, talks were given to the physicists as well as to the astronomers, it was both. X-ray astronomy definitely was not part of the mainstream of astronomy.
There were people at NASA that were originally not willing to support the field. Giacconi got the funding for the first rocket from the Air Force ostensibly to look for the possible fluorescence of X-rays from the moon as powered by solar particles. The argument would be that it might be useful for understanding the sun and monitoring storms and weather. And clearly, they wanted to look at more than the moon and they did look at the sky and discovered Sco X-1 plus the all-sky diffuse background. But NASA originally was not interested in funding this research . They were pressed. They (NASA) really, I think were more interested in ultraviolet astronomy, extending the optical astronomy that you could do from the ground into the ultraviolet. There was certainly much interest for the x-ray people in talking to the physicists about emission processes, how could you make these X-rays and what processes were generating the higher energies, whether they were thermal or non-thermal. Now there had to be dialogue. Eventually, there was a lot of dialogue with the traditional astronomy community. Through the 1960s, what Giacconiís group did among other things—they, and Friedmanís group, and Fisherís group at Lockheed, and a couple of others did—were scans of the galactic plane which picked up another couple of dozen sources.
They were trying to refine the locations and Giacconiís group was particularly interested in Sco X-1 because it was the brightest source. Minora Oda came over from Japan to work with them for a little while and they were developing these modulation collimator techniques to improve the position of Sco X-1. They worked with Alan Sandage and got this optical counterpart, 13th magnitude, give or take. I think 13th or 14th magnitude. Very blue star. It was clear that one of the ways you had better understand these X-ray sources is by locating them well enough, and figuring out why they were relatively bright in the X-ray, and perhaps not nearly so bright in the optical or radio. From the optical, you could do spectroscopy and determine if it was an object in our galaxy or perhaps another galaxy. We didnít think they were other galaxies because they were concentrated primarily in the galactic plane. Sco was a bit off, 15? or 20? degrees off the plane. If you put in reasonable distances, youíd find for stuff towards the galactic center, you must be getting close to 1038 ergs/second out. How did you generate five orders more luminosity than you get from the sun in all wavelengths combined? So there was really some intrigue and this was a great time as a graduate student to be thinking about these mysterious objects. You come into a field in which there were a dozen and then eventually two dozen objects; you could read the entire literature in a few nights. You could name all the sources by their first and last names. They had the potential (to be important), because they were very mysterious, so it was a wonderful opportunity. Plus the fact that MIT was in the middle and you could build something as a graduate student. It was just a great time to be there and to get into the field. Then to be fortunate enough to make a discovery as part of your thesis that sources did, in fact, vary.
McCray:Can I stop you for a second to talk about AS&E? As you were working as a graduate student, did you have any particular attraction towards the theory side of things or the instrument building? It sounds like you had a little bit of each.
Tananbaum:I was probably not interested in the details of the theory of the emission processes. I was perfectly happy if someone else did some of those calculations. I was certainly more interested in the building of the instrumentation and then observing. Primarily what I was interested in, maybe if you asked me today, still the most fun, the thing that I find the most interesting is using the equipment that is built to look at something and to try to figure out how it works. So the observing, the strategy to observe, the actual observing, the analysis of the data: those are the things that I enjoy the most and thatís what I was interested in even back then.
McCray:You graduated in 1968.
Tananbaum:Yes. I had applied for several jobs and there werenít too many openings in X-ray astronomy. We hadnít solved any of the mysteries; they were just a little bit deeper. So now we had these sources, we didnít know how they worked, and they could probably vary. I went over to AS&E and gave a talk on my work. I learned—maybe from the talk itself, from some of the questions that were asked and certainly thereafter—they still werenít convinced that sources could vary. There was still a healthy skepticism. But there were several opportunities to work on different projects at AS&E. They needed people and somebody who had already done some work in X-ray astronomy was very attractive to them, or attractive enough. I received a nice offer.
McCray:Did you have any hesitation about leaving an academic setting?
No. Thatís a very good question. Thatís an area where there is a small amount of sensitivity. In the time that I was there, Jim Overbeck had decided that it was important for him to have full control over the use of funds that he was getting. Either because he wasnít tenured faculty or because originally it was done as an amendment to George Clarkís grant, Iím not really sure. But his money was set up to flow through George Clark as separate, but part of Georgeís larger grant. It may have been that reasonable for Jim to have independent funds in that he was doing his independent projects and part of what happens in this wonderful world in which we still live, your opportunities for promotion and permanent positions could be affected by your ability to get independent funding. Maybe Jim was concerned that the things that George was doing were going to run into some difficulties and what if George needed some of Jimís money to do Georgeís work. It was technically under Georgeís control.
I donít believe anything at all happened other than these kinds of what-ifs. And Jim asked that his funds be separated from Georgeís and this created quite a flurry of discussion within the physics department and the MIT Center for Space Research, which did not, I think, even exist quite yet, but maybe it was coming in as a fledgling organization. Eventually, I think the grants couldnít be separated, but Jim was given a separate budget, separate from Georgeís money within the single grant. Their personal relationship became unfriendly and they were working in the same general area so they were probably already a little bit competitive, but they became even more competitive. And George had most of the graduate students, and there was Alan Womack and myself working with Jim. I only learned subsequently that George, for example, had regular discussions in a reading or journal club with his students where they discussed what was new and what they were going to do next.
We werenít invited to participate. I have been friends with George and worked with George through a lot of projects since then and so there is no long-held, strong, negative feeling on my part. But I think it was also a statement to me that the academic environment isnít quite the ivory tower that it is perhaps made out to be by people. I really wanted to do research. I didnít want to teach and AS&E was an opportunity to spend all my energy doing research. I mean, building things as part of research, of course. I had struggled a bit at MIT in terms of the academic aspect of things. There was never any interest on the part of the Institution to see if I was interested in staying on. I wasnít encouraged to apply for positions at other universities. And so, it was very natural for me to move out of the academic circles. In some sense (maybe in every way), it was the best thing for me.
McCray:What was the environment like at AS&E? If you walked into the building, what kind ofÖ?
Tananbaum:The building was a converted warehouse. So the digs were not exactly— it wasnít like the high corporate end. I think this was the Spring of 1968. Riccardo Giacconi was the intellectual and dynamic force at the place. He was Executive Vice President, and Martin Annis was the President and nominally ran the company. A lot of the companyís original funding had been involved in weapons testing; looking at fallout from nuclear weapons. By the mid- to late-1960s, more of the funding was coming in X-ray astronomy and there was a possibility of looking at commercial applications of X-ray instruments.
Tananbaum:We did get heavily into the airport baggage scanners using the detectors. If you think about it, detecting X-rays from halfway across the universe is probably going to need the most sensitive detectors. If you want to have low dose devices so you donít either kill the passengers or the people operating the equipment, you need sensitive detectors. So AS&E got into that. They were interested in remote readouts of electric meters. They never really made much of an in-road in that area. But security inspection devices was a good area. They did a lot subsequent to the time when we left. They did a lot in medical instrumentation. Again, low dose, sensitive and imaging X-ray devices.
McCray:Were these projects you ever worked in?
Not really. Riccardo certainly had some attention that he gave to them and they certainly were mixed up with the overall corporate financial picture. When I came to work, I had an interesting experience in that I was told when I was hired that there were three openings. One was to work on the Uhuru satellite which was still called Small Astronomy Satellites, SAS-A. Second was to work on what became the X-ray telescope on the Apollo Telescope Mount on Skylab. And the third was to work on some particle data solar wind particles or cosmic rays Ė I donít really recall Ė with Frank Paolini, who was one of the original four on the Giacconi, Gurksy, Paolini, and Rossi discovery paper. Paolini was leading that area. That was, to me, a no-brainer in the sense of what I was interested in. I had already been looking at these sources and so working on the first X-ray astronomy satellite was great. The thing that happened was, I was over there and available to start in April. I had finished my thesis up and I could start a month or two before my graduation in June. And the two people who came after me worked on the Apollo Skylab project and on the particles data.
And so, I donít know. If those people had come in before and had made different choicesÖ You see the chances, the things that influence the way things work. So I got there in April of 1968 and we launched SAS-A December 12, 1970. I worked on the hardware for about two-and-a-half years. The design was pretty well done. We were struggling when I first came to make proportional counters that didnít leak, that would go up in space, and could last a year or two as opposed to five minutes. And we didnít like the idea of flow counters because regulating the pressure could be problem, so they would be sealed counters. It had been decided to make the housing and the windows out of beryllium so as to deal with thermal co-efficients, and mismatches, and not have the windows crack as the temperature changed. The first thing I think I was assigned to was to see if we could sort of paint some sort of a passivation or a sealant on the counter and then put it in a humidity test. It had already been decided that we were going to launch this thing off the coast of Kenya from a former, oil drilling platform the Italians ran, so humidity was a concern.
McCray:Was the testing done in-house?
Tananbaum:Yes. We had an oven and we could put water vapor in there. The counters died in a few days to a week. They would turn white. It wasnít just the windows. You didnít want to put a lot of goop on the windows because then you would start cutting out the low energy X-ray transmission. But beryllium, if you put it in a warm, moist environment, you get a white powder which, no doubt, is beryllium oxide. Beryllium corrodes and the counters leaked. It just takes a pinhole, a little bit of oxygen inside the windows or through the body to poison the counters. So we tried a few times, and we didnít have a lot of counters, and we were ruining the counters. I mean, we were killing the counters.
McCray:Where were you getting the counters from at this point?
We had two vendors. One was called LND which was located in Rockaway, New York run by Bob Lehnert; it was a one-person company. The other company was called Reuter/Stokes in Cleveland. Harshaw was the scintillator company that made the sodium iodide and they were in Cleveland too. I believe they were in Cleveland. Reuter/Stokes, their counters were never as good as Lehnertís. When you got them, the resolution wasnít as good. These were large counters for those days. The counters, the cross-section at the end was a couple inches by a couple of inches and I think it was about 16 inches long. But, in any case, a bank of six counters gave us an effective area of about 800-square centimeters in my recollection.
So the effective area of each of the individual six counters on one bank and the six on the other (we had two independent banks of counters, back-to-back), more than 100-square centimeters. And it was challenging to make that big of a unit without contaminants, without leaks, with thin beryllium windows with a sort of grid work that supports the window, a sandwich above and below Ė it mattered how you epoxied it to seal it. If the counter was bad, you could tell because you would look at the iron 55 source that you held in front of it in the lab. Instead of seeing a nice, sharp peak you would get broad, broad responses that would tell you the counter wasnít any good. So we were trying to get counters that were sealed and that could hold their seal. We were trying to decide between two different fill gases.
One was argon with CO2 as the quench and a trace of helium. The helium trace was put in so that if the gas was leaking you could put it in a vacuum chamber and sort of sniff. You pulled on a vacuum chamber to see if there was any helium. The only place helium could be coming from was inside the counter Ė leaking out. And helium is inert so it didnít affect the performance. And the other option involved methane quench with argon as the primary detection gas. We found that the methane, if you put high count rates into the detector, the methane could sometimes breakdown or polymerize in some way and you would get carbon deposits on the anode which would cause breakdown or poor resolution. So we ended up picking the argon/CO2 choice although the basic energy resolution may not have started out quite as good, because the lifetime of the counter was better. We solved the humidity problem by deciding we would keep the counters in a nitrogen purge and not let them see the humidity at the launch site.
McCray:How were they transferred?
Tananbaum:They were sealed in a bag. The whole satellite was bagged, I think, when it was shipped and probably dry nitrogen was flowing through the bag. It was like that until hours before launch and the purge was separated or cutoff just before the launch. We never could figure a way to avoid this corrosion issue.
McCray:While you were doing all this work, how were you keeping up to date with what was going on in the astronomy side of things?
It wasnít too hard because there werenít that many experimental groups. Ours was the biggest; ďoursĒ now being AS&E and we were focused on building the satellite. So we werenít flying that much. There other was another rocket flight and the spectrum was measured for Cas A and another source. Again, I think I would have to go back and look at the literature, but Iíd be surprised if there were as many as ten papers written in the year of 1968 or 1969; maybe twenty. But it meant that you read a three or four page paper every other week or once a month. You would look at how the statistical analysis was done because the big debate was over how you measured the background and what constituted a statistically significant detection.
Or, what was the statistical uncertainty if you fit a temperature. These are problems physicists work with all the time. But there were interesting problems. Some of the groups would say that the result from another group was a little bit questionable because of the way the data had been analyzed. You couldnít just take the lowest background and then subtract. The background varied over the course of the observation. You had to deal with that some way or other. So it was not that difficult then keeping up with the literature. At the end of the series of interviews, we will talk about keeping up with the literature now. Itís impossible. Itís all there Ė with the astroph data server and the Astrophysics Data System and the journals Ė itís just that there is so much Ė the field is so vast that there are several new papers almost every day. Back then, people sent you preprints. You didnít have to wait for the stuff to be published. You got on the mailing list and people sent you preprints. (And one file drawer would hold them all).
McCray:Richard Hirsh has written a book about the early years of X-ray astronomy and spends a considerable time talking about Uhuru. In order to save time, there will be some things I am skipping over, just so youíre aware, because it has been talked about before. I guess Iím more interested in focusing on your role, of course. But before we talk about Uhuru, I wanted to know if you had any particular anecdotes about Rossi or Giacconi while you were at AS&E that are worth sharing.
