Interview with Joseph Weber

0:00 We had a library in the city, and my brother was in high school taking physics, and he used to talk to me about it. And there was a wonderful book in the library on radio. So I read the book on radio and started to build radios. And as Richard Feynman says, the radios of that era were quite easy to understand. Old components were big. The integrated circuit hadn't been invented. And I could repair every radio, any radio, when I was 11 years old. I got a job after school for a buck an hour. So reading the book on radio increased my income from a dollar a day to a dollar an hour, which was more per hour than my sisters were getting as bookkeepers. and that was quite a bit of motivation to learn about technology and become educated. But we still didn't have money and the idea of my sisters who were working as bookkeepers for $15 a week sending me to college didn't seem too good. And one day I saw an advertisement in the post office for a civil service exam. so I took the civil service exam and three weeks later there was a knock on the door and we opened the door and were astonished to see standing in the doorway the United States Senator from New Jersey and he said I was number one on the list as a result of this civil service exam and he was prepared to appoint me to the United States Naval Academy to get a free college education Well, I think the reason that the United States Senator came to the House, in those days, blacks weren't welcome at the service academies, and women weren't welcome at the service academies. He wanted to be sure that I was neither black nor female, that I had the right number of arms and legs. When I did, of course, he appointed me, and I borrowed $125 for the uniforms and went to the academy and got my undergraduate education there. Well, I walked right into World War II

2:30 and I should have been killed at Pearl Harbor. My carrier was the target of the Japanese, but my carrier left Pearl Harbor a day and a half early and the ship that was tied up at the carrier's dock was completely destroyed. Well, also during those two years, the carrier visited places like Seattle. And when I went ashore in Seattle, I bought a book on electromagnetism and a second handstand and worked all through it. And I was the only one aboard ship who knew anything about electronics, so I was put in charge of the radar. And by the end of the war, I had pretty good qualifications as a radio engineer, and of course, in those days, electrical engineers learned about DC machines and electric power, and Maxwell's equations were essentially unknown and the University of Maryland wanted to hire someone who could teach the engineers Maxwell's equations and I was recommended so they offered me a full professorship when I was still in the Navy in electrical engineering to teach the students Maxwell's equations. So I resigned from the Navy, took the job and taught microwave engineering at the University of Maryland, but I also taught graduate courses in antenna theory. Electromagnetic antenna theory was a fascinating boundary value problem field of really applied mathematics. And I was told I had to get a Ph.D., And you couldn't get PhDs in engineering in the area of other schools, so I got a PhD in physics, and then got two fellowships, got a National Research Council in Guggenheim fellowship in 1954, and went to Princeton.

5:00 and each fellowship needed a senior advisor and Oppenheimer offered to be advisor of one and Wheeler offered to be advisor of the other. It wasn't a bad choice, actually. Stop and think about it. Well, both had been interested in general relativity and gravitational waves and that was Wheeler's primary interest at the time but Oppenheimer was also quite interested. Well, since I had been teaching electromagnetic antenna theory to make a living, the idea of building a gravity wave antenna was sort of attractive, you know, a radio antenna going on, a gravity wave antenna. So that was the first motivation that got me into the thing. Well, of course, Of course, the standard formula that was available for gravity waves in those days was that the power was Newton's constant of gravity times the square of the quadrupole moment times It's the angular velocity to the sixth power divided by the speed of light to the fifth power. I think there's something like 32 over 5. It seems to me that Einstein and Eddington argued about this factor within a factor 2 or something. Well, if you're a sort of hard-headed electrical engineer, as I was in those days, this formula is very easy to understand first your intuition and the books tell you that in weak field approximation gravitation is like electromagnetism while there aren't negative masses and then electromagnetism you have a multipolar expansion first you don't have pole radiation a spherically symmetric oscillating thing doesn't radiate, my Ph.D. oral questions. And dipole radiation is in lowest order. Well, if there aren't any dipoles, then quadrupole radiation is the next order.

