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.
See the catalog record for this interview and search for other interviews in our collection
Interview with Richard Garwin
Aaserud:I would like to fill in on some of the questions from the last time and then go on in chronological order from there. I've also looked at a very tiny part of your correspondence here, and some of the questions relate to that as well. First of all, beginning with your youth, I did notice in some of your correspondence that you had at least one brother. I would like you to fill in a little bit about what you would say about his or their influence, if there were more siblings, on your early life, your career decision and interests and that kind of thing.
Garwin:Well, my brother Ed was born five years after me, on March 22, 1933, and so he didn't affect my career decision very much. He was my only sibling and he eventually, like me, went to Case Institute of Technology in Cleveland to get a Bachelor's degree in physics. He then went to the University of Chicago and got his PhD in physics there -- with Val Talegdi, I guess. I think he started off working for, or he wanted to work for, Fermi, but by the time Ed got to Chicago, I think Fermi lived only a year longer; Fermi died in l954. Then my brother went to the University of Illinois in Urbana-Champaign for a couple of years; he joined a small company in Los Angeles for a couple of years; and ever since l960 or 1962 he has been at SLAC, the Stanford Linear Accelerator Center in Palo Alto.
Aaserud:So it was more your being an influence on him than the other way around probably.
Garwin:I suppose. He has a very good memory. He remembers all kinds of slights and terrible things I did to him. But aside from being a long distance from one another, we're quite good friends now.
Aaserud:You don't remember any particular instances or cases of collaboration or having some kind of common interest in your early youth that helped you along? I guess the age difference was too big for that.
Garwin:Well, one thing was important. He had a couple of good friends, and his friend Howard Levy came to our house a lot because Howard's parents worked. They had a grocery store, and Howard's older sister would call up all the time and ask us to send Howard home for dinner, and I got to know this girl, who is my wife now. So there was this very important influence.
Aaserud:Has he had the same kind of broad approach or development of interests that you've had, from physics to broader concerns?
Garwin:He has been much more usual. He has worked at SLAC primarily. Although his PhD was in particle physics (I don't know what his thesis was actually) but he has worked largely on instrumentation or surfaces, vacuum systems, outgassing -- things which are non-particle physics oriented, but are fundamental in building and operating accelerators.
Aaserud:Also relating to our conversation in the car, are there other influences in the early home or environment that you could point to? We didn't go too deeply into that. I don't know how deeply you want to go into that or can go into that.
Garwin:Well, did I tell you my father had a Bachelor's degree in Electrical Engineering from Case from 1921? He then went to work as a high school teacher in East Technical High School and as a motion picture projectionist, which was compatible with that because they had usually two shows a day, with even the afternoon show beginning after the high school classes were over; and there was shift work and what not. There were college books and engineering handbooks around the house, so I read those when I was a kid. When my father was building things for teaching other projectionists about the introduction of sound films and so on, then there was equipment around the house -- old movie equipment. I would take it apart or try to understand how it worked. That's essentially all. We didn't have any other friends active as an engineer or a scientist. I wasn't very active socially and I wasn't very good at sports and I didn't want to work for anybody, so I figured I would be a scientist; you didn't have to work for anybody then.
Aaserud:The influence was more getting it in into the fingers than reading yourself theoretically to it.
Garwin:Yes, that's right.
Aaserud:I have a number of miscellaneous questions. First of all, patents -- your first patent was rather early, wasn't it?
Garwin:Well, I had invented a lot of things as a child -- that is, the cascode amplifier and some other things which were also invented by other people maybe even earlier. Naturally, when you invent something, you want to get rich from it, and so very early I thought about patents. I guess I tried to patent the pulsed homopolar generator, which I wrote up for my Bachelor's thesis in 1947, and also a projection television system, which I worked very hard at the first summer I was at Chicago, (so it must have been the summer of 1948). I came back to Cleveland and I worked at Case. They gave me some space to work; I wanted to have a diamond faceplate for my television tube, but couldn't get diamonds that big that were transparent. So I used Z-cut quartz which has a pretty good thermal conductivity. Of course now everybody is rediscovering that diamonds have thermal conductivity which is better than that of copper at room temperature because of the high Debye temperature. In fact, in 1952 at the University of Chicago I began to work on the growth of crystals in regions of the phase diagram in which they are unstable in respect to other phases. I began to work on sulfur and then (at IBM in 1953) on diamond. Probably one of my first patent efforts was this projection television system, and then there were some other things I tried to patent. One of the early ones was an outgrowth of my work on spin echoes. The wonderful thing about spin echoes, compared with steady-state nuclear magnetic resonance, is that in steady-state NMR, you have to be looking for the spin signal in the presence of the inducing signal; therefore you need all kinds of balancing and what not, because noise on the inducing signal contributes a problem, and you have to balance it out. For spin echoes, you can actually turn off the transmitter, and wheel it away. Then the echo comes back a second later or so, with no radio-frequency noise in the vicinity. I was thinking about this, and I was involved with anti-submarine warfare and things like that. It occurred to me that by putting a loop of wire on the ground -- or into the ground for that matter -- and putting a current into the ground for some seconds, one could have the eddy current spread to considerable depth, depending on the conductivity of the earth. Then, by cutting off the current and looking at the resultant voltage versus time induced in a loop, one could map the resistivity versus depth of the ground, magneto telluric sounding. We had a patent search, and I don't remember whether I had a patent issued on that. I think not, because there was some prior art, and eventually it gets rather expensive, so sometimes I gave it up rather than spend the extra thousand dollars or so. So there are many ideas in communications or in low-cost improvements in hydraulic brakes and stoves, temperature controls for stoves, instrumentation of various kinds.
Aaserud:Has that played a significant role in your career in any way in relation to your other activities?
Garwin:I've never made any money (directly) from any of my patents. Actually, I did; there was a patent with Jerry Wiesner and Dave Sunstein which grew out of some air-defense work in 1953 and 1954. This was essentially not a packet-switched network, but a multiplex radio channel communication system, with speedup of the individual packets. But eventually IBM decided that my employment contract forbade me to have this independent patent. I argued with them about it, and eventually they bought it from me for a thousand dollars or whatever. I had a lot of complaints about that, because each of the inventors has the right to practice the patent, so that reduced to zero the value of the patent to Sunstein or to Wiesner. But like most of these things, the patent expired before it was ever used. I think now it is used. And the rest of the patents -- I gave you a list of them, didn't I?
Aaserud:I don't think so. Not a separate list of patents.
Garwin:I'll give you a list of patents. And I have a list of Technical Disclosure Bulletin items in IBM, where we want to be able to use the invention ourselves but not to keep other people from using it. We publish it in an IBM Technical Disclosure Bulletin.
Aaserud:You mentioned that in general terms last time. I don't think I got a list of that either.
Garwin:I'll give it to you because that's interesting.
Aaserud:Has application for patents provided any kind of motivation in your work in any way, or is it just something you do on the side altogether?
Garwin:Well, it's really very difficult to apply for a patent; it takes a lot of effort, depending on whom one is working with. I have applied for only one or two independent patents in recent years, most recently for a device for washing shellfish; with a long-time (since 1947) chemist friend, Harold Friedman, I have a patent on a mussel washer, and it's really quite a good thing. We built many generations of them and they worked fine. If you're into gathering mussels from the sandy bottom, and you don't like to have the sand grit in your teeth, this is a very good way to do it.
Aaserud:How high tech is it?
Garwin:It's not very high tech at all. It has floats -- styrofoam floats; it's a perforated can. It floats on the water so that when you're gathering mussels you put them in, and it has a crank by which you turn it. It doesn't have any bearings. There's nothing to get out of order. It's very satisfying to put 10 liters of mussels into one of these things, turn the crank, and you see positively black water coming out. All of the sand and muck and some of the roughness on the shells gets removed. Then you can have a whole pot of mussels, and the last spoonful of liquor from it has no sand in it. So it's very successful. But we haven't made any money out of it.
Aaserud:But that's something for the individual amateur mussel picker, so to speak.
Garwin:That's right; you could copy it. But anyhow, it was a lot of work and something which we shouldn't have done.
Aaserud:How typical or untypical do you find that kind of activity for a physicist?
Garwin:Well, many physicists disdain patents. That is, they want acclaim from their fellow physicists. But you know, there are people, especially in commercial enterprises, who have many patents. Luis Alvarez, a good friend of mine, is always inventing things and patenting them; I think he has never made any money from his patents either, except maybe recently. I remember when he was on my Military Aircraft Panel of the President's Science Advisory Committee, his wife was ill with malaria on a trip that they had taken, and so he invented a variable power lens. [Actually, I think that's when he invested his stabilized binoculars] If you take two surfaces which have cubic terms in their surface profile, and move them -- one past the other -- then the resultant can be arranged in the region of overlap to have a spherical path length versus position from the axis as a function of relative motion. The power of that sphere can be arranged to be linearly dependent on the relative motion, so you can have a lens which goes from -2 diopters to +2 diopters as one moves this. He tried hard to get people to use that, and I think his patent has expired. Now Polaroid uses that (as their "quintic lens") in their Spectra System camera. Another interesting thing that he invented is a stabilized binocular system in which you don't hold the whole binoculars, which would require a big gyro. He recognized that if you have an optical system with an angular magnification of two, then if there is some place in that optical system a mirror, which is stabilized in inertial space, then, no matter how you move the rest of the optical system, the light beam that comes out is not going to move. He and his wife and some others have just recently set up to manufacture, not stabilized binoculars, but they figured they could get more money by putting a stabilizing element into a zoom lens for TV use; and they've been selling those. So it's not very usual. I don't think that I would do it to make money, mostly because I'm too busy doing these other things in arms control or whatever. Whatever I want to do seems to conflict with that.
Aaserud:In general terms, what motivation drives you to apply for a patent?
Garwin:It's part of my job. The whole patent system was created early in the history of the United States, and of course in other countries too, with the idea that if you give people a grant of monopoly for a period, then they will have an interest in perfecting techniques or inventions, and putting them into use. In exchange for revealing everything about it -- enough so that anybody can do it -- you get the grant of monopoly for seventeen years; then after that anybody can practice the invention. IBM, of course, in order to be able to do these things, doesn't have to have a patent, but if somebody else gets a patent, and we cannot use it, we're in trouble; so we need to publish, as in the Technical Disclosure Bulletin. But then other people could keep us from using their valuable and important patents, so IBM uses its patents for trading, cross-licensing and so on. In fact, IBM has always had a very generous licensing procedure.
Aaserud:So it's mainly IBM-related.
Garwin:Oh, sure. Applying for patents is a big nuisance, because the first thing you have to do is to make an invention disclosure. That's not so bad because that helps to get your thoughts in order. You can see that from one of Luis Alvarez's invention disclosures, or mine are OK too; it's really a lot of fun, and in the process you think things through. But then after that you have to work with a patent attorney for preparing the application. One problem is the search; there's no sense filing a patent application if the information is old, published or patent applied for. You make a patent application, and depending on the patent attorney you have, that may be a joy or a big pain, and that's pretty much the end of it in IBM. Then the review comes back, and you may have to argue with the Patent Office, and then there are foreign patents and what not. So it takes a lot of effort. When I say I've never made any money from any patents, that's really only from my personal patents. Many of the IBM patents have expired before they were actually used by anybody; however, they were very useful for trading purposes because they were I think all good patents. More recently, a couple my of patented inventions have actually appeared in the IBM product line -- for instance, our piezoelectric touch screen. And we have some laser printer technology things that are coming out in the IBM product line, and some other things. Even those that don't appear in the product line, IBM has a system for rewarding -- or compensating people for the effort of applying for patents, because it is a nuisance. Although it says we're paid a dollar, we don't get that dollar, but we get something like $500 for a patent, and a few hundred dollars for a Technical Disclosure Bulletin publication or whatever; and I've had some Outstanding Invention Awards on some of these things, like $l0,000.