Tananbaum:Iím sure as we talk, a number of things will come up that are about Riccardo. Bruno I knew quite well. And often my interactions were when he would stop by in the Spring before going off to give a series of lectures for the Summer in schools in Italy and he would ask what was the latest result from Uhuru and even Einstein in the early days. We would just talk about the science. He was a nice man. It was a nice relationship. I have an anecdote about the person who married my brother. My brother has been married twice and at the second wedding ceremony they had to sort of give me an evil stare to stop giggling. They asked why I was giggling and it turned out for the serious reason that the Rabbi had one brown and one black shoe on. I found that to be quite hilarious. Even if we had anecdotes, I think our sense of humor may not exactly excite people. If they think the shoe story is funny, I donít know Ė maybe other tales would be interesting. There are anecdotes about Riccardo. Either you or somebody has and will be interviewing Riccardo.
McCray:He has been interviewed for work on Hubble and a variety of different things.
The X-ray stuff in particular, I think, is very important. No doubt, somebody would have decided to do X-ray astronomy whether Riccardo had been that person or not. But who knows when it would have been done, and how it would have been done, and whether it would have been successful. Would we be talking about doing the first X-ray observation now as opposed to forty years ago? There is a proposal that Riccardo wrote in 1963, a year after the first discovery of Sco X-1 and the background. I didnít have a copy, but he sent one to me. I think when we won the competition for the Chandra Science Center, it was still the AXAF Science Center then. And with a congratulatory note, signing the 1963 proposal for me. But in any case, in that proposal he talked about an X-ray satellite which was what became Uhuru. He also talked about a 1.2-meter telescope with arc-second resolution. It was the observatory which we started out to do as Einstein, but as the HEAO program was cutback, Einstein only became part of that. Now weíve actually done it with Chandra. So whether itís an anecdote or not, in 1963, a year after the first couple of discoveries, Riccardo wrote down a blueprint, as it were, for the field.
And to some extent or other, Iíve spent over thirty years of my life building some of the things in that blue print. Some of them with him and some since he moved on to Hubble and now radio astronomy too. But thatís an amazing vision. You could also say itís a statement about the competence of the people who subsequently built the things, that it took us thirty-odd years to do it. I donít accept that, of course, because what we have built has been tremendously successful. Riccardoís vision, I mean, itís an amazing capability of seeing what you would want to do, and how to do it, and laying it all out. So I worked closely with Riccardo from 1968 to 1981. Of course, weíre still good friends. But in the early days it involved interacting on an almost daily basis as opposed to seeing each other every couple of months which is the current case. Thatís a very different situation.
McCray:What was the style of work like while you were working on Uhuru at AS&E.
Tananbaum:Was there a method to the madness?
McCray:Well, how did it vary through time? There was a team involved. Was there a particular style of running the team? How were decisions made?
Tananbaum:There was a style, and Iím sure the style evolved while we were working together, and probably has continued to evolve since. You know, if you could write it down and bottle it to the extent that itís been successful, itís useful. But probably whatís most useful is to try to look at it and see what things worked, realize that there are many different ways to do things. Some of it depends on personalities, some of it depends on the circumstances. So Iím sure it has evolved. When I came on board, I probably worked for six months to a year with Ed Kellogg as my supervisor. He was the Project Scientist. He worked for Herb Gursky and Herb worked for Riccardo. But Riccardo was the Principle Investigator so there were, Iím sure, regular reports that were made to Riccardo on how the project was coming along Ė how that new kid was killing all the detectors in the water chamber. They were also worried about what was called the star sensors or how we were going to figure out where we were looking at the sky with Uhuru. I didnít just work on the counters. I also was working on the aspect system or star trackers.
McCray:This would be pointing and tracking sorts of things?
Yes. And I also worked on some of the electronics to reject background pulses. So I was working with the electrical engineers. The nice thing at AS&E was that the science and engineers worked in the same building, under the same roof, as a team. And that is something we donít have exactly right now, in that we have lots of outstanding engineers, but at SAO weíre spread over a couple of buildings. So the frequency of the interaction and dialogue are not quite as high. I think that was an important ingredient. It takes scientists and engineers to build these satellites. There is a contribution from each and it doesnít mean a scientist canít do engineering; it doesnít mean that the engineers donít think about the science. There is a lot of great overlap. But in any case, working together with the engineers was very important. We had a technician, a junior scientist who worked with us.
We had the equipment discussed already, and we had sun sensors too. And so, we were going to calibrate the sun sensors. Well thatís easy. You can actually go outside, and look at the sun, and do that. And he put the thing on a cart and he rolled it out onto Carlton Street, which is where AS&E was located. He hit a bump. It was a prototype so it wasnít the unit we flew. But it was an important piece of hardware. He hit a bump and the damn thing fell off the cart and landed in the street. So I mean, I think, you know, we were kids to some extent Ė enthusiastic but sometimes making mistakes. I remember one little lab counter, one that we probably had already ruined. Looking at it, I had a screwdriver in my hand, and I poked the screwdriver through the thin beryllium window. The beryllium window was like an eggshell, that is how I can recall the screwdriver going through. So the engineers, probably some of them that I still work with to this day, a few of them will laugh and tease me about being allowed to have a screwdriver and being anywhere near any of the flight equipment. But what Iím saying is I guess it was an environment that without my even knowing it, it was open, there werenít long lists of rules. People were encouraged and enabled or empowered to sort of get things done. There werenít sharp boundaries; there werenít a lot of turf or turf wars. Within about a year, I had done well enough that I got promoted to the Project Scientist role.
Ed Kellogg was interested in taking a couple of other things on and recommended to Riccardo that I be promoted. So I became the Project Scientist. And again, it was just a wonderful opportunity. What happened then is that I began to get involved in more meetings that Riccardo would be participating in as well. I donít remember the exact dates and probably I could be off by months. But at some point, the Crab Nebula was discovered to pulse in the optical and then in x-rays too. It could have been that it was discovered in 1968 and reported in 1969. We were part way through the building of the flight electronics for Uhuru and the designs were certainly done. Riccardo wanted to modify the electronics to have a millisecond or a few milliseconds timetagging in capability so that we could use Uhuru to look at the Crab pulsations and also search for other pulsars. Now pulsations, notwithstanding some people at AS&E still didnít believe that sources varied in the macroscopic kind of a way Ė i.e., by large factors and irregularly. I remember sitting in some meetings where first there was an internal assessment as to how you might accomplish this fast timing with Uhura. We could go to NASA and say we wanted to do it, what it would cost, what the impact on the dollars and scheduling would be.
I remember in one of these meetings—it may have been involved with the electronics, it may have been involved with some other aspect—Riccardo asking one of the senior engineers or managers what the individual thought about some aspect of it. The person, who probably had an opinion, was a little bit intimidated by Riccardo and so he started to say something in one particular direction. Letís say he said, ďwhiteĒ and Riccardo said ďgrayĒ or ďblackĒ. And the guy said, ďOh yeah, definitely black.Ē And then Riccardo said, ďWell, white.Ē And the guy came back and said, ďWhite.Ē I donít even know if Riccardo necessarily did it consciously and he certainly didnít do it out of any mean-spiritedness. But he may have been doing it to test how much the person had thought about it or how strongly held the conviction was. Perhaps, just to argue both sides because we surely did things that way all the time. We would argue both sides of any argument until we could essentially be convinced that the direction in which we were headed was the best direction or at least the best that we could come up with. But in any case, I felt a certain amount of discomfort, and a little bit of the stomach churning, and was glad that I wasnít that individual. I felt badly for somebody I knew and liked. Iím not sure how much the individual realized what was going on, but it was very clear to me and it was clear that in my mind whenever I was going to be quizzed by Riccardo, I was going to tell him what I thought and why I thought it. I would say it probably took until some time later for me to be able to able to articulate exactly what I saw and how it influenced me. But it influenced me from that point on. I was never intimated by Riccardo. It doesnít mean I didnít have this almost awe. Well, I didnít hold him in awe; I respected him tremendously. But I always felt comfortable saying what I thought and I have tried to do that with everybody that I have worked with ever since.
To get the best from people, you need to be willing to trust them and they need to trust you. And what Riccardo did, which was the best thing I think, is he trusted us to do what we were capable of doing and then some. And I think the trick for us was to realize when you needed advice or you needed to discuss something, when you needed help. You could ask for help and Riccardo would be there to give that help. And so you would be encouraged to take something on without even necessarily being sure you could do it. Then if you did it, great; and if you didnít and if you needed to discuss it or look at options, then the resource to help was there. So that was an experience that has obviously left some impact. We decided we could, in fact, modify the electronics. I think the cost was less than one hundred thousand dollars or of that order. We took it to NASA and they turned it down. They were concerned about our staying within the overall budget, staying on schedule. That is not to say that they didnít require us to do other things. We had to pull all sorts of electronics apart later and put a crimp in the leads on the resistors because there was a sense that there might be too much stress without this little crimp or curve. We spent money doing things that Riccardo would have said—and his engineers and his experts would have said—were not needed. We spent more money fixing things that didnít need to be fixed.
He was quite unhappy, letís say, about not getting the approval for the fast timing capability. But I also learned that even Riccardo didnít always get his way. So that was probably a useful lesson. We did look for ways in which we might be able to use Uhuru to look for pulsations. For example, we realized we had a different time resolution on the two different banks of detectors and we actually had the option of using the faster time with the broader field. So it would allow us to stay on a source longer and with more ďbinsĒ than the basic design had intended. The best resolution we had, as I recall, was 96 milliseconds. It was very relevant to some of the mysteries of Cygnus X-1 when we get to it.
McCray:Did you go to the launch?
Tananbaum:Did I did go to launch? I did. I was 1/3 and then 1/4 the launch team. Launch was the San Marco platform in Kenya. I, by now, had a little boy. My wife took my son, and they went to stay with her parents and my parents in Buffalo because I was going to be gone a month. We moved into our house which my wife and I still live in; we moved in on Halloween day. The only house on the street with no lights. No trick-or-treaters either. The day after Halloween, she went to Buffalo with Ken and I went to Africa.
McCray:That was a pretty busy time, buying a place and moving into it, and going to Africa for a month.
Tananbaum:Maybe we went a few days after we moved. It probably wasnít the very day after because I was in Africa for about five weeks. We launched the 12th of December. So I may have gone the 5th of November, give or take. But in any case, we had one electrical engineer, one mechanical engineer, and one scientist. Our team preparing for a launch was the three of us. Stan Mickiewiz was the electrical engineer; Dick Goddard who still works with us, was the mechanical engineer; and I was the scientist. Riccardo came over about two weeks before launch, but we had unpacked, and integrated, and tested that everything had survived the transport. Also spent our times on the beach. Once I had been, at twilight, mistaken for a local because weíd had that much sun exposure. Rode a rubber raft out from the launch camp to the platform and got more suntime.
McCray:How far out was the platform?
Tananbaum:Iíd say it was either three or six miles offshore. I donít remember which. Had steak for breakfast at 6:00 at our hotel and spaghetti at 7:00 at the base-camp, before starting the workday. It was an interesting way to live your life. My recollection, Iím sure the numbers arenít right, but the hotel was in Malindi, several miles from the base camp. It was the Sinbad Hotel. The meals came with the room, but I think we had a food allowance of seven dollars a day. It was 1970 so it is possible that our per diem was only seven dollars a day for food. But our food was all covered with the hotel room rate and rum and Coke was fifty cents a drink. On any given day I was never able to drink my full seven dollars worth of per diem in the form of rum and Coke.
McCray:That was probably a good thing.
Tananbaum:It was a good thing. I met the drummer, Charlie Watts from the Rolling Stones.
McCray:No kidding. What were they doing there?
He was on vacation with his wife and a young baby and stayed at our hotel. The local entertainment consisted of movies that were shown at two or three different hotels—or maybe just our hotel—on a big bed sheet. And if it rained, there was no movie that night. There was a guy who was a DJ who went from hotel to hotel, there were four hotels, and he had a stack of records. And he would play one side forward one night. He played us the flip side the second night. He played the flip side backwards (bottom to top of stack) the third night and played the original side backwards the fourth night. From top to bottom and bottom to top. I guess we were there too long because he tried to sell us his house before we left. That must have meant he knew who we were. But in any case, we checked the payload almost daily to see that things worked and we also had special tests assigned. We might have a fourteen hour workday and other days we might just check that everything still worked from one day to the next and we would be through in a couple of hours.
We had a couple off days on one weekend that Dick Goddard and I toured in a Peugeot and saw some of the game parks. That was a splendid place to be. It was a lovely, lovely trip from the point of view of, that we worked very hard, but also had a little bit of time to relax. The launch itself was incredibly chaotic. Riccardo decided to stay out at the control room. There were two platforms. Santa Rita was the control room; it was in a separate platform and then San Marcos was where the thing was launched. I was in the telemetry van on the shore so I could monitor some last minute tests to make sure the detectors were good before we launched. The countdown went up to a certain point. It was supposed to launch just after midnight, or maybe at dawn. Anyway there was a holdup. Some stuff was checked. Some stuff was checked further. By 11:30 AM, we were still at T-minus-2 hours or something like that and holding. It could have been T-minus-30 minutes. And somebody noticed that the temperature was about to go off scale; theyíd been sitting out in the equatorial sun cooking for several hours, and the battery was getting up to an unsafe temperature level as were a few other critical components. All of a sudden they said, ďT-minus-30 seconds.Ē
The go ahead came from the group in charge of the launch, the Italian launch director and the advisors Scout rockets. It was LTV that made the Scouts and they were based in Texas. They had a person there who said it was good to go and it just blasted off. I ran outside of the trailer to watch it go. It was this interesting contrast which I had noticed days before. This base camp was a hundred yards from a native village where there was no plumbing, no electricity. And here we are sitting in a computerized van launching a rocket into space. So it was this amazing set of contrasts. The launch went smoothly. We werenít supposed to turn it on for several days because we wanted to wait for it to out-gas and then the high voltage connections that were potted would hopefully, not have any arcing. A couple of hours from launch, I was headed back to the States because I was supposed to go to Goddard Space Flight Center and help oversee the turn-on.