7:30 So in gravitation, you'd expect quadrupole radiation. Well, if you take Maxwell's equations and replace the squarely charge by the constant of gravity times the square of the mass you get that formula so there's nothing mysterious about it yeah so that if you again were brought up on electronics and electrical engineering and saw this formula and you thought in those terms the formula is perfectly obvious and you'd sort of never question that. Now, the thing that's discouraging about this formula is that you have the constant of gravity, which is very small, speed of light to the fifth power of the denominator, which is very big. And so that tells you that, you know, the gravitational, of course, if you work out the ratio of the gravitational attraction of a pair of electrons to the electromagnetic interaction ratio is just incredibly small. So this formula stood around. Still, my intuition told me that one should barge ahead and get money and build an antenna. And so we did that, and it turned out my intuition was pretty good as far as picking frequencies and things like that. I thought the telehertz region was right. Well, I returned to the Institute for Advanced Study in Princeton in 1962, and when I got back at the end of the year, I found a letter waiting for me from Freeman Dyson. and Dyson said, Dear Joe, he said, when you started coming to the Institute and talked about looking for gravitational waves, he said, I really thought you were insane and should have your head examined, he said, but Oppenheimer thought you ought to do it and other people were optimistic, he said, so I didn't say very much. He said, but since you were here the last time I thought about it, I've decided that you're probably not insane after all. So here's what Freeman said.

10:00 He said, suppose the sun burns up its nuclear fuel. Well, the sun rotates about once a month. and if the sun burned up its nuclear fuel there would be no radiation pressure to prevent its collapse so it would start to collapse while angular momentum is conserved and peripheral velocity would get faster and faster eventually the peripheral velocity will approach the velocity of light as it contracts which won't exceed it and so Dyson said okay, as soon as that happens So c is roughly omega r. You put this in here, and you get g m squared r fourth. And after omega to the sixth power, you put c to the sixth power over r to the sixth power. And then you have still c to the fifth power. But at least you've got c to the first power in the numerator. But things are looking up. and you have gm squared over r squared apparent here what I said is the Sun continues to rotate it will almost certainly break up into I'd say two pieces breaks up into two pieces two pieces are gravitationally bound and balanced by the centrifugal force with a peripheral velocity, the velocity is right, so that's mc squared over r, which is the centrifugal force. So that m over r, for this case, is c squared divided by g. If you put that in here, you get GC m squared over r squared, which is c4 over g squared, and the end result is c fifth divided by g. And that proves what everyone knows, that Freeman Dyson is a genius because only a genius could change G over C fifth to C fifth over G. Now, C fifth over G means you radiate away the entire rest energy of the sun

12:30 in less than a microsecond. And that means that when you go from the most discouraging problem in the world, G over C fifth, and if you've got the intelligence of Freeman Dyson, you can see that the effort is worth looking for because of radiation. It's not all that small, it can be enormous when the system goes relativistic. And so that was It's the start of the present, the optimism, you know, being there for all other detectors and so forth, but that didn't occur until, as I say, my second trip to the Institute for Advanced Study. But I continue to be alone in the field until about 1970. Anyway, one should give Freeman Dyson credit for this perhaps, it's hard to simple-mind it, but one of the last profound analysis. Yes, it was interesting. I was interested, I noticed in your book the other day, a reference to Dyson, around that time he had come up with the idea of neutron star-binding, and he made a book called Interstellar Communication, which he discussed and then mentioned the possibility of gravitation that was coming from him. And was that about any role in discolored? Of course, I didn't read those books, but I did get that letter. And of course, it turned out that if you took a second look at the program in terms of these equations, that there wasn't anything one would change because my intuition had been pretty good. and that the bars, the frequencies, reception frequency, and things like that were entirely consistent with this model. So really nothing had ever changed. And, of course, the third time I went to the Institute for Advanced Study, the Institute paid my salary. I guess I'm one of the few experimenters in history to have the salary paid by the Institute for Advanced Study.

