History Home | Book Catalog | International Catalog of Sources | Visual Archives | Contact Us

Oral History Transcript — Richard Garwin

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.

Access form   |   Project support   |   How to cite   |   Print this page


See the catalog record for this interview and search for other interviews in our collection


Interview with Richard Garwin
By Finn Aaserud
In Yorktown Heights, NY
October 23, 1986

Listen as Richard Garwin discusses working at IBM and his thoughts on corporations that were created for a specific purpose.

Transcript

Session I | Session II | Session III


Aaserud:

You said something about your archival material, which may be IBM confidential, government files, all that kind of thing.

Garwin:

Yes. One problem is that they're not quite separate. All the government classified materials of course are totally separate, and in a safe.

Aaserud:

At IBM?

Garwin:

Some of the archival material is in IBM archives. I don't know exactly where that is but the IBM archivists do know. And then a lot of it is still around, and I have a lot of very interesting correspondence in regard to the President's Science Advisory Committee — all of the substantive things: automobile emissions and fuel economy; automobile safety; air traffic control; the evolution of the Northeast Corridor transportation system, in which my Aircraft Panel of the President's Science Advisory Committee played a substantial role; the background of satellite technology for air traffic control, that is, for communications surveillance and navigation, which was the most important and productive thing we could see in our 1971 Air Traffic Control Panel activity. So all of that is around, but there is just so much that it will never be of interest to any individual historian. Let me talk about what I'm doing first, because it has evolved to some extent, and then point out a couple of the problems. For many many years all of the outgoing correspondence has been produced with the help of computers — one kind of computer or another — and initially the actual files were not available. The computer files were destroyed, but the "hard copy" has been kept.

Aaserud:

The paper copy.

Garwin:

The paper copies. Now, since about 1977, I guess, maybe earlier — I could look — the actual computer text has been saved on magnetic disc or tape or whatever, and so some time soon I would like to look at the prospect, first, of transferring these computer files to video discs, that is to CD-ROM (Compact Disk — Read Only Memory) where they would all fit for the cost of just a few thousand dollars. We could make the first CD-ROM and then make copies of it for $30 each, and would want to do that just from the point of view of redundancy so the unique copy would not be lost in a fire or whatever.

Aaserud:

Yes, but you still have the hard copies of those.

Garwin:

Right. I still have hard copies some place. Actually, some of the things I don't have hard copy of any more, because we have had a program of transferring these things to microfiche.

Aaserud:

Yes, but from 1977 or whatever they're more easily accessible on the computer.

Garwin:

Yes, on the computer, and even though it may have been transferred to microfiche and the hard copy destroyed, really the right thing to do is to keep the digital text, so that at some future time one could scan through it with soft copy, with a search procedure on a display, and print out only those things that are of interest. One could of course print from the microfiche as well. It seems peculiar that you would have transferred from paper to microfiche, destroyed the paper, and at some later time reproduced all of this onto paper, but I have no room to store it, and I probably have not made the right decision about what to do with the hard copy which has been microfiched. Now when the fiches come back and we check —

Aaserud:

So that's been recently done?

Garwin:

Yes. We check that it is all present on the fiche, then we typically destroy the hard copy or send it to the archives. I don't know which. That's a good question. That's only of course the outgoing correspondence, and in fact there are some things where there are sketches and what not. I tried to keep them down to a minimum because it's harder for me to create those things. But the incoming correspondence is not computer readable, and that is kept around so I can refer to it, and then eventually it is transferred to microfiche and then either destroyed or archived — I forget which.

Aaserud:

There are procedures for that?

Garwin:

Well, there's a procedure in my office which is written down, but I forget what it is.

Aaserud:

Has it been done up to a certain period?

Garwin:

There are different categories. There is incoming and outgoing correspondence, and what we have been doing is to microfiche the incoming and the outgoing on the same schedule. Recently I have been thinking that really it would be better to postpone the transfer of the outgoing to microfiche, because it may be unnecessary. That is, one of the things I can do or somebody can do for me is to find a computer output-to-microfilm (COM) machine. These exist.

Aaserud:

Optical reading?

Garwin:

No, no. The computer output microfilm (COM) is something which takes the computer file, produces the script file which is about to be printed on paper, but instead modulates a cathode ray tube and exposes an image on a microfiche, so the masters, instead of being done with hand work, would be done in this machine fashion. There are such machines around, even within IBM. It's just that the people at Yorktown Heights don't do it; we don't have one; they don't know how to find me one in the area.

And over the years, when I get a very good secretary, after a few years they tend to move on to something else, because I try to give them training which will make them suitable for other jobs. I try also to explain to them that being my secretary is, first of all, something which will pay as much money as a lot of other things they can do, and second, that it is a more important job. But unfortunately secretarial jobs have a general aspect in this society, so smart women who are in those positions often tend to say, "Well, if I go to a party and people ask me what I do, I say I'm a secretary, and they go and talk to somebody else; whereas if I say I'm in charge of payroll, accounting, or whatever, it may pay less, it may be more frustrating, it may be less socially productive, and yet you can talk about such a job."

Aaserud:

Yes, it's a stigma.

Garwin:

So, anyhow, the material is organized, and I don't know whether I sent you the portions of my office "How To." Did I send you that?

Aaserud:

No.

Garwin:

It is a big book which says how the office is run, which is a useful thing, and has had a lot of investment — tens of thousands of dollars — in creating the system which is recorded there.

Aaserud:

Records management?

Garwin:

Among other things. It's also how to answer the telephone, take messages, file things, use the computer tools and so on.

Each of the outgoing items gets a log number automatically supplied by the computer, almost without exception, and so everything that is created has a unique log number which contains the date. And then, since they're all created by me, the log number usually contains the initials of the addressee, rather than my own. If they're not unique, at the end of the week or two weeks when these things are logged, there's a conflict-resolution procedure so new log numbers are assigned. So in case I send two letters to a given person on the same day (or if I sent it to Jeremy J. Stone and James J. Smith who are both JJS's), one of them will get a different log number. So that's good, and the incoming correspondence — and I keep probably more of it than I should — is also given a log number in the same way — by the date on the document, and then by the initials of the sender, or if it's an authored article in a newspaper or whatever — and filed away. Some of it is put into a file called "Others," which are papers by others which I found of interest, and I send out copies to people, but most of it just goes into the incoming correspondence file.

Aaserud:

Lists of all correspondence, incoming and outgoing, are computer accessible, is that it?

Garwin:

Yes, the list and one-line descriptions, in which the log line contains the log number, the date, the person as well as the log number and how many pages approximately. If it's an outgoing correspondence, it says whether there are enclosures, and gives a one line description. Then the file "Others," and the file "Bib" — my bibliography — contain several line descriptions in human readable form, as do files titled "Pats" and "Id," which contain descriptions of RLG. U.S. Patents and RLG. Information Disclosures.

Aaserud:

Yes, that's what I've got.

Garwin:

You have that. So these things exist. The log file goes back to 1979. Now what we are doing is to use the same logging techniques for documents that we come across and are putting onto microfiche. They're all piling up in the earliest 1980 log file or whatever, but some of them go back a good many years earlier. So as we come across the real documents and put them on microfiche, we give them log lines, we give them a log number, and we record the image location on the microfiche, so if we ever want the full document, we can get it.

Aaserud:

Is this something you're doing in collaboration with the archives or is this your own effort?

Garwin:

Just my own.

Aaserud:

Because in the correspondence with you, there was an agreement with Rofes, the IBM archivist.

Garwin:

I have to look at that; I don't remember what the agreement is.

Aaserud:

The agreement was that at least for the time being, all non-current correspondence, all non-current papers, will be transferred to the IBM archives, so that there will not be a question of something that will be lying around (letter W.L. Rofes to S. Weart 21 November 1978).

Garwin:

OK.

Aaserud:

It's a possibility for transferring everything to there, even if it's not IBM related, perhaps.

Garwin:

OK, that's good to remember; I will look that up.

Aaserud:

I brought a copy of the agreement.

Garwin:

Great, another copy. It's very good bureaucratically to have these agreements, because then people just have to follow the agreement, whereas otherwise they have to get decisions from people as to what to do. All right, we'll do that.

Aaserud:

It's stated very very clearly.

Garwin:

Now, what are the problems with this? Well, the problems are that we have not been very careful to keep IBM Confidential material out of the microfiche which are of interest in the science policy sense, and so the materials cannot be made available generally.

Aaserud:

And a person cannot come in and look at the microfiche?

Garwin:

Oh, they could, but they would have to sign an agreement. I think a person could sign an agreement that he would not look at or make use of IBM Confidential material. There might even be a few personnel letters in there, that is, commenting on the performance or problems of individual IBM employees, and those should be marked "Personal and Confidential," but they might not be. If somebody was going through the file and found a very interesting letter which had to do with some individual, we should have a rule that that would not be accessible either. Now, eventually these rules presumably have to expire, and I don't know what the decisions would be there.

Those are all the problems I know. There are various full-text searching techniques which of course will only become more and more powerful as computers evolve. I can show you some of them, so that if one is working on a particular topic, and wants to find all of the documents which have these particular words, phrases or whatever in them, that can be done quite rapidly, especially if all the documents were available on the same CD-ROM or whatever. If there's something you can't find in the incoming letters, I can remember many of these things. We made the decision long ago to use dates as the primary identifier, so I can remember the approximate time and I can find something related to it, and if I remember the name or find the name, then it brings the item back. But if I'm not around, or if I'm otherwise busy, that doesn't work.

Aaserud:

You will certainly get a rush of historians, which I'm sure you're beginning to get, and you would hardly have the time.

Garwin:

Well, I know, and some of them are not very good. For instance, we have this Gregg Herken person who writes things. He takes interviews, and I was complaining bitterly about his first book. I forget what it's called. It came out a couple of years ago, and it's really extremely —

Aaserud:

Counsels of War; that is his second book.

Garwin:

That book anyhow. It had the opportunity to be very interesting, do a great job, but in fact it is extremely poorly written and inaccurate, and I complained to him about it, and he protested. He sent me actually the transcript of the interview he had with me, and that made me even angrier, because he had gotten the transcript right; but it was as if somebody else had written the book from the transcript, because they got the words wrong — things which I never said, things which I said, they got backwards. It was just a very bad show, so I wrote him an even angrier letter complaining about that, so he's writing another book.

Aaserud:

I think it's on PSAC.

Garwin:

I guess so. He sent me galleys, I guess manuscript, of the sections which involved interviews with me, and I wrote him back about something that didn't involve me at all. For instance, he had Mr. Nixon vetoing the appointment of Franklin Long as president of the National Academy of Sciences. Well, the government has nothing to do with the presidency of the National Academy of Sciences! He was proposed as head of the National Science Foundation which is a totally different thing — it is a governmental body.

Aaserud:

Just sloppiness, yes.

Garwin:

Very sloppy. There are historians and historians, and in general, I am very much in favor of democracy, of people understanding their problems and the mechanisms — really taking the future into their hands. That's what this country is all about, and few enough people are in favor of it. People do seem to take the short-term view that the less the people know, the easier it is for the people who have to run the country. Sure it is, but they're supposed to be running it for the People, who are supposed to be fully informed. I try to do my part.

Aaserud:

You've become increasingly frustrated with that, I suppose. The last expression of that that you sent me was that Bulletin of Atomic Scientists article, "Who Proposes, Who Disposes, Who Pays?"

Garwin:

That was 1983. It was a speech that I gave the same night as the President's Star Wars speech. I had prepared my speech; I hadn't written it out, but I had prepared it. But while I was waiting for the dinner — it was in California, the President's speech was at 8 o'clock, as I remember, on the East Coast, which is 5 o'clock in California, so I had a few minutes — I happened to turn on the television, and heard this portion of the speech. I even had a few comments in my talk about that.

So, I joined IBM in 1952. I was born in 1928 in Cleveland, Ohio. My father was an electrical engineering graduate from Case School of Applied Science, now Case Western Reserve University. He never worked as an electrical engineer. He took a position as a motion picture operator, a moving picture projectionist, and because of his technical background, he was much involved in teaching the other operators in the union these skills of the trade. And then in 1928 or 1929, when talkies — sound motion pictures — came in, then he was much involved in instructing them in sound projection and helping to install such things in the motion picture projection booths and so on. But he also was a high school teacher at East Technical High in Cleveland. He taught electricity there, and he had two jobs for most of my childhood. I guess he gave that up around 1940 or 1941, when we moved from Cleveland proper, at 932 Paxton Road, to a home in the eastern suburbs, University Heights, Washington Boulevard. He was still in motion pictures.