Aaserud:From within IBM.
Garwin:From within IBM, yes. So I do all right.
Aaserud:How many are personal patents, how many are IBM patents?
Garwin:I'll give you the list. I have, I think, 33 patents altogether. Two of these are US government patents; that is, one of them was done with Carson Mark and Ted Taylor, I believe it was. It's for a neutron source -- an initiator for nuclear weapons -- but aside from the title, the patent is still secret. And another Department of Energy patent -- I believe this was turned over to the Department of Energy -- was a superconducting inductor for fast-switching storage of energy for all kinds of purposes. So aside from those, and aside from my mussel washer patent with Harold Friedman -- a chemist friend of mine -- and this Sunstein-Wiesner patent, where my interest eventually was sold to IBM, the rest are all IBM patents. There may be a couple here and there which are something else.
Aaserud:So it's a question of corporate strategy more than anything else.
Garwin:Yes. IBM would like to have more good patents, but many scientists would rather do their science than do that.
Aaserud:And the Disclosure Bulletin is another strategy, right? There's the patent strategy and the disclosure strategy.
Aaserud:How important a part of your body of publications are your patent applications, would you say? How would you compare that with your open publications?
Garwin:Oh, I have hundreds of those, maybe 300 or 400 by now. My biography says 200 but I just haven't counted them recently.
Aaserud:So there's no comparison there.
Garwin:Not in number. Actually in my publications there are really two kinds. There is the kind which satisfy scientific standards; that is, when you don't publish unless you have something new to say. And then there are the others, some of which are also like that. For instance, an article on Reducing Dependence on Nuclear Weapons -- a chapter in a book -- which is a reasonably new thing; but many of the others are just restatements of things that everybody should know. They're more akin to instruction, to teaching freshmen; you've got to do it every year; there are new people coming along. That's something which people in the policy area really don't seem to understand. But scientists in policy are unwilling -- I am unwilling -- to say the same thing over and over again. It's a pain, and it's not what scientists do, and yet it's necessary to do it.
Aaserud:You're forced to, yes. So that the most natural distinction may not be between patents and non-patents at all, but between different kinds of published literature.
Garwin:Yes. I don't have very many scientific publications. Maybe there are 50 or something like that. I really haven't counted them.
Aaserud:There was kind of an end point, but we can get to that later. The next thing on my list here is, you changed from academic to industrial physics, and as I understood you, that was at least partly motivated by a desire to avoid the big science that was coming up.
Garwin:That's exactly right.
Aaserud:Do you think that's an indication of a trend? Are there other examples of that? How unique are you in that kind of a decision? Do you have any reflections on that?
Garwin:Well, I don't think many people go in that direction for that purpose(to do individual rather than group research), because they have the wrong idea of science in industry. They think that it's all teamwork.
Aaserud:It sounds paradoxical to an outsider perhaps.
Garwin:Quite right. I really don't know. The other aspect is that in industry you don't have to apply in advance for support for your research; you don't have to make all kinds of promises. In exchange, of course, you give the local authorities a lot more power over your career and your research, and the difference too is that in some cases, your audience -- your peers -- are the outside scientists, but in many cases, it is a matter of delivering things of interest to the company. In IBM research, we have a dual standard. That is, if you do world-class research, then you don't need to worry whether what you do is of interest or even understandable to IBM, although it would be a real shame if the organization didn't have people who were capable of evaluating these things themselves. But if you don't do absolutely first class research, and in that way contribute to the advancement of the science and technology on which everybody depends, and also enhance the reputation of IBM, then you'd better be doing things that are relevant to our products or our future products. Some of that can be very esoteric science, ranging from optimization theory to fluid flow, to new numerical techniques which are extremely important, no matter who is trying to do what. But their motivation, unless they are recognized outside, is for the application that they can have -- for the profit that they will bring us. And I think that's perfectly OK. I think some scientists regard the profit motive as less than respectable, and I don't know how they think they're going to get their money or why people should support science instead of the study of Greek, or music. I believe I've always had arguments with many people on the President's Science Advisory Committee, including my good friend I.I. Rabi, who always argued that science should be supported just because it's beautiful; I say it should be supported because it is intensely practical. That doesn't mean that you support only science that's immediately practical, but because it has the potential for producing practical things eventually, that's why we should support science. I think we probably still have this kind of disagreement.
Aaserud:That goes deeper than your respective academic and industrial backgrounds.
Garwin:Yes. This was always my view.
Aaserud:In the industrial sphere, how many positions like yours are there? To what extent are you still an anomaly, if you ever were, in the industrial sphere as a physicist?
Garwin:Well, there are two aspects. There are a lot of people who by strength of character, just plain cussedness or whatever, do what they want. They may or may not contribute to the local group. They may or may not do great work. But they do what they want to do, and they're tolerated or encouraged to do that. Then there are some people who are recognized and are also in that category. In IBM we have the IBM fellow program -- about 30 IBM fellows in the Research Division, about 50 in the whole company; I think about 65 were made altogether, but some have retired, a few have died and what not. I think I invented the IBM fellow program. In l959 or 1960 I wrote Manny Piore, who was our director of research and then our chief scientist, and I said, "You know, we ought to have a rank called IBM Leading Engineer or something like that, and we ought to select such people because they can contribute more by their individual efforts rather than be directed." And a couple of years later, came the IBM Fellow Program. I was supposed to be among the first, but I and my family had arranged to be off skiing for the first time in 3 years or more, because I'd been doing so much work for the government. When Piore told me of my selection for this award. I asked him whether they couldn't give it to me in absentia. He said, no, this was really a great honor, and I said, "Well, tell them to give it to me next year," which was a very foolish thing to do. It took two years before they got over the slight, because it's given by the chairman of the board, and with the first set of fellows, it was of course a tremendous new thing. The ultimate irony, of course, is that, we had no sooner arrived at the ski lodge (and perhaps spent half-a-day on the slopes), than I received a telephone call that evening regarding an urgent meeting in Boston the next two days of one of my most important Government activities, a panel chaired by Dr. Edwin Land. Not only was this an urgent meeting, but an important meeting, and we immediately returned from our skiing vacation. No IBM Fellowship; no skiing with the family!
Aaserud:A very expensive vacation. Does it make sense to distinguish between industrial and academic science at that level, do you think, at all?
Garwin:Oh no. No, we have people here who are equal to the very best in academic work.
Aaserud:Not only in terms of quality but also in other respects.
Garwin:Topic? Well, we do astrophysics. But in general, yes, because in fact the amount of non-applied and totally fundamental science being done in US industry, in my opinion, has decreased considerably. There have been really only two, and at times three, major centers for that -- initially Bell Labs and maybe General Electric and a little bit at Sylvania or elsewhere; but in recent years, IBM I think has had most, and then Bell Labs; and General Electric have a little. AT&T Bell Labs has this terrible problem, due to the breakup of the Bell system, that formerly whatever they did could be found applicable across a broad scope, including in the local operating companies, and it was supported by the revenues from the local operating companies as well, so they had not only more money but they had a broader range of applicable work. Now people there are quite worried about the pressure to do practical and profitable things.
Aaserud:It's already hurting the basic science part of the industry?
Garwin:I can't make that judgment.
Aaserud:No, but that would be a natural assumption.
Garwin:That's what you hear; I haven't talked to many people at Bell Labs. I talked to Phil Anderson, who left Bell Labs; he had a joint appointment there and at Princeton, and went to Princeton full time. But it's a problem; of course IBM didn't have a very good profit picture last year, and you might expect a push to save money, and to do things which are more helpful to the profits. But in the near term and in the long run too, that's not necessarily served by getting rid of good fundamental work, and there's no effort to do that.
Aaserud:To get back to your more immediate experiences, and to go back a little in time again -- we talked some about your Los Alamos experience the last time. I would like to know a little more about how important that experience was in generating your science policy concerns. Or was it?
Garwin:Well, you have to start somewhere. I don't know that it was important, but it probably gave me a lot more confidence and credibility, because I had worked in the area of nuclear weapons, and I got to know in that way a lot of people whom I would not have known so intimately otherwise. I had known Fermi and Teller and Szilard, but this way I got to know Bethe and Rabi and all kinds of other people. I paid attention to what they said, talked to them, and they had more confidence in what I had to say. It was clear to me that after a while there was not adequate control of these dangerous and important things we were working on; if I didn't do it, well, there wasn't any other person who would do it effectively. So somehow a lot of people -- a considerable number of people -- had to spend a good part of their effort specializing in this field; it's not something for amateurs, although everybody should understand that their vote and their understanding is required; but people cannot make up these views -- they're too subject to manipulation. And what is disappointing is that too many people won't do even a little bit in involving themselves in thinking about and judging these questions on the basis of what's available to many people. Even many scientists also just feel that they can have too little influence, and so they won't do anything, and let the powers that be handle it.
Aaserud:How well defined a group with shared interest was there at Los Alamos at the time?
Garwin:Oh, very little; in policy, hardly any. None. None.
Aaserud:So you wouldn't say that was the best background for that kind of involvement.
Garwin:Oh no, all those people, you know, had left at the end of the war, and formed the BULLETIN. Initially I didn't have a lot of sympathy actually for the BULLETIN, and these "disarmer type" people. Only in recent decades have I developed for them the respect that they deserve.
Aaserud:There was no common project either at Los Alamos that led to that kind of involvement.
Aaserud:You would say that Los Alamos was almost entirely technical in that respect, and that the science policy input on your part came from different sources.
Garwin:That's right. For instance, I worked with Marshall Rosenbluth and Conrad Longmire Marshall Rosenbluth on the details of the hydrogen bomb, after Teller and Ulam had their idea, and Bethe was in charge of the theoretical work. We had meetings of the Theoretical Megaton Group every week, and I was there whenever I was in residence at Los Alamos, which was in the summer time -- three months, four months, five months, for lots of summers. But of course the hydrogen bomb work was all done in 1951, because it was detonated October 30, 1952. At one time, though, Marshall and I were involved in two activities. One was that we had observed that there was a problem with a design of a nuclear test, and he and I did an analysis. We showed that something had been forgotten, so we put a paper on Norris Bradbury's desk, and the test was delayed, I believe; there was a substitution made for the primary fission bomb. The other thing that we did was to calculate the fallout from one of these shots -- Shrimp or whichever -- in 1954. It had also in our opinion not been properly calculated, and we put that on Norris Bradbury's desk one night, probably either one or two days before the shot; that, I think, was the case in which the Lucky Dragon was irradiated. I don't know the decision procedure, but it was perfectly clear that there was lots of room for people to calculate these things that hadn't been thought about. I had early on looked also at the question of fratricide -- of the interference of the explosion of one nuclear weapon with another -- and that had also been totally neglected. We didn't at that time think about nuclear winter; we didn't think about the destruction of the ozone layer. We did think eventually about world wide fallout, and these are important things. But no, there was no significant effort at Los Alamos in such matters.