McCray:Thatís where that was being done from?
Tananbaum:Yes. Thatís where all the data came back, to Goddard Space Flight Center in Greenbelt, MD. So I was on some kind of a car, or a bus from Malindi to Mombasa, and then I was supposed to fly to Nairobi. There was some problem with the flights and everything. I was half awake and half asleep. Everything got jumbled around. I got back. I was in Africa one day and I got back to the States the next in a snowstorm. It was the 13th of December. It was clear I was going to be at Goddard for more than a few days. My wife brought our son to Maryland since I hadnít seen them in several weeks. I was busy re-integrating into the family and getting the satellite turned on. Now back to Riccardo and Marjorie Townsend. One of the things that was notable was the project was managed by a woman which was very unusual in those days. Marjorie and Riccardo, I think, went into the telemetry station outside of Malindi that was connected to the launch site and they turned the satellite on for a couple of minutes a day or two after launch. Maybe an hour or two. I donít know which. It is probably written down someplace. They couldnít contain their curiosity and they peeked. I donít know how much data they looked at. I donít know if they turned the high voltage on. I doubt that they did. But they did look at a few things.
McCray:Very hard to resist not to.
We almost lost it a few weeks after launch. The telemetry became intermittent. The battery got very warm. The transmitter got warm so we re-oriented to cool things down and everything was fine. Six weeks after launch the tape recorder, actually did fail and we had to scramble to setup a network of stations on the ground so that we could get about 60% percent of the orbit covered with real time coverage. I mean, it was very fragile, in a sense. It was pretty amazing the amount that we ended up getting from the satellite. In some form, part of it operated for over two years with reduced capability. We had the opportunity to guide the science program. We could pick a circle in the sky through which we would scan and there were some limitations because we had to be at a certain orientation relative to the sun. There were thermal considerations and solar sensor considerations. One of the things that I put into the program I think in the first or the second week was a scan that went through Cygnus X-1 so we would get a data point every 12 minutes for a day.
Sure enough, Cygnus X-1 varied very substantially within those data. And that was one of the reasons we went back to it and looked at it quite substantially in the next few months. That led to a big discovery eventually involving ourselves and other X-ray groups and other astronomers. Between the fast time pulsations, the optical identification, and the mass, it was the first strong evidence that black holes actually could exist. Evidence that there was really such a thing as a black hole and thatís what Cygnus X-1 was. Something we didnít do particularly well was that I donít think we realized the full significance of the discovery. They didnít have NASA Space Science updates back then. The interaction with the media wasnít the same. There were discussions within the science community to get radio people to observe, to get optical people to observe, and the story got built up piecewise.
McCray:Were you connected well to those communities so you could get to the radio and optical people onboard?
Tananbaum:Some. There were people who were interested and you would put an announcement out. IAU telegrams were out on minor planets and comets and asteroids, and you could put astronomical discoveries there. You went to meetings and you gave talks. Things didnít happen overnight quite the way they do now. Somebody had to have telescope time and you couldnít switch times easily or change the program. But in any case, everybody who wants to write the history of the first discovery of a black hole, including ourselves, can write it as they choose. Parts of the story came from our X-ray work, parts of it came from the optical determination of the mass function and the connection of the X-ray to optical was only secured because of the radio. Actually, I was also involved in that connection. Christine Jones, who still works in our group, did some of the work as a graduate student with me showing the high probability that X-ray source was in fact the same as the radio source. And in turn, the radio source was at the same position as the star for which the mass function was being determined. I would say that the result is that nobody directly got prizes for discovering the first black hole. Nobody is written up in the textbooks as having made the discovery.
McCray:Why do you think that is?
Tananbaum:Because we didnít manage or control the distribution of the data or the news releases. We, in part, didnít completely realize the trail that we were on. We were working with other members of the astronomy community and not in a closed shop kind of a way. So it was a very natural way for things to get done. But because it wasnít orchestrated in quite that way, there is no individual or group formally credited for the first discovery of a black hole. I mean, Riccardo, of course, has gotten a number of awards for lots of the discoveries (including the 2002 Nobel Prize for Physics) and other people have gotten awards for outstanding work that was done. I think the work on super-massive black holes and quasars also was beginning to get discussed in this same time frame. And within a few years it began to get accepted from Donald Lynden-Bellís work and Martin Reesí work that very massive black holes might exist in the cores of galaxies. See, one of the keys is that with Uhuru the problem that we were more interested in solving was how are these sources powered. That problem was solved as we were following up on Cygnus X-1. We first realized that we could get this 96-millisecond timing by switching the electronics to the broader field; that really was a combination of Riccardo, Ethan Schreier, and myself who were working on these sources.
McCray:Was Ethan a student?
Ethan was already a Ph.D. He had come over from MIT about a year after I did. We had worked together on the software which is another whole area we should talk about at some point. In any case, we realized that we wanted to see more of Cyg X-1. We had these pulse trains and there was the question on whether they were periodic and what they were telling us about the source. So we decided we could slow the spin rate of the satellite down, not in a controlled way, but make it go much slower. So that as the satellite went across the source, it would stay on the source longer and you would see more of the source with the best time resolution. I think we picked a trajectory that went through Cyg X-1 and Centaurus X-3. So it probably was pretty much along the plane of our Galaxy. What we discovered in those data which probably we took in the March time frame, March of 1971 was really key.
This was just three or four months after launch that all of this is happening. We came in one morning and the plots were waiting Ė I donít know Ė we may have left a job overnight. But anyway, there was a plot in the morning and we saw these regular pulsations from Cen X-3 Ė Ethan, Riccardo, and myself. It was very clear within seconds that this was an X-ray pulsar, with a period of nearly 5 seconds so we could see the pulses even without the millisecond electronics, which NASA had not approved. What we wanted to then do was to figure it all out Ė , you know, we didnít have an optical counterpart. We didnít know anything about it (Cen X-3). It was in an obscure direction in the plane of the sky in the Southern Hemisphere. It took some time to find out optically what the corresponding star was. But we realized, well, we could see if the period was stable. And we had some data which is what is plotted behind you in that golden frame. We see the overall source intensity at a low level and then over a period of about an hour, the intensity goes up by more than a factor of 10. And then more or less wiggles a little bit, but stays up for the better part of a day or whatever the time frame is. [Tananbaum shows plot from original data collection.]
McCray:What are the dashes above this spot?
Tananbaum:Maybe, I donít know. Whatís the scale? Maybe spectral hardness.
Tananbaum:Chi-squared. So that was for each of the observations, either the deviation from the short-term sinusoidal fit to the intensity or more likely something to do with on the X-ray spectrum.
McCray:Now, is that the original plot?
Tananbaum:No. This is a hand-drawn plot. It is in pencil on graph paper drawn by me. It is an original plot in that it was made in the Spring of 1971. But each of the data points was the intensity computed as Uhuru crossed over the source, each crossing of the source every 12 minutes or so, would give us an intensity. And the time of the crossing in seconds is probably what is plotted as the x-axis here. So this is the observation at 78,582 seconds. This is the observation at 79,285 seconds. So 700 seconds later. So this is when the satellite was spinning at its nominal rate. This is not from the slowed down data. But when we had a longer stretch of data and this plot is over almost two days, it turns out the period of this source is just over two days, and if we had stayed on the source just a bit longer, it would have come back down. But first of all, the baseline is not a zero. There was another source in the field, but we only realized that later which complicated our interpretation. This plot of Cen X-3 is actually the first transition out of an X-ray eclipse that we recorded from an X-ray source, but we didnít know. The paper that we wrote says there is an intriguing intensity variation in the source. Riccardo did the analysis of the period and says the period is not constant. We can fit it to a series of linear regimes in which if you look at those linear regimes but if you look at those linear regimes after the fact and offset the zero, you can make a sine wave out of the three straight lines. And this is because we were still physicists and we were not interacting yet all that regularly with the astronomy community. The idea of an eclipsing binary source is trivial to an astronomer and it is trivial now for the X-ray sources, but it was not something, we realized at first Ė I mean, we knew sources could eclipse. We knew there were binary stars. But we had never really looked at the data from an eclipsing binary before.
McCray:What was required to make that connection between those two things?
Tananbaum:What happened was is that we went and got more data with the satellite on the period and on the intensity. And what we found was that we began to see a regular pattern of the intensity going up and down. Had it gone to zero, it might have been a better hint of an eclipse, but donít forget, this other source prevented it from going to zero. So we plotted. Ethan took the lead and Rich Levinson was working with him. They plotted the intensity and they were fitting the pulsations over as much time as they could. In a second paper that we wrote, we perceived clear sinusoidal variations of the period data. And we could actually see the period speeding up over time.
McCray:Because of the binary accretion?
Tananbaum:Yes. Because of the accretion, which tells us it is not a pulsar losing energy, but in fact it is a gravity that is generating the energy that weíre seeing. You can measure all sorts of parameters like the size of the orbit, the speed, the masses of the stars, and you start getting into the real details of the system. But, in fact, the Astrophysicsí Journal included in its Centennial issue, the paper by Ethan Schreier et al.
McCray:Yes. I have the paper.
Tananbaum:I was asked to pick a paper of special importance, and I picked the paper by Schreier et al., which is the paper where we actually explain what is happening. As opposed to the first paper which is Giacconi et al., that says we have this mysterious periodicity that changes in an unexplained way with time and also the light curve goes up and down. It took us a couple of months to put all the data together and understand.
McCray:How did astronomers react to this?
Tananbaum:Everybody reacted very well in the sense that it was clear that the solution was correct and that we now had a very good handle on the basic process by which the X-rays were being generated. That gas was accelerated by the force of gravity, falling into a deep, potential well.
McCray:It is a very elegant solution.
It had even been one of those proposed in the mid-1960s when the sources were first being described. But there was nothing to favor it over any of a half-dozen other competing possibilities. The astronomers, they knew binaries, so they started to get very interested in these sources because you could start to calculate and observe and measure some of the key parameters. Include the spectral types of the normal stars and start to get into the question of the mass of the collapsed start. Look, one of the things is that the prevailing thought was that when a supernova exploded and created a neutron star, whether it was in a binary or not, clearly the more massive star exploded. No. Let me say it right. The more massive star would evolve most rapidly and therefore when it exploded, it would disrupt the binary because more than half of the mass would be lost from the system. So you wouldnít expect to find neutron stars in binaries. Now, when you look at close binaries you need to consider the mass transfer processes. Ed VandenHeuvel had been doing calculations about this point in time. Lo and behold, he could show very clearly how the originally more massive star transfers material to the original less massive star.
The star that has the most evolved core losses its atmosphere and transfers it to the other star. Itís not an X-ray source at that point because both stars are more or less normal. But the more massive star as it is evolving actually transfers material to the less massive star. The core of the more massive star continues evolving and can undergo a supernova explosion. But the more massive star is now the one that was originally less massive and is less evolved. So more of the mass is in that star and even with the explosion of the more evolved, but now less massive star, the system isnít blown apart. It goes into an elliptical orbit which can recircularize. And as this now more massive star evolves, it has a neutron star or black hole companion. As it starts to lose mass, the mass falls on to the collapsed star and thatís how you make the X-rays and it can be in a binary system. So it affected ideas about stellar evolution and mass transfer. And things that traditional astronomers were certainly very interested in. You have to realize our excitement over the discovery of Cen X-3 and subsequently Hercules X-1 and determining the energy source. I mean, Cygnus X-1 was very important. But we werenít focused on the black hole so much as understanding the process by which the X-rays were produced. Thatís the best way to say it.
McCray:In retrospect, the black hole seemsÖ
Tananbaum:Well, itís more glamorous. No. I donít think itís more important until we start getting into, hopefully, the extreme physics around black holes. That is still to the future as opposed to the present or the past. I think it also made it easier, though, for the people looking at super massive black holes, and AGNs, and quasars to talk about the accretion processes for those in a way that after Cygnus X-1 was much more acceptable in terms of explaining the energy source in quasars and AGNs. You would have to actually ask Lynden-Bell or Rees, you know, how much were they influenced at all. And they would probably say their calculations were already ongoing, that there were papers in the 1960s that said that these energy sources could be of that sort. I think whatís key is that the discovery and the explanation of the binary X-ray sources certainly made it easier to accept that explanation that there are these super massive black holes. So that may be a good place to stop.
McCray:I have one other question on Uhuru. Were there any downsides to the success of the mission, in terms of all of a sudden, now, X-ray astronomy has all this visibility and interest among a broader scientific community?