Aaserud:

Was he born in America?

Garwin:

Actually, I found out most recently he was not. My parents had always told me that they were born in the United States, their parents having come over just before. But it turns out, according to my mother, that she was really born in Hungary. I knew her family came from Hungary, of course, and came to this country when she was two years old or so, and my father was born in Poland and came to this country at a very early age as well. But my parents, I suppose in order to make my life easier, lied to me about that. My family is Jewish. My mother's parents were quite religious. My father's parents I think were not. In fact, my father's father died when he was about seven years old, I believe. I only know what I'm told. I was told that he was shot in a shoe store he had in Chicago, but it's not clear whether he was shot by a robber or by his partner. I haven't straightened that out. My father died five or six years ago, so I'm not going to learn it from him. My mother is still alive and she's in good health. She's 86 years old.

Aaserud:

Shankland referred to your parents when you went to the centennial anniversary for Case?

Garwin:

Yes, I guess I gave a commencement address there, that's right. So there were always technical things around the house, that is, old electrical engineering technical books and what not. And then to supplement my father's job as a high school teacher, he did have this job as a motion picture projectionist, probably till 1955 or so — I don't remember. But he developed a business installing and repairing sound and motion picture projection equipment for schools and industry in the Cleveland area, so he worked at that. His brother Joe joined him in the business, although his brother did not have a technical education, and then in my spare time I helped out with such things as well. He had two older brothers, Abe and Lou. That's how I learned electronics when I was 10 or 11 or 12 years old, and I did a little chemistry and I was interested in glass blowing, and managed to find some old surplus glass blowing equipment, and I had quite an installation for making, not artistic things, but scientifically useful things. I had the usual darkroom. I was interested in photography and what not. Then I graduated from high school in 1944, which was during the Second World War. There was an accelerated program, you could go to school in the summertime, so I was 16 when I graduated.

Aaserud:

Which school was that?

Garwin:

It was Cleveland Heights High School, and I went to Case for my Bachelor's degree in physics and received that in 1947. There too they had an accelerated program. I had gotten some scholarships — a full scholarship at the small Alleghany College in Pennsylvania, and one at the University of Chicago. But money was sort of short, and in any case I wasn't really ready to leave home. So I had a half scholarship at Case, but there I could live at home and just take the bus to college. So that's what I did, and then when I was in my senior year, it was a question of deciding where to go, and I applied to a number of universities. I think I got scholarships or fellowships to all of them, to Brown maybe, certainly to the University of Chicago. Bob Shankland, who was in the physics department at Case, knew people — I guess he knew Arthur Compton and he had met Fermi — and he thought that Chicago was just the very best place.

My math professor, Sid McCloskey, thought I should go into mathematics, but my father wanted me to go into engineering, but it turned out that he really didn't know what modern engineering was about. When I did go into physics, the background that I had in experimental physics was very useful; what I really like most to do is to make my own equipment, make it work, find out why it isn't working, and so on, whereas in high class engineering in universities, they never touch any equipment. They do theoretical design of things, and even very good engineers — I find it in all of my work at IBM, my consulting with the federal government, mostly in satellite systems or weapons or whatever — really for the most part do not understand that after you design the thing, then you really have to build a model or simulate or whatever or actually build the real thing and see how it works. So we have too much — both in social engineering and physical engineering, mechanical, electrical — of people designing things and having perfect confidence it will work — to disastrous effect.

Aaserud:

That was not a conscious decision on your part.

Garwin:

No, I didn't know. I went into physics for all kinds of new things, and of course many of them were new to me because I didn't really know much about physics. I spent two years at the University of Chicago.

Aaserud:

Excuse me, is Shankland's story, that you had a choice between Chicago and RCA, correct?

Garwin:

That was later. I was getting my Master's degree in 1948, I guess, and had married my wife — still my wife — just before going to Chicago in 1947. She was working to support me while I went to school. My fellowship paid some living expenses and paid tuition but really not enough, and she hadn't finished school. She'd only finished two years at Western Reserve University, so she went to work as a clerk for the Blue Cross Hospital Insurance Company in Chicago, and then I think when we came home for a couple of months in the summer of 1948, she got a temporary job with the telephone company in Cleveland. My uncle was an accountant — I guess he was the chief accountant for the Ohio Bell Telephone Company. The telephone company was a very good employer, and in those days of course if you worked for the telephone company in one city and you moved to another city, then their predisposition was to hire you, so she got a job with the telephone company in Cleveland and worked for them. Then being trained by them, she got a job with the telephone company in Chicago, and that paid better, was a better job; so she did that until our first child was born in November 1949 — our son Jeffrey.

When I was about to get my Master's degree in 1948, I wanted to know what was the best thing to do. Should I get a PhD in physics? What were the job opportunities for supporting a family? So I decided I would go and look, and I went to RCA and a number of other companies, I guess, and I was offered a job — maybe this was when I had my PhD — at International Minerals and Chemicals or something like that to be director of research. They wanted to know how much money I wanted, and I told them some vast amount, maybe $15,000 per year. I don't know that they actually offered it to me, but the problem was, when I went to the RCA research laboratory, I think they offered me $3,700 a year, which was very very little. So instead, I stayed at the University of Chicago. I got my Master's degree, then I chose my thesis with Enrico Fermi.

I'd worked for him, starting about six months or a year after I went to Chicago. I just went to see him. I'd had some courses, and I said, "I'm very good in the laboratory, and probably there are some things you're doing that I could help with." So he put me to work with Leona Marshall, who was his technical assistant really, and Jack Steinberger was doing his thesis on cosmic ray muons in the laboratory. Then Fermi and I and Leona Marshall did some experiments together involving the lifetime of positrons in matter. We were doing this by soaking cotton threads in dilute solution of positron emitting material, putting the threads in Geiger counters so that we would have 100 percent efficiency of detecting the beta decay, and then detecting later-on the annihilation of gamma rays with one inch diameter brass Geiger counters.

And then while we were doing this, we learned that Martin Deutsch at MIT was using a new technique. He had an agreement with RCA to get a transparent window, that is, end-coated photomultiplier tubes. He was using scintillation crystals for his detection, so he could do these things with much higher counting rate than the 100 microsecond dead time or so that the Geiger counter allowed — clearly a better technique. When Fermi came back from an American Physical Society meeting, he told me about it, and so we got scintillation counters and I started building apparatus, and looked at the coincidence circuits we were using, which were sort of 1930 vintage Rossi coincidence circuits, which had a few microseconds resolving time. I decided that there was a very much better way to do that, and introduced what in the semiconductor business is now called emitter-coupled logic. I built these vacuum tube circuits which had a few nanoseconds resolving time. Those "GARWIN coincidence circuits" were used for 20 years as the standard in particle physics and nuclear physics. We eventually took them over into transistors as well.

Obviously, when the time came for me to do a thesis, Enrico Fermi said that there were some things having to do with fission which were of interest. There was also this business of gamma-gamma angular correlation, but nobody had done beta-gamma angular correlation, so we talked about that and he thought that would be a good thing for me to do. I'd worked out all of the techniques, so it was only a matter of using one of my photomultiplier tubes, slicing some scintillation crystals thin, making a vacuum chamber which would allow the radioactive material to be deposited on a very thin film — micrograms per square centimeter — to be supported in view of this beta counter, and then having the vacuum chamber thin enough and homogeneous enough so that a thick-crystal gamma detector could be used, and I did this.

The best place to find equipment is in the junk pile, so I went down to the shop and here I found a bakelite cylinder about 10 inches in diameter. I put it over my shoulder and took it up to the lab, where we had all our own equipment — our own lathes, our own hacksaw and what not. Fermi used these tools; I used these tools. So I machined O-ring grooves in this and in the lids and we had a discussion, because Fermi thought the chamber would collapse under atmospheric pressure. But I pointed out that it was end-stiffened and it wouldn't collapse, and of course it didn't collapse. Anyhow, I had designed it according to some formulas in Strong's Experimental Physics.

I used a number of radioactive materials, which I made by irradiating things with neutrons in the CP-3 heavy water reactor at the Argonne National Laboratory. I learned to do separations and what not, and took all these data. Typically, you would count for a few minutes at some angle and write down the number from the scaler count for more minutes, and write down the number, having duplicate registers; you could never count on these things — if they make a mistake, your data is polluted. And then I did Fourier analyses for the cosine, and cosine-squared theta terms in the distribution. I remember a hot summer day my wife would come in with the baby — this was not with the baby then, this was summer of 1949, she was only pregnant with child — and read me the numbers from the manuals and I would punch them into the desk calculator, which was fortunately motorized, and do these multiplications for the Fourier transforms to see what we had.

So I wrote up the thesis, and then the question was, what to do? So probably it was then that I looked again at RCA and this International Minerals and Chemicals and a couple of other jobs. But I'm not very good at making decisions, so I was very glad that the University of Chicago broke its rule and invited me to stay as an instructor. Instructors in those days were paid $4,700 a year for nine months, which was not really enough to live on. But the other three months would be up to me, so as spring came and I was looking for something to do — some way to earn money in the summer time — Fermi was going back to Los Alamos to be a consultant to the laboratory, and I decided that would be a good thing to do also, so I was made a consultant to Los Alamos. The first thing I did that summer was to read all of the classified library material on nuclear weapons — how these things were done during the war, and what had happened afterwards.

Aaserud:

How did you get a clearance?

Garwin:

Well, because I had this job. I guess I also had to get a clearance to work at the Argonne Laboratory — to irradiate things in the reactor. It wasn't so hard because I was so young — I hadn't done anything. And of course, my parents were born in the United States (which turned out not to be true — just recently I found out), so there was no difficulty.

Aaserud:

To backtrack a little bit, you took your Master's in one year and then your Doctor's in one year?

Garwin:

Yes. Probably I should have taken longer because had I taken more courses, I would have learned more physics probably. But I just couldn't see wasting time. They had an arrangement whereby in the spring of the year — in April — the people who were beginning their graduate work could take a Qualifying Exam. The Qualifying Exam would admit you for the study of the PhD program, and then typically the next year you would take the Basic Exam. The Basic, I believe, was to show that you had satisfied all of the requirements of the PhD except for your thesis. Since they had old exams available, I looked at these. I decided that I could do the Basic Exam and so I would skip the Qualifying Exam. I negotiated with them, and I got an agreement that I could do this, but that if I failed the Basic Exam, then I would have to take the Qualifying Exam the next year and the Basic Exam the year after, so I would lose two years compared with passing the Basic Exam, and I would lose a year compared with doing the routine thing — taking the Qualifying Exam when I was supposed to do it.

Aaserud:

It was a gamble.

Garwin:

So I studied hard, to the extent, in fact, that I had some stomach problems which they thought was an ulcer or whatever, but it was probably just nerves. But I took the Basic Exam and I got the highest score in the Basic Exam, as I recall. T.D. Lee was in the Basic cohort. I got a higher score than anybody, but of course that's because really the Basic Exam doesn't test specific knowledge but really problem solving ability. Of course they thought you had solved the problems the way you had learned in the courses, but I knew different ways of solving those problems, so I solved the problems my way. So in a year I got the Master's degree I guess with having passed the Basic Exam. Then as soon as my thesis was finished, I wrote it up briefly and submitted that. Then as I say they had a ruling against hiring their own PhDs unless they had been away and had experience elsewhere. But they broke the rule, so I didn't have to move to another place and find another apartment.

It wasn't easy living in Chicago in the postwar era. It was very difficult to find a place to live, and I think in the first year we were there, we moved 12 times. We had just sublets, people who would be away on vacation, and when we were without an apartment to sublet, we moved back to a residential hotel three or four times, so I think we stayed in nine different places and moved 12 different times. Then the parents of some friend of ours whom we had met there were friends of friends of my mother's, and we grew to know them. They went to California. He had gotten his PhD in chemistry with Henry Taube and went to the University of Southern California — Harold L. and Edith Friedman. We're still good friends. They live in New York now — SUNY Stony Brook. So we inherited their apartment on South Shore Drive. It was a basement apartment, that had been a laundry room or something, but for us it was heaven because it was a room and a little bath and a tiny kitchen — kitchenette. It was a place where we lived for a year.