Aaserud:These are the efforts that Teller refers to in that talk; I don't remember the exact source of that, but he relates at some point your involvement in the development of the hydrogen bomb.
Garwin:Oh, yes, that was at the 1981 Erice Conference. That was mostly a considerable argument between us. I don't know exactly why it was that he felt compelled to tell that story. Now, really, he regrets having told it, because it gives me a lot more credibility than he would like me to have.
Aaserud:It was not politically very wise.
Aaserud:Continuing a little bit into your graduation into science policy, so to speak -- I came across some correspondence when I was here last time. Of course I was able to go through only a very little part of it. But you were employed with Midway Project in Chicago at some point, doing some Air Force project work.
Garwin:Yes, I think I was consultant to the Midway Laboratory when it was first created. That was probably as a result of Joe Mayer -- Professor Joseph Mayer -- professor of chemistry there, and my going to Korea and Japan for a month in the fall 1952. And I had thought about night vision equipment. Well, I worked also for some weeks one summer with Al Hiltner at the Yerkes Observatory, because I had had an idea for an image intensifier which might be useful for astronomical purposes. This was to produce grids of metal coated with secondary emitting material, stacked one above the other. I would have a photocathode and then the electrons would be imaged, constrained, focused by a magnetic field on successive grids. This transmission multiplier could be used to intensify an image from the focal plane of the astronomical telescope. We built a vacuum system, and I tried making some of those things and coating them. I didn't have enough time. Eventually Al Hiltner and other people did go on to use such techniques for doing polarization mapping of interstellar dust and things like that.
Aaserud:Was that with Midway?
Garwin:No, I had proposed it to them too for night vision, but this was at Yerkes.
Aaserud:But your trip to Korea was connected with Midway directly, correct?
Garwin:I don't know whether it was actually done through Midway. That's all a blur. It was done because Joe Mayer was part of a senior scientists crowd with Willie Fowler and Lauritsen and Zacharias at MIT. In order to help create a Tactical Air Command -- the Strategic Air Command had too much influence -- nuclear weapons, or maybe non-nuclear weapons, for application to ordinary combat were of interest to this group. You had to go and see first hand. I think these people -- I didn't have much association with them at that time -- had a study. I know there was a study, and different people were supposed to come back and report on different things. Joe Mayer was making this inspection trip, and he asked me to come along; I invented the fan-jet engine then -- I'm sure other people had also. I invented a pulsed-light ranging system; it was before the days of lasers, but we certainly did have both gas-discharge pulses and modulated zirconium arc lamps. So when a pilot was looking at a target on the ground or aiming with the machine gun, then the pulsed light would go out and be reflected and the range be determined -- just as with radar, but with a very narrow beam. And I some communication techniques -- a large number of things like that.
Aaserud:So at that point you were more drawn into these problems than pursuing them independently.
Garwin:Yes, that was probably my first exposure, aside from nuclear weapons. In nuclear weapons there are a lot of other things. I was very active in devising totally new techniques for nuclear testing -- to determine with very fine spatial resolution what happens in the interior of nuclear weapons when they are exploding. I had a lot to do with the firing systems for detonating the high explosive, and many many things like that.
Aaserud:That was Los Alamos work, was it?
Aaserud:I'm bringing up the Midway connection and the trip to Korea in part because I came across a letter that you wrote to David Griggs at the University of California at Los Angeles on 21 October, 1952, where you express your frustration about, I think, the use of radar in Korea.
Garwin:This was for the bombing of ground targets.
Aaserud:It was regarding the impotence of radar against low flying planes; I think that's the way you expressed it. That letter was forwarded by David Griggs to Al Hill at Lincoln Labs at MIT, so that indicates that you weren't on the top of it then, and that it was part of your getting into the larger world of science advising.
Garwin:That's right. Al Hill and Jerry Wiesner and Jerrold Zacharias organized the Lamplight Study, and they requested -- this was just after I'd gone to IBM -- of Tom Watson, Sr. that IBM contribute my services to this; it turned out to be half time for a year, maybe even two years. I'd go up to Cambridge (Massachusetts) or to Lincoln Laboratory. That was probably because they had heard of me through Dave Griggs or whatever, and Dave Griggs was a good friend of Bill Libby. Bill Libby was at University of Chicago at that time.
Aaserud:So that was a direct response to that letter, perhaps?
Aaserud:Does it make more sense to see that as a graduation towards science policy problems then? The Lamplight thing must have involved some discussion of the implications involved.
Garwin:No. In fact, they tried not to have any discussion of implications. The whole idea with Lamplight was to extend the air defense of the United States to the sea approaches to the United States and Canada, and we worked on that technically. I had some ideas for mounting radars on ships, so that they would be automatically stabilized instead of having to fight to stabilize them, and many many other things, like a moving target indicator, and radars. The Canadians had brought in an automatic detection radar that would look at eight successive pulse returns, and if there were five or more returns in a given range-bin, then it would have a digital output that there was a target there. I think I mentioned that they used digital logic; there were 700 vacuum tubes or whatever that were required to do this, because there are many ways you can get five out of eight. I replaced that with one vacuum tube, because we just quantized the return and added them linearly and set a threshold at five eighths. So there were a lot of interesting things. That's when I learned about meteor scatter and other kinds of communication systems. So that's a lot of fun. But there was no question whether one should do it, and finally I complained to Jerrold Zacharias. I said, "This is all wasted effort; by the time we have this kind of defense that we're talking about, the threat will be ICBMs." Of course, it really did turn out that way. When I was complaining, he said, "Well, you know, if I don't get the result I want from this study group, I'll convene another one until I do." Jerrold Zacharias was a very interesting person. It was from him that I learned the saying, "Don't get it right, get it written." That's true, because if you write something down, you can pass it around, and other people can criticize it or add to it or whatever.
Aaserud:But actually you were discouraged more than encouraged toward that kind of broader discussion.
Garwin:Well, I wasn't much interested in the broader discussion, either.
Aaserud:It's as simple as that, at that point.
Garwin:I really wanted to get back to IBM and do my work. Also, it was difficult, because my wife and baby were here, and so I was spending several days a week away, and I didn't like that.
Aaserud:No. How long did this Lamplight involvement go on?
Garwin:It must have been about a year.
Aaserud:Continuously, more or less?
Garwin:Well, the middle of the week was spent away.
Aaserud:Throughout the year. What connection if anything at all did you have with the Berkeley Loyalty Oath? I came across a letter to M. Kasha from the 15th July, 1951. Of course, these are just letters I happened to find; it may or may not be representative. But you indicate there that your decision not to accept a position at Berkeley at the time had something to do with the Loyalty Oath and the handling of it there.
Garwin:Yes. I think I mentioned, too, that it wasn't something that I had any particular trouble with. But, since many of my friends were leaving Berkeley because of the Loyalty Oath, I wasn't going to go there at that time. And also, I didn't think it was appropriate to have people, in addition to their duty to uphold the Constitution, take another oath to uphold the Constitution. It's the beginning of the end. But, unlike many of my friends, I was not politically active so I had no particular problems with signing, except I didn't want to as a matter of principle.
Aaserud:It was more that people were leaving than any high-flying idealistic principles, at that time. You have a large correspondence in the mid-fifties with E. N. Adams who I think was at Midway with you.
Garwin:He actually worked for the Midway Laboratory. I think he was an assistant professor at Chicago, and spent his summers and consulted at Midway. He's around here. You could talk to him, actually.
Aaserud:In that correspondence there is some indication -- I could pull out a couple of letters if you want to, or you may remember -- of some kind of conflict between the corporate IBM and the academic Columbia University, mostly in terms of the professors at Columbia seeing corporate IBM as somewhat less worthwhile.
Garwin:There was some of that.
Aaserud:This was in 1954. It may have blown over.
Garwin:The Chemistry department was very reluctant to have their graduate students do their work for their PhD any place except with people in the chemistry department on campus. That's because they feared that their best students would be attracted to our laboratory or elsewhere. The physics department, having played a major role in the staffing of the laboratory, was very open about this; half a dozen of our first eight or ten people came from Columbia. Seymour Koenig, who's just down the hall, Sol Triebwasser, Gardiner Tucker, Bob Gunther-More, Haskell Reich -- all those people were graduate students who then went directly from the Columbia Radiation Laboratory to the IBM Watson Laboratory; I was sort of the only one, with Erwin Hahn, and eventually Peter Price, who had come from outside. Naturally they regarded these people as students rather than researchers in their own right -- your children are always your children. I think that accounted for some of it. Anyhow, I never had any trouble with the Columbia people.
Aaserud:During that year, I don't know how actively you were applying for other positions. Certainly you got offers -- Lockheed, Berkeley, Los Alamos. To what extent was that an active job search on your past?
Garwin:It wasn't active at all. When I came, I was disappointed, because I had been promised that the laboratory would be ready for research when I showed up December 15, 1952, or whenever it was; I'd left a place where I could do work, and I was involved with good things. I think Seymour Koenig had written me the various particular letters about physics. Wallace Eckert is the person who had been the director of the laboratory and who hired me. But he was not a physicist, and the outfitting of the laboratory was really the responsibility of this young physics crowd, who had never done anything in renovating a building or whatever before. I don't think they were lying to me; I think they were just optimistic about the schedules. But I got there, and there was no experimental work possible for several months; my desk was behind a door in the stairwell or whatever, and who needed that sort of thing? That colored my views for a while.
Aaserud:But that was all taken care of in time. That might have something to do with your applications for positions.
Garwin:I didn't apply anyplace else. Whenever anybody offered me a job, in my early days, I thought about it. And then I discovered that thinking about it really was a big effort, and so I no longer think about it.
Aaserud:By November you write to Rosenbluth that you have refused both Lockheed and Berkeley and you're not interested in Los Alamos because you want to do experiments, so it blew over pretty fast, it seems.
Aaserud:The McCarthy period, generally speaking. Again, I came across a letter in that same box to Chairman Libby of the Atomic Energy Commission, regarding Jack Steinberger, who was a professor at Columbia.
Aaserud:He had been refused clearance to work at Brookhaven. That's one indication of the McCarthy era, I suppose. Do you have any comment on that specific incident and whether that is representative of your general experience in that period?
Garwin:Well, there wasn't much. I thought McCarthy was a terrible person, and obviously, when the bureaucracy would keep really fine scientists like Steinberger from working because of the necessity for clearance, that was a miscarriage of efficiency as well as justice, so I did what I could. I wrote people and tried to reverse these things. But I didn't write very many letters like that, because I don't know a lot of people.
Aaserud:It was more an isolated incident than a general campaign for that kind of case. Around l955, you were involved in something called the Accord Project, isn't that correct?