15:00 I guess at that time, the general period of the late 50s, even the clear of the 60s, some people liked infall and rolling orders. How do you express that about whether gravitational radiation actually cracked any energy at all? Well, I think you've probably read Infeld's biography, in which he talks about the time in, I guess, the late 30s, or when he said that Einstein was so famous that if he wanted publicly, no physicist could get in the door, that the rooms would be full of news media people, and no scientist could get near it. And so Einstein's colloquia were never publicly announced, as there was a searcher saying that there would be a colloquium, and the person giving it wasn't named, and people and I simply assumed that it would be Einstein. So Einstein had convinced himself that gravitationally bound objects would never radiate and wanted to give a colloquium, a Princeton colloquium on the subject. Well, H.P. Robertson, former Caltech professor, or heard about it and asked Einstein to see Einstein's arguments. And when Einstein went over the arguments with Robertson, Robertson showed him that it was incorrect. And so Einstein showed up at the colloquium and he bowed and said that there wouldn't be a colloquium because H.P. Robertson had shown him that the analysis was wrong. But later, it seems to me that Infeld went to Canada after the war and at one point had claimed to prove the same thing, and Andre Traublin showed him it was incorrect.

17:30 Andre Traublin, I guess he's still alive and somewhere in Poland, you know, he's going Those analyses... I guess yourself and John Green wrote a paper also addressed... Rosen's argument that based on, I guess the outgrowth of Einstein being corrected by Robertson was that he and Simon Rosen came up with the metric for cylindrical gravitational waves. Right. And Rosen then in the 50s came up and put out a paper and said that it seemed that the waves didn't have any energy because you could transform a way that the energy went to see what's going to happen. And then a number of people, Simon Bondi and on the paper Ryan, yourself from Heeler, pointed out that the, since the Riemann tensor didn't transform away, that he would still have motive. Yeah, of course, in my text, I simply wrote all the gravity wave formulas simply in terms of the Riemann tensor and wrote the energy density in terms of the Riemann tensor. It always seemed to me that the renal tensors would be a really important thing there, and build apparatus to measure it, and that the gravity wave produces a wave renal tensor, and that's the absurd part. Do you remember if a couple of people that I talked to implied that the, well, I know that there was a big emphasis on the Rayman tensor particularly as an important thing from certainly the Chapel Hill Conference in 1957. Do you remember there being a particular point at which it began to appear that that was the key thing, or was that fairly obvious from you? Well, certainly obvious to me. From the very start, there's a whole chapter in the model textbook on the idiosyncrasies

20:00 of the stress-energy pseudo-tensor. It's one of the features of the general theory that's never been, in my opinion, quite cleaned but uh it seemed to me from the start that you could avoid all those problems if you just uh worked with a remote tensor because that's not going to be transformed away or won't be changed a lot much by important transformations i guess you got a copy of this uh paper that uh published in the Let's see, this conference in Pakistan, somewhere in the right, I guess, that's from here. Superdome 1987A, Gravitational Radiation. This is a short, about 12-page paper that is a pretty good summary of the 37 years of work. And I think what's important is that these correlations here, these are the correlation functions. Two antennas on the baseline of about 8,000 kilometers correlated with neutrino detectors in the same way, and there's a 67 second period. You can do a fast Fourier a transform of the work. Early on, the Japanese Kamioka people sent me a copy of their tapes with no restrictions as to analysis or publication. And there are several, I think, important things about these observations. One was that if you talk to a professor of probability and statistics,