Tananbaum:No, I donít think there were any downsides. The fact was that that was how support increased for a telescope to extend X-ray astronomy even further. Certainly the successes of Uhuru, the discoveries helped us to secure the support for Einstein which eventually led to AXAF/ Chandra. It helped that with Uhuru, we were up to 350 sources, give or take. What was more exciting was that besides the binaries and the black holes, we discovered the extended emission from clusters of galaxies. Just before Uhuru launched, there were a couple of rocket flights that talked about X-rays from M87, Perseus, and 3C273. So maybe one galaxy, one cluster and one quasar. But the idea that there was extended emission associated with the clusters, and the hot gas in the clusters, and all that subsequently follows from that, was a major discovery from Uhuru. The fact that a couple more quasars plus a couple of dozen Seyfert galaxies were X-ray sources Ė Ariel 5 and Uhuru made those discoveries. The importance of those sources is that they relate to the All Sky Background, explaining the background as discrete sources actually. It all leads to Einstein and that leads on to AXAF. It was a wonderful time to be doing science. There were discoveries sometimes literally daily. Being part of thatĖ mean, we were kids. When Uhuru launched I was twenty-eight years old. To be in the middle of that; to have that kind of responsibility. Of course, I was too young to realize the importance of it or I would have worried about the responsibility.
McCray:How much did the mission cost?
Tananbaum:It was, I think the whole satellite, meaning the spacecraft and the experiment and operations, about $20 million dollars; of that order. It was significant at that point in time.
McCray:You left AS&E in 1973 and came here to SAO. Was it the Center for Astrophysics?
Tananbaum:Yes. The Center for Astrophysics just formed a few months or up to a year before. I think George Field had agreed to become the Director of the Harvard College Observatory and the Smithsonian Astrophysical Observatory. So the idea of the CfA was probably there for about a year. But George had just started and I think when we came July 1st of 1973, it was about then that the CfA really became the CfA as a real organization comprised of the other two pieces.
McCray:What brought aboutÖ?
Why did we come or how did it come about? I think this was something which was really totally driven by Riccardo and Riccardoís situation. There was probably a mix of both professional and personal considerations. The professional I know, of course, more about. The situation was that we had started studies on the HEAO, High Energy Astronomy Observatory program. And at that point I think there were going to be four HEAO missions in which a focusing X-ray telescope would have been the third. It was called HEAO-C. Again, A, B, C. A very original labeling system. That had been brought about, in part, in a response to a proposal that Giacconi and Leon Van Speybroeck, who was really already quickly recognized and accepted as the worldís expert on X-ray optics in telescopes, had put together with a larger team. A group of them had put a proposal together in 1968, í69, í70 time frame. So about the time we were busy getting Uhuru ready to go, they were thinking about an X-ray telescope beyond those that had been used up until then to look at the Sun. The new telescope was to look at the stars, the quasars, and so on, and so forth.
Riccardo was already envisioning that such a program probably would need to have some kind of a general observer component where it wouldnít just be for those who had proposed it and built it to be the only ones to use it. That there needed to be some kind of an opportunity for the community at large to use it. And he envisioned there being some issues, perhaps, of trying to do that under the leadership of a group at an independent, private corporation as opposed to a university or some kind of a federally funded research center. That there was no simple reason that the group couldnít have done that, if we had we not done it here. There is no reason that the group couldnít have done it at AS&E, but since there was a small fee or a profit associated, there was some concern. I mean, AURA charges a fee for things that they do running National Observatories. But thatís among universities as opposed to among stockholders. So I think Riccardo correctly perceived that there would be problems at the government and at the agencies and with the community accepting a general observer program that somehow came out of private industry. So he had it in his mind that there might be issues in terms of the direction in which he saw the field developing. This was a major reason for considering the move to CFA. There were other issues that dealt with the direction in which the Company (AS&E) might go. There had been a period of time where the President of the Company, Martin Annis had been, by reasons of personal choice, sort of more in the background and Riccardo had been running the Company day to day. And I think, of course, he enjoyed that, but then Martin Annis felt he was ready to take a more active role and resume being President of the Company.
Iím not sure Riccardo at that point was comfortable with the idea of taking a bit of a backseat in the sense that he had ideas about directions in which the Company might go. And so, there were probably personal reasons that made it attractive for him to move or at least be open to the possibilities about leaving. Then he was offered a Harvard professorship as well as a Smithsonian civil service position, and probably was attracted by the prestige that comes with being a Harvard professor. The tenure and the independence associated with that, the opportunity to do new things in an environment where students would be interested in working with members of the group would be certainly an interesting thing to be able to do. So I think there were a variety of personal and professional reasons. Then as part of that, Riccardo negotiated, is probably as good a word as any, that he would start with a group of about ten people here, most of whom were members of the group at American Science and Engineering. He was able to obtain four openings for federal/civil service tenured kind of positions. And funding out of the NASA money for Uhuru and the HEAO studies to support another half-dozen people. There was an administrator, a couple of secretaries, a newly minted graduate student who had just finished his Ph.D., Bill Forman. And I think about eight scientists altogether. I could count them off: Giacconi, Gursky, Gorenstein—see, the Three Gís are easy to remember—myself, VanSpeybroeck, Steve Murray, Ed Kellogg, Ethan Schreier and Bill Forman the new Ph.D. (Makes 9 scientists altogether).
McCray:So a pretty sizable team.
It was enough people that you could imagine that we could continue doing the things we were doing and maintaining sort of a lead role. We still had a significant amount of activity involved. The engineers, for example, that were working on the studies and eventually the construction of much of the HEAO, what became the Einstein part of the program, were still at AS&E. And many of the scientists who had worked on our different projects were also still at AS&E. I knew that you or we or someone would have to ask Riccardo if he had a particular perspective on how things might subsequently develop. The fact is, is that, of course, we know that months after month, if not more often, some scientist or other, and eventually some of the engineers announced their intent to leave AS&E and apply for a position at CfA. Now, we didnít always have a position at CfA. In fact, generally we didnít, but if the work needed to continue, then you were in this awkward situation where maybe you would want to decrease the subcontract at AS&E that was supporting that individual and move the money to CfA. CfA didnít have legal status, so they would be moved to SAO because it was the funding channel. This led to letters back and forth between Martin Annis and some of the senior CfA management team. Some of the people at CfA, George Field, the Director, and John Gregory who was his Deputy were concerned over whether we were, in a sense, raiding this expertise and stealing the people from AS&E.
There were no instances whatsoever of which I am aware where anybody was recruited. I mean, it was delightful that some outstanding people joined us. It was difficult in some instances to decide if it was really a plus to have certain other people move. But there was no intent to hurt the Company, or steal, or otherwise entice the people to come. But thatís also a very naÔve statement because, of course, where Riccardo was, and where some of the rest of us were, was where the real action was perceived to be, and where the real opportunities for the future were perceived to be in the field of X-ray astronomy. So it is not surprising that some people wanted to move and eventually large numbers of scientists and engineers whom weíd worked with at the various stages wanted to move as programs completed and new projects started. As openings were posted, lots of people Ė I wouldnít hazard a guess Ė I mean, I will hazard a guess, but it could be as many as fifty people eventually from AS&E, a combination of scientists, engineers, administrators, and others eventually joined us here over a several year period.
McCray:What was the effect, thinking back to Uhuru and then on with HEAO, of these missions on the X-ray astronomy community as it existed? How did it change it in terms of size and composition?
There are probably two sets of effects—probably fifty—but two that come to mind. One set of effects is what actually happened to the community and then probably the other set deals with the astronomy community in the larger sense. So the X-ray community, Iíd have to go back and look at which groups existed and were publishing say around the time of Uhuru was launched in 1970. There were groups at MIT; there was a group at American Science and Engineering; there was a group at the Naval Research Lab; there was a group at the University of California San Diego. There was a group at Lockheed (also in private industry) that was Fisherís group. They were doing rocket flights. Herb Friedmanís group at NRL and the MIT group and the AS&E group were probably the three largest.
There was a significant capability at the University of Leicester in England. Ken Pounds was the leader of that. Minora Oda, who would come back and forth from MIT and AS&E, was building a group in Japan which subsequently under his and Tanakaís leadership has launched four or five very successful satellites, Japanese satellites. Joachim Trumper was really still doing more High Energy stuff, but eventually got the group in Germany to the point where they built the very successful ROSAT satellite. So the groups though that existed in 1970 that were the forces in X-ray astronomy, if you consider our group here at CfA to still be the AS&E group, you could argue that we still do regular X-ray astronomy any more. Because, of course, there was a continuity, but the AS&E group doesnít exist. The Lockheed group might. They do solar astronomy and solar X-ray astronomy, but they donít exist. The San Diego group is still there. There was a big group at Goddard Space Flight Center. I forgot about them. And for the AXSF/Chandra program a group at Marshall Space Flight Center. So Goddard, MIT, CfA. Those are probably the largest groups in the U. S. right now. So a couple flourished; a couple have disappeared all together. For the ones that didnít get into the satellite business, only flying a rocket once a year, wasnít enough to attract and retain people. It often wasnít the cutting edge of the science. So the satellite, in some sense, the satellites you could argue, started to drive a few of the instrumentation-oriented groups out.
McCray:I appreciate that the satellites were just much more expensive than the rockets.
Tananbaum:There werenít as many opportunities and the ones that didnít get a part of it, didnít win at least an instrument, if not a whole satellite, eventually struggled and tended to disappear. In terms of the community as a whole, having satellites instead of a 5-minute rocket with a strip chart of data is great. With a satellite, you could figure out what you wanted to look at, what science you wanted to do. You could plan it ahead. There was, starting with HEAO, some very substantial guest observer, general observer programs.
McCray:How did peopleís proposals get picked?
Tananbaum:Peer review. You would write proposals starting with Einstein. Fred Seward was actually hired by us from the Livermore group, which was another group that was very successful early on, but eventually went away. We hired Fred, in part, because he had a very, very strong reputation of being very independent and fair. And we didnít want the community to say, ďOh, well theyíre running a general observer program, but theyíre just picking their friends to get data.Ē And so, I tell you, in the very beginning, Iím not sure what was done Ė whether people were just encouraged to apply and maybe the over subscription wasnít so great. Anyway, eventually we did go to a peer review on HEAO. We also had peer review for the U. S. part of the program on ROSAT and ASCA the Japanese satellite, and now with Chandra. People wrote proposals and the best proposals as judged by a panel of peers got picked for observing. Thatís fairly comparable, I think, with what ground-based optical and radio astronomers do now, what Hubble does, what the physics community does for a lot of the things that go on there. So Fred was brought in, though, to make sure that it (the Einstein Guest Observer Program) was run the way the community would want. He was tough. I mean, he didnít hesitate to tell Riccardo or tell me that such-and-such needed to be done because that was the right way to do it. I guess we were all successful. There were probably a few instances where—and Iím laughing now—we wished Fred wasnít quite so tough. He did what he was asked to do.
McCray:In terms of, what?
Tananbaum:In terms of how the time was going to be given out. Maybe we were interested in doing something and somebody on the outside was and Fred said, ďWell, the outside person should get to do it.Ē And weíd say, ďWhy them and not us?Ē But then Fred would point out how much data we were already getting and end the debate, usually in favor of the Guest Observer. Guest Observers ended up with 25% of the overall observing time on Einstein; more than 400 proposals were selected and observed, over 2 plus years.
McCray:When you came to CfA as it was starting or SAO—
Tananbaum:Well, SAO was already here and so was HCO. So it was definitely CfA that was starting.
McCray:When you came here, then, what tools did you see yourself bringing to the astronomy community that existed already?
Tananbaum:Do you mean the community that existed here or the larger astronomy community?
Again, I think we could definitely do an analysis. I mean, coming here, I didnít know anything about the place and I had to make a decision. I was going to work for the Smithsonian. The only Harvard position that was available was for Riccardo. I was five years out of grad school. I didnít have any interest in going in an academic direction. I really, with the Uhuru experience, was loving the opportunity of building things and doing research. I couldnít have imagined any necessity for tenure anyway because we were really going to need money from NASA to do things and if we didnít have money, who would want to stay?
Again, I was all of thirty-one years old by now, right. So I had to choose between whether to accept the federal position that was offered or whatís called the trust fund, which means the support comes from the contracts and grants. Itís the first of the month (July 2002), Making it twenty-nine years where weíve been successful from an entrepreneurial or business sense to have enough funds to pay our staff and to do our work. So it turned out that the perception that my choosing one over the other of those options in 1973 wouldnít have made a big difference was probably true. Ultimately, I analyzed a lot of detailsĖ you know, if they had given free parking, one way or the other, I might have chosen on that basis. I mean, youíre talking to somebody who is a very deep thinker here, when it comes to big picture decisions in life. But in fact there was one benefit that I found. On the federal side if I had died unexpectedly after eighteen months, as long as I had worked eighteen months, there was nice annuity plan for my wife and now my two children. And so, there was a reason, but it wasnít a reason that it was intended for, but there was a reason for me to say being on the federal side would be best. At about the time that those discussions were going on, Martin Annis, the President at American Science and Engineering approached first Herb Gursky, who was the Chief Scientist under Riccardo and asked him if he would be interested in staying there and helping to keep the science going. And ultimately Martin Annis approached me and maybe one or two of the other people. I think Herb was offered a significant raise and a company car. It was actually, in those days, a pretty attractive situation. Herb decided that he didnít want to stay there.