Then some other best friends of ours, Harold Agnew and his wife Beverly, had come to Chicago with Fermi in 1945, and he finished his PhD there (I went to Chicago in 1947, after I graduated at Case). In 1949 he went back to Los Alamos and eventually became director there for ten years. We inherited their apartment, which they had actually built. Mr. Beaudry owned the house on 50th Street when Harold rang his doorbell in 1945 (the Agnews couldn't find a place to live either). Harold, smart fellow, very handy, got around to seeing this porch and said, "You know, if I enclosed this with glass, that will increase the value of your house and you'll have an apartment you can rent to us." So when they left in 1949, Harold told Mr. Beaudry, the owner of the house, that I too was very handy and reliable and that they should take us as tenants. We lived there for a couple of years until we moved to a university-owned apartment which was only half a block from the accelerator laboratory — the Institute for Nuclear Studies — where I worked from 1950 until I left in 1952.

The reason I left Chicago, although I was promoted to assistant professor and so on, was that the cyclotron — first the betatron, 100 million volts, and then the 450 million volt cyclotron — was a shared resource, and you had to think in advance what you were going to do and present your proposal. There was an informal scheduling committee, and you had to decide six weeks in advance what you were going to do, and then the cyclotron would be pumped down; some of the most interesting things required internal targets, and I was really quite good at that, but I just didn't want to compete. I didn't want to feel that my time on the cyclotron was at the expense of somebody else's, and I also didn't like the idea of writing proposals for money so I could get money from the government for doing what I wanted to do. Thinking of these things in advance has never been my style. So I decided that I didn't like the sociology of high energy physics. It was, you know, teams, and really too competitive. I don't mind being judged on what I do — I think that's just fine — but I really don't like to make arguments for what I do. Let somebody else recognize it.

So I decided that low-temperature physics had not benefitted from the great infusion of analysis and talent after the war that had gone into nuclear and particle physics, and that's what I should do. Chicago had a good low-temperature physics laboratory, with liquid helium available and so on.

I came to IBM through a peculiar circumstance. When I decided I would change fields, I thought also I should look around and see whether there was another place, because Chicago was not a very good place to raise a family. There was all kinds of street crime and break-ins and rapes and so on. It was just not good, so I thought a smaller town would be better. I guess that Florida State University offered me the chairmanship of their physics department and the opportunity to appoint some people. I wasn't interested in that either. I don't like administration. I just like working on things I want to work on. I had offers from a couple of other places.

I guess I visited Berkeley, but IBM had a laboratory since before the war really. It was in Columbia University buildings, founded by Wallace J. Eckert, an astronomer who introduced the punched card to scientific computing. And now in 1952, having gone into scientific computing in a big way at the Watson Scientific Computing Laboratory in 1945 on 116th Street, they wanted to move from vacuum tube computers to solid state physics. Wallace Eckert, a very far seeing person, was faculty member in the Columbia University Department of Astronomy and good friend of I.I. Rabi and Polykarp Kusch and others. They decided that solid state physics was a good thing to do. There wasn't any at Columbia University, so the Watson Laboratory was going to have a strong push into this. They had, through contacts, asked Emilio Segre to come see whether he would like to head this laboratory. He was at the University of Illinois on sabbatical from Berkeley that year, which was a time of problems in California, with the loyalty oath and so on. I went down to Illinois from Chicago to see Segre. It turned out that Segre thought this was a very good idea, so I visited Wallace Eckert in New York. Eventually Segre decided not to do this and I decided to join the laboratory, so Wallace was director for a while and then I was director. So that's how I got here and that's why I got here.

Aaserud:

Was the development towards team research, competition and all that you talked about in Chicago something that actually developed while you were there, or was that something that you came into?

Garwin:

I think it was developing. I think there were more block grants up to that time. Emanuel Piore — Mannie Piore — was head of the Office of Naval Research immediately following World War II. He deserves a lot of credit, in my opinion, as the prime mover in the decision to put the leading edge of research in the United States into the universities together with graduate education, as does the Office of Naval Research, which gave money for the University of Chicago cyclotron. At the same time, people were trying to create a National Science Foundation. Bill Golden, who was at the New York Science Policy Association meeting today, was very instrumental in that. But money has to come from some place, and it shocked me at Chicago to find out that of the million dollar annual budget of the Division of Physical Sciences, only $50,000 came from university funds. The rest you had to go out and get by yourself. I hadn't had any training in that, and I didn't like the idea.

Aaserud:

I suppose the structure of the funding and the structure of the team work changed during this period.

Garwin:

That's right. I liked working with people. There was no problem with that, but people had formed their own teams. Herb Anderson was a particular problem for me because he was very aggressive in searching for money, and dividing up the research areas; they were going to take positive pions and somebody else was going to take negative pions, and I would have to think of my own topic. Well, I did look for strange particles, V-particles or whatever, and I devised a lot of other things. But anyhow, I like to think of things, and if I think of something, then I want to work that night and do the experiment the next morning. You can do that when you don't have such a large shared facility to use, and also if you don't require a lot of support. What I've done over the years has been very limited by these preferences. I've never had to apply for money. IBM provides funds to its research division. A few percent of the money does come from the federal government. In the old days when I was involved in such things as Director of Applied Research or whatever, I know our policy was to get 5 percent or something like that just to show we were competitive. Otherwise the management of IBM might say, "How do we know how you stack up against academic research?" And if we can show that where we wanted to, we could get money to support our research from the government, that was one way of showing; and of course if one gets Nobel Prizes or recognition, that's another way of showing, so we're very happy we got our share of the physics Nobel Prize this year.

So I had two lines when I came to IBM in 1952. I had already been working for three summers at Los Alamos — I think for three months the first summer, maybe four months the second summer, five months the third summer. I'd spent a month in Korea and Japan consulting for the Air Force, and in 1950 it was clear that all the talk about building a hydrogen bomb was just based on sand, because the knowledge of the reaction cross-section — what is your field?

Aaserud:

Well, I come from physics.

Garwin:

What kind of physics?

Aaserud:

I graduated in theoretical physics at the University of Oslo.

Garwin:

I asked you that before. Well, anyway, the reaction cross-section between deuterium and deuterium and deuterium and tritium was really not very well known at all. It had been measured a decade earlier by Tom Bonner, mostly at higher energies, and we were dealing with great extrapolations. There's a resonance in the D-T reaction cross-section; the extrapolation to low relative energy is not easy, not reliable. I decided I would create an experiment to measure this down to 10 or 20 kilovolts, which isn't an easy thing to do, and I designed such an experiment. I started to build it, and then clearly it was more important to get it done quickly than to get it done by me. So Fermi helped in staffing it with various people — Harold Agnew, Stovall, Sawyer. He got Jim Tuck to come from England to the University of Chicago, where he stayed while he was waiting for his clearance, and then he led this experiment which did get good numbers for the cross-section. It was published in Physical Review eventually.

Aaserud:

We're in Richard Garwin's office at IBM. We're out of the car, where you ended with your move to IBM. But maybe we should talk a little bit more about Chicago.

Garwin:

Well, I think what I was doing was to talk about my early work at Los Alamos, setting up to measure the reaction cross-sections of deuterium deuterium on tritium — identifying problems by which one nuclear explosion could impede another. The early weapons were very vulnerable to that, so when I wrote a paper at Los Alamos it caused quite a stir. Some of the early tests in Nevada were coming up. They went out and they made these measurements and they verified that indeed this problem existed, and they fixed it. They had to redesign nuclear weapons as a result.

There were many other things that I did. Looking at the use of nuclear explosions in testing nuclear explosives, I introduced new techniques for getting detailed information on the behavior and verifying the design. Then in May of 1951, when I returned to Los Alamos for the second time, Teller and Ulam told me about their approach and I designed not the "experiment" that Teller wanted but an actual hydrogen bomb which as he says was controversial but which was built and worked just as designed. About this, Teller said in 1981, "In the early 1950's when I had the first crude design of the hydrogen bomb, Dick Garwin came to Los Alamos and asked me how he could help. Actually the design I had in mind was not that of a real bomb but of a model for an experiment. I asked Garwin to change this crude design into something approximating a blueprint. He did so in a short time — a week or two. That experiment was carried out. Garwin's blueprint had been criticized by many people, including Hans Bethe. In the end the shot was fired almost precisely according to Garwin's design, and it worked as expected."

As Herb York says in his book The Advisors, which information I didn't know had been declassified, there were several air dropable hydrogen bombs made using liquid hydrogen and liquid deuterium, which I designed also. They were available for use in that era, to be dropped from B-36 aircraft.

After that I made a couple of other interesting little inventions which were widely applied in nuclear weapons, among them an initiating device which provides the neutrons for initiating a nuclear reaction. I did more things at Los Alamos years after. The reason I mention this is that when IBM and Wallace Eckert offered me the job at the Watson Laboratory in New York, I had been involved in these things for three years; I thought it was very important and yet there was no way in which IBM could decide whether a particular activity of this kind was beneficial to the country or not, or help make the decision. I thought the very best thing would be just to leave it up to me, but in order not to either have the reality or the appearance of a conflict of interest — which you all too often see in regard to a person who is a policeman, gets a salary, and then keeps money which he gets in tips or graft on the side, or a physician who has a position at a university hospital and then uses the time and facilities for private practice and keep the income — I didn't want to either be tempted by that or have it said that I was having such things sway my decision. So I proposed that I would turn over to IBM any money I got for travel reimbursement or consulting fee.

And that's how it has been. In the recent era — the last ten years or so — with JASON — because that is an extended period and because I had not been using up my IBM vacation — I changed so that I would take vacation for JASON activities, and for that I keep the travel reimbursement and consulting fees, but all the others go to IBM. So that's the arrangement. And with IBM I think I've had a beneficial influence on the national security or the rationality or the economic well-being of the country. IBM has shared in those contributions, in the sense that they have knowingly made this arrangement and almost never made any questions or problems about what I am doing or how much time I spend on doing it.

Aaserud:

Both IBM and JASON, in that sense.

Garwin:

mp3

But IBM is my employer, principal employer. And so it requires a lot more tolerance on their part. There's a problem with organizations which are created for a purpose and then are involved in controversial things with a small part of their efforts. Mostly it is prevented, because people on the board of directors or board of trustees say it's our business to see that our principal purpose is carried out; that is, the stockholders benefit and the customers get good computers and so on. So in addition to the three principles that Tom Watson Sr. I guess enunciated for the company — that is, service to our stockholders, service to our employees, benefits to our customers — I encouraged people to add a fourth, and that is, to be a good corporate citizen, that is, to have service to society in general. Of course you can't do that to the extent that the company isn't profitable, but if you find some way in which it is cost effective to help the rest of the country or world, then you ought to do it. The calculus that I use to justify IBM's supporting me in these activities is simply to say that IBM is more than one percent of the Gross National Product, and so if I can save a billion dollars here or there, then one percent — 10 million dollars — of that is money that doesn't have to be collected in taxes from IBM. It's not that IBM benefits by selling computing machines to the people that I consult with. In fact it might be quite the opposite. I may never know. But it is this other. And so whatever rationale IBM uses — and there must be times when people complain to them, but I hardly ever see those complaints — they continue to do this, and I think there ought to be more and not less such activity.

Aaserud:

If you compare the work at Chicago with the work at Los Alamos, to what extent was it the same collaborators, the same kind of problems, even though one was classified and the other was not? To what extent did it expand your way of working and your view of the world to go to Los Alamos?

Garwin:

Well, at Chicago I didn't really have any collaborators after I got my degree. I would do everything myself. I would make drawings which would go to the shop. I had one graduate student, I think, whom I neglected shamefully — Maurice Glicksman who is now dean at Brown University. At Los Alamos, mostly I worked theoretically. That is, the first summer I was there I was in the Physics Division, and I was building a little low energy accelerator and ion source for the deuterium cross-section measurements. I think there may have been some troubles with the people in the Physics Division, so the Theoretical Division was my home after that, and Carson Mark was always very friendly. But I would write the papers myself. I think maybe I had one paper with Enrico Fermi on this question of fratricide between nuclear weapons. But I would talk to people quite a lot. And I guess the patent that I have on initiators is with Carson [Mark and John Reitz] no Ted Taylor.

What was different really is that at Los Alamos you could see immediate applications for what you were doing. At Los Alamos I also learned how difficult it is to get people to do things. If you have an idea, really quite a good idea, the majority of people won't touch it, either because they didn't invent it or they will think of all kinds of reasons why it won't work. So on several things which I proposed, it was a much harder job explaining and persuading than I thought. But that's the way the world is, and it's good to learn it.

Aaserud:

Do you have examples of that, of projects that weren't accomplished because of that kind of reaction?

Garwin:

No, I got them all done. But it takes sometimes a kind of guerilla warfare. If the people whose job it is, you would think, to advance the techniques of nuclear testing don't pick up the ideas, then you have to go to other people who will influence them, or you have to do more of their work and consider alternatives and work it out for them.