22:30 he'll talk to you about a priori and a posteriori assumptions. And the standard example is if you're stopped at a red light and you look at the license plate of the car in front of you, you say, what is the probability that I will see those digits? And the probability is incredibly small. that you'll see the digits that you're seeing, but you're seeing them, and you're not seeing a miracle. It's just that first you see them, and then after you see them, you worry about the probability. Well, if you wrote the digits down first, before you started, and then started working around totally different building well in the case of the supernova data the Japanese neutrino detector clock was incorrect and the Japanese neutrino data didn't correlate with anything. So IMB and Boxon and Montblanc detectors had data, particularly IMB, and so they started analyzing the data and they discovered that if you corrected the Japanese clock, they got Well, it turns out that if you start with that clock correction, and then look at the gravity wave data, the gravity wave data correlate very well only if you put in the clock correction. If you don't put in the clock correction, you don't get correlations. So I think that's important to remember the fact that, you know, initially the data don't correlate with anything, including the bars. But if you correct the clock so that it correlates with IMB, it also correlates with the bars. That's important. The other thing is this one and a half hour period during the supernova, there are, I think, events on the Japanese tape.

25:00 If you write down the corrected times of those 192 events, and look on the Rome and Maryland bars, and write down the power outputs, do a fast Fourier transform, you get this large peak at 67 seconds, and so that gives a nice model of what happened when the supernova the supernova grows up into two orbiting neutron star objects. And every 67 seconds, they get very close. Now, if they're neutron star objects, they're very dense, and they will have internal oscillations onto the kilohertz region. But every 67 seconds, they get close enough so that the strong gravity forces induces a quadruple more given the burst of gravity waves. Also, every 67 seconds, the strong gravity field changes the shape and volume of the neutron star fragment. And the thing that keeps the neutron star from decaying into neutrinos, protons, and whatnot, is that the Pauli exclusion principle tells you that all the quantum states are filled and the neutrons can't decay. But if you change the shape and volume, because they're very close together gravity fields, lift the Pauli principle restriction on neutron decay, permitting a burst of neutrinos. So that explains why you get simultaneous bursts of neutrinos in gravitational waves. And why, instead of getting one pulse, there were over a hundred. Now, if there were one pulse, one wouldn't. At least I would tend not to believe that data But with one pulse, people say, well, how do you know it wasn't a cosmic ray shower? And you don't know it wasn't a cosmic ray shower, but if there are over 100 pulses with a periodicity of 67 seconds and they're periodic with a neutrino detector,

27:30 that's pretty strong evidence that you're seeing gravity waves from the supernova. So these data, I think, are quite important. And since supernovae don't occur frequently, you know, I think that the features of this, which are important, are the fact that if you start a field like this, and you build gravity wave detectors and you operate them, I would be the first to say that you can never be sure your feet are on solid ground or that you haven't made some stupid mistake or some more subtle mistake. But here, for the first time, the bars correlate with six elementary particle detectors. And so, you can be sure that the things you're seeing aren't due to earth vibrations, passing trucks, or a variety of other things, that neutrino detectors in Japan, in Cleveland, in Russia, in Switzerland, get pulses and those pulses currently in the bars. Also, the theory published in 1986 gives a good account of what the pulse heights were. And each pulse height, each pulse amounts to less energy radiated than 10 of the minus 2 solar masses. You mentioned the skepticism expressed by people early on, but that largely based on what people perceived as the likely source of strength. To some degree, I think that was part of the problem.

30:00 Perhaps I should give another statement about Peter Bergman, who, as I say, at the first international conference, pointed his finger at me and said, I have done nothing right. Well, about four years ago, So I was invited to be a speaker at a celebration in Europe honoring Bergman on the occasion of one of his birthdays, 75th or 80th or whatever. and so I got up and presented the quantum theory of the bar and when I as soon as I finished Bergman got up very angrily and said Joe, he said when you talked about this 20 years ago the things you said made sense but what you just said makes very much sense. So he'd forgotten that 20 some years ago he said that what I said had made no sense and in the intervening 20 years he had decided that perhaps the 1960 cross-section was right after all. So all I can hope and pray is that he'll be alive 20 more years and find the The first one was in, I think it was in Warsaw, I think it was 1962, right. I remember I was a visiting professor at the University of Colorado at Boulder, and I had to leave for ten days to go to the Warsaw Conference. I remember my first wife being terribly afraid she would never see me again. It made me fuss about my going, you know. So in fact, there was, as you say, a few people felt that the fear of the bar that you don't mind.