He would rather come along with Riccardo to CfA. I was offered, I think, a five thousand dollar increase which, based on salaries in those days, was about a 22% to 25% percent jump in pay. I remember coming back and saying to the CfA that I wanted five thousand dollars more, which they did offer which really means Riccardo must have okayed it. So hopefully not too many people here knew, because I came in, for a relatively young person, at a pretty senior level. What we brought here, though from our perspective was a certain, we used to laugh and call it high energy. But there was a certain enthusiasm, a certain energy which we brought. A certain excitement from some of the discoveries that were going on. A certain ability to work with some of the engineers and build things. Also some of the money that comes along pays the rent and some of the overhead expenses as things grow over a period of time. So I think there was a certain, a polite term might be, liveliness. A certain set of characteristics. I think that there were probably those who said, ďGee, they got a car, and it runs fast, and if you happen to step in front of it, theyíll drive it right over you.Ē Was it true? Who knows.
McCray:Did you notice that in terms how this influx of High Energy people, as youíre describing, literally and figuratively arriving here, how it affected the overall program at CfA?
Tananbaum:Well, there was plenty of space originally. But at a certain point, space became a problem.
McCray:This building was here?
Tananbaum:Yes. (Leo) Goldberg got the Perkin building. This building (B - in which my office is located) was built by (Fred) Whipple in the late 1950s, tied to the satellite tracking stuff. That was money that Whipple raised. This office or part of this office is part of what was once Fred Whippleís office. That wall behind you didnít exist and the little office to that side was his back office. In the afternoons when he wanted to do research, he would go in the back office, close the door, and instruct his secretary if anybody asked if Dr. Whipple was in the office to essentially say no because his main office was empty. This allowed him time to get away from the bureaucratic entanglements and do research.
McCray:So the satellite tracking program then was based here?
Tananbaum:Here at the Smithsonian. Well, parts of this building. I donít know because it was already, by then, it was started in the late 1950s and by 1973 when we came in, it was winding down. So I think to the extent that we brought in contracts and we needed the contracts people to negotiate the contracts or to proof and sign them. To the extent that when we built stuff, we needed purchasing to go and get things. To the extent that we were used to working with engineers in a certain way and a certain style. Iím sure we were, maybe disruptive is as good a word as any. I think we were able to compromise in instances. We were able to negotiate. We were able to share certain services and so on. But Iím sure we were quite disruptive in terms of people who had a fairly set way of doing things and a certain pace in which things were going on. We paid for what we needed, but we needed a lot. And when we were able and willing to pay for it, we expected to have the support provided and generally it was. Obviously, we have been quite successful here. I think there were probably people who, if they had to vote, wouldnít have voted for us to come. They might not still vote for us today.
McCray:On logistical grounds or intellectual grounds?
Tananbaum:I think intellectually, if they were able to just be unbiased, on intellectual grounds, I think what weíve done has been exciting and we contributed to the field of astronomy in a very substantial way. So I donít think it would have been intellectual. I think it would have been logistical. It would have been just the demands that we put on the whole place, the system. Of course, as I said, weíve contributed a lot to it. They used to have, and they still do, they have a set of Associate Directors. Originally there were eight different divisions into which the science activities were placed and divided. A few years in, two were merged and so there were seven. So Riccardo was one of seven Associate Directors. His area of responsibility was High Energy Astrophysics. Sometimes there were turf wars between those Associate Directors with the Director. I mean, it could have been through the 1970s that there were some contentious times. By the time we got into the 1980s, Riccardo had moved on. Herb Gursky had gone to NRL to replace Friedman as Superintendent of the Space Science Division there. George Field stepped down as the director. Eventually Irwin Shapiro was hired to replace him around 1982 or 1983.
McCray:Yes. Right after he did the decadal survey, I think he stepped down.
Tananbaum:So some of the Associate Directors left or stepped down; some younger people were promoted. I became the Associate Director when Riccardo left so I was responsible for High Energy. I guess after a certain amount of time, with the players changing a little bit, people were here a decade. We were used to being here after a decade. So the health of the place overall becomes an important element as well as the health of your own activity. So there is a certain amount of interest in the place overall. In terms of the impacts I have been trying to describe, some of the impacts on the place, probably took a decade for everything to sort of settle in. I wouldnít call it as happening in a comfortable kind of way, but in a way in which at least there werenít battle lines regularly drawn. There was a certain sense of direction. I mean, we could in fact support certain initiatives that might help a different part of the place or vice versa. There was a broader kind of perspective of the place.
McCray:You became Associate Director in 1981. I have some questions about Einstein, but we can come back to that. When you were an Associate Director, how many people were there in the High Energy division?
Tananbaum:I donít know. By that time frame I would say we probably numbered fifty to seventy-five people. One of the things was that we had hired a number of the engineers that had worked on Einstein and they moved over from AS&E. We had hired some new engineers a year or two before the launch and we were starting studies for what became AXAF and were using some of that expertise from Einstein for AXAF. In the early 1980s, a centralized engineering group was formed, it was called Central Engineering. So all the engineers were put into a separate department and reported to an engineering manager and could be tapped on as a sort of a matrix to provide support to all of the divisions.
McCray:Whereas before engineers would be in one division or another?
Tananbaum:Whichever division had the funds and the needs and hired them. So in a sense, overnight we lost about fifteen of the engineers and then had to pay for to have them back on our projects. That is the way some overhead departments work. So we were paying, we might have been paying seventy-five salaries and counting overhead, it was a little bit more. Today, again it is hard to do an exact count, but we probably have on the order of three hundred people.
Tananbaum:High Energy which accounts for Chandra and non-Chandra parts.
McCray:When you first became Associate Director, just ballpark, how did you roughly divide staff between scientists and engineers? Can you give a sense of the ratio?
Tananbaum:There probably were twenty-five or thirty Ph.D. scientists. Probably half-a-dozen programmers and another half-dozen to a dozen people with a Bachelorís Degree providing technical support to the Ph.D.ís, four or five secretaries, a couple of administrative people. Probably in the division, fifteen to twenty people who were programmer managers and engineers. So probably about seventy-five overall.
McCray:Again as Associate Director or maybe just when you arrived here, I am curious how particular areas of research were emphasized and how those changed over time. Can you give me a sense?
Tananbaum:Do you mean within High Energy?
McCray:Yes. Within High Energy.
Well, when we arrived here, we had the Uhuru data. We were still working very hard on the details of sources like Cen X-3 and Cyg X-1 so binary sources with neutron stars and black holes. We were working hard on the clusters. We were just starting the work to show that some of the Seyfert galaxies were detectable in the accumulated data set. We were still making the third, and ultimately the fourth catalog of Uhuru sources. We were refining the source positions. Deciding with greater rigor which sources were real. A few sources went away. A couple of dozen more sources were added to the catalogs. The positions, certainly, were refined. We had a telescope by the late 1970s, so even before I became Associate Director, starting with Einstein which was launched in November of 1978. We will talk more about HEAO and Einstein because there were four HEAOs that became three. And the third one, the telescope, became the second one and that launch was really Einstein. So Einstein launched, almost from the get-go, had the ability to detect X-ray emissions from the corona of a relatively normal star. One or two stars had been possibly seen as X-ray sources somewhat similar to the sun, but maybe 10 or 100 times more luminous in their X-ray output.
But now we were routinely detecting all classes of stars and able to conclude that magnetically driven dynamo processes were the energy source for the X-rays and not acoustical waves, which was the previously accepted theory. Thatís driven by the luminosity in the X-rays relative to the other wavelengths and the fact that the stars that are brightest in X-rays arenít the ones that the acoustic models were predicting. Einstein also led to the ability to detect all sorts of extra-galactic sources. We had a number of clusters where we could see they were extended X-ray sources with Uhuru, but now we were actually able to make these images with a couple of bright central galaxies and diffuse gas and seeing gas fall off the radius and measure its temperature. In the case of quasars and Seyfert galaxies, we were able to image enough what we call active galactic nuclei, the combination of Seyferts and quasars, to say that they probably were the origin of most of the background seen wherever you looked on the sky. So moving into this imaging age, there was a solid state silicon spectrometer that Goddard had built for Einstein and a Bragg Crystal Spectrometer built by MIT. So for some of the sources, we were actually seeing lines in the spectra, the emission features. A few lines had been seen in rocket flights and with earlier satellites, but we were getting the idea that spectroscopy could be an important part of the field, which it certainly has become with AXAF/Chandra. We had a grating on Einstein that had a very, very modest throughput equivalent of a few square centimeters of the effective area.
So people looked at Sco X-1 and a few of the bright galactic sources that could show some absorption features. But I donít think a tremendous amount of breakthrough science was achieved with the gratings on Einstein. It was much, much different science from Uhura, however. So there was a shift. Certainly the mix moved from the bright galactic sources to the fainter and more often than not extra-galactic sources. People again began to think more about X-rays from various classes of objects. Stellar corona, of course, are galactic sources so producing X-rays by different process. Einstein provided super images of supernova remnants, my goodness. The details of these shells and their possible point sources in the interior were fantastic for the time. I think that it was a key point where Riccardo had the vision of Einstein to open it up to observers all across the board in astronomy. Other X-ray astronomers and other astronomers in general. People began to realize that you could start to make meaningful contributions to the mainstream of astronomy with the Einstein. With Uhuru, it was these very weird and different objects with the basic mystery of what they were. And then, it was obviously clear that they had things to say about stellar evolution and about mass transfer, and the properties of collapsed stars.
Tananbaum:And with Uhura we detected large quantities of hot gas in the extended X-ray emission from electors of galaxies. There was a much or more likely significantly more mass in the hot gas than in the stars in the galaxies in the clusters themselves. So from the point of view of the balance of the material, the hot gas or what you could learn from the X-rays, was clearly important. So I think with Uhuru, that was beginning to impinge on peopleís awareness. And a few people even did collaborative studies.
McCray:Tell me about some of those collaborative studies.
Tananbaum:Well, it would be the optical identification of the X-ray sources or a theoretical modeling of the hot gas in the clusters.
McCray:Now, would that all be done between different divisions here or were these collaborations—?
Some within different divisions here and others with people around the world. Both. But then with Einstein, the telescope, the images, the ability to work in almost any area of astronomy then led to lots—hundreds—of people writing proposals interested in using the telescope time through the general observer program. We definitely were interested in many more of the same kinds of objects. And then that changed even more dramatically with Chandra because with Chandraís resolution at a half-arc second while not as good as Hubble or not as good as you can do with a radio interferometer, is of the same order of the angular resolution you can do with a good ground-based telescope using the optical or the radio. So with the quality of the images and because of the sensitivity increase that comes with that, we can study almost anything in our galaxy. We can look at individual supernova remnants or individual globular clusters in M31 or other nearby galaxies. We can detect individual X-ray sources in galaxies out as far as the Virgo cluster or even twice as far away. So you begin to get into the populations showing where the sources are distributed in those galaxies, what do the populations of stars that led to those X-ray sources look like. Then there is the science with the clusters, with the quasars, in snapshots of just several thousand seconds with Chandra, we detected three red shift 6 quasars that the Sloan Digital Survey had just discovered, which were the three most distant at the time. But I think what is really impressive is that now weíre getting nearly a thousand proposals a year.
A couple of thousand different people involved as Co-Iís and PIís of those proposals. Peer review takes a hundred astronomers to review the proposals. The satellite is up. Itís long-lived whereas Einstein lasted a couple of years. Chandra has now exceeded that with reasonable expectations of lasting a decade, hopefully longer. No guarantees because itís in space and we canít get at it and fix anything that breaks, but itís working beautifully. So I think itís this progression from Uhuru to Einstein to Chandra—and of course, many of the missions in between. But those are the ones Iíve been primarily involved in—you can see the larger part of the astronomy community getting progressively more interested in these capabilities and astronomers who study a particular object or set of objects. If you want to study an AGN, you really need to know about its infrared properties, its optical, its X-rays, its gamma rays, radio. To the extent that you can get high quality data across the electromagnetic spectrum, it is pretty important for developing, letís say, a viable model to describe how such a system works.
McCray:So you need a multi wavelength picture?
Tananbaum:Yes. And so Chandra for some people just incorporates and fits into a piece of what they are doing. For other people, discoveries that they find with Chandra may drive all sorts of things that they will then do with other telescopes and other wavelengths.
McCray:In your own research, were you having much interaction with observations being made at other wavelengths? Or is that not something that you have gotten into?
Well, with Uhuru in the beginning we were doing primarily X-ray studies, but it was very important to find the optical counterparts. We didnít have specific individual collaborators. We would put the positions out where we had an interesting source or an important source. When the results would come back, we would try to assimilate the new information and incorporate what we were learning. It was very hard, at first, for the people who found the optical counterpart to Cyg X-1 because it was an eight or ninth magnitude B0super giant star. And based on Sco X-1, we expected a faint blue—kind of faint, relatively—Sco X was13th magnitude. Scaling the counterpart for Cyg X-1 should have been a 16th or a 17th magnitude. That was the case, until we understood the binaries and that the mass transfer and the object unto which it was falling dominated the X-ray and the optical might be dominated by the donor star. And the donor star could be a variety of different kinds of stars.