It's the same way in business. The job of the innovator is not done when he or she thinks of a new way of doing something. We've done exactly the same thing, as I pointed out, in the Laser Printer. I spoke for years with the people in the office products division about making the IBM electrophotographic copier into a laser printer, and there were all kinds of arguments why they couldn't do it. The market wasn't there and so on. Well, the proper market of course is for a computer output printer, not an office product printer, and there the reason good people didn't pick it up is that they were in the wrong division. If they had made this thing it wouldn't have been used in the computer division, and the computer folks when I talked to them about it said the copier didn't have the reliability and longevity required of a computer room printer. But that's only a matter of engineering, and I could have done that. There was no reason to discard it from that point of view.

Later when I was proposing the laser rotating mirror for the electro optical transducer in the IBM 3800, the ones in the proper division didn't want it because they had their own solution, they thought. So they looked for not any kind of balanced assessment, but arguments against it and in favor of theirs. Many people don't even know they're doing this. So they said, "Well, it's mechanical, it has rotating parts, this rotating mirror." That might be a very good argument if the rest of the system had no moving parts, but a printer that has to handle paper and feed it in a complicated path and stack it has hundreds of moving parts, and one more uniformly moving part is not a problem. It's just like one of the fans in the machine. So eventually, if it matters, you can, with considerable delay, get these things done, and really get them done, not just cheer when somebody else does them. But in order to do that, you can't do it to benefit yourself, because first of all, there's no way to benefit from most of these things.

So in Los Alamos I started working on these defense matters. There's a whole outside community there. I like to consult like this; it's something I'm most interested in doing. When I was leaving Chicago, I guess I talked with Sam Allison, now dead, and he said he thought that probably at IBM they would value what people in universities tended to resent, when I would stop into their lab and tell them better ways to do things that they were doing; they didn't want to hear that. Well, Sam Allison said, "Nobody wants to hear that anyway," but you have to be either very tactful about it, or you have to point out to them that their job is not to enjoy themselves but to do the company or the society some good. It's very rare that people know that, because they're mostly organized according to Adam Smith, that is, to get people who are interested in doing what they're doing, and you count on their doing that as benefiting the larger organization. And to some extent that's right, but it would be even better if then they took into account the things that would directly benefit the larger organization.

The laser rotating mirror printer is one of them. The mass storage system that we have, the IBM 3850, which uses cartridges about that long and that fat with a roll of wide magnetic tape, is another; the people had in their minds doing it one way with 800 longitudinal tracks on the tape, and you might do it that way, but my experience in development programs is that you would be far down that line before you found out you weren't doing it, and then you would have a terrible catastrophe, because this product niche would not be filled. It would be much more conservative to wrap the tape around a big drum and use a video head, which comes across at a diagonal — one head instead of 800 or 50 or whatever. There was a big fight over which way to do it, and finally — this was when I was on the Corporate Technical Committee — after all kinds of crazy arguments, we won that one and that's the way it's done. And I've done that with the government in many many programs. And just recently, in introducing the touch panel on the IBM InfoWindow display — I'll send you the article from Think Magazine about it — we were working on that here. We were looking for applications of this touch panel, because it was clearly the lowest cost technology available and also the most rugged — just a glass plate with four piezoelectric pills. We talked to the people at IBM who we thought would have an application for it — the copier division, the people who make the automated teller machines. We wrote them many letters, gave them operating models of these things, and yet they were never able to put it into their investment plan. Finally, another group at Raleigh, North Carolina, I guess it is, was making a fancier thing and they had, since they put it together from other concepts, an externally supplied touch panel. We showed them ours, and they said, "But nobody's manufacturing that; nobody's using it." But ours was clearly cheaper and it was really much easier. We could manufacture it. It was very simple. But it took a very great effort and took a lot of bad experience with the other manufacturers for them to realize that they were taking less of a risk to go with the new technology than with the other one. So you have to be somewhat relaxed about the tactical question, but you have to be persistent strategically in order to do these organizations some good.

Aaserud:

But that was something you learned much earlier.

Garwin:

I learned that when I went to Washington to consult. When I came to IBM, I was involved — Jerry Wiesner and Jerrold Zacharias at MIT had asked Tom Watson Sr. for my half-time services on air defense, on Project Lamplight, and so I worked on that. I didn't want to at first, but the company really encouraged me — the only time they ever did such a thing, I guess there was another time — to do that, and I had to spend about half-time for a year in Lexington, Massachusetts. But there were interesting people working in a technical task-force mode, so I learned a lot from that.

Aaserud:

When was that, precisely?

Garwin:

1953, 1954, I think. So just as soon as I had come.

Aaserud:

That was a defense effort.

Garwin:

Expanding the air defense system to cover the ocean approaches to the continental United States and Canada. So that's when I learned about radar and systems and things like that.

Aaserud:

What about theory versus experiment? Did you have any problems with that, or did you do both as easily? I notice on your publication list that they are very interspersed.

Garwin:

Well, there is not much theory there, only minor theories. No, I'm an experimenter. The way to be an experimenter is to be able to design an experiment, to see in your mind by analysis how things are going to work. So there's a lot of what would be called theory or anyhow applied physics, applied mathematics, in the design of experiments. From that point of view, I have quite a good theoretical competence, but not from the point of view of pure theory, of field theories or of inventing new theoretical approaches.

Aaserud:

You said that the Los Alamos work was more theory oriented than the Chicago work. That's theory in that sense, is it?

Garwin:

No, I don't think I said that. It was not theoretical in that sense. For the most part at Los Alamos I didn't do anything with my hands, except for the first summer when I was building this accelerator, so that was mostly the same kind of design of experimental systems. Now, some people might call that theory. I don't really. What I call theory is to invent a theoretical approach and then to look at the approach, whereas what I call applied physics or applied mathematics is to try to either create a physical system to do a specific job, or to analyze how some particular thing is going to work.

Aaserud:

Well, let's return to IBM-Watson. We talked a little bit about this when we walked across here. What was the relationship between the IBM-Watson Laboratory and the physics community at the university?

Garwin:

That was very good. We hired a lot of graduate students with fresh PhDs the first year. There were good relations. Seymour Koenig; Bob Gunther-Moore, who was then in semiconductors at IBM, and staff; Gardiner Tucker who became director of research here and then went to be Principal Deputy Director of Research and Engineering at the Defense Department; and I think a couple of other people. Oh yes, Haskell Reich, who died a couple of years ago, unfortunately, worked very closely with me, and Sol Triebwasser, who's still here on staff. Those were all Columbia, mostly molecular-beam PhDs. And we had some collaboration; Charlie Townes worked on an ammonia maser relativity experiment with By Havens, one of our electrical engineers. We had students from the physics department, and we taught solid state physics courses; I was an adjunct associate professor for a while. Of course in 1957 Leon Lederman and I, although I wasn't in particle physics any longer, did the pi-mu-e non-conservation of parity experiment. That was an anomaly, but then I worked very closely for several years with the Columbia folks at the Nevis cyclotron on such things. And we helped them back and forth a little bit in some of the low temperature physics they were doing. So we had good relations. I used to have lunch every day with the people in the math and physics departments at the Faculty Club. That was nice.

Aaserud:

You wouldn't go as far as saying that it was in effect a department of the university, the way the work went?

Garwin:

It was. But it was not a particular department, because we had solid state physics as well as electrical engineering as applied to computers and we had chemistry. The chemists were very reluctant to share their good graduate students with us. That was very different from the physics department, which was perfectly happy for us. Our graduate student stipends were the same as those at Columbia University, and most of the people really just did pure physics. Maybe they were capable of inventing things but they weren't interested in such things. On the science side, really it was mostly myself who has a knack for and interest in inventing things. I would devise delay-line storage systems using cesium beams or whatever, or better ways to have bearings or superconducting power lines and things like that. So there are many patents, and some of these things even work and are even useful. But many of them, truth be told, only came into actual use more than 17 years after the patent was issued, so patents aren't all that valuable.

Aaserud:

Of course, one difference within IBM was that you had unpublished reports within the corporation or within the lab, so to speak.

Garwin:

Not very many. There were some on the engineering side.

Aaserud:

I'm thinking about your work.

Garwin:

Physics and chemistry — almost all of them saw the light of day either through scientific publication or through the Technical Disclosure Bulletin, because things that we don't patent, very often we want to be able to use. Nobody else can patent them if they've been published, so it's not worth our patenting. We have a Technical Disclosure Bulletin where we write things up in half a page and explain how we know how to do something, and then anybody else can use that. Now, there are occasional things that get written — if I study somebody's project, I find that they're having a lubrication problem on a particular magnetic tape, and here is the analysis and here is a proposed solution — that may not see the light of day, but there is very little of that.

Aaserud:

So you didn't feel much of a restriction from the corporation on what you could publish compared to what the physicists did.

Garwin:

No, not at all.

Aaserud:

It went smoothly and there was no questioning on the part of the university people of having a corporation in the midst of it, so to speak, or corporate research in the midst of it.

Garwin:

For the most part not, and really only by people who didn't understand it.

Aaserud:

Yes, but it came up once in a while.

Garwin:

When I moved to IBM, I guess in 1952, I think I had been earning $5,500 a year for a nine-month academic year at the University of Chicago. I was going to get $10,000 a year at IBM, so I thought that was a good raise. But I found out I had less money than when I was at Chicago, because first of all the $5,500 was for nine months and I earned at a higher rate for the other three months; and at Chicago I could be a consultant to some outside organizations. Furthermore the cost of living in New York was much higher even then, and there was also a state tax. So really this was not an advantageous thing to have done.

Aaserud:

Not from that side, but you did feel that it was a good transition in terms of what you had experienced in Chicago?

Garwin:

Oh yes, I thought it was a much better working environment after we got our laboratory fitted up and I could do what I wanted to do.

Aaserud:

How did you work? How did the working relations go? Did you work with people or work mostly by yourself?

Garwin:

For the most part I worked by myself. I had a technician, and we had a very well equipped machine shop and electronics shop. For the most part, when I want something built, I design the whole thing so it just has to be machined. I can go into the machine shop and machine it, too; I'm a good machinist. But once you get it to a point where the drawings are adequate, somebody else can do that, so that's what I would do. Somebody else would build the parts, and I would test it out and make sure it would work and modify it and what not. Then of course I had these other interests, and one thing which was difficult, and a peculiarity with me, is that if I'm away for a few days or an extended period because of government consulting, while working at Los Alamos, then the work would stop, if it were just myself and the technician. So it was clear that I needed somebody to work with me, and after about a year, I guess, Haskell Reich received his PhD from Columbia. He worked with Polykarp Kusch, I guess — a very capable young man at the time. We worked together for many years on liquid and solid helium and helium 3, so we have many publications. He was a fine person to work with, also a very good experimenter, a tenacious and warm and open person. He and I and my wife and his wife were good friends until he died.

And then during the late 1960s I was so very much involved with Washington. I was here at IBM Yorktown Heights for a year, in 1965-1966. I was Director of Applied Research, during which time he continued with our work on helium. But then, I guess, I wasn't doing so much physics any more. Around 1970 I went to corporate headquarters at Armonk to be on the Corporate Technical Committee. That was at the urging of Mannie Piore, who was the IBM Chief Scientist. I really didn't want to do that. I thought it would not be good for me, and wouldn't be of unique benefit to the company. I don't know whether it was Mannie's idea — that he needed help, somebody to talk to — or whether it was Tom Watson, Jr. or whatever. But anyhow I did that for a year. I'd been there for a year and I hadn't gotten any word that I was doing good work. I hadn't got a raise and so on, so I told him (he had told me that I only had to work there for nine months or twelve months and now it was three months over that) that I was just not going to show up for work on Monday. I was going to go back to the lab or whatever. So eventually I was through with that, because I liked to do things. For me, the CTC membership is just too much being driven by schedules, by meetings and travel and what not, and that's incompatible with the important work that I have outside.

In 1970 the decision was made to close the Watson Laboratory in New York City, and so I came here. Most of the people from there did come to Yorktown. A few did not. Philip Aisen, a physician and biochemist, went to the Albert Einstein College of Medicine with which he had been associated. I think Al Redfield did come up here briefly, and then he went off to Berkeley to learn biochemistry, and has been at Brandeis I guess in biology or biochemistry ever since. But aside from those people, for the most part they did come here. But at the Watson Laboratory, I worked by myself with a technician or with Haskell Reich and a technician, and a graduate student occasionally, not always. It's good to have ten graduate students or five graduate students. When you have that many, they teach one another. But I didn't want to run a program of that magnitude, and so we had an occasional graduate student.