32:30 Yeah, of course, I think the things which are in dispute now are something like this that probably have a bar and if you treat it classically then all mass elements are the same and there a lot of people, some guy last week kept arguing with me in Aspen, Colorado, that the bar, he said, simply has to be described by classical physics because it's a big microscopic object. And my answer was, if this room is filled with a superconductor, that's a big, massive object, and classical physics doesn't come anywhere near explaining superconductivity. What's even more impressive is if that wall were covered with a metal, and you shine a flashlight, photoelectrons come out instantly. And classically it would take over a hundred years for the first photoelectron to come out of the wall. and so there are examples of macroscopic objects classical physics isn't slightly wrong, it's wrong by orders also if this whole room is full of uranium then classically all the atoms should decay at the same time and quantum mechanically you should get radioactivity that you see so you have that well anyway think of the bar made up of atoms i like to think of planes of atoms because if you drive a plane of atoms you can excite the normal mode of the bar by driving the plane at the right frequency see anyhow the quantum theory says that you can get energy showing up in any mass element statistically classically the energy has to show up at all mass elements at the same time

35:00 you could get a burst of phonon showing up any place in the bar Now, the period of that group of phonons doesn't have to bear any relation at all to the normal mode frequency of the bar, which is like 1,000 hertz. The period of this oscillation can be as high as the bi-frequency, which is 10 to 14 hertz. And so you have a group of phonons show up. Well, as soon as the excitation is over, the group of phonons don't disappear, they're there, and they oscillate back and forth from end to end. If you have center crystal instrumentation, it's a non-resonant structure, and center crystal sees the group of waves every pass and so the electronics is going to see a group of waves and then half a cycle later that and half a cycle later that so what i'm saying is it doesn't matter what the period of these waves are the bar itself converts as a result of the oscillations back and a substantial fraction of it into the low-frequency fundamental that the electronics is tuned to. And that means that the bandwidth of the bar is much wider than the bandwidth of the micro-interferometer, Contrary to Ray Weiss, Thorne, and Treeder, and everyone else who felt from the start that bars were worthless because the bandwidth is very narrow. But even if the original classical theory of the bar is right, the bandwidth is bigger than any of these guys thought. because if you treat the bar classically and write the equations like maxwell's equations and equations of elasticity you're an engineer you have a pair of terminals which is the crystal and then you have the remo tensor which is a voltage generator

37:30 and you have an inductance which is the mass equivalent and you have a capacitor which is the stiffness equivalent what's important is that you have another generator which is the Einstein-Brownian motion generator and then you have the output Now, if resonance, the Riemann tensor gives you a big voltage output, but the Einstein-Brownian noise generator couples to the normal motor the same way as the Riemann tensor. It also gives you a lot of noise. You go off resonance and the Riemann tensor gives you a lower the noise voltage is smaller so in the absence of external noise the signal of noise ratio of this is considerably is good over considerably greater bandwidth than the resonant bandwidth because the well let's say einstein was wasn't just the genius who gave us general relativity he was the genius who gave us the theory of the brownian motion first and so Even the classical picture of the bar says it's a lot wider bandwidth than 1 100th of the hertz that people normally quote. But if you think in these terms, the bandwidth is, of course, orders greater than that. Now, if you say, if you follow the late Professor Fairbank, he said that he didn't like this, he would put a highly resonant structure on the end of the bar. and now he's adding a degree of freedom and a forward well if you work this thing through when this thing oscillates back and forth it takes this resonant structure i have a long time to absorb significant energy from that back and forth but the center crystal instrumentation sees it immediately and i think that's one of the reasons that bars instrumented this way I have never seen anything, including the LSU crew that are still getting lots of money.