In the case of Sco X-1, the blue light youíre seeing is probably not the primary star at all or the donor star. Itís probably some of the optical light from the accretion disc and maybe some of the light from the donor star. So Sco X-1 was a lousy prototype. The other sources werenít scaled from it. And so the suggestion was made, ďGee, thereís this bright star within the arc minute error box that Uhuru had determined. Thatís the brightest star. Why wouldnít it be a candidate?Ē ďWell, based on Sco X-1, thereís no way such a star could be the X-ray sourceĒ. So there was certainly a lot of interaction, and we were dependent on each other for putting the picture together and getting the information. With Einstein, we were very dependent on good samples of quasars. My own research involved a lot of the X-ray properties of quasars. In one case, we worked with Maarten Schmidt and Richard Green using a subset of whatís called a complete sample. Which means over a certain piece of the sky down to a certain optical magnitude limit, they have done a comprehensive job and found just about all the quasars that met the properties for belonging to their sample. There were 120 and we were able to look at about 60 of them as a random subset, just about half of their sample with Einstein. We detected X-rays from more than 50. Our work involved correlating the X-ray and the radio and the optical properties, extrapolating those data to say: if all quasars behave like this, would they produce enough X-rays to explain the X-ray background? Of course, we were looking at the brighter end, but you could say if they were further away, and they behave the same, or if they were further away and they evolved with time, how much you would be able to see. What could you say about the evolution? I worked with Gianni Zamorani from Italy and Yoram Avni from Israel on these projects. Gianni mostly on the observing part of the program.
McCray:Who is the second person?
Yoram Avni. Yoram was a master of innovative statistical techniques and was able to reinvent—although we didnít know it was a reinvention at the time—whatís known in some circles as Survivorís Analysis. If you give a bunch of people a particular medication and so many die per year, youíre trying to make predictions about the whole efficacy of the medicine and the life expectancies of those that survive based on those who died. So in the X-rays, we would look at a sample of optical quasars. We would detect X-rays from some, but others were below the threshold of Einstein and werenít detected. Now, there is some information there. The fact that we knew that the X-ray emission was less than a certain amount, didnít tell you what the exact amount was, but it told you a range that it was and it also excluded the range that it no longer could be. Yoram developed statistical techniques to use the information. Assume you have 50 or 60 objects and you have detected half, and half you didnít detect. So if you took the half you detected and said the average brightness is such-and-such, youíre obviously getting a biased answer because you are taking the brightest half and you are ignoring the faintest half, or the relatively faintest.
So he invented a mathematical system called Detections and Bounds. The Detections being the things we saw and the Bounds being the things for which we had limits. He developed the whole mathematics and methodology of it. About a half-dozen years later or more, Eric Feigelson, because he was interested in the methodologies, discovered that this whole independent literature existed in the biological medical sciences called Survivorís Analysis. And many of the same techniques, not all, but many of the same techniques were there. In a sense, they were a virtual world to us because if we didnít know about their existence, we couldnít tap into that to help analyze our problems. But there was this whole statistical revolution. Getting people comfortable using the technique where they had to incorporate the information from what was seen and also what wasnít detected or was unseen. So I worked with some of the optical people: Schmidt, and Green, and with a Zamaroni, and Avni. I think Giani Zamaroniís background was originally radio astronomy. So it was a group of collaborators on a set of projects.
McCray:Would there be particular people that you tended to collaborate more with than others if you were looking at your publication list? I didnít see a pattern.
Tananbaum:Sure. In the first half-dozen years, there was a period in which for a year or two after launch of Uhuru, Giacconi, Gursky, Kellogg, and Tananbaum had their names on all of the Uhuru papers. They were responsible for building it, for operating, for writing the software, and analyzed the data. So if you did an analysis, those names would show up more often because of that rule that was there for a couple of years. But the reality is that even within those days, with Riccardo and with Ethan Schreier, there was a lot of work on the binaries Cen X-3, Hercules X-1, and Cyg X-1 among others. There were other co-authors on all those papers, but in the earliest block of research, which I would consider sort of the first phase of my research career, the emphasis would have been with Riccardo and Ethan. The next stage with Einstein, which was primarily the quasars, mostly with Zamaroni and Avni. I would have to stop and think. There were individual projects I worked on. There was an extended project in the early 1990s also with quasars with Belinda Wilkes, and I worked on some other projects ver the years with Wallace Tucker who is a theoretician. There are a number of different people. With Chandra, I have one object for which I have some observing time scheduled at the end of the summer 2002. It will be the first science after almost three years with Chandra that I would consider my own, personal interest. Separately, I am responsible for 5% percent of the time, which is called the Directorís Discretionary. So, I have to certainly stay alert and aware of whatís exciting, whatís interesting, whatís important, whatís happening. I stay very involved in the press releases and the Space Science Updates of whatís newsworthy and what is a significant discovery. I have been giving a significant number of talks on highlights from Chandra, but I havenít actually done my own research; I have been reporting on other peopleís research.
McCray:Is that frustrating?
Not really. I have had a number of wonderful invitations and opportunities to go places. The results are spectacular. Being able to, in a sense, synthesize some of the different things and try to give some kind of a semi-coherent presentation of some of the highlights, which of course, is impossible now. Three years, almost, into the mission. I was invited to Princeton. I was there in May for a series of lectures named after Lyman Spitzer, the Spitzer Lectures. Lyman, of course, had been the founding father, the pioneer of what became the Hubble Space Telescope. So it was really an honor to be invited to give this series of lectures named for him. And it was a hell of a lot of work. I had a general talk that I could update that Iíve given every few months; sometimes more often, sometimes less often about what Chandra is, how it represents advances over what we had previously, and what is exciting. Supernova images, clusters, Seyferts, quasars, some spectra, and some images. Stuff that even if I added some new stuff, I can get comfortable fairly quickly in giving a lecture. But I had to give four other lectures in addition and I didnít have a body of my own independent research with Chandra upon which to draw. So I tried to think about what would be the most interesting and also what I might enjoy learning a little bit more about as part of the process. I gave a lecture on supernova remnants and pulsars.
I gave a lecture on galaxies and clusters of galaxies. I gave a lecture on surveys with Chandra and active galaxies and quasars observed with Chandra. So this one was mixed because what you find in the surveys are AGNs and quasars. So it was looked at from a couple of perspectives. And I gave a fourth lecture on what we call an X-ray roadmap. It started with a little bit about the instrumentation on Chandra. I talked about the Constellation X-emission, which is the next significant X-ray mission weíre working on already. And then an X-ray interferometer, which may be beyond my professional lifetime. If things went well, we might build it around 2020; nearly twenty years down the road. Even if Iím still walking the halls, I donít think theyíll let me hold a screwdriver to any of the hardware because the hand will probably have a few shakes to it. So those talks all drew on different topics. For one on surveys I had been pretty actively involved on a project called CHAMP which stands for Chandra Multi-wavelength Project.
McCray:What is that?
Tananbaum:Anytime an X-ray telescope looks at a particular target, there is a patch of sky that is observed. With the CCD imager, which is the most commonly used instrument on Chandra, that patch of sky covers an area about one-quarter of the size of the full moon. Itís about a 16 arc-minute by 16 arc-minute patch. And what we see depends on the duration of the exposure, the exposure time. If we look for a few thousand seconds, there will be a few sources. If we look for a day, there will be 50 or 100 sources. So as Chandra has pointed to these different places in the sky, weíre picking up X-ray sources all the time. But what those X-ray sources might be is not so easy to determine. I mean, the target that you looked at, if the X-ray source shows up, you know itís that particular galaxy or cluster of galaxies or quasar or star. The others, if you run an algorithm across the field and detect all of the X-ray sources that show up and get their positions, you can cross correlate them with existing catalogs of stars or galaxies. Generally the exposures on AXAF are sufficiently sensitive that most of the sources will have optical counterparts fainter than the existing digitization of the Palomar Sky Survey, for example. CHAMP, the optical follow up, I think the average or median optical counterpart might be about 23rd or 24th magnitude in the V band. Some of them are getting down to the Chandra Deep Survey equivalents in which case the optical counterparts are 26th or 27th magnitude. So youíre beginning to get to a pretty good sized optical telescope to find the little speck of light at the right position and then you need an even bigger telescope if youíre going to try to do spectroscopy.
McCray:Which telescopes on the ground have you been using?
Tananbaum:Weíve been using a combination of some of the national telescopes, the 4-meter at Kitt Peak and the 4-meter at Cerro Tololo in Chile. Weíve used 1- or 2-meter class too Ė I think it is actually a 1.2-meter Smithsonian SAO telescope.
McCray:Is this the one at Mount Hopkinsí?
At Hopkins, yes. And we use one at Kitt Peak which has a similar size. So we have used those for the shorter exposures where the counterpart should be brighter objects. We have done limited follow up in the radio and infrared. Itís been mostly optical so far. The idea is that because the relative Chandra positions are really accurate to better than an arc-second, if you identify a few objects on the field, the field is really registered precisely from the X-ray to the optical. You can actually put the X-ray sources and youíre sure the location is at a precise place. Then unless thereís a little elasticity in your image, your whole image is registered. And if you get an object, if itís more than an arc-second off, it probably isnít the object except when you get out to the corners of the field and then the positions arenít quite as good. So we are in the process. We have picked about 150 fields over the first two years of Chandra observations. The field canít be filled with a big galaxy that fills the whole field, or a cluster that fills the whole field, or a cluster of stars. It has to have a region around the target which you exclude from the analysis because that obviously, we were biased when you pointed there to see what you would see and what you care about is this other random or fair sample of X-ray sources. What weíre doing is weíre piggybacking. Instead of saying we need more Chandra time for some of these surveys and that itís important to look, we use these15 fields and I think weíre covering something like 14 square degrees over the two years.
And if you wanted to look at 14 square degrees, 150 fields, it would take several months of Chandra observing time. So what weíre saying is those sources have been found. That time, we really wonít have to spend it again. And now we should go and follow up in the optical and find out what we can. And now, of course, weíre using optical telescopes as the lever arm. So what we hope to do is by looking deep enough in the optical, get the identification. We look in 3 optical colors or bands. From the optical colors and the X-ray intensity, we can pretty much tell if itís an ordinary galaxy, if itís a quasar, if itís a star. And so, we can do a rough classification and we can do whatís called a photometric redshift so we can get an estimate of the distance of the object once we have some idea what kind of an object it is by its colors in these different bands. Weíre going to do spectroscopy, which means even bigger telescopes. It means the Gemini 8-meter telescopes and if we can get some Keck time, 10-m telescopes for which weíve done a little. It also means the 6.5 m converted MMT at Mount Hopkins.
McCray:So this collaboration to do the Chandra project, is that strictly within this organization? Or is it a collaboration with others?
Tananbaum:No. There are astronomers in Chile, at Kitt Peak, at Ohio State and Michigan. We basically get people who either want to contribute to the analysis of the data or help get access to the telescope time by writing better observing proposals for the telescope time. So Belinda Wilkes is leading it and Paul Green is the optical team lead. Theyíre in-house as part of the Chandra team.
McCray:How many people overall?
Tananbaum:I would say there could be thirty-five to forty people involved.
McCray:Research astronomers or—?
Tananbaum:Yes. Research astronomers working on it in some aspect or other.
McCray:How do you coordinate all these different people?
I donít have to. Thatís what Belinda and Paul do now. Isnít it great that either some of my organizational skills or some of my disasters have provided enough lessons? No. Iím just being facetious. We use the Web. Itís incredibly well-coordinated by Belinda and Paul. We use the email and the Web and we have regular meetings where as many of the team members who can come in-house and from the outside. But thereís a CHAMP Web page thatís internally accessed. So they post lists of positions. They put a sample, ďHereís analysis we did on the X-ray field. Here are the sources we found. Will some other people look at it and see if you think weíre missing sources or if thereís a problem with the positions. And hereís the optical image for the same field.Ē So we put the data through what we call standard pipelines. We have a couple—I call them kids—of the younger people in it. Of course, thereís no law that says there has to be younger people, but typically some of the people with a Bachelorís degree will run these pipelines and will eventually process all 150 fields. Process them uniformly to find the X-ray sources. Reduce the optical data with a standard set of tools to get rid of the cosmic ray background hits and look for the real images. Weíll have, by the way, a catalog of additional objects for which we have the optical magnitudes, possibly a classification, but no X-ray source. So thereís a reverse to the CHAMP or complimentary in some sense for which those X-ray upper limits may be very interesting for some of the objects. If somebody drafts a section of a paper, they put it up on the Web and everybody can comment.
Those who feel theyíve contributed will be co-authors. Those who want to can contribute to another paper in another area. Thereís quite a bit of flexibility. We had one paper where we published early because one of the grad students as part of this thesis, had found a quasar with redshift just under 5. And at the time was the highest redshift quasar that was originally identified through X-ray means. So a little paper was written just on that. I did not co-author that paper even though I read it, and critiqued it, and told him I thought he needed to do certain things in a better statistical way. He then explained what he had done, and he improved one or two things, and the others had been done just fine. I didnít want other than an acknowledgment that he had gotten some comments from me, and I didnít feel I merited being co-author. I donít need to lengthen my publications list at this point. So he might have six or eight or ten co-authors. Iím not quite sure. But those people, it was John Silvermanís paper, those people who contributed were the co-authors.
McCray:How do people feel about being part of such large collaborations?
Tananbaum:People are of several minds. Those who didnít want to probably chose not to join it. So the call went out and said, ďThis is what weíre going to do. Whoís interested in participating?Ē Anybody who signed up was aware that it would probably be a pretty big effort. You know, keeping track of several thousand sources; it is not the Sloan Digital Survey, which is millions of sources. But it is still a pretty daunting undertaking.