Aaserud:

You mentioned Lamplight and you mentioned Los Alamos. You had also other consultantships.

Garwin:

Oh yes. I was a consultant to AVCO and to General Dynamics on the re-entry aspects of ballistic missiles, and then on a number of other things like that, so I didn't confine myself to such things. Around 1952 I recognized solar radiation pressure could be used to drive satellites not only in toward the sun but out, and how it could be used to go in orbits expanding from the earth's surface, or the other way. I tried to interest General Dynamics, and there was absolutely no interest in the government community. I tried to interest IBM. I tried to publish it in the first issue of the IBM Journal of Research and Development, and they found it somewhat frivolous. Eventually I published it in 1958 in the Journal of the American Rocket Society. The Air Force chief scientist had a press conference (he was a person from Westinghouse) and he said, "No, no, maybe you can go out from the sun with solar sailing but you can't come in. You don't have a keel in space."

Aaserud:

What was the origin of that?

Garwin:

I was thinking about these other things. In 1951 or so there was a paper published by Primakov and others I guess on thermal positronium, and the fate of positrons as they slow down in metals. They had published an argument that when the positrons drop below a few volts kinetic energy, they would no longer be moving faster than the electrons and so they couldn't lose energy to them. You can see how that is in one dimension, because if you have an electron which is moving faster than the positron, the positron doesn't catch up to it, so there are no collisions, and if it's going the other way and it's moving faster than the positron and there is a collision, then they just exchange velocities, since they have the same mass, and the positron heats, except that those collisions are quenched because of the Fermi sea (there's no room for the electron at lower energy). But they were wrong, because they were looking at this only in one dimension, and metals are three dimensional, and in fact the positrons cool really very nicely, and this has an effect on the formation of positronium and what not. But when I was looking at that — and I published a little paper about it — then I told Fermi, at Los Alamos — it must have been in 1950 or 1951 — that you could use the same thing (he had also looked at an analogous question, namely at the equilibration of energy between clouds of magnetized matter, plasma, and individual protons. The protons would be batted back and forth between the soft mallets, and they would pick up twice the velocity of the magnetized cloud each time, so you could get relativistic energies out of these clouds moving at a mere hundreds of kilometers per second). So I pointed out that you could do this also in escaping from the solar system. That is, you could go out and have a tour around Jupiter and pick up some fraction or multiple — up to twice — of the orbital velocity of Jupiter, and thus escape. Of course, the Grand Tour really does work very well. That's how we explore the outer planets with the gravity assist, rather than having to provide all of the kinetic energy from earth to compensate the potential energy of the gravitational well.

But I always worked by myself or with one colleague, and here at Yorktown, after moving here in 1970, I guess I had directed the setting up of my laboratory but I didn't have anything to do. The question was, would I go on in liquid and solid helium? There was a lot of advance and there were very good people in the field by now. So I looked at the work that Joe Weber had done in gravity waves, and decided to take it up in view of the theoretical interest. People were making all kinds of models of the universe that were just insane, because the universe would convert all of its energy into gravitational radiation in 50 million years or so, if one were really detecting what Joe Weber was detecting. I decided that it was important to look at that, and I and Jim Levine within a few months put together an apparatus at very low cost — a few thousand dollars. We did experiments which were much better than Weber's and more sensitive and easier to analyze, and at the same frequency, which showed that there were no gravity waves at that frequency. Whereas everybody else in the field said, "Now that Weber has discovered gravity waves, let's look for them at different frequencies. Let's make gravity-wave telescopes" and what not. It's stupid, because nobody had looked to see whether the phenomenon was really there.

Aaserud:

Yes. That still isn't resolved, is it?

Garwin:

Yes, it is. Weber is just such a character that he has not said, "No, I never did see a gravity wave." And the National Science Foundation, unfortunately, which funded that work, is not man enough to clean the record, which they should. So it's not a bright spot in the pages of modern physics.

Aaserud:

It's an interesting episode. That was in the sixties, that started in the sixties.

Garwin:

No, it started in the seventies, after I came back here. I went to the Corporate Technical Committee in the fall of 1970, and probably came here in 1971, and that's when we got this aluminum bar from the stock room and we got a couple of glass bell jars and stuck them together and did this work. So that was published 1973, I guess, with further work published 1974, and then in 1975 I published a detailed description of the electronics, as well as some letters of commentary. So that's really the last experimental physics work that I've done.

Aaserud:

While we're at this, that was quite a new pursuit for you to follow up. I don't know how dependent this was on gravitational theory and complicated by it.

Garwin:

Well, it's really very easy. We didn't have much problem with the theory, because I had long ago in the 1950s worked out a very simple way of looking at antennas for electromagnetic signals or acoustic radiators or whatever.

Over the years I've been astonished that not everybody looks at antennas the same way. The last few years, in some JASON work, I've decided that a lot of smart young people, really very well trained, know more techniques than I will ever know. There are many more of them than there are of me, so the best thing that I can do is to look at the problems that we're working on, and just in the first few days of the JASON summer session, try to say how I would solve these, what tools I would use, how I'd look at them, and how the answer will come out. A lot of people don't like to participate in such a mode. But there are a couple who do, and who are sufficiently self-confident that they don't mind hearing how somebody else would look at a problem. And of course, they don't slavishly follow my ideas. They look at it themselves. They may even find something wrong with my understanding.

So one of the things that you do is to understand antennas, and the way you understand an antenna is to know that the radiation field starts at a distance — from the source, if the source is small compared with (???). And so you have to determine from the local configuration of charges, velocities and sound or moving masses what the local static field is at that distance. You project that onto various propagating electromagnetic waves. So you have two problems which are quite separate. First, the static local field, so that if it's a dipole, it goes down like 1/R3. Second, the propagating field, which, starting at this boundary sphere, goes down as 1/R, because the power density is 1/R2 then, and the area of the successive shells is R2. That's a good way to design antennas, or to determine what the Q is of an acoustic transducer. That's how you do the gravity wave problem, too, because the gravity wave propagates with an amplitude going down like 1/R beyond this static-to-radiative transition sphere. And obviously you can't have a gravitational monopole, you can't have a gravitational dipole, because there are no external forces acting on the system; the center of mass is fixed. The lowest radiating motion has to be a time varying quadrupole. The field of the monopole goes down like l/R2, the field of the dipole at l/R3, and the field of the quadrupole like l/R4. So that gives you the whole answer to the problem. And at the other end the detector is just another antenna.

What was unique in our experiment is that we said, "This is going to be no use at all unless we can calibrate it." And what is the calibration? It's a change in the metric. We can't really do that with gravity waves, but we can stretch the bar with a very weak remotely applied electrostatic force. After a little bit of thinking about how you measure the amplitude of vibration of a half-ton bar, we would like to have a fixed pillar nearby, and measure the distance between them. But a fixed pillar without any noise in it is going to be very difficult, so one contribution that we made was to recognize that you could suspend from the end of the half-ton bar a small "proof" mass of a few kilograms. If the frequency of vibration of the mass on the tiny piezoelectric pill — just a couple of millimeters in diameter, which was our transducer — is below the resonant frequency of the bar, then the proof mass would move in opposition, and so the sensitivity of the transducer — the signal that you would get for a given oscillation of the bar — would be a greater signal by virtue of having this five kilogram proof mass than if you had a million kilogram anvil there. I can show you the brass proof mass in the laboratory. The absolutely essential calibration came from having a plate on which we put a few volts of alternating voltage at the other end of the bar for a few cycles. We made sure that the energy into the bar was proportional to the square of the time and the fourth power of the voltage and things like that. Then the question comes, how you adjust this electrostatic calibrator and measure its spacing, and it turns out that you can hang it from the end of the bar also. It takes a little subtle reasoning but you can think about it and you can also prove it out by experiment. And so we did those things.

I had for many many years been involved in rather big experiments. One big experiment I really did was at CERN 1959-60, where I was in charge of a group. We looked at a precision measurement of g-2 for the Mu meson. I had about six people working with me, and we had to design an 80-ton magnet which had a year and a half construction time, and a polarization analyzer and a vacuum chamber, which was really quite an effort for a small-scale person like me.

Aaserud:

That was unusual for you.

Garwin:

Oh yes, very unusual. But what was usual was, when the magnet got delivered, I had the crane put it down next to the cyclotron which was going to give us the Pi mesons coming in, and I had designed some air bearings for the magnet. We put some thin steel plates on the floor. And so that night after my magnet had been delivered, I went in and connected the tygon tube to the air bearings, and opened the valve, and now I could push the magnet around with one hand. Of course, the first thing I discovered was that the floor was not level. And if the floor is off by a milli-radian, that is, twentieth of a degree, that 80 ton magnet — 80,000 kilograms — is pushing you with a horizontal force of 80 kilograms. So you levitate it; it's very nice, it sits there, it's obviously not stationary any more; and then you notice that it is springing on you, accelerating like an enormous boat, and you try to push it back and it doesn't stop, and then you realize that all you need to do is turn off the valve and it stops. So then we did that.

But what I was going to say is, one of our people, I think it was Francis Farley, was in charge of developing the computer program which would take the individual counts from the polarization analyzer (which was done by George Charpak), and deliver to me the value of g-2. But I had no confidence that this would be done correctly, not because Francis Farley isn't a good person, but because people make all kinds of mistakes, whether in the algorithm or in the coding or in the sign or whatever. And so I asked myself whether I should have somebody else do this analysis independently, and decided I would be unable to check until too late. The right way was to make a simulator; so I made a computer program which would take incoming pi mesons, convert them to mu mesons, including random elements and what not, propagate the mu mesons down the magnet, trigger the polarization detector, eject the mu mesons and flip their spins and determine counts up and down. And the whole thing could be done as a function of a parameter A which is g-2, and I could put in any A, not just or something like that, which is the answer we expected to get. And I programmed the simulation. I gave Francis Farley the computer file with the pulses up and down at different times, and he gave me back the A, which was not the A I had put in, so I knew there was an error, either in my simulation or in his analysis program. So I told him, and he went back and looked at the program and fixed something, and then he gave me my value A. No matter what A I put in, he gave me the same A out after going through this enormous long calculation on my part and his part. So I knew that was right.

Soon after that, when I was involved in some satellite systems for the government, I insisted on a similar simulation in these complicated systems. You put a satellite up there and you send it controls, and if you have a misunderstanding of what's connected to what in the satellite, you can just destroy it, turn it the wrong way, as the space shuttle turned in the experiment with the laser, and so on. So we began, as a routine, to introduce the requirement for a simulation. People, before they launched the satellite, had to have a good simulation of it, and then the simulating computer would be connected via telemetry with the controlling computer, and you would see whether the whole thing worked.

There was an airplane, helicopter, which was being tested at Messerschmidt Belkow-Blum, I guess it is, in Munich, in 1966 or 1967, when I led my Military Aircraft Panel on a tour of Europe. Here, they had some stabilization system. Fortunately they were testing this thing on a test stand because they had the stabilization system connected with the wrong sign, and instead of holding the rotor horizontal, it just flipped it over immediately. It happens a lot of times. When we did our gravity-wave experiment, the other thing we did was to simulate gravity waves, and determine the efficiency of our detection algorithm in picking these things out of the noise. So it was a real experiment, unlike that of Joe Weber, unfortunately. We published it, and nobody has ever asked for our primary data which we offered, or criticized the experiment. Everybody else presents their data now in the form that we chose.

Aaserud:

So what triggered this experiment in the first place?

Garwin:

Well, my seeing all of the theoretical interest in this new "fact" that the world was full of gravity waves of an extraordinarily high intensity, and that nobody was looking at whether the fact was true. First I decided I would talk to Joe Weber and tell him how to fix his experiment. So I went to visit him, and it soon became apparent that there wasn't anything that I could tell him that he would do. It also became apparent, there were much easier ways to do his experiment. He had a subcontractor who manufactured many kilograms of this piezoelectric ceramic, which he then glued in the least efficient way around the belly of his bar, which doesn't move at all. The right way to do it is to put the piezoelectric between the end of the bar and a "fixed anvil"; and the difference is the difference between tens of kilograms of piezoelectric ceramic and a few milligrams.

Aaserud:

Let's go back to the consulting thing perhaps. Did that come automatically, that you had these consulting jobs? Was that something that physicists were generally asked at the time?