40:00 My understanding is the Stanford operation has been shut down by the NSF. Anyway, welcome to the club. and the real world and all these possibilities. But I think what's of primary importance here is that again, one of the things that I would strongly recommend you read, and I think you can find it in the Caltech Library. A thin little book by Max Born, called Theory and Experiment in Physics. There's a retrint of it published by Dover, and I have found it in libraries. And, of course, the enormous amount of progress that's been made in physics in this century has been a result of, of course, theory and experiment. I mean, one wouldn't have had the Palmy exclusion principle if people hadn't looked at the periodic table and realized that ordinary quantum mechanics couldn't explain the periodic table without the common principle requirements of, of course, electrons or ferreons, and you can't put a border to two in one quantum space, it's a half spin in half, and you follow that up, and of course you get the periodic table, but without the observational fact but you have a periodic table and an algorithm of coding quantum mechanics. And I think that these experimental data are critical in understanding the theory of the bar and the other things. And I continue to be shocked at the large number of people who work in the field as theorists

42:30 and years and years, thousands of papers, no experimental data. And this is the only supernatural data. And, of course, it's been ignored by everybody. You really have to have experimental data. And that was the progress that was made, starting with Galileo, before Galileo, and he believed with Aristotle that the human intellect could give you a precise description of how the world was put together without making observations. You have to make observations. Yeah. Yeah. I'm going to ask you just a quick focus, and you mentioned that your book was critically reviewed by Meg Vitti. Do you remember where the review appeared? I remember when the review appeared, she was concerned that I didn't make a lot more of the cosmological costume, and, you know, he didn't think anything in the book was good except the chapter on the tensor calculus, and, well, I sort of wondered about his intuition about that because the chapter, I won't say the chapter on the tensor calculus was written but I was heavily influenced by Marcel Rees, who was the greatest mathematician in the world. He retired to the University of Maryland in the mid-'50s, and he took an interest in me not because I was intelligent, but because I owned a car. He was a cripple, and he couldn't walk from the place where he lived to school. and I would show up in his office about an hour before he had to go and he would give me a private lecture

45:00 on Imanian geometry and tensor calculus and argue with me and give me an oral examination driving home through traffic and that toughened me up quite a bit Of course, the treatment of tensor calculus is essentially what I got from Marcel Reiss during those years. Perhaps that was one that he might get better. Actually, I can't remember the details of the paper in the 50s. which was cited as evidence against the physical location. I see. You said that he couldn't have claimed that. Yeah, well I remember there was a paper written on that subject. Yeah, I would say in the middle 50s when I got into this field, I wouldn't say the scientific community was equally divided, but there was a sizable fraction, I believe, that there weren't gravitational waves. Again, being an engineer type, it seemed to me the thing to do is to build an antenna and go out and work. you probably have

47:30 copies of this just to be sure as I say this is I think a 12 page even 12 pages summary of the 37 year effort a 9 page summary So I feel proud of myself being able to sum up 37 years in nine pages. Yes, that's impressive. And the quantum theory is pretty trivial here. And I think that this treatment answers the criticism. against the cross-section theory so I give you a copy of that now you do a number of things I won the probably make a copy of this this is the paper that has this analysis that, like everything else, this analysis has been questioned, but I didn't do the analysis, the room group did the analysis, and as I say, I've checked it with three mathematics I guess the other thing I should probably give you is, uh... If I remember, I guess, if I remember correctly, that even the alternative analysis of the coincidence in the Academy of Opinion and Gravity Barham-Pennel case still left it as a likelihood that it was a... He could prove using Monte Carlo methods that it couldn't have happened.

50:00 Well, I'm always flabbergasted when you build a machine and you operate it for, in this case, 30 years waiting for a supernova. You finally get supernova data, and first the federal government tries to get you to avoid even reading tapes and then after you read the tapes and someone else does the analysis you find the chance of correlations a few times 10 to 11 is 8 and you see things like I suppose that could be accidental I suppose these could be accidentally. Then someone on shoots. And another one is Rubenco, who proved that it couldn't have happened. Perhaps I'm not sure. One should reach a stage where one isn't there.