McCray:How do theoreticians fit into this picture for CHAMP? Because what you have been describing so far has—sounds fairly observational driven.
Yes. Right now, it probably is. If eventually, I mean, one of the things that we will be able to do with CHAMP is weíll be able to determine some key properties of quasars. There is speculation out there based on a limited sample from ROSAT that at redshifts of 3 or greater, the numbers of quasars selected through the X-ray technique, those discovered by X-ray emission are significantly more than what you would expect based on optical or radio samples. So it comes to the question, how many quasars are there and how bright are they? How many of them were there at earlier times at higher redshifts? So right now, the Chandra X-ray data donít tend to give us any conclusion better than or different from the optical data. But theyíre tentative, with limited numbers which will improve as CHAMP progresses. Well letís say with John Silvermanís project and other peopleís involvement that we find that, gee, there are significant numbers of these X-ray sources, which has been suggested to be the case for redshift greater than 3. Is the number really different from the optical?
And if itís different, what are the energy mechanisms which make the optical only begin to show up in significant brightness or luminosity in significant numbers at redshifts 2 to 3? Whereas in that case the X-ray is already very luminous at redshifts 3 to 5. Itís possibly, by the way, that because of the density of the accrediting material, or the taurus or the disc that surrounded the black hole, or the overall geometry system, or the mass flow properties, it is possible to put a larger fraction of the energy into the X-rays. Perhaps the optical is somehow being obscured and the X-ray isnít for quasars at high redshifts. The damn things are very dusty early on. Gas will absorb low energy X-rays; dust is less effective. But in the optical, dust will obscure or absorb and the gas is less relevant.
McCray:So itís okay with infrared then.
Tananbaum:Exactly. What might happen is that these things that show up on X-rays and are fairly dim in the optical might, in turn then, still be very bright in the infrared. So theoreticians would get involved in trying to develop the models to try to describe that. Iím not sure. Maybe this is a particular project which is more observational and less theoretical.
McCray:How is it funded, this project?
CHAMP is funded maybe in two to three ways. The scientists who are working on Chandra as part of the Chandra X-ray Center, that includes Belinda, Paul Green, some of the others have a fraction of their time available for them to do research. This was part of the proposal we submitted to NASA when we competed for the Chandra X-ray Center in 1990. We said that, on average, 30% to 40% percent of the peopleís time would be available for independent research. That is important if you are going to attract a first class staff to a science center and institute. You want people who actually do real research and the only way youíre going to attract and retain them is by having a research opportunity. Itís also important to have expertise with staff using the software you are going to put out there, if for helping people answer problems with a help desk for issues that they have with the satellite or the software for reviewing proposals, for anything you are going to do. If you donít do anything with any of the data, youíre not going to be a very good resource to help the rest of the community to use the instrument and telescope. So research opportunity is both to attract the people and so they have and keep the skills they need to be useful in service to the rest of the community. Sometimes you can have a situation where people can fund their research time. If they write a proposal and it is selected then they can do research. But weíre not big enough at the Chandra X-ray Center to depend on this. If all our schedulers were successful in a given year, and they all went to do research 100% of the time, well, who would do the scheduling that year?
Or if the calibration team was all successful and next year we donít have any people to work on calibration. So our thinking was that it would be best to say that people can have up to this amount for research. If they want to do less research, theyíre certainly allowed. I donít put 30% or 40% percent into independent research. And each year we review with each individual their plan for the upcoming year and, have they had time to do research, and how have they done, and are there issues or problems? So for some of the CHAMP people, their basic salary is already covered. We wrote archive proposals for Chandra because the data that weíre analyzing do come out of the general set of Chandra proposals. So some of the research assistants that we want to hire, some of the people working on the specialized pipelines and software are getting their salary from this archive proposal. It has to go through peer review to compete with other Chandra proposals. We have been fairly successful up until now to obtain some of the support. We have applied to the National Science Foundation because we do use ground-based optical telescopes. To date, we havenít been successful to get NSF funding.
McCray:Any idea why?
Tananbaum:Itís very competitive. Iím sure they think weíre getting some money from NASA, from Chandra. Youíd have to look at comments of the panels that reviewed the proposals. Maybe they donít think itís particularly exciting although I donít really believe that. Itís just very competitive.
McCray:I was thinking about the data archive. Is the data that Chandra produces, is it done in the same way that Hubbleís is with the proprietary period and then made available to the community?
Tananbaum:Yes. The standard data rights for Chandra observations allow exclusive access to the person or team who proposed and received the observation for one year from the time we deliver the data, which is usually several days after the data are obtained. The data go through a standard processing and the results are placed in an electronic archive or library along with the raw data. The investigator is given an access code and has up to one year to follow through on his/her ideas behind the original proposal. After one year, the access is opened to any scientist (or other individual) interested in the data. Scientists with new ideas for the data, may also propose for modest NASA funding as ďarchival researchersĒ.
Tananbaum:Some investigators waive the right to proprietary time, but most donít. If theyíve applied for what we call Directorís Discretionary Time for an observation, they can get, at most, three months proprietary time because it didnít go through the same process.
McCray:Through the Time Allocation Committee?
Tananbaum:Right. So the feeling is that the ownership of the data should be either zero or considerably smaller. So, zero to three months. And a lot of people in that case donít request proprietary time; some still do. Archival data are public at the end of a year so the CHAMP is working to generally update with data already in the public domain. What CHAMP is doing thatís different is a more detailed and specialized, standardized analysis, although Chandra does a standard analysis on all data. CHAMP is carrying it a step further and then comparing the optical and the X-ray data in a very standard way.
McCray:Is there any thing that has to be done differently to reduce X-ray data that you donít do with infrared or optical data, or is it fairly straightforward?
Tananbaum:Sure. There are similarities; there are differences. On the ground, the telescope has guide stars it locks onto and it is assumed that it is stably pointing. Up in space, we slide over to a target. We have a star camera that locks on to stars, but we allow the thing to wobble and weave a bit during the time weíre taking the data. And we use the star images plus the gyroscopes to reconstruct where we were looking after the fact. And for the X-ray data, any time an individual X-ray photon is detected, we look at where the detector was pointing and where the telescope and the detectors together were pointing at that instance in time. And we put that in to a sky bin, and we rebuild the maps on the ground with this aspect correction that corrects for the pointing, and also a little bit of breathing within the whole observatory that is allowed for.
That means that the telescope, and the bench, and the instrument can, as temperature changes, they can expand or contract just slightly as a system or as a subsystem. We monitor with a set of relay lights. We actually have light emitting diodes that bounce from the detector off the mirror or off the periscope mounted to the mirror and into the star camera. So the image of the stars has these lights, and the X-ray image on the detector focal plane has that image related to a fairly rigid location relative to the lights, so we can reposition or correct the data if the thing is slightly expanding or contracting. And it does drifted. Over the first couple of years, it drifted about an arc second or two as one of the instruments basically relaxed over time. So we are able to monitor that and correct for it. That is a little bit different for Chandra. We count photon by photon whereas the optical/infrared/radio typically integrate.
They donít have to, but typically thatís how they do. So an optical telescope has to stay stable. If it integrates for thirty minutes, it has to stay on that target, exactly rigidly located, for thirty minutes. Hubble has a very severe pointing requirement stability because it integrates. But since we can figure out where we are, and we only integrate with the CCDs typically for 3 seconds, we only require to stay steady to a fraction of a pixel for 3 seconds. With our CCDs, the analysis of the image data from the CCDs because we count each photon, also allows us to get its energy precisely. The X-ray comes in and stops in the CCD, creates a little charge, the charge then spreads and gets captured into the proper set of checkerboards; the rows and the columns that are a corresponding position. For Chandra, events, sometimes all the charge goes into a single pixel and sometimes it can spread over as many as a half-dozen to 9. So we collect the charge—the region around the event—onboard we decide if thereís enough signal to make it worth sending it down to ground. If it is, we either send a 3x3 region or a 5x5 region of the CCD. And then we have to find a centroid on the ground and we have to put the data back together to give us the best energy information about the event. And then when we look at all the events that correspond to a source, we can make an X-ray spectrum. So that is different from what you would do with an optical telescope.
McCray:It sounds really different.
Tananbaum:Some of the things like source detection and estimating the background, some of the problems are certainly similar conceptually.
McCray:We kind of skipped over the Einstein satellite. Let me pause this for a second. [Tape cut] Do you prefer to call the current satellite AXAF or Chandra?
McCray:One of the questions I had which is something you mentioned before I went to lunch was how did the problems with Hubble affect the Chandra program?
Tananbaum:Good news, bad news. Through the 1980s, we had been recommended in 1980 by the decadal survey chaired by George Field. AXAF/Chandra was recommended—it was still AXAF then. See, I will go back and say AXAF—it was recommended as the highest priority large project for the 1980s, ground or space-based. But it was very difficult to get to the stage with NASA of getting the program approved and funded. We had OMB, the Presidentís budget, and NASA. We also have Congress. There are lots of hurdles that you have to go through. And there was a certain sense, well, we are already building this large telescope. It wasnít Hubble yet. It was Large Space Telescope; then it was Space Telescope. And it was running behind schedule. It was costing more. It was an incredibly challenging project. And we donít want to start another telescope. The original name, AXAF, thereís no ďTĒ, thereís no Telescope.
McCray:Facility versus Telescope.
Right. Because it was considered politically a liability to use the word ďtelescopeĒ. Not by me, by NASA management. But in the 1980s, a lot of other things were going on. We had inflation with large deficits, and the budget, and the buildup for the Defense Department for the Cold War. So economically, it was just difficult to get more money for the space projects. In 1986 we had the Challenger calamity, so the whole space program almost got put on hold for a couple of years. The people at the Office of Space Science and Applications at NASA headquarters found themselves with an ever increasing portfolio. Earth observations were becoming more important; atmospheric, oceanographic and satellite communications were of interest. So it wasnít just the traditional planetary astronomy/astrophysics/space physics kinds of programs. But the pie wasnít enough to match the large number of new fields that were of interest. So it was tremendously challenging to get the support and get over—even with the Academyís endorsement—to get the endorsement to build the project. The problems with the Hubble optics were not even known at this point. The thing that did happen, now sticking strictly to your question in regard to Hubble, the problem with the Hubble optics, of course, was embarrassing for some of the people involved, but more than anything it was incredibly disappointing. People had worked for so long and had these great expectations. When they figured out with the costar, how to correct for the optics, it was a tremendous victory. One of the things that people will ask now sometimes, ďChandra, Iíve heard a little bit about it, but, you know, why doesnít it get the publicity that Hubble gets? Why isnít it on the tip of everybodyís tongue?Ē We give explanations in that, you know, we collect fewer photons.
A lot of our images are spectacular, but perhaps with fewer photons. Theyíre not things that, they are harder to colorize and to project. X-rays are something people donít know about as much. But the reality is that the single most important thing is that Hubble had the worst of times followed by the best of times. So, you know, Letterman would point his garage door opener to the sky and make some joke about fixing Hubble with garage doors or something. The late-night guys made a living for better than a year. And when NASA did this marvelous servicing mission a lot of people, probably more than since probably man walked on the moon, turned on and tuned in, in the middle of the night to watch the Hubble servicing in progress. So it had this publicity. People had an awareness of it for bad reasons and then for a wonderfully good reason. And then it has returned just marvelous, marvelous data. So, it set an incredibly high standard in that sense of public awareness. One of the things that happened after the problem with the Hubble optics is that I think the managers and the engineers never wanted a situation to arise in the future where they would have the science community saying, ďWell, you never asked us.Ē Or, ďWe told you, you needed to calibrate it and you didnít.Ē Or, ďItís your mistake. Itís your fault.Ē We already had a decade or more of history with Marshall Space Flight Center. They managed the project with TRW, the prime contractor, and Perkin-Elmer which became Hughes-Danberry was one of the companies involved in polishing our mirrors. Kodak was where the mirror was assembled. So there was a long history with the companies.
Many of us knew the optics, and the mission overall. Dan Schwartz as the head of our mission support team, some of the engineers, Leon VanSpeybrock the Chandra telescope scientist. We had personal and corporate relationships, including NASA, that were together on AXAF for more than a decade. And counting HEAO and AXAF, two decades. So we had certain credentials. Then after Hubble there was no temptation for management to handle a situation with, ďYou scientists, be good boys. Go play in the sandbox.Ē Nobody wanted to do that anymore. NASA management wanted us out there helping and sharing the responsibility. The result is whenever there were difficulties or problems on AXAF, we were consulted. We were involved in the decision making process. We were on the inside of the team working to build it. I think that contributed tremendously to the success of the mission. There were times where we had a requirement and somebody was going to have to break the bank to meet it. And one of our scientists could say, ďOh, we can relax that requirement because in this other area, we have done better than expected.Ē Or, ďEven though it was written that way, it was written that way because thatís what everybody expected we could do. If we make the number 4.3 instead of 4.0, thereíll be no noticeable effect on the performance.Ē But there was a huge difference because the guys building it could do 4.3, but they couldnít do 4.0. This kind of give and take. There were times in which having some sense of what the telescope was going to be used for was a very useful thing. Scientists did have a little bit more insight on that than the engineers or the management team. Scientists did not drive the thing bankrupt by insisting on some new requirement day in, day out. There were times where we sat at reviews and we looked at things with a different critical perspective and Iím sure contributed to finding and resolving problems that might have otherwise not been found and resolved.