Garwin:

Well, I never looked for one. I did look for a way to make money in the summer of 1950, and that was the Los Alamos consulting job. I guess Hans Bethe was a consultant with AVCO. He was a long time consultant, and as a consultant helped to build their laboratory. I told them how I thought about these problems, and Bethe told Kantrowitz I guess, and Kantrowitz hired me as a consultant. General Dynamics, I guess, was a similar thing. Charlie Critchfield who was at Los Alamos, had gone to General Dynamics. They had an advisory committee, and so they wanted help, and so put me on it, and I suppose they wanted ideas or analyses or whatever. So those were the ones I had early on, and those were really the only commercial consulting arrangements I ever had. The rest of them were for the Argonne Laboratory or the Department of Energy or the CIA or the National Security Agency or any of these things, or Los Alamos. There I was just trying to help with government work, whatever it was.

Aaserud:

It seems to me that this kind of consulting or summer studies like the Lamplight or whatever were quite a common route for the early JASONs.

Garwin:

Well, the summer studies were really invented by the MIT crowd. There was the Charles River study, Project Charles, I guess, and various others.

Aaserud:

Lamplight was an extension of that.

Garwin:

Yes, there was nothing that couldn't be done, according to the MIT crowd, by a summer study. Lamplight was an extension in time and also went throughout the year, so it was a bigger study than any of these others, some of which were on antisubmarine warfare, or aircraft or whatever. I didn't participate much in the others.

And then when JASON was considered, I had been working directly as a consultant for the government. I'd been working for Los Alamos since 1950, and I spent every summer there through 1957 or 1958. In 1957-1958, as part of the PSAC assistance to President Eisenhower's response to Sputnik, I served on the (W.O.) Baker panel which had to do with our national security intelligence capability and activities — an effort in which I have been much involved for the U.S. Government ever since.

Then at the end of 1958 I was on the US delegation to the Conference on Surprise Attack. That was at the same time as the conference of experts on nuclear testing. It was housed in the same building, and since the Conference on Surprise Attack didn't have a lot to do, and since I was intimately familiar with nuclear testing and nuclear weapons, I went over and helped the other folks some. I remember once; we were proposing a certain characteristic of seismometers, maybe with a peak response at a tenth of a Hertz, and the Soviet group proposing one with a peak response at l Hertz — don't hold me to that. The question was, which would be better? And I decided I would use the computer. There was a computer at CERN, I knew that, a Ferranti Mercury. I would take a signal, which I would generate as that of a ten megaton explosion, and I would put it through the two seismometers, and I would add noise and so on. So I took the characteristics of the seismometers as they were specified. I didn't have the tools in mind for using direct calculation in spectral space and then to come back to the time dependence. Rather than do that, I built a mechanical model, that is, the equations governing a mechanical or electrical model of the seismometer which would have the specified frequency response. Then I fed back with the propagation behavior and the assumed signal, and added noise, and I had some graphs. I had never used the Ferranti Mercury; I didn't have permission to use it at CERN, didn't have a user ID or anything like that. This was in 1958, I guess. So I went out with an Air Force lieutenant colonel who had never been to CERN. I'd been there just once, I guess, and we found our way out there, and walked across the lawn, opened the door — there was a door from the lawn to one of these laboratory buildings — marched in, and here was an electronics lab; I could see the person was putting together coincidence circuits or something like that. So we looked around, waited until he came back. I introduced myself. He knew my name. And I asked him please to take me to his computer, and show me how to use it. We got a user's manual for the Ferranti Mercury and I coded this program, and brought it into a session of the test-ban negotiations, which of course had never before seen any such things.

I was not a member of the delegation, either, and so towards the end of the meeting — weeks later when things were not going to be happening — we were allowed to bring our wives to the meeting, just a few at a time. So Lois was there with me, and all of a sudden I hear my name. The Russians are speaking. They mentioned my name. So they were responding finally to the calculations which I had shown, so I had to come up to the table and answer them. So that was a lot of fun. Now, what question was I answering here? Oh, consulting, great. So that was government consulting, and that's all I'd been doing.

Aaserud:

That was in 1958.

Garwin:

1958 or early 1959. It started November 15, I guess, of 1958.

Aaserud:

Was that a particular group of physicists that went to these kinds of consulting efforts? Was that a well defined group?

Garwin:

Well, there are a few, relatively few; Panofsky for instance has been in this for a very long time. Then Spurgeon Keeny, who has a Master's degree in physics from Columbia University, was deputy director of the Arms Control Agency, but more importantly was on the National Security Council staff and the President's Science Advisory Committee staff throughout the 1960s; and a few others — Pete Scoville, now dead, and Albert Wheelon, now at Hughes Aircraft. Bill Baker of Bell Labs is a grand old man of such things. And John Tukey has been involved; he's at Princeton.

Aaserud:

What I'm aiming at, I suppose, is the origins of JASON for one thing, at any rate. Do you know whether that came out of a group of physicists with that kind of experience who wanted to do something more independent?

Garwin:

Well, first there was a need. I think JASON came from the need. In the 1950s, President Eisenhower realized finally that he was captive to a lot of majors and lieutenant colonels in the Defense Department who had their own hobby horses. There was no way to force them to be quantitative, to be fair in their evaluation of the impact and what not. When Sputnik was launched, the President's Science Advisory Committee, which was a sleepy committee in the Office of Defense Mobilization, was brought into the White House and attached to the President himself. I.I. Rabi had been head of the President's Science Advisory Committee in the Office of Defense Mobilization, but this was clearly a real opportunity for science helping the country which could be seized only if the chairman were a full time person, and Rabi was unwilling to move to Washington full time and leave his scientific research. So Jim Killian was selected, did an excellent job, even though he isn't a scientist himself. Have you talked to Rabi about these things?

Aaserud:

I have not.

Garwin:

That would be a good thing to do, because he's 88 years old, you know. He's not going to be here for a whole lot longer, and he'll tell you — he told me this just recently — so he'll be glad to tell you such things.

Aaserud:

But he's still perfectly able and willing to be interviewed?

Garwin:

Oh yes, that's right. I suppose if you talked to him too long, he might tire.

Aaserud:

Of course.

Garwin:

But his mind is still clear. So the President's Science Advisory Committee was formed and studied things. Where do we stand relative to the Russians? What is it they can do with these new technologies? What is the relative future? Jerrold Zacharias, of MIT was the spearhead of the secondary-school education reform, the Physical Sciences Study Committee, Biological Studies Committee, Chemistry Revolution and so on, and that was all very good, in my opinion. And also people asked, well, what is it technology can do for us? And it was I guess at that time, or a little bit earlier, that various people were involved in making the U-2 program, which was a simple recognition, long resisted by the Air Force, that the 35-mm camera exists, and that you don't necessarily get better pictures by going to larger film size and very long focal lengths. Those same people, Edwin Land and Ed Purcell and others — myself to some extent — made possible satellite observation as well. President Johnson said about the whole space program, that if it had just given us satellite photography, it would have paid for itself many times over. In fact that wasn't the "space program" at all. It had nothing to do with NASA. It was people, Land and Purcell, trying to do something specific for the country, rather than what NASA tends to do, which is to do things for NASA and hope that something will fall out of that which will help somebody some place — it's a totally self-serving bureaucratic enterprise.

Aaserud:

Were you involved in the establishment of PSAC as such?

Garwin:

No, but immediately after Sputnik I was involved as a consultant on intelligence matters and then on the Strategic Military Panel in the late 1950s. So from the beginning I was involved as a consultant and then became a member for the first time in 1962. JASON was a result of the government looking beyond the ad hoc advisory committee or the standing advisory committee of six or twelve wise men on a particular topic. There was a real need for people who would do some work, and furthermore it was clear that nearly all the people who were involved except for myself had been through World War II either in the radar program or the atomic bomb program, so we needed to bring up new people with a familiarity with the defense programs and with clearances, so that they would be able to help the President's Science Advisory Committee. So in part it was conceived as a kind of training ground for panels of the President's Science Advisory Committee. In part it was designed to provide better studies which could be reviewed, and in part it was to get the Defense Department to do a better job itself of management in getting proper programs, because it would have these people with whom to interact. JASON has rarely been a policy-oriented organization. When it tries to do that, our sponsors in the government explain very carefully that that's not what they want. In fact they don't want it at all, telling us that we're not good at that and so on. Well, that is not true. We're at least as good at that as anybody else is. But that they don't want it, is very clear. I guess Charlie Townes, when he was vice president of IDA, had proposed this JASON as Project Sunrise or whatever.

Aaserud:

Yes, that's the original name of it. I acquired some correspondence on that, his original letter of proposing JASON, and that kind of thing.

Garwin:

Yes. So I was, I think, listed in the original list of people who might be usefully involved. But because I was so much involved as a consultant and member of the President's Science Advisory Committee, I did not want to do this other thing until I saw how it worked; I thought it might be a conflict of interest. On the one hand, here I was trying to solve these things for the President, and on the other, I would be actually doing the work which I would then be reviewing, so I figured it would be better not to be involved in the work if I were going to be involved in the review. I joined only after JASON had been around for quite a while, and I guess I was no longer for a time on the President's Science Advisory Committee. Did I join JASON in 1966 or 1967?

Aaserud:

I'm not sure.

Garwin:

Probably 1966 or 1967.

Aaserud:

As I said, I have had some difficulties with the membership list. So I don't have any precise dates for that. The first of your Science Advising reports was, as far as I can see from your Bibliography, the Air, Sea and Space Traffic Control Report from 1958. That was the panel of PSAC, was it?

Garwin:

No, that probably wasn't. That was at a time when President Eisenhower was talking about Open Skies, I guess, or something like that. I just decided I would do this. But I was connected with the President's Science Advisory Committee, so I'm sure that I would have given that to Jerry Wiesner or whoever. It was just in response to something I'd read in the newspaper.

Aaserud:

The reason I pull it out is that it's the first statement that I found on your publication list that could be called science policy. That's why I'm concerned, that's why I'm interested in it. But what kind of forum did it reach?

Garwin:

Well, it actually reached the people on the President's Science Advisory Committee and so on, but because I remembered it and because it's not classified, I was able to send it out at various later times when I was trying to get my Air Traffic Control Panel, for instance, to do satellite navigation and so on. It didn't have a lot of resonance, and the problem is, there are very few people who are interested in policy and technology and have anything to do with the decision process. If I look it up on my computer and I say, to whom did I send this, I will know to whom I have sent it in the last six years or so. So I'll be able to answer that question recently. But in fact, it has a lot of useful things in there about this pulse compression and various things, and if Gerry O'Neill and company ever really do rent time on satellites for their commercial — [interruption]

Aaserud:

During the period we're talking about, of course, even if you were beginning to get into advisory things, both for government and other wise, you were mostly dedicated to research. Advising didn't become a full-time thing for you until a few years later.

Garwin:

No, really in the 1960s — that was my busiest time in the advising role. That is, when I was a member of the President's Science Advisory Committee, 1962 to 1965, and then I was consultant and then a member again, 1969 to 1972. So I had two four year tours. And from 1961 until 1967 or 1968, we had many panels, and they worked very intensively, say two days a month each. In that period, I was chairman of two of them, the Naval Warfare Panel and the Military Aircraft Panel. We also had the meetings of the committee itself and I did a lot of work for the various agencies and in preparing for these panel meetings. So more than half of my time was taken in this government role. Now, that was all secret; that is, was working on things which were the job of the President's Science Advisory Committee and its panels, and for the purpose of advising the President. So there was no occasion to speak publicly about such things, although in my various public policy speeches — to universities in general — I tried to handle questions in more generality. But it wasn't until the supersonic transport program that this became a serious issue. When Lee DuBridge first came in as Nixon's Science Advisor, he told the press that he had two panels that were looking at questions of some urgency to the administration, and that this administration was going to be very open with the press. DuBridge really believed what the Nixon Administration and staff were saying. He explained that there was a panel under Marvin Goldberger, which was looking at the Sentinel system or the ABM system, and a panel under me that was looking at the supersonic transport — two questions that were before the nation at the time — and that he would keep the press informed, and he hoped that it would be possible to give them copies of the reports. Now, the administration didn't like the reports, and so they refused to make them public. Having announced them this way, DuBridge set in motion all kinds of public interest. The public now knew that there would be or was such a report, and so I received requests from the Congress to testify on these points. I didn't for more than a year, but eventually when there was in the testimony all kinds of things from the government — the Department of Transportation, the FAA, Mr. Beggs who had led an inter-agency committee — which were deceptive, I decided that I should testify. Of course, I would not discuss our report, but only speak on the basis of what information the Congress had in front of them and which for some reason or other they weren't able to interpret. So that's what I did.