Garwin:Accord was the name of the patent that Jerry Wiesner and Dave Sunstein and I got. It stands for Accordion, where you take speech -- a second of speech, maybe. The whole idea is a multi-user radio communications system, where time of travel is important. Let's say you are having a naval task force, maybe a few hundred kilometers across, that is a time of flight of a millisecond, and everybody's always listening to the wrong frequency, for instance. We thought what we should do was that instead of having nets which are distinguished by frequency, we should have them distinguished by time, and we should have voice communication possible over these nets. You have a choice. You may divide it up into very short pulses, and you interleave these pulses, but then the pulses will overlap at different points in this broadcast system; it's not like a time division multiplex system from one point to another point, where the travel time is the same, and you can interleave these things permanently. So the interleaving has to be with an allowable time variation, which is bigger than the transit time over the task force. What we thought of instead was to take a second of speech and compress it by a factor of 100 or 200 to, say, ten milliseconds. You can have 100, or with guard bands 70, speakers simultaneously, and you assign slots and you send these things out. Then they're received by a person who either looks at the header on the message, or (for net traffic) they're assigned to a particular slot, captured and then stretched and fed out to an earphone or loudspeaker. Now, in the early 1950s, we didn't have digital storage technology, and so I devised a number of mechanical systems which would do this -- which would write slow and read fast, and write fast and read slow -- and that was Accord. [Interruption] That was that patent. We proposed that in our study to solve some of the problems of naval task forces, and broadcast systems. It has to do these days with network communication, in which there are nodes in the network with different transit times. It's an important problem, and we looked at the patent agreement which MIT, who ran this study, had with the members. We decided that, aside from uses by the US government, we owned the invention. We paid our money to have it filed, and then as I say IBM disputed it. IBM disputed my right to have this thing, because even though it wasn't in IBM's field of interest, it was done while I was working for IBM and paid by them.
Aaserud:So that's a case of an interface, so to speak, between IBM interests and government interests, perhaps.
Garwin:No, not government interests, because there was no doubt that the government had every right to use this. The conflict was between IBM interest and my interest. But it was just a misreading on my part of the agreement.
Aaserud:How unique was that type of conflict?
Garwin:That never happened again.
Aaserud:During our last session, we talked a little bit about the Primakoff origins of your solar sailing idea in the l950s. Am I correct in stating that Primakoff's work in part provided you with a motivation for doing this proposal?
Garwin:Let me see which came first. My paper on thermalization of positrons in metals was published September 15, 1953, and it must have been submitted a few months before. There are three things that are mixed up together here, and what you're talking about is what became known in the l960s or 1970s as the grand tour, the gravity assist. I had been thinking about the exchange of energy between positrons and electrons in metals, and pointing out that this calculation by Primakoff and others, which said that you couldn't lose energy as soon as you got down to the Fermi energy and so on, was wrong; it required two dimensions. Then of course I realized that this had a lot to do with what Fermi had been doing in his theory of cosmic ray acceleration by collision with magnetic clouds, in which the equipartition would come when each of the charged particles had as much energy as the whole moving magnetic cloud. I pointed out to Fermi that you could escape from the solar system in this way and use the heavy planets and what not. That was just an observation; I told people about it, but never published it. Solar sailing was something else; I published on that around 1958. But I had first told people about it. I first thought of it in the 1940s and first told people seriously about it in 1951 or 1952 when I was a consultant to General Dynamics and tried to get them involved. Kraft Ehrike and others there were interested in space flight craft. They had their ex-Germans who were interested in space. They were talking only about chemical propulsion and maybe ion engines, and not at all about solar sailing. I pointed out to them that you could do this; not only could you go out, but you could go in toward the source of light, which was not recognized at that time. At first I thought the government might use this, so I didn't want to publish it. Then after nobody in the government showed any interest, I tried to publish it. I guess Bob Evans, who retired from IBM two years ago, was a staff person then. It was a much smaller IBM, of course, and we started the IBM Journal of Research and Development in 1956 or 1955. I submitted this for the first issue, but Bob Evans who was in charge of it said we ought to patent this and not publish it. I told him I thought that was ridiculous; and I think now I was rather unfair to him. In fact it might not have been a bad idea. But then they responded by saying they thought it was not the sort of thing which they should have in the first issue, because it gave the wrong idea of IBM's interests and views and what not. Then I decided to publish it outside, and I did. But it was an old idea with me.
Aaserud:OK, so the suggestion to Fermi is entirely different.
Garwin:That was different. That came from positrons and it had to do with gravity assist. It was not solar sailing at all.
Aaserud:That's what I wanted to clear up. You went a little fast for me the last time, I suppose, because then you went on to the Yorktown gravity waves work.
Garwin:Oh, that was much much later in the seventies.
Aaserud:I know, but that's the progression of the interview, so I just wanted to recapitulate the relations there, and to see how we came from one to the other.
Garwin:Well, if I look at my interests here from the bibliography -- let me just get a copy of this bibliography. [GARWIN gets bibliography of his patents] Perhaps the government patents are not in here. The first patent listed is on spin echo memory devices that we worked on here at Watson Laboratory.
Aaserud:For the secret patents, for example, the titles would be unclassified.
Garwin:I think so, yes. Most secret patents have unclassified titles.
Aaserud:But they're not on that list?
Garwin:I don't think so. So, let's see, the patent with Carson Mark is not here. Let me just see whether the one with Dave Sunstein is there. Yes, that's 1964; it took until 1964 for that one to be patented. This is the Accord patent. And the one with Mark is not there. So there may be a couple missing. But they're mostly here. So that's the patents.
Garwin:And then here's a list of publications in the IBM Technical Disclosure Bulletin.
Aaserud:Are those in addition to that general bibliography?
Garwin:They're not in the bibliography. I think they're not.
Garwin:So here I have the bibliography, and these were my scientific interests, except for a few which didn't get published. My bachelor's thesis employed computers to calculate the current distribution for axially symmetric magnetic fields, and then air-core betatrons, and homopolar generators. Then it's my PhD thesis on experimental beta-gamma angular correlation; then a vacuum tube differential analyzer for the Schroedinger equation; then some pulse generators and coincidence circuits and servomechanisms and liquid scintillators; then search for V particles; then some more electronics; then thermalization of positrons in 1953; and then in 1954, the excitation of subharmonic resonance. John Von Neumann had an idea to use nonlinear elements as computing systems, which have essentially infinite gain between degenerate subharmonics. So I have two papers on such things; then a paper with Marshall and Ariana Rosenbluth on the Pinch; then some spin echo memory papers; then a proposal of 1956 to polarize hydrogen and use polarized hydrogen gas as a source of polarized internal proton beams; and then all this stuff -- subharmonics, low noise amplification. I recognized that you could calculate heat flow in a medium which had a conductivity depending on temperature; it looks like a nonlinear problem but can be cast in the form of a linear problem by a transformation of the temperature axis. I think that's probably the sixth time that has been published, but it's been done in different contexts. And then it was the parity stuff starting in 1957 -- a whole lot of such things. Then in 1958, we come to the solar sailing publication finally. Then I had put some of my students to work on superconducting penetration depths in 1958. In 1958 also on July 20 came the publication on international air, sea and space traffic control using radio beacons on the vehicles. That was in response to things published, speeches of President Eisenhower, and complaints by the Russians -- they didn't know where our bombers were. Then there are some publications of instruments we had used in our parity work -- these precision current regulators which I had used for my nuclear magnetic resonance work and adapted for the parity work; then some of the results -- self diffusion in helium 3 with Haskell Reich; more helium 3, more parity stuff, more superconducting stuff in 1959, and the scintillation counter. A speech I gave in 1961 at the Ten Outstanding Young Men convention was my first public policy speech. I do list in here some of the IBM Technical Disclosure Bulletin entries; if you're looking at 1962, for instance, I've put in here recently this item from May 22, 1962. "Command Enable Switches and a Possible Alternative"; permissive action links is what they were called after that. That was a secret document that I declassified recently. Then there is an IBM Research Note, which is just a little think piece I wrote at the Watson Laboratory on the minimum diameter of filaments obtainable by drawing out molten metal and glass -- "Increasing the Speed of Cryogenic Circuits"; that's looking at thin film cryotrons. Then, in 1964, I invented a two-photon laser which turns out to have giant pulse characteristics.
Aaserud:That was your first involvement with lasers, at least in publication, right?
Garwin:I had much earlier looked at recombination lasers from nuclear explosions, but that was in a report in Los Alamos. Then in 1965-66, I was director of applied research here. Here's one about two-dimensional scan lasers. Anyhow, you can read this. And then more and more it became publications, although I had been much involved without publishing these things, for the President's Science Advisory Committee and the government agencies. I began to publish them. In 1969, for instance, there's a paper on advanced air-to-air systems, that was originally written as a secret paper, but I declassified it before actually presenting it.
Aaserud:Maybe we'll get back to these publications a little bit in my chronological questions, if we have time for that. You know, you did send me some of those papers.
Garwin:They are all here.
Aaserud:When I asked you this question about the order of things with respect to Primakoff and solar sailing, it was just to make clear whether you wanted to indicate some kind of interrelations there, or whether it was separate projects that you went through in your mind. It was the latter, it seems.
Garwin:There was a relation between the thermalization of positrons and the Grand Tour. I just wanted to point that out. But then you got me to thinking about other things.
Aaserud:I've come to the last item on my list here of filling in questions. I started out asking about your brother, and I want to end with a question asking about your wife and children and their role in your life, just generally speaking. You were married to Lois in 1947. Jeffrey was born in 1949, and you can tell the rest, perhaps.
Garwin:That's right. My wife actually worked, I mentioned. I had a fellowship that paid my tuition at Chicago, and also some living expenses. It wasn't really enough to live on, so she left college after two years at Western Reserve University and went to work at the Blue Cross (Health Insurance) Company in Chicago. And then, as I mentioned, when we were home in the summer of 1948, she was hired as a trainee at the Ohio Bell Telephone Company. When she returned to Chicago in the fall of 1948, she went to work for the Illinois Bell Telephone Company, and worked until a short time before Jeffrey was born in November, 1949. Then she stayed home and took care of the baby. Actually, Chicago was a very rough place. There was a lot of street crime and robberies, and that's primarily what decided me to leave University of Chicago.
Aaserud:To peaceful New York.
Garwin:Well, we didn't live right next to Columbia University, and in New Yorktown commuting was then common, whereas in Chicago it was not common. If you didn't live at the university, you were rather out of things.
Aaserud:Was that because of the differences in transportation?
Garwin:I think it was possible to live at the University of Chicago, because it's not high rise, but anyhow, that's how it was; maybe during the war it had been easier to live there. I don't know. But we didn't want to live in a densely populated city, so we lived in Riverdale, just north of the Henry Hudson Bridge.
Aaserud:You moved in there from the beginning when you came here.
Garwin:Right. And lived there until May 1955, when we moved to our present house in Scarsdale. Tom was born in March 1953. Tom was born shortly after we moved to New York and he was two years old when we moved. We moved from the Bronx because we wanted to have a home rather than an apartment, and because we wanted to have better schools than were available in New York City. So we looked in Westchester County, primarily in Hastings where there was a large number of Columbia people. It is or was a more direct commuting shot from there than from Scarsdale and things were cheaper there, but Hastings is very small. We didn't find a house there. Then Laura was born in 1957, and by then I was much involved with traveling -- President's Science Advisory Committee things. She was born at the end of July 1957, and a few days after Lois came home from the hospital with Laura. (Of course we had two other children), I left for a trip to Europe. I remember installing a dishwasher as I left as a parting present, before I went away on this trip. This was 1957, and I think I was gone a whole month in Italy and elsewhere traveling with Leon Lederman and T.D. Lee. We went to a conference in Venice to report on our parity experiment and various other things. I always had plenty of time and solitude to do my work. I've always done my chores around the house -- helped with the dishes or the cleaning or whatever. Sometimes if the children at night would wake up and their diapers would need changing or whatever, I would do that, because I found it easier to wake up than my wife does. But she took care of the children and we always had dinner together in those days. But there was plenty of time to do my work; it's amazing how much time there was. And we traveled two people and three children and skis for everybody, and it still seems to be possible. I guess when people are young, it's a lot easier.