When we were assembling the optics at Kodak, I think as we put the third parabola on, we found that relative to the alignment of the first two, there was a 1/3-arc second shift. And it was not that badly outside of spec because nominally, it was Ĺ-arc second to a1-arc second mission. We didnít know where the shift was coming from. It didnít agree with anything we expected. The Kodak people were very sharp and spotted it. There was a discussion as to whether we could proceed. Our assessment, since we didnít know what was causing it, how do we know that it was only really 1/3 of an arc second and not a much bigger effect? Because we were not seeing what we expect to measure, until we understand what is happening, we canít proceed. It took a month until I think it was finally determined that the lights in the chamber provided just enough heat as they hit a black surface of a support cylinder. The cylinder didnít expand, but it heated the air near it, changing the density and therefore the index refraction in the air. And the laser light that was being bounced up to do the alignment was being thrown off by this change in the index of refraction. It was actually 1/3 of an arc second misalignment. The number was valid. We did not insist they unglue the mirror that had been glued into position. It would have been risky. But we then aligned the remaining five surfaces with the lights out. We set everything up, and then turned off the lights, let it stabilize, and made the measurements. It wasnít a hard solution once we knew, but the key was that we, collectively because of the Hubble experience, realized we needed to understand. Turns out, we didnít avoid a complete disaster because it wouldnít have been, but we didnít know that, so we avoided what I would call the potential. So this openness of communication between industry, NASA, and the science team was important.
McCray:Do you feel that it was more going on with Chandra than what you were hearing with Hubble.
Infinitely more. And partly because of the long history amongst the participants. But more because of the fact that that was pointed to as one of the problems on Hubble. The lack of that kind of communication or dialogue. I think I would like every project that I work on in the future to be done in a similar way. The relationships between the entities. There were things that we certainly couldnít do here at Smithsonian. There are things that NASA couldnít do without us. There are things that industry can do that neither of us can do, but that they canít do everything either. I mean, a big project, theyíre not mass produced. Itís not even like an automobile. You canít go down to the corner store and buy one. Itís developmental. Even if youíve done all the technology work beforehand, the things that youíre inventing almost for the first time or putting together as a system for the first time. The collective heads of the people that are working on it are much better than just any one individual group or subgroup. So that was definitely a benefit from HubbleĖ recognizing the need for dialog and teamwork. A big thing is that of getting the money and getting it approved. Hubble finally got close to the point of delivery. There was a new head of the Office of Space Science and Applications, Len Fisk, joined NASA headquarters. I actually went and visited Fisk shortly before he took the position. He was in New Hampshire so it was just an hourís drive away.
I hadnít met him, but he was a friend of a friend, so Josh Grindlay took me up there. We talked about AXAF. One of the things he would have to do shortly after he got to headquarters was to set his first priority, new project amongst half-a-dozen competing ones. I didnít break any laws by getting there first, but it was kind of a little bit proactive, maybe even an aggressive step. In any case, one of the things we spent that time in New Hampshire, that afternoon talking about were differences between Hubble and Chandra/AXAF. We didnít need to point to a fraction of a hundredth of an arc second in order to integrate our data and conserve the image quality. Hubble already had a catalogue of guide stars which we could use to figure out where we were pointing. We werenít going to try to operate it in real time from the ground, which was still a Hubble requirement at that point in time and dominated the specifications for the communications through the ground net and up to the satellite and back down. Pointing had to be amazingly stable and you needed the ground software to instantly show you what you were getting and then react to it. Our thermal requirements, because these little fiducial lights, our thermal requirements were measured in the range of a degree or two rather than tenths or hundredths of a degree.
So I had, because of all the questions that I had been asked for the previous half-dozen years and the engineering work that had been done, I was in a position to point out these kinds of differences. Not necessarily in a negative with regard to Hubble, but just indications of why Chandra, while certainly the optics would be very challenging and some other things would be. Why we thought we could we could really do it. I think it made an impression to make this kind of direct comparison with Hubble and take the issue on head-on. And so, those are the kinds of things. Itís a real world of many variables. I think itís in the Tuckersí book anyway, but when we finally got into the NASA budget with Fisk being the leader, OMB zeroed us out in the passback. And then NASA went back and said no, theyíd really like it. OMB said no. And this was happening between November and December of 1987, I guess it would be. The fiscal 1989 budget. Then in January, Fletcher was the NASA administrator. I guess he was convinced by Fisk to appeal one more time. Weíd go to the President of the United States who was Ronald Reagan. We got a call. Of course, certain things, you know, youíre within the family. Because I worked for the government, there were certain things that NASA was willing to discuss with me that was very marginal that, you know, OMB doesnít allow an agency to discuss stuff outside of the agency.
What rules were we or werenít we violating, for example. So a lot of discussions were, ďWell, if somebody did this, what would you advise,Ē or so on and so forth. In any case the reality is, we were working on a single chart that would be shown to the President of the United States explaining what AXAF was or why it should be funded. Why should he overturn the OMBís decision to not approve it. I was at an AAS meeting in Houston, Texas when I got a call from people in Washington. I didnít hear much astronomy for that day of the meeting. I was mostly on the phone. And Charlie Pellerin and Art Fuchs at NASA headquarters and myself and then Fisk ultimately were involved. And we all take credit for making the chart which Iíve actually never seen. My recollection is that it was primarily my idea. It doesnít really matter because, again, it was a team effort. We made a viewgraph that had an American flag and had a Japanese flag. The Japanese were hammering the U. S. economy at that point. The chart also had a Russian flag, which the Russians were still our Cold War enemy. And the Russian and the Japanese have substantial programs in X-ray astronomy. So did the Germans and the English, but the Russian and Japanese were among the big players. And under the three flags it said, ďWho shall be first in X-ray astronomy?Ē or some equivalent set of words.
McCray:Was it shown to Reagan?
No. It never got to him. Thereís a board that vetted things still one more time. Regan, Baker, Baker, Miller. Some set of those four people. Three or four of those people. Chief of Staff, head of OMB, and one other advisor. Three of them looked at it and they probably said to Fletcher, ďWell, if you want it that badly, why didnít you just say so?Ē Which, of course, we had said. They said, ďOkay. It will be in the budget.Ē The flags were clearly a winner, I guess, as things go, right? I mean, otherwise, why would they have agreed to go along? Iím sitting here laughing a little bit, right. It is part of the evolution, my personal evolution and certainly us as a team of people working. We began to realize, you know, even if the decadal survey and the National Academy and the science community thought this was a great project, stuff happens. Having a lousy project unless you happen to live in the right section of the right state or something, a lousy project, your chances arenít very good. But having a good project, you have to have a little luck and a lot of perseverance. It also helped to be politically astute enough as a group to figure out how to put something on a chart.
Now one chart. How are you going to explain what an X-ray telescope was, or what it was going to do, or why it was important? We didnít deal with the X-ray telescope at all. I think it said AXAF on it and then, who should be first in X-ray astronomy? But we basically analyzed the audience to whom it was going, and what was out there, and how can we somehow put that together. It seems simple to sit here and talk about it, but I donít think they got too many charts that looked like that in those days, right. It was outside the box. Maybe thatís exactly how the military got every new toy it wanted. Maybe it was exactly within the box. It was outside the box for astronomers.
McCray:Iím curious, have you read the Tuckersí book? I mean, since that book exists and weíll put a footnote to it in the interview transcript. Are there particular parts of AXAF/Chandra history that were not included in the book that should have been? Or anything that you would like to add?
Tananbaum:There might be, but I would have to think. The Tuckers. We realized around 1990 that one of us, namely me, had worked on this project for about fifteen years and was beginning to, not so much forget, but some stuff was fading and some significant events could get lost. I kept notes for a while. I donít have notes anymore, for example, that I took at the time we made that viewgraph for the President.
McCray:Thatís too bad.
None of us, neither Pellerin nor Fisk nor I seemed to know who actually has the viewgraph. I donít have it. It wasnít a viewgraph, of course, it was a hardcopy as you wouldnít show a transparency in those days to the President. It would have been a single sheet. Maybe it exists someplace within the NASA system. I never saw it because, of course, it was not something that could be distributed. And in principle, I never knew that it existed. But in any case, we started each summer. The Tuckers would come by and they typically spent summers on the East Coast and the rest of the year in San Diego. We would work on some science projects, Wallace and I, while Karen and I would be working on some of the politicking and generating support. One of the things we always did, we spent a few afternoons with a tape recorder, not all that different from this. We would go over the events of the previous year and then we would also dig out files on relevant events from a ways back.
So they did a pretty great job of capturing the AXAF/Chandra story up until the time, of course, where the book stops a few months after launch. And Iím sure the story about the flags is in there; I would guess that itís in there. I am very biased in the sense since I worked on this thing for a long time, itís very near and dear to my heart. And obviously that doesnít take much to induce me to talk about it. But the fact is there are things that are a little bit of history there, and things that occurred, and there are things that we learned. One of the things we learned was that the science community, had many individuals very eager to help once the thing got to Congress and to see that Congress would support it. Itís fine, itís very helpful if they talk to the Representatives from their state or the Senators from their state, particularly those that are on Authorization and nowadays Appropriations where almost all the action is. And they can be very helpful. But thereís also the Appropriation Committee itself has professional staff.
McCray:So you talked to people like Dick Malow?
Tananbaum:Dick Malow and his counterparts.
Kevin Kelly. And his predecessor Tom Van der Voort. And you wouldnít want the average scientist thatís excited about the project, you wouldnít want that person to go see Dick Malow. Dick Malow doesnít have time to talk to a hundred people that want to talk about AXAF anyway. But Dick Malow is going to ask you some tough questions. ďWhy arenít you building the mirrors a different way?Ē Or, ďHow do you know such-and-such?Ē And, ďWhat about some technology spinoffs?Ē It could be any of a lot of different questions. But you really have to have people that were the most knowledgeable about the program, which was the TRW Washington rep whose name was Hank Steenstra, and Kathy Lestition who worked for us, and myself, and a couple of the NASA people to the extent that heíll ask them for information. Of course, Pellerin, and Fisk, and Fuchs were allowed to provide it.
They couldnít go over there uninvited. We had to compare notes, we had to exchange information, we were prepared to answer questions. There were prepared rebuttals almost in the sense that if somebody said to you, ďWhy do we need another telescope?Ē You would show an image of a cluster of galaxies with Einstein showing you the diffuse gas and none of the galaxies while the optical is showing you these little dots that were the galaxies. Most of the materials are in the hot gas. You canít see it without an X-ray telescope. Sixty people had done Ph.D. theses using Einstein data. Having that information, whatever the number was, at your fingertips. And, ďCould I name all sixty?Ē ďWell, I bet I could name forty-five of the sixty,Ē if challenged. And if he asked me, ďWere there any in your state?Ē I probably knew if there were any in your state as well. So we were very prepared and maybe just the right amount. We got our ears boxed often enough and had had enough setbacks that we had the proper mix of humility and eagerness. So we tried to listen to the question the staffer or the member was asking. Mostly it was staffers. The standard thing was, ďWell our committee deals with veterans, HUD and other matters.
McCray:I have interviewed Malow before. He can be tough.
Tananbaum:That wasnít so much from him; that was from a staffer for Senator Leahy who did not want this discussion to continue. He was busy. He really wasnít very interested. I remember trying Technology, trying Education, and just hitting a stonewall. It was a pretty important person to get at least some sense of them saying, ďThis was a good project and if there is a way to support it, I will try.Ē Not necessarily anything more than that, but—
McCray:A flicker of interest.
Yes. I finally went off into a dialogue which was, ďI have two teenagers at home. It costs me a hundred dollars a week or whatever to feed them. Thatís great. Iíve got a good job. I can afford to feed them. My wife goes to the grocery store Friday and then again on Monday. A big station wagon. The food disappears. When they go to school, I give them a half-a-buck each for lunch. How can that be?Ē He said, ďWell, whatís your point?Ē I said, ďWell, I canít feed them a meal for fifty cents. I canít begin to come close to feeding them a meal for fifty cents. Theyíre big kids. And they donít come home hungry. They donít come home passed-out from hunger. One of them goes and runs track after school. They get a lot of lunch for fifty cents. I got a good job. My wife works. I could pay for their lunch. Why are we subsidizing all of the school lunches?Ē ďOh, weíre not doing thatĒ, he said. I said, ďWell, thereís no way youíre feeding them on fifty cents.Ē I didnít really know. I never bluff on stuff like that, but I was so desperate. I didnít really know.
So he called another staffer. He got off the phone and said, ďWell, apparently we are.Ē At that point, it was the mid-1980s. ďI guess we are paying for school lunches.Ē I think the reason was that they didnít want some kids to come to school with a voucher or a different way that some paid and some didnít. It was a social view that has some merit. They got the surplus and so some of the lunches, of course, cost less. But even with all the surplus milk, I donít think they were feeding them on fifty cents.
So he said, ďWhat is it you wanted to talk to me about?Ē Now, it was silly that I had to get his attention by talking about the kidsí school lunch. And it was a bit of a gamble and if it turned out I was full of beans, probably the people at NASA wouldnít have been too happy with that. But the point is, I had done enough of this and I could see where the problem was and maybe could I react or respond.
Letís break for the day. This is a good place to stop.
Session I | Session II