Earlier, my public aspects of my policy role began as a rather routine thing at the Christmas meeting of the American Association for the Advancement of Science in New York, as the session on ballistic missile defense ended. It must have been 1967, when Mr. McNamara had announced in San Francisco in the fall that although a defense against strategic ballistic missiles didn't make any sense, we were going to deploy a "thin" area defense against the Chinese nevertheless.

Aaserud:

That was the best argument ever against anti-ballistic missiles.

Garwin:

Yes, so there was a lot of interest in that, and the AAAS arranged a session. I think Marvin Goldberger was chairman, Hans Bethe and I were two of the speakers, and Gerry Piel, the publisher of the Scientific American at the time, told us as we were getting off the platform that if we could write him a paper right away, he would publish it right away. So Bethe and I wrote the paper in Scientific American. He actually wrote most of one part of it, I wrote most of the other part, and I took charge of putting it together. That was probably my first real open publication in science policy, but there had been all these other speeches and what not. And of course it's something which I had been doing ever since 1950, not only trying to improve air defense, but deciding whether we needed it. I told the leaders of Project Lamplight in 1953, Zacharias and Wiesner, that I didn't see why we were worrying about extending the air defense to the oceans, when, by the time we could do any of that, the threat could really be ICBMs and not aircraft. So Jerry Wiesner and Jerrold Zacharias said, "Well, let's do this first and then we'll do the other when we get around to it." But you have to look a little bit farther ahead than that.

Aaserud:

But your argument, was that introduced into the Lamplight conclusions in any way?

Garwin:

I don't know. I didn't have much to do with writing the report. It's not something that interested me very much. I guess I would be more interested now in doing such a thing, because it's the report which has the influence, and not just the analysis that goes into it.

Aaserud:

Before your public public role, so to speak, you did a lot of internal advising. It didn't go out into Scientific American, the New York Times and what not.

Garwin:

Yes, and probably that's because I had then a higher regard than now for what the government would be able to accomplish. It was a different era, in which the Administration had more influence, relative to the Congress, and so you didn't have to educate the Congress or the people in general in detail. This paper in my bibliography from 1962 on command enable switches is just an indication. It was originally secret when I wrote it, but I reviewed it recently and declassified it, and it's just an example of the things that I was doing in that era.

Aaserud:

Was that part of your PSAC activity; was it a panel?

Garwin:

No, it wasn't a panel. It was when President Kennedy came into office. Jerry Wiesner had been most interested in nuclear weapon security and arms reduction and so on. That was early in 1961. At the same time, there had been a trip in just recent years to Europe, with Harold Agnew and others, which found that nuclear weapons on German operated airplanes or other Allied airplanes were really not properly secured against misuse. If you dropped them on somebody, they would go off with a nuclear yield, so it was only procedural safeguards that we had. With the combination of Wiesner and Keeny (and myself in the President's Science Advisory Committee, or at least in its family, with long experience in nuclear weapons and electronics), with Don Cotter from Sandia and Harold Agnew from Los Alamos, and with favorable views from the Joint Chiefs of Staff and Secretary of Defense McNamara, this was proposed and done in a matter of months. We managed to lock up all the nuclear weapons in Europe and all the other ones except the ones that are on submarines. So that was a good thing.

Aaserud:

That was something that really had an effect, that was acted on.

Garwin:

There were many things like that.

Aaserud:

There was one project that wasn't acted on. That was saving energy by superconducting lines. Did that ever come out?

Garwin:

Yes, I did that for the President's Science Advisory Committee in 1961. We were looking at the problem of unemployment in Appalachia, and Jerry Wiesner said, "Well, how about setting up coal burning power plants at the mine mouth, so you wouldn't have to transport the coal?" So I looked at various transmission prospects, and decided superconducting power lines was really the way to go — at large power transfers, at multi-gigawatt levels. It doesn't pay below that because the cost is going to be the same no matter how small the amount of power. In fact, I have a patent on this thing. So a lot of design is involved, and a lot of things I published then have appeared since then in cryogenic systems, for the transport of cryogens over long distance, and in superconducting systems.

But then when I tried to get the Department of Energy to work on it, the people wanted to work on difficult problems, mostly physical problems. So people at Brookhaven went to work on AC superconductivity, and therefore, if you could get a good cheap AC line, they could use it for entries into cities, where real estate is very costly, and it would be possible to use it at a lower power level (with higher competitive costs per KW). When I complained that the AC problem was a much bigger problem because of the vibration and squeezing and losses and so on, Forsythe I guess said that when they solved the AC problem they would have solved the DC problem too, so why was I complaining? Well, that's like saying that when they solve the problem of eternal life, we will have cured AIDS, but one of them can be done sooner presumably than the other, and would have some benefit.

So the folks at Los Alamos who were interested in DC things were just really too sleepy to do it. It's just that there was no particular motivation. General Electric and other companies that sell to the electrical industry don't have much of an incentive to pioneer a new technology. I kept track for a while of the status of inverters and rectifiers, because you need some inverters which are priced at $1 or so per kilowatt. There's no reason for them to cost more, but they weren't going to cost that little because the companies were already selling equipment that was considerably more expensive and if they reduced the price to that level, then their profits would fall too. So it would only be some other company not in the business that could be relied upon to make that innovation. That's why that hasn't happened yet. But there's no doubt that it's feasible. Now we have in particle physics establishments miles long superconducting cables with enormous numbers of magnets in the middle, and miles long liquid helium chambers. So all those things do happen, and it's only a matter of deciding that one wants to do it, and asking about the other policy questions of terrorism and so on.

In the last couple of paragraphs in the 1958 paper on Air, Sea and Land Observation by Cooperative Means I talk about the problem of jamming satellites. A few months ago we had the "Captain Midnight incident" in which somebody who worked at a TV station in Florida was upset about the scrambling of the Home Box Office communication. So for several minutes at midnight of one day, he put on his own legend that said, "Worried about scrambling? You should be," and so on. They tracked him down recently. But what I had proposed long ago is that when you have important satellite communication systems which are vulnerable to jamming, then you should use them in order to locate the jammers extremely accurately by time difference of arrival. I had noted that in my 1958 paper.

So there are many many things. Sometimes only little things get in. For instance, in our military aircraft activities we tried to get people to use helmet mounted foveal television, so that you could just take out all of the switches, gauges, indicators from the cockpit of the airplane. Wherever you looked there would be the simulated switches, gauges, indicators and so on generated by a computer, with very reasonable resolution requirements and data transfer requirements, because it wouldn't give high resolution, or a lot of pixels (picture elements), to the eye where it can't use them. The fovea of the eye is only about that big, about half a degree, and so very little information need be presented at high resolution to make the eye think that it's all there at high resolution. I tried to get the military to do that, then I tried to get IBM Federal Systems Division or any other contractor, then IBM Commercial. People kept saying "It can't be done." So Jim Levine and I then demonstrated all of the relevant technologies, and we tried to get the Air Force Aerospace Medical Center to pick it up. Now we've made available all of our knowledge and technology to the University of Virginia where they have a crew of people. These things — making a whole system and selling it — are fairly hard if they're beyond doing just by yourself. Finally, with the perfect resolution display in the implementation that we were considering, it was going also to have three dimensions, because we present different images to the two eyes, and full color. So it's obviously a pretty interesting sort of thing. We showed that you could track the position of the eye fast enough in a non-intrusive manner. We demonstrated this by making an eye-controlled terminal. And we demonstrated some of the other features. We made some inventions of multiple laser beams being modulated by the same crystal, and of compensation for facet tilt on the rotating mirror. Many of those things had previously or simultaneously, one way or another, been put into military systems through my consulting for the government.

Aaserud:

Which time are we talking about now?

Garwin:

Well, we were doing this work 1979 to 1981, so the helmet-mounted display work for the government had been long before.

Aaserud:

What about conflicts of interest or accusations of it? The SST argument of course could be seen as being against IBM; IBM might have some interest in it and you were against it. The high power superconducting lines conceivably might be something IBM would be interested in.

Garwin:

I didn't have anything to do with IBM in that regard. I published it. I did it for the President's Science Advisory Committee. And after it's done, my rule has been that I'll talk with anybody with a government contract. I will try to improve their effectiveness in doing something. I'll talk to people by publishing so that everybody has equal access. So I deal with IBM just the way I deal with anybody else. And if I know some thing in IBM which is useful to the government, I will, if it's not IBM Confidential, tell them, and if it is IBM Confidential, I will try to make it available to the government by getting approval to do that.

Aaserud:

It's a difficult position that requires some balancing, I'm sure, at times.

Garwin:

It really doesn't. I don't get involved in anything in which IBM is interested commercially. That is, if it's a matter of choosing this kind of computer or another for some application, that's not my business. If somebody has a non-IBM computer and I see a way to improve its effectiveness for government purposes, I just go ahead and do it.

Aaserud:

So in general that hasn't been a problem at all as far as you're concerned.

Garwin:

No. There was one highly classified Air Force study that they wanted me to look at. I think this may have been in the Carter Administration. We had an early briefing on it and I filled out the various possible conflict of interest forms. They just didn't find any way in which it would pass the legal requirements, because there was "too much of a possible conflict of interest." What was involved may have been at most about 25 million dollars worth of computing equipment. I wrote them a letter. To imagine that my views would be swayed — in a company which was probably 20 billion dollars at the time — by whether IBM would or would not get the contract for 1/1000 of that in computing equipment was ridiculous. After all, if a salesman can sell a 25 million dollar item, that's a good thing and is worth a lot to the company. But to think the company's view of me would be modified imagines that they would know what I had recommended, which they wouldn't. But I was just as happy not to do it.

Aaserud:

Have you been involved in many classified reports like the declassified one on command enable switches already discussed?

Garwin:

There are many, many classified reports. In fact, some of them I have here in my safe. I'll see whether I can just give you a list of them, then I can talk to you about it. All too often people like Gregg Herken for instance seem to believe that the universe is what they see of it. If they will find a few documents, they think that's a person's work or a person's view, and never mind that it's only a tiny portion of what that person has done.

Aaserud:

Well, it's a problem, it's a genuine problem of course.

Garwin:

That's right, that's right.

Aaserud:

If you don't have access to it, it's very hard to know what isn't there.

Garwin:

Yes. (Garwin is reviewing index cards — one for each secret document in his safe) So I have several categories of secret documents. One is Department of Energy things, and another is National Security Council consulting, mostly when Henry Kissinger was advisor to President Nixon from 1969. Some of these documents pre-date that substantially. I have notebooks of 1966 and so on, and that was involved with the PSAC work and the JASON work.

Aaserud:

Was that technical advice or does it also include political advice?

Garwin:

Well, it all has a technical component. It was advice about MIRVs and things like that. Some of that is in the Herken book, but not much. And it was discussing the role of the MIRV, of the ABM system, of some specific verification techniques — what the United States ought to do in this regard. As I go back here to 1962, I have a note from me to members of the AICBM Panel, the Anti-Intercontinental Ballistic Missile Panel — the ABM panel it would be called these days — of the President's Science Advisory Committee, and to Panofsky. It's on Nike Zeus. I went to Bell Telephone Laboratories in Whippany in September 1962 to discuss their fancy radar. I proposed some new pulse compression schemes and what not, so this was technical. Eventually our reports would say, "You know, Mr. President, don't get the idea that all this technology is ready for deployment because here are the difficulties." Then I have here a document from 1963. This is by no means all of the things I did, because this is 1963-29, so there were 28 other documents which have been destroyed by now. This is a report of the Tactical Aircraft Subpanel of the PSAC Limited War Panel; "technology for limited war"; 1964 Anti-Submarine Warfare Panel, "comments on the recommendations of the report of the Tactical Aircraft Subpanel"; three memos to Don Hornig in 1964 on ICBMs; a memo to the Anti-Submarine Warfare Panel, "the choice between mines and nuclear submarines"; a memo from me, chairman of the Military Aircraft Panel, in 1964 regarding the program — that's a self-contained surface-to-air mobile missile system, which we cancelled — a report by the Military Aircraft Panel on the C-X HLS — that is the Heavy-Lift Support aircraft which caused a big argument in 1964.