Aaserud:You found time for that every year essentially during those years?
Garwin:Yes. We would go home to Cleveland to see our parents, and then of course we traveled to Europe for six weeks in November of 1958 for the Surprise Attack Conference; I was on the US delegation, and we took our three children. We stayed at a pension. Then we had an apartment for a while, and we had a maid. In fact, we brought her back eventually, toward the end of the next year. She was with us in New York and then spent the whole year with us again in Geneva when we went over in the summer of 1959 for a year for me to work at CERN. Eventually children grow up, and Jeffrey is the appropriate age for a person born in 1949. He has a PhD in biochemistry and an MD. He's assistant director of clinical trials for McNeill Consumer Products in Philadelphia. He's married to June Bossinger, who has a PhD in biochemistry too, and who's working in Boston, but she's going to be moving down to the Philadelphia area with their baby who's a year old. And our second son Tom, who was born in 1953, has a bachelor's degree in history from Harvard, a Master's degree in public policy from the Kennedy School and has done all of his coursework for his PhD. He works for the MacArthur Foundation. He worked previously for the Office of Technology Assessment for a couple of years, and for Brookings for a couple of years. Before that he was in graduate school, and before that worked for the Defense Department in international security affairs for a year.
Aaserud:Did he bring you into the Kennedy School, or did you bring him into the Kennedy School?
Garwin:It was independent. In fact, he has been in some of these fields before me, and I in some of them before him. It's quite confusing, because many people think I wrote the paper with John Steinbruner on Soviet Attack on Minuteman, but he did. Laura, who was born in 1957, now has a PhD in geology from Cambridge University. She was in the first batch of women Rhodes Scholars. That must have been about nine years ago, when she went to England for two years -- to Oxford and the rest of the time at Cambridge. She has before that a Bachelor's degree in physics from Harvard. She plays the trumpet and the piano, and she does a lot of trumpet playing on a semi-professional basis in England. She had been vice president of the Hasty Pudding at Harvard, played in orchestras there, and now she works for Nature Magazine for the last year.
Aaserud:Is that her on the wall?
Garwin:No, that was a student who worked for us part time last year. Actually Laura is the one there in black, and that's Jeffrey -- that's I guess about five years ago; and Tom and Sally -- I think Jeffrey was not married then. Those are the children at a younger age. Sally is Sally Ericsson, Tom's wife. She has a Master's Degree from the Kennedy School at Harvard and worked several years in Washington on the staff of the Steering and Policy Committee of the House of Representatives (for Thomas P. O'Neill). Now she is head of the staff for Sam Gajdenson, Congressman from Eastern Connecticut.
Aaserud:Did your wife take up work in New York ever again?
Garwin:Yes, after the children were at a certain age, I guess probably around 1961, when we came back from Geneva. In fact, she was quite depressed. Geneva's a much nicer place to live and you have plenty of local transportation and what not, and we had a maid. She decided that she would finish her education, make progress with it, and I thought it would be good for her to have a better specialty than having worked as a clerk. In case something should happen to me -- I should die -- she could get a job. She finished her bachelor's degree in French from Columbia University School of General Studies and then she has a master's degree in education from C.W. Post College on Long Island. She has taught part time as a substitute teacher, initially all over southern Westchester County, and for the last few years only in Scarsdale. When we're not traveling, she works about half the time. She worked Monday and Tuesday. She's working this morning, and she'll not work on Friday because we're going away. So that's how things are.
Aaserud:Now I come to my pile of cards with new questions. I don't know how you want to do this. Following up on the first session, I was planning on becoming chronological again. I was thinking of talking a little bit about IBM, in relation particularly to the publications you sent me, just to have something to grab onto. Then we could talk a little about internal policy advice, if I can call it that -- the pre-PSAC period and PSAC and other activities during the PSAC period and after the PSAC period. We could then take up outsider critique in terms of publications, in terms of teaching, in terms of relations to Congress. I don't know if that division of things makes sense to you. But that's the way my cards are.
Aaserud:So maybe we should talk a little bit about your relationship with IBM through the 1960s. I don't know if you have anything to add in relation to the style of research at the IBM Watson when you arrived there -- how the leadership was, how the collaboration was, and how it was in terms of a continuation or a break with your previous experiences in science.
Garwin:Well, it was not much different, primarily because these people had all come from academic life of one kind or another, and our purpose was to do academic-like research. [Interruption] The style of work. Well, most people found it no different. That is, they did their own work. Sometimes they worked together. What was different was that one didn't need to ask anyone else for money. I organized a lot of things. We had weekly meetings which I styled to follow the Bill Libby seminar at Chicago; people reported on what they were doing. We had conversations. We had a daily coffee hour. I arranged to provide free coffee, tea and cookies, which I think is the way to get people to meet together. We had stockroom committees for organizing things, a library committee, and things went really very well. Wallace Eckert was an admirable person -- a person who had done research and was doing research. He worked on the theory of lunar motion, and, in fact, I was just given in recent weeks by Martin Gutzwiller here the residue, that is, what Martin and a colleague had done to complete Wallace's life's work, which is an expansion approach to the motion of the moon. So he was a very understanding person and a person of principle. He trusted people and we did our work. He demanded high standards. Together with the physicists, there were engineers. In the previous incarnation, Watson Laboratory started out before the war as Wallace Eckert and a few people introduced punch cards into scientific, mostly astronomical, computing; it was the IBM Watson Scientific Computing Bureau, or something like that. Then after the war it was housed on 116th Street. Wallace Eckert had been at the Naval Observatory during the war, putting out the Naval Almanac by punch card computation and automatic typewriting techniques. After the war, he hired a number of people from the MIT Radiation Laboratory -- Bob Walker, Byron Havens, John Lentz, who had worked on electronics. They were helping IBM enter the electronic computing era, compared with the electromechanical, relay-based punched card era. At the laboratory, when I came in 1952, there had been a well advanced effort to bring electronics into IBM. Of course for the most part that was being done also in the main development plants. There wasn't any research division or research laboratory in IBM at the time. People were making vacuum tube calculators, vacuum tube punch card machines, mostly a line of punch card machines, with the exception of the card programmed calculator, which was around at the end of the war, and the Selective Sequence Electronic Calculator, the SSEC, I guess. Anyhow, there was a vacuum tube machine which was built and operated I guess for the first time in 1949 or 1950 at IBM corporate headquarters at 590 Madison Avenue. It was built, among other things, to run analyses of the implosion device. When I was at Los Alamos in 1950 or 1951, Bob Richtmyer from Los Alamos had been spending some months or a year or more in New York, running the implosion of the Trinity bomb on the SSEC which had vacuum tube. What kind of storage did it have? I guess it had Williams Tube storage. [No, probably it had vacuum-tube latches]. Wallace Eckert decided that they wouldn't really get very far without having a better understanding of the fundamentals -- and solid-state electronics especially, because the transistor had been invented then in 1949 or whenever. IBM didn't know much about that, but they were building transistors. So a couple of the people from Columbia were not hired to re-do their thesis; they were hired to work on germanium and silicon and to understand the solid-state things. I was interested in those things, but I wanted to work on liquid and solid helium and on superconductors, which I said had been sort of untouched by the development of experimental technique and theory since the war. The engineers continued to work. In 1954 By Havens group completed the construction of the fastest electronic calculator in the world, the Naval Ordnance Research Calculator. John Lentz made a personal automatic computer -- the 604 or 605, I forget what it is. Bob Walker worked with the physicists, that is with Art Anderson from Poughkeepsie, an IBM person who was coming to Watson Laboratory in New York to get his PhD at NYU. I think he was a person who had been selected by the IBM hierarchy as somebody with management talent and put onto a fast track or whatever. Such things apparently happen. He worked with Erwin Hahn, and then a bunch of us applied some of the spin-echo techniques to memory. We could store a thousand pulses in a drop of water, and had we used electronic spin resonance, we could have stored millions of pulses. There are a lot of interesting things which have to be understood in order to do that -- fundamental problems of coherence and interference and things like that. I helped on that. L. H. Thomas and Bob Walker worked together on a kind of delay line storage system -- self-sorting delay-line storage -- in which the most recently used information is in a short delay line, recirculating, and gradually it ages and gets put into longer and slower delay lines. Thomas invented a new kind of cyclotron, a strong focusing cyclotron, which was eventually built. L.H. Thomas is a very ingenious person, but a person who had great difficulty communicating any of his ideas. That's the sort of thing that went on in the engineering field too. It wasn't great. People had the ideas, of course, of automatic voice recognition typewriters, but things weren't ready for them. I began a program on superconducting computers. John Lentz I guess, or Bob Walker, had come back from a meeting at MIT. A graduate student there Dudley Buck -- this must have been in 1956 -- had thought about superconductors, as young people do -- as Brian Josephson did, for instance, much later. Josephson said, "You know, if I really believed this stuff, I should be able to get persistent current through an insulator at no voltage and when I apply voltage, I should have these microwave oscillations," and so on. He worked out the theory and interpreted it. What Dudley Buck did was to recognize that if he used fine niobium wire, which has a high transition temperature and a very large critical magnetic field compared with tantalum, and wound a helix of insulated niobium on a fat tantalum wire, he could control a current in the tantalum by switching it superconducting or nonsuperconducting by the magnetic field of the solenoid. He could control a current which was much more than was required to be put into the control helix, and so he would have gain, and he could have two of these cross-connected as as latch. They would use absolutely no power at all, because the currents would be persistent. He actually built counters and what not from these "cryotrons" as he dubbed them. They were limited in speed to milliseconds because of the decay of eddy currents in this fat tantalum wire, so I immediately invented thin film cryotrons. A consultant to Arthur D. Little, John Gaunt, had looked at them and decided they were infeasible, but he had made an error in symmetry; he just had thought about it wrong. And so we had a project. I had a hundred people working for me at various IBM locations by the end of 1956, to build superconducting computers out of thin film cryotrons. I had organized that and directed it for three or five months when the parity opportunity came along. I went to Manny Piore, poked my head into his office, and said, "I am going to go and do parity experiments; there's this marvelous experiment." It was after I did the experiment, because it only took four days to do the experiment. When we had the result, I went to see Piore, who was director of research, and said, "I'm sorry. I have better things to do. You should appoint a new director of this program; [But he did not appoint a person recommended by me.] When you're doing a program like that, what you need to do is to continually review the things that you can accomplish, and the things that you cannot. You work on the things that you cannot, in parallel, preferably, so that you have something that can be applied. The new director Don Young, who still works for IBM, had much more a physicist's view in that you work on the things that are interesting -- the fundamental things. That's important for the long run, but when you have so many people -- a hundred people or more -- working on something, you cannot afford to keep them all busy pushing back unnecessary frontiers. Anyhow, there was a considerable disappointment on their part that I wasn't going to do this any longer, and disappointment on my part that it wasn't going to be directed properly. John Lenz from my laboratory was much involved in this program. He was doing things, in my opinion, right. We introduced photolithography for the production of these zigzags, planar circuitry, and what not. Other people were machining masks and evaporating lead and tin and things through them, but we wanted to do this with wet processing by photolithography. I had Miriam Sarachik, my graduate student, make these masks and we reduced them and did a little bit of photolithographic work. Later on, when IBM did get into planar silicon technologies, many of the people who were involved were those who had had training on the thin-film superconductors in superconducting computer program, which ran for several years. As I say, it was not very well directed, and eventually finished without building a superconducting computer. Then for a long time I very largely worked in, as my publications show, parity. But it was at that same time, following Sputnik, that I became involved with the President's Science Advisory Committee.
