Notice: We are in the process of migrating Oral History Interview metadata to this new version of our website.
During this migration, the following fields associated with interviews may be incomplete: Institutions, Additional Persons, and Subjects. Our Browse Subjects feature is also affected by this migration.
We encourage researchers to utilize the full-text search on this page to navigate our oral histories or to use our catalog to locate oral history interviews by keyword.
Please contact [email protected] with any feedback.
This transcript may not be quoted, reproduced or redistributed in whole or in part by any means except with the written permission of the American Institute of Physics.
This transcript is based on a tape-recorded interview deposited at the Center for History of Physics of the American Institute of Physics. The AIP's interviews have generally been transcribed from tape, edited by the interviewer for clarity, and then further edited by the interviewee. If this interview is important to you, you should consult earlier versions of the transcript or listen to the original tape. For many interviews, the AIP retains substantial files with further information about the interviewee and the interview itself. Please contact us for information about accessing these materials.
Please bear in mind that: 1) This material is a transcript of the spoken word rather than a literary product; 2) An interview must be read with the awareness that different people's memories about an event will often differ, and that memories can change with time for many reasons including subsequent experiences, interactions with others, and one's feelings about an event. Disclaimer: This transcript was scanned from a typescript, introducing occasional spelling errors. The original typescript is available.
In footnotes or endnotes please cite AIP interviews like this:
Interview of John Clauser by Joan Bromberg on 2002 May 20,
Niels Bohr Library & Archives, American Institute of Physics,
College Park, MD USA,
For multiple citations, "AIP" is the preferred abbreviation for the location.
Clauser discusses his father's influence; early interest in electronics; undergraduate study in physics at California Institute of Technology in early 1960s; graduate study at Columbia University in