It was all contained within the committee, because the government was honest about it. Harold Brown was Secretary of the Air Force, and our panel had done a thorough study. You might call it a policy study as well, because we not only looked at whether the C-5 airplane could be built — no problem in building it — but it was proposed for a specific task: to deploy five or eight divisional equipment sets to Europe in 30 days. We said, if that's the job, we can do it seven times more cheaply with ships, and we designed the ships and we got them (government contractors) to run the calculations, and sure enough it was so. So eventually, after all these reports, Hornig had gone over to McNamara and discussed it with him. We got what I really wanted, which was a chance to argue this before the committee with Charlie Hitch, the controller of the Defense Department, and Harold Brown. We did, and sort of in the middle Charlie Hitch got up. The room had 20 people or so in it, and he stamped out saying, "I still say airplanes are better." But his arguments did not persuade.

And there are things about European trips and what not. Here's one, from 1966. My panel was asked to review the military applications of supersonic transport, and we said that there really aren't any. We have supersonic airplanes. Here's one report on helicopter supported radar. For many years our committee had been trying to get the Air Force and the Army simply to lift ordinary perfected ground radars to a high place with out a mountain — just hold them from a helicopter — and have them work. They weren't really interested in doing that. They were interested in making AWACS, jet-speed aircraft and so on, which is much harder. So eventually, when Al Flax (a member of my Military Aircraft Panel) became Assistant Secretary of the Air Force for R and D, he did get a million dollars together to test this, and it worked very well. There's no interest in it though, because it's too cheap a solution to the problem. And of course, the periscope radars, detecting periscopes from submarines — these are all from 1966 — and so on.

Aaserud:

Is that a PSAC file you have there?

Garwin:

Yes, almost all of these in that 1965 to 1966 era are President's Science Advisory Committee work. We also worked on accurate convenient navigation. We were trying to get what is now the Global Positioning System (GPS) deployed. For many many years I was trying to do that and kept it alive in the Defense Department, even in the Carter Administration. I would talk to Harold Brown because the chief of staff of the Air Force, Lew Allen, said, "Well, this is a high priority with us but we just don't have the money." So I got the money.

The C-5 aircraft; the proper mix of air-lift and sea-lift; in the aircraft business, the desirability of replacing essentially all manned aircraft, except for transport aircraft, by unmanned platforms, vehicles and so on — those are long-term policy issues in which certainly we have not totally succeeded, although when you create these things and people really need them, if they have enough time to introduce them, they will succeed. We dealt with laser-guided bombs, which we didn't invent. The Air Force in its operational development provided the kind of funds that are used to fix up a system. Somebody at Eglin Field had a good idea, and they got Texas Instruments to make an add-on kit for an ordinary bomb, thereby eliminating the problem of having to redevelop things that are perfectly satisfactory. The whole thing cost about $4,000 per bomb, something which, under special development, would cost $100,000.

Acoustic gun locations, and counter battery radar — those are things that we did too. The question of vertical take-off and landing aircraft, and the role of such things in the air traffic control and civil airplane systems are interesting — the competition between short take-off and landing. People kept saying that for passenger transport you ought to have a generation of short take-off and landing before you go to vertical take-off and landing aircraft. That's like saying, as the Air Force did, that before we go to ballistic missiles or unmanned satellite surveillance, we ought to have people in orbit taking pictures, or for that matter, riding bombs, directing bombs to their targets. That's just dumb. It's much harder to do it that way than it is to do it in a totally unmanned system.

So there are those military technology questions, and policy things like the amount of freedom that a laboratory director should have in order to set the course of the program of the laboratory. The counter argument was, "You don't have very good people as laboratory directors; they don't have very much power, so maybe that's why, and maybe it's good that they don't have very much power if they're not very good people." We tried do break that chain and in fact at one time were even told by the Assistant Secretary of the Navy for R and D that he would give us (our PSAC Naval Warfare Panel) five million dollars or something that we could spend in his laboratories, but he rescinded that. It was not a role we were looking forward to.

So that's the sort of thing that happened in the 1960s, and of course the 1960s was much involved with Vietnam too. There the problem was that people in high positions really have to work very hard in order to keep their minds and their eyes and ears open to views different from their own, because many people get along in this world by finding out what their bosses want to hear and then telling them; and that's not the way to run a company or a democracy. It may be good for the individual, but it's bad for the system. In our military panels, we would get routine briefings on the Vietnam War, from the people who weren't on the military panels but were very smart people. We would ask difficult questions just to be helpful. After McGeorge Bundy and others had sort of lost heart on the war, they brought in Walt Rostow, who was gung-ho for doing new things. But there weren't new things, and the reason that he was so enthusiastic was that he didn't have the benefit of experience. Chester Cooper, a very capable person, was on Walt Rostow's staff, I guess. He came in and briefed the President's Science Advisory Committee. I remember writing him a letter and explaining that we wanted to help him, because we had a panel on the Vietnam War; we had all this experience, which he didn't know about. He had been for months across West Executive Avenue in the National Security Council staff, and he had no idea that there were people who had good channels for finding out, who had a lot of information, and who had made recommendations. It's just too bad that the government doesn't make better use of the information that is available to it. So anyhow, there was a lot of science policy, but it was just not public.

Aaserud:

How much of the PSAC work at that time is unavailable or restricted now?

Garwin:

I don't know.

Aaserud:

Do you have any guess, good or bad?

Garwin:

No. I never tried to look up any of the PSAC work in libraries one way or another. I don't know where it is. I suppose it's in Presidential libraries but I don't know.

Aaserud:

It would be reflected in what's in your safe and what's open with you, I suppose.

Garwin:

Well, no, because when something becomes obsolete, I throw it out from the shelf, but not from the safe from where I can't throw it out. I can send it back to somebody to be destroyed.

Aaserud:

So your documentation on PSAC in those boxes is much less full than what is in the safe.

Garwin:

Many of these classified things have become unclassified with the expiration of time, because many of them were marked, "Downgrade after 3 year interval," "Declassify after 12 years," and so on.

Aaserud:

Have you kept up with that?

Garwin:

No.

Aaserud:

That's still in your safe so that's safe, I see. So it's you who make the decision as to whether something is still current.

Garwin:

Well, if it has a marking like that so that it has a fixed expiration date for the classification, then I am willing to declassify it, if I look at it and see that it is properly classified in the first place. That's a lot of work, so I don't do that except in rare cases. For instance, I chaired a committee for the Defense Science Board in 1968, I guess, which produced the Advanced Tactical Fighter Task Force Report. I just came across this two volume report in my safe a couple of years ago, and I noted that its classification had expired in 1980. Now I send it out to people who can benefit from it.

Aaserud:

I'm also talking about the unclassified things. Is it you who decide personally what material you don't need any more?

Garwin:

Nobody else decides for me. But mostly I don't have time to decide, so it's sitting there in boxes.

Aaserud:

It's more there than you need for use now.

Garwin:

Right.

Aaserud:

But you have thrown out some things.

Garwin:

Probably. In recent months the only thing I've thrown out I believe are old expense accounts. And travel things and what not. I haven't thrown out any substance recently. Well, I'll show you some of these boxes before you leave.

Aaserud:

Yes, I would appreciate that. Since I'm interested in JASON, and since there obviously was some overlap of personnel between JASON and PSAC, how was the division of labor between PSAC and JASON? Was it entirely different kinds of work that JASON took up, or were they asked the same things in a different way? How was that?

Garwin:

For the most part the JASON sponsors — originally it was ARPA, the Advanced Research Projects Agency — had a few big questions on ABM like discrimination of warheads from decoys. In the Strategic Military Panel of the President's Science Advisory Committee in 1965 or whatever, when we were talking about those things and looking at the results from radar discrimination, probably I said there are two kinds of payloads which can't be discriminated at all — and one is an all-decoy payload, the other is an all-warhead payload. So why don't we give up this problem of making the decoys look just like the re-entry vehicles, and consider these two approaches. One was a MIRV, the other was just spoofing, and no matter how good they were, they'll never find anything which is different from anything else in there, and if they think there's one warhead, then they will have to shoot them all. The other approach that came up at the same time was anti-simulation, i.e., make the decoys crudely, make them give off ionization from salt pockets or whatever in their skins. Then you do the same thing with the warheads, rather than try to make all the decoys perfectly match the warheads. When we said that, the interest somewhat went out of all of these discrimination efforts, because we knew it could never succeed except if you could actually weigh the warheads, so there was a lot of work on weight detection and what not.

So JASON was mostly the technical work — very smart people doing actual analyses, some of data but mostly theoretical. PSAC was mostly talking. We would occasionally have equations in our reports but it was fairly rare. And PSAC would deal with the real problems, whereas JASON would deal with the problems as they were said by the agencies to be real which were very often details or bureaucratically driven problems. I remember a meeting in 1960 of the Strategic Military Panel, when Dick Latter and Al Latter came in and talked about the newly perceived vulnerability of re-entry vehicles to X rays. This is something people hadn't thought of before — that the X rays into which the space exploding nuclear weapon gives 70 or 80 percent of its energy come out in a very short time and are absorbed then in a very thin layer on the surface of the weapon or its ablative coating. If they have enough total energy per square centimeter (fluence) to raise the temperature above the boiling temperature, then a thin sheet of material comes off at high speed. That doesn't seem bad because it's very thin, but it sends a shock, its recoil, into the material. This is discussed in our Scientific American article of 1968. So you have to take these things into account.

And I remember thinking about it, and pointing out that this could be mitigated; that is, if I have something which is coming down and slapping here, it will hurt, but if I put on the table under the slapper some foam plastic, then I can absorb the momentum, and reduce the peak force. So the next time I came to PSAC, I brought from home a keyhole saw blade and a piece of foam plastic half an inch thick. If you bend up the blade and then let it slap down on the table, it really sounds quite fearsome, and if you put your hand under it, it really stings. But if you cover your hand with the foam plastic, and you do the same thing, it doesn't hurt at all, which is obvious from our experience. But it really works, and that's how you protect reentry vehicles against X rays, and that's exactly when that technique was discovered. I did it. I was talking to some very smart people in the nuclear testing business years later. One of them I had known since 1951 when we were devising some of the test equipment for the hydrogen bomb experiment. He said he should have thought of that, because for many years in the underground tests we had used such techniques for catching samples and not destroying them. I don't know whether that example is policy or technical, but it certainly has big impact on what kind of warheads you build, and what you decide about having ICBMs and how much you fear the defenses on the other side. So it's this combination that PSAC did, whereas JASON was much more lengthy calculations for the most part.

Sometimes JASON has a brand new idea that had not been discussed previously. SST had done some kind of study in which they ridiculed the idea that the exhaust would do anything. But they had looked only at water vapor, not nitrogen oxide, and it's the catalytic action of the nitrogen oxide that is crucial. So Rowland and then Harold Johnston looked at this, and it came slightly too late for our PSAC committee report; we didn't need it to draw our conclusions. But then the Department of Transportation asked JASON, or JASON suggested to the Department of Transportation that this was a reasonable thing to study. The late Henry Foley and Mal Ruderman began to study it. They had some theory, and looked at the theory. They weren't great chemical theorists, and then they said, maybe there's some experience. They knew that when you have nuclear explosions in the atmosphere, the fireball, which is in equilibrium with large amounts of nitrogen oxide at high temperatures, would in equilibrium give up all the nitrogen oxide at low temperatures. But because it's out of equilibrium — the rate of breakup of nitrogen oxide gets to be very slow at low temperatures — the equilibrium concentration at 2000 Kelvin or so is frozen in. That's about 1 percent. And these fireballs are very acid — one percent nitrogen oxide — and as they rise, they deposit tens of thousands of tons of nitrogen oxide in the ozone layer. They looked for the image of the Soviet 1962 test I guess it was — a 58 megaton thing — and they didn't find it. But then they got to thinking, OK, so you can't see one nuclear explosion, but what would happen if you had thousands of them? So that was a very unpleasant surprise — of the possible damage to the people and the ecology from nitrogen oxide in a large-scale nuclear war. That was then looked at later by the National Academy, and it's really true, it's a real problem, which of course has been dwarfed by "nuclear winter," due to smoke from forest or urban firestorms set by nuclear weapons, which was a totally unexpected phenomenon which should have been thought about by all of the experts but wasn't, even by Foley and Ruderman. So JASON is not usually asked to do policy-relevant questions, although in recent years they have been asked to do C02 reports and I guess acid rain reports and things like that, which are very largely policy-relevant, even though they haven't been asked for the policy recommendations.

Aaserud:

But they do have policy implications, of course.

Garwin:

Yes. So, tell you what, I think the best thing would be for me to show you some of the boxes and let you poke around in there, because there are things I ought to be doing, and we can continue this at a later time.

Aaserud:

Sure.

Session I | Session II | Session III