Aaserud:Yes, we'll get to that.
Garwin:I spent a lot of time working on intelligence and technology for defense and education and so on. There was also the Physical Sciences Study Committee which produced the new high school education curriculum.
Aaserud:When you first came to IBM, that was part of the conscious transition, wasn't it, to solid state physics?
Aaserud:Were you part of the planning of that?
Garwin:That's why they hired me. They didn't want me to come in and do particle physics, although I think the recommendations were such that if I had insisted on doing that, they realized that I had broad enough interests that they might still have hired me, but I didn't want that. It was to have solid state physics as background for the IBM business.
Aaserud:This was much different from what you had done before.
Garwin:I had much more freedom, because I didn't have to teach, although I did eventually teach some. I didn't have to worry about getting money. I didn't have to schedule my experiments on the joint use facilities. So it was really much better.
Aaserud:What about the problems themselves that you were working on? How different were they, and how difficult was it to go into a new field like that?
Garwin:Well, they were problems which I chose for myself. I don't know that I did a very good job of choosing them, or very good job of working on them, necessarily. But it's easy to go into a new field, at least for me, because I build all the apparatus myself. So if you understand how things work together, it's not difficult. For the helium 3 work -- I guess that was my primary interest -- we had to build a helium liquefier, or at least get one; primarily we worked with Columbia. I didn't want the responsibility of having a liquefier, so we got helium liquid from Columbia. We helped them operate their Collins machine. Eventually I guess we did get a Collins machine ourselves, because helium was quite expensive and not very well available. There were shortages and what not. So we installed a recovery system for our helium, put the boil off down there; it would get re-liquefied. Eventually we could buy helium liquid, just delivered in Dewars which is the way we get it here. I think the country and the world are doing a very bad thing, in venting all their helium. I worked on that quite a lot, over several different year periods in Washington. There's not much you can do with society to have them take care of resources, while there are still resources around. So there's helium, and then helium 3; you bought your helium 3 from the Atomic Energy Commission in those days. I remember my first helium 3 was bought at $l50 per cc of STP and my last helium 3 was bought at I think ten cents per cc of STP, so the first experiments were done on a micro scale; you asked what techniques you're going to use. I then learned about spin echoes from Erwin Hahn. That's just a marvelous technique, so I applied it to study the diffusion and relaxation of helium 3 -- in helium 3 crystal, in alloys, and in solutions of helium 3 and helium 4; it just leads lots of places. We demonstrated the two phonon relaxation -- Raman effect essentially. Andre' Landesman came for a year and worked with me. He was the person who translated Abragam's book on nuclear magnetic resonance from the original English. Abragam wrote it in English, so Landesman translated it into French -- very fine person. I had some ideas on superconducting penetration depth and I invented some new superconducting materials. I had people work on penetration depth measurements. We did some of the first very accurate penetration depth versus temperature measurements. I decided that one should explore the use of superconducting materials for radio frequency accelerators, and I tried to get the folks interested. I guess Panofsky was about to build his first linear accelerator, the Stanford Linear Accelerator.
Aaserud:At SLAC, yes.
Garwin:So it was important to see whether this could be made superconducting. I had a Florence flask blown with another tube at the other end of it. If you get a Q of ten billion or so -- ringing time of many seconds at microwave frequencies -- in the absence of any coupling, as soon as you put something in to couple into this cavity, then you spoil the Q. The easy way to measure such a thing is to use waveguides beyond cutoff for the coupling. In that way you can have a coupling loop in one of them; a tube which is maybe 2 centimeters in diameter for a wavelength of 5 centimeters is beyond cutoff. As you move it away, you can get any attenuation, and so minimize the leakage of microwaves out of the bottle, if it has otherwise perfect reflectivity. And so we could with very simple equipment demonstrate evaporating lead on the inside of these flasks and tubes at very high Q and measure these things. I remember making a special trip to California to talk to Panofsky about it. He got Bill Fairbanks into his home, and I explained to them what we were doing -- how you for various reasons should make this material not the purest material, but probably the dirtiest material possible in order to get the highest Q under these circumstances. They got started on doing such things, and I dropped it, because it was perfectly clear that with one graduate student and 2 percent of my time I couldn't keep up with those folks. It was also clear that it was premature, so I was happy enough for them to go with their copper cavities.
Aaserud:But you were involved in the start of it.
Aaserud:We're back after lunch; or I'm back after lunch. We were talking about your early years at IBM essentially. I have more questions in relation to that, but one in particular relating to the period before the parity work. To what extent did you purposefully direct your work to support the Corporation, generally speaking? Did you try to do research that had input into the Corporation as such?
Garwin:No, but I tried to be familiar with what they were doing, because my strength really is to understand quickly what people are about, and to see whether there's something that I already know which could solve some of their problems, provide alternatives, or make it not necessary to do what they're doing. I would poke around and write people letters or make file memoranda on different kinds of memory, or means for assembly or programs or whatever -- what are called pipeline multipliers. I think the electronics that I used in my nuclear physics -- particle physics coincidence circuits -- were the first so-called "emitter-coupled" devices that one has in the transistor world. I tried to tell people about these things, but it was really not very useful for me in general, because I like to work by myself, to take on something which would be directly applicable. Sometimes I would prove a point, for instance. This was somewhat later, maybe 1968 or thereabouts. I looked at intensive cooling of semiconductors by immersion in boiling freon. I told people about that, and they said, "But we look it up in the DuPont literature and we see that it has a corrosion rate of copper of less than 1 mil per year, and we need only a 100th of a mil in order to destroy our circuitry, so we can't possibly consider that." But I knew that there shouldn't be any more corrosion than there is of appropriate materials in air. We set up some thin film resistors in a bath of boiling freon, ran them for a long time, and kept telling these people about it. So eventually the people in Poughkeepsie, whose job it was, actually took up on a much grander basis than we could the investigation of immersion in boiling freon. It was the same when we built our closed cycle liquid helium refrigerator for superconducting computers. When I left the program, I decided that I would deliver this machine, because everybody said it couldn't be done and that's why you couldn't have such superconducting computers. I said I didn't know about the rest of it but the one thing I was sure of was that we could make a low-cost refrigerator and I would do it. We did that, and occasionally we would prove other points. I think I mentioned to you the laser printer -- the IBM 3800. Although I couldn't persuade them, I got them to deliver me an engineering model and we fitted it with a laser and rotating mirror, and that's what won out. So, I tried to consult. I went in the very early days to Owego, New York, or Binghampton; or Endicott, where IBM was working under license from Carlson, the inventor of xerography, on a "dick strip printer" for magazine labels. I helped them with solvent fusing of these labels and a number of other things. I think I mentioned to you the system whereby -- I may have told you something wrong -- we could not only communicate over the power lines in a building, but we could get a specific answer back. After you had 10,000 individual controlled sensors or valves in a building, and had commanded something to operate, in your central control -- timer control in those days, not computer control -- you wanted to know whether it had operated. You would have kilowatts of acoustic power, that is, acoustic-frequency power. A few kilohertz would go out over the power line to broadcast, and some of it would be absorbed by this one detector, which would then operate a motor. But you cannot provide every one of those remote locations with the ability to generate kilowatts of power for signaling back. So I designed a communication system for doing that over the power lines, which used the technology of that era, and I have a patent on that. That may be the Accord system; I may have misled you. That may be the IBM answer back system for building control; I'm sure that is. The other one that Dave Sunstein and Jerry Wiesner and I worked on is Centipede. The reason that was confused is that it really is like an accordion in the Centipede system, but Accord is the answer-back over the power lines. That's the sort of thing; there were just a lot of things I was involved in. Whenever I would look at something, I would tell them what I thought was an alternative or a better way. Sometimes it was also very useful to tell them, "Well, if you're going to do it that way, you will run into this or that problem;" it's better to be told in advance. A lot of times people don't believe it, though. "Here's the answer when you do, but you would be better off taking this different approach which doesn't have that problem."
Aaserud:That was an internal advising function, so to speak.
Garwin:Yes, and that's what I'm most efficient at. That's my comparative advantage, so that's why I should do it.
Aaserud:You had another function. I don't know how typical this is, but one of the things that you sent me was this press release of showing a group of teachers a powerful IBM magnet. I don't know how indicative that is of IBM public relations in general. It's another aspect of the kind of involvements you had, at any rate.
Garwin:Right. Bell Labs had discovered this particular superconductor. It was a rhenium; well, it tells about it here. This was niobium-zirconium, but I had one a little bit before that which was rhenium alloy. [Interruption] Well, I just thought it would be useful for people to understand the meaning of these things. What I did was to make a simple superconducting shunt which could be controlled, not by a magnetic field, as in a cryotron, because we would have very big magnetic fields around anyhow, but by this niobium-zirconium wire. I just put a little loop of it up into an inverted test tube, and put a resistor in there. When I wanted to unshort the coil, I had these two terminals coming up to room temperature and I could put a dry cell across it. With this niobium-zirconium wire in the normal state -- that is 30 degrees Kelvin or so, because of the heater -- the current coming down would go into the coil and charge it up. The voltage would produce a rate of change of current V = L dI/dt, and after a reasonable time, I would have 10 amperes of current in the coil from my little dry cell. At that time I would turn off the heater, and now there would be a short circuit across the coil. When I took away the dry cell, the current would close in the superconducting shunt. That was a reasonable thing to do in order to demonstrate how simple it was to get these very large magnetic fields; I mentioned power generation, and there's also magnetic levitation -- shielding space ships from radiation. Let's see, when was this done? I don't think there's a date on this, is there?
Aaserud:15 February 1962.
Garwin:Yes. Do they have a date any place on the document itself? I don't think so.
Aaserud:Well, it's in the bibliography.
Garwin:Oh yes, down here; it was covered by my finger. It was believed to be the highest field yet produced by a current persisting in a superconducting coil after the input power leads had been removed. If you do such things, why not use it to show physics teachers at the American Association of Physics Teachers, and get some beneficial publicity for IBM? Shielding space ships from radiation is a real application. You can put up a sufficiently large magnetic field so that certainly the few MEV electrons in the Van Allen Belt are totally turned away without producing any local radiation -- X-rays -- as they would if they were stopped by material. Even very high energy protons, and many cosmic rays are moved away, too.
Aaserud:I think you mentioned thermonuclear power generation there too as a possibility.
Garwin:That's right. I'd been working myself since 1950 on some aspects of magnetic confinement fusion, but it was perfectly clear that superconducting coils would convert a scientific feasibility into power generation.
Aaserud:Was it practical?
Garwin:No, not necessarily economic, but it would help a lot.
Aaserud:We've mentioned this laser work before, so I think we have covered that essentially. Then there's an article in Physics from 1965 with Landesman. You mentioned the collaboration with him.
Aaserud:Is this article significant for a turn towards international collaboration in any way?
Garwin:Let's see, how did this happen?
Aaserud:That's the question.
Garwin:I guess Andre' Landesman wrote me. He said he would be interested in coming, and I arranged for him to work with me for a year at the Watson Laboratory. We did experiments together; he was a very good, very helpful, very capable person. He knew a lot more theory than I did, so that's why I was particularly happy to have him come. But I was away much of the time. This was a very critical time; I guess this was received in 1965. This paper was a very long time in genesis. I think Landesman was with me 1962-1963. Then we started to write this paper, or to do the theoretical work on modeling the interaction that we had measured, so here are the measurements. We measured the T1 and T2 for helium 3 solid, in the crystal and alpha and beta phases, as a function of molar volume. Helium 3 is very squishy material, so with the modest pressures that we could have in our good old laboratory-built equipment, we could change the molar volume from 23 cc per mole down to 16 or 17 -- a factor of 1.5. Later on I think we changed it by a factor of 2. There's this enormous range in exchange interaction that one sees by doing so, and we tried to understand that. First we looked for built-in anti-ferromagnetism, or something like that. Then I realized that as we squeezed the crystal, the wave functions of the atoms overlapped; and that's what gives you the fundamental explanation of the exchange interaction. As you squeezed them together, you would think they would overlap more. But the wave functions are obtained by some kind of self-consistent adaptation to the local potential. As you squeeze the atoms more, the potential wells get steeper, because it's not a harmonic potential, it's a Leonard Jones potential -- it's 1/r12 repulsive core minus 1/r6 attractive portion. So, as you squeeze, the wave functions narrow faster than they are squeezed together. The wave functions narrow, they are moved in closer together, but they sample the wave function of the neighbor at a lower amplitude; that's the idea. Now the question was, is there some way to calculate these wave functions in crystals? So we plunged in to do this, and it turned out to be rather difficult for us. We were exchanging correspondence back and forth across the ocean, and finally I decided, in the summer of 1963, that I should pack up my computer programs and go to Paris and work there for a month at Saclay with Landesman and write this thing; I would have less distractions than I had going to Washington. And so we did so; we wrote the first draft of this in summer of 1963, but I guess we didn't get around to finishing the calculations until 1965.
Aaserud:He was here too, you said?
Garwin:He was at the IBM Watson Laboratory in New York City. He was here for a year, 1962 to 1963. After that year we tried to continue the collaboration on the writing things up, and on the theoretical work, by mail. That worked fine except that we really had to have this month or six weeks when I was in Paris in order to do it with any efficiency.
Aaserud:How typical or untypical was that kind of international collaboration?
Garwin:Untypical. [Interruption] Well, I mostly don't collaborate with people. I find it extremely difficult to collaborate with people, because of my schedule and what not. The only people I have collaborated with on my research was, first, Haskell Reich, in the laboratory. He worked with me. I was his manager. He was a senior staff member, but he worked with me. More recently, Jim Levine did. If there are people with whom I am permanently associated and are senior enough so that they run the whole thing, and can accept contributions of the kind that I can make, that's fine. Otherwise I just work by myself.
Aaserud:The parity work also involved some collaboration, didn't it?
Garwin:It was the equipment which had been put together or inherited by Marcel Weinrich for his thesis. One idea was from Leon Lederman, the other idea was from me. Leon and I worked for those four days. After that, of course, you bring in people you need for extra work; there's a lot of equipment and things. That's inherently a kind of team effort, if you want to do it as quickly as possible, and why not do it as quickly as possible? You have every reason to want to do it that way, and no reason not.
Aaserud:But it's a whole different way of collaborating than the collaboration with Landesman, for example.
Garwin:That's right, and it's a field which is highly competitive, whereas in the fields in which I work, for the most part it doesn't matter whether I work on things sequentially with very short publication time, or whether I work on things in parallel with longer publication times. It's only a matter of efficiency; which is the better way to work. If I find people who are willing to accept suggestions or proposals and so on, I just give them to them. In fact, I would rather not be involved in the publication, because then I have to be responsible for the paper and worry about it and what not.
Aaserud:It's better to be thanked.
Garwin:Or even not to be thanked. The best is to have people steal your ideas. That's the very best.
Aaserud:At this stage, yes.
Garwin:Take for example this thing at CERN -- the muon G minus 2. I had gone to CERN in 1959, having been involved in all of these things with the government for a couple of years. I just wanted to read in the library, so that was the idea. I wouldn't do anything. I would just read in the library and find out what was going on, and sort of polish my skills. After about two days of that, and walking around, these people had gone to Gilberto Bernardini. They said they needed a leader for this experiment that I had been involved in discussing with them, and that I was the leader for the experiment that they wanted. I said no, and Bernardini tried to persuade me. Eventually I decided that I would do that. It was very sad, because I really had wanted not to do this work but to study. It turned out well, but I didn't really accomplish what I wanted in going there. So for the most part, these are all single author papers. Well, the physics more recently is with Jim Levine because it's just not possible for me to keep technicians in the lab busy, when I'm not here. As for writing things with people, that's also very difficult. There's just a lot of negotiation to do. It's an anomaly -- for instance, the recent collaboration with Kurt Gottfried at Cornell University.
Aaserud:You were Director of Applied Research from 1965 to 1966 at IBM. We have touched upon that. I don't know if you have anything more to add to that.
Garwin:It came about in a rather strange way. I think Gardiner Tucker was Director of Research. He was one of those people who had come from Columbia to Watson Laboratory. He wanted to do his job as best he could, and so he talked to me about being Director of Physical Sciences. He felt that I would be more closely involved in the company's technology that way than I was with the Watson Laboratory or whatever, and I tried it. Maybe it was also a way to make more money, although I don't think so. I don't remember that. (At the moment when the announcement was going to be made, my office was an office like this, and there was a secretary's office somewhat bigger than my secretary's office. Then across the way there was another office with the door opening into the secretary's office. That was Ralph Landauer's. He was Director of Applied Research, and it occurred to me that I would be a much better Director of Applied Research than Ralph Landauer, and he would be a much better Director of Physical Sciences; so we changed. So I never was Director of Physical Sciences. I was Director of Applied Research.) That was a time when I had efforts in germanium, high speed circuitry and silicon integrated circuitry. All the magnetic recording was under me. We started work on magneto-optical recording. I was in charge of the San Jose Laboratory as well, where we had a good deal of photo-physics. Cliff Herrick at that time was working on technologies for copiers, so, as I think I mentioned, I chose the technology for the IBM copier. But then I gave it up. I was at a Defense Science Board Task Force in the summertime at Woods Hole, and every day I would get a shipment by airmail of my work from Yorktown. I just decided that Yorktown deserved somebody full time working at this who could do it as well as I or maybe better. It was just cutting too much into the things that only I could do. So I went back to the Watson Laboratory at Columbia, and they urged me to be Director there, which I was for a year. But then that takes time too; I decided that that wasn't important enough for me, so I resigned from that. Seymour Koenig I think succeeded me as Director.
Aaserud:There was a year in between, wasn't there?
Garwin:I don't know. I may have gone back.
Aaserud:According to your vitae, you were Director of Applied Research 1965 to 1966, and Director at IBM Watson 1968 to 1969.
Garwin:I may have been deputy or associate director. I guess I was associate director.
Aaserud:When I look at your bibliography, there seems to be a rather abrupt transition around 1965, when your main concern with science and its immediate applications kind of goes into the background, and you turn more full time to science advising. I don't know how significant that is, or if you have any comment on that. Is that a reflection of a change of interests or affiliations?
Garwin:Well, it was a change of my view of what the most important thing was that one could do. I felt at that time that one could not really count on the governmental system to handle these problems without more education of the public. I took the responsibility of talking to people about my views of priorities and solutions. Some of these are anomalous; for instance, the two 1968 publications on fighter aircraft were secret reports, of which I have many, many. These are just two that happened to be done for the Defense Department. They had appropriate stamps, "Declassified after 12 years." I declassified them, and now they appear in my bibliography; but there are a lot more of them. Some of these others are unclassified reports, for example on vertical takeoff and landing aircraft, which I had done for my Military Aircraft Panel, and on satellite systems, some of them done with the Air Traffic Control Panel. So this is not so big a change as one might think, because there are many of those starting from 1960 or 1962 or so, and only a few have leaked out with a changing view on the possibility of publishing these things in the open literature. But then frankly, beginning in 1968, it's an expression of my concern that the governmental process is not adequate for making these decisions, and that we had to educate people from the outside so that when they came into the government, they could do a decent job. They should also be able to make demand on their government, with an informed skepticism.
Aaserud:So that's when you turned from an insider to an outsider, so to speak, in the science advising business. Before getting to that, I would like to have you say a little bit about pre-PSAC science policy and advising activities. That is, if you have the time now; this is a big new area we're getting into.
Garwin:My involvement with PSAC began in 1957, with a PSAC panel on intelligence under W. O. Baker (of Bell Labs). Soon after I was involved with the PSAC Strategic Military Panel. I did a lot of that, but I had been involved at Los Alamos. Then around 1951 I guess it was, since I had been working with Hans Bethe, he put me in contact with Arthur Kantrowitz. I was hired as a consultant to Avco and to Convair on ballistic missiles and ballistic missile re-entry technology. I told them what I thought about those things, and I talked with Los Alamos about everything they were doing there; I gave them my views about such things as well. Then the Air Defense Work at MIT that I talked about in 1953 or 1954 -- Lamplight. I guess that's about all that I was involved in, until PSAC came along.
Aaserud:What were the most important forums at the time for science advising, in your opinion?
Garwin:There were others than PSAC I didn't have anything to do with. There was the Science Advisory Board of the Air Force -- the Air Force SAB. Then there were the Army and the Naval Research Advisory Committee, what was it called? I forget. Then there was the whole Academy structure. I wasn't a member of the National Academy of Science. Many advisors are not. But I wasn't involved in those things either. Then for many years, when I was so much involved in the White House, anything else would have been a lesser use of my time, so I tried to stay away from that, so it was very largely PSAC. Some of the things in which I got involved with the CIA and the National Security Agency were individual contributions as a consultant, or small panels in special areas.
Aaserud:That was about at the same time?
Garwin:It began probably in the late fifties, right.
Aaserud:You mentioned before your participation in the 1958 Conference for the Prevention of Surprise Attack.
Garwin:Jerry Wiesner was the chief backup staff person for Jim Fisk, who was the head of this panel, I guess.
Aaserud:And it was Wiesner who brought you into that?
Aaserud:Did that have anything to do with your PSAC activities at all, or the panel of PSAC, directly or indirectly?
Garwin:Yes. Well, Wiesner was a member of the President's Science Advisory Committee -- very important member, a very activist type person. PSAC, with President Eisenhower, were quite interested in this activity and provided the staffing for the working group, for the US contingent. Bud Wheelon, who is now president of Hughes, was eventually to become Deputy Director of the CIA for Science and Technology -- a job that was offered to me, I guess.
Garwin:When was that job offered to me? I think, when Wheelon took it, after Pete Scoville had it.
Aaserud:So he took it after you had been offered it.
Garwin:Yes. I think that was the chronology. And Sid Graybeal was there from CIA and John Tukey from Princeton, Hendrik Bode from Bell Laboratories, Frank Press -- some of these people. Frank Press was probably on the test ban talks, but the others were on the Surprise Attack business. Although there's some confusion. Because Surprise Attack wasn't doing very much, I spent a lot of time with the other delegation too, and I explained how -- the computations at CERN and what not.
Aaserud:That you mentioned the last time, that's right.
Garwin:So I tried to stay away from other advising. I thought there would be a conflict of interest with my PSAC activities. That's how it was until 1966 or 1967, even though JASON had been formed in 1960. I certainly encouraged their doing that. So then we come to the JASON era, I think. JASON was formed following some thinking and correspondence on the part of Charlie Townes, among others. I was on the list of people proposed to be members of this new group, but I believe when it actually came to the bureaucratic process within the Institute for Defense Analyses, I was regarded as ineligible because I worked for a profit-making company. I believe the only other non-academic person at the beginning was Bob Lelevier, who worked for RAND. But RAND was a not-for-profit. Perhaps it was only after a good many years of experience on PSAC (and obvious acceptability to the Government in this area), that JASON and IDA felt it suitable for an industrial person to be a member of JASON.
Aaserud:What about consultantship for the Arms Control and Disarmament Agency? I just threw that in here because I don't know when that started on your part.
Garwin:It didn't start very early. I was a consultant during the SALT I era. I don't know how that came about; it was probably through the PSAC activity. The ACDA took me on as a consultant, and I was ever after, essentially, except that every new administration fires all its previous consultants at ACDA, because the Congress hedged ACDA with many restrictions to keep them from giving away either our secrets or our weapons. Anyhow, in 1966 or 1967, whatever, I went to the JASON meeting. Or maybe earlier, 1964, I went to the JASON meeting at Bowdoin College in Maine to talk to them about re-entry physics and things like that, from my own rather homey point of view. I spent a couple of days there giving them some lectures and talking about other aspects of military technology. In 1965 or 1966, I guess, I was involved with them, in discussing possible weapons or sensor systems for Vietnam. I don't think I was a member yet, but I wanted to make sure that they were aware of what my Military Aircraft Panel had been considering. We worked closely together in this, so many of the weapons and concepts were things in which I played a role, although I don't think my name is on any of the JASON papers from that era. The drone relay aircraft and things like that were things in which I was involved. And then after that had been briefed to McNamara in 1966 or whenever that was, there was formed at McNamara's order, to implement the system, the "Defense Communication Planning Group," headed by General Alfred Starbird. I was a member of the Scientific Advisory Panel to that which played a very active role in trying to help them with their work. I went in February 1968, for a week. We started off going to Vietnam, but it was the time of the Tet Offensive, so they wouldn't let us land in Vietnam. We could only go to Thailand, so we went to the main operating base for this effort at Nakon Phenom in Thailand, and reviewed what was going on.
Aaserud:I had a few questions about the basis for your involvement in PSAC. I don't know what you remember about that or what you care to say about that. How were you approached to become a member? What was your reason for joining?
Garwin:I think Manny Piore first asked me to come to the Cosmos Club where he was staying in Washington. He told me that the President's Science Advisory Committee, which I wasn't really aware of, had created two panels. One was a Panel on Technological Capabilities, and the other was a Panel on High Level Intelligence, dealing with our information on the Soviet Union. He told me that these were very important to President Eisenhower, and that they needed to get the very best people; which one did I want to work on? He was really my boss, and I figured, I'd come to work for IBM and they'd given me a third of my time or whatever to work on things helping the government; here was a new opportunity. I felt that nuclear weapons and air defense were probably not so important as some of these other things. I chose the Intelligence Panel, so I went to work with Bill Baker on the Intelligence Panel. I guess that was in 1957, so I worked on that. Then I was asked in a year or so to join the Strategic Military Panel. I think Hans Bethe or maybe Panofsky was chairman -- I don't remember who. I was on that, and then in 1962, having been on many panels -- there was a Tactical Warfare Panel and several others -- I was called to Washington. We had a nuclear test series, and the Starfish shot -- the high altitude nuclear explosion -- had caused the death of several satellites, not all of them ours; -- there were a Canadian satellite and a French satellite. It had all been reviewed, I learned, by a panel under Panofsky of the PSAC. Panofsky was off on his vacation trekking in the desert with his wife and maybe some children and could not be reached. They needed somebody to investigate what was going on, so I moved to Washington for two weeks.
Aaserud:That was in 1962.
Garwin:On July 4, I think, was the shot. Within days I was in Washington and put in charge of finding out what was going on and what could be done about this. Among other things, President Kennedy had made a commitment to put an American on the moon within the decade and bring him back safely, and here we had these radiation belts which were very much enhanced. The question was, could we launch through them? What would be their duration? The radio astronomers were up in arms, about the radio noise from these electrons. Satellites were dying here and there. It was just awful. The Russians had a space mission up, with astronauts, and the question was, would those astronauts be killed by our trapped electrons? So it was as rather hectic heady time. Jerry Wiesner, who was Science Advisor, took me in to see President Kennedy. I had to tell him, could we launch through this? What were the options? I had ways of sweeping the Belts, and of course we knew that they were decaying. We didn't know exactly how long it would take. We could also launch from the North Pole, but it is not a very happy thought to have to move your space launching facilities up to the North Magnetic Pole, or South Magnetic Pole.
Aaserud:That was actually the circumstance for your getting into the PSAC proper?
Garwin:That's right. Then after I did this and reported to PSAC, they asked me to join.
Aaserud:Were you involved in the advice leading to the decision for those tests as well?
Aaserud:You came in when Panofsky was in the desert essentially, that was your first exposure.
Garwin:I knew about them, I guess, because of my involvement in Los Alamos. There had been some secret tests in 1958 -- the Argus series of three shots, launched by rockets fired from aircraft carrier, I guess, in the South Atlantic.
Aaserud:This was the Dominique series; wasn't that what they were called?
Garwin:I don't remember. Do you have a reasonable stopping point in another card or two?
Aaserud:Well, I wouldn't say that, because I have quite a number of questions on PSAC, to tell the truth. We could run quickly through them, of course, but I don't know if there's a point in that. The mode of work in PSAC, members of PSAC that you collaborated with -- we did say something about projects the last time, so we needn't go into that again.
Garwin:You have the book on Presidential science advising, so that's available.
Aaserud:Yes, I have your article right here.
Garwin:Not only my article, but there's an article by Dave Beckler who lists the public reports.
Aaserud:Yes, I have the members here.
Garwin:All right, and the public reports from the committee and all that. The committee worked, you know, two days a month. It met in full session. We had these panels, and I thought that was a great innovation. We'd get the experts but we'd also get young people; we made a studied effort to bring in young people. All the people we had for the most part were those who'd been involved during the war. Some of the young people really proved to be adept and flexible, like Charlie Slichter, and some of them were very good but narrow, and some of them didn't work out at all. But that's a good thing to do, bring them in and send them away if they don't work too well. And the idea of having a two-stage review, that is, a focused review by the panel and then a presentation to intelligent people under rather informal auspices, but having read the written report was very good. The committee would then either adopt the panel report and make it a PSAC report, or forward it with comment; we had the interaction of the Science Advisor, in the good days, with the Secretary of Defense or the technical people in agriculture or wherever, and that was very good. I mentioned that I brought in the Rachel Carson articles. That was the beginning of the PSAC involvement in insecticides and pesticides and such things. But my own work was in intelligence, military space and military aircraft and naval warfare -- and a little bit of land warfare but not much. I was never on the Land Warfare Panel, I guess.
Aaserud:Who did you interact with at those meetings or briefings?
Garwin:There is the PSAC panel, and my panels always met two days a month also, for two really full days. I had, as I say in my article, as many military officers as possible. They would be allowed to sit around the room. These were very big rooms, with tables that were probably eight feet wide and 30 feet long. The room was probably 20 feet wide by 60 feet long, so you'd get a lot of people in there. We would deal mostly with Assistant Secretaries for Research and Development of the services, or of the Department of Defense; or the Under Secretary for R and D; or Alain Enthoven who was the Comptroller. Then these were the chief technical people in the services, military people on the development side, operational types, and the contractors. So if we wanted to find out about some precision radar bombing system, the services would provide us with an overall briefing on the application -- the need, the studies -- and then we would talk to the contractors themselves.
Aaserud:So there was a back and forth kind of interaction all the time at those meetings.
Garwin:That's right. We tried to help them technically, so they wouldn't think we were just critical, and we also killed a good many programs. We killed an early infra-red surveillance satellite system. We killed the Mauler -- a General Dynamics semi-active radar homing system fired from a transportable launcher which was ahead of its time. And various other things. But we also pushed hard for some things -- for gunships, to try to get C-130s from the Military Air Lift Command to be given to the fighting forces, and to get infra-red shields available for helicopters to protect them against homing infra-red guided missiles, and such things. That was the mechanism of interaction. Then we would write reports; we'd sit around and write reports. In the early sixties, it was easily possible to send copies of secret reports to people who had safes. Eventually, when the universities wouldn't allow that, it became more difficult. We'd have to get in to Washington to review the secret reports. And eventually we would agree on one, brief the Science Advisor and the PSAC. We would find out that we had used too many acronyms or whatever, and that it was not understandable, and then rewrite the report.
Aaserud:What was the degree of success? Relatively speaking, how many projects that you discouraged were taken away, and how many that you recommended were acted on? I guess that changed in time as well.
Garwin:It was pretty good, but I think that's a bad way to put it. Mostly what you do is to interact with people so that they can do a better job. Jack Ruina for instance, who was Assistant Secretary of Defense for Strategic Systems or something like that for McNamara, would use the threat of PSAC to get his folks in line. He would have management reviews and technical reviews which were a lot better, because he would tell them that they're going to be murdered if they go over and tell PSAC that. So a lot of the influence simply comes from people being honest with themselves, and trying to think deeply about it, because they're going to have to see somebody else. We started a lot of things, and we stopped a lot of things, but there were many things which went on endemically or even into full scale operation which we opposed -- like the AH 56 armed helicopter, which really did not have a cost effectiveness edge properly calculated over its predecessors, and things like the NavStar, Navigation Satellite System, (GPS) that went very slowly. But we had a major influence on it; the air-launched cruise missile, the Captor mine, and several other concepts came right out of PSAC activities.
Aaserud:Were you involved in the choice of members at all?
Garwin:Oh, sure. The committee would forward a slate to the President; we had l8 members and four-year terms. Four or five members a year would have to be replaced, and so we would prepare a slate of people with whom we were familiar, and the Science Advisor would discuss them with the President; sometimes the White House would add somebody. In the Nixon Administration, President Nixon added Pat Moynihan, who's a very intelligent person, but not appropriate for the Science Advisory Committee, especially as he didn't come most of the time. When I criticized him for it -- chastised him -- he said, well, he had to give speeches to support his family. I told him he should do it on some other day, because we all had to do such things too, but we put PSAC first.
Aaserud:I guess the kinds of people in PSAC changed over the years; it broadened over the years.
Garwin:It used to be physicists to begin with. That was good, because people could communicate readily with one another, and they had a shared background. Eventually there was criticism: "The physicists couldn't know everything. You had to broaden it to have chemists and social scientists and economists and things like that." That's not necessarily good because, although you have more talents represented, the difficulty of communication among the members increases. And if you have one economist, you have to have two economists, because there are schools of economics, unlike schools of physics. Nobody would believe that the physicists really organized panels and got people of other persuasions and backgrounds on them. So that is a problem. The bigger problem comes when people want geographical diversity; then they start looking at political affiliation. For instance, when Mr. Nixon appointed me to PSAC, the local Republican organization sent me a copy of a letter they had sent back to the White House saying that I was a Republican in good standing and that they had nothing bad to say about me. I do happen to be a Republican; I don't always vote Republican, but I had voted for Nixon against Humphrey.
Aaserud:We can stop here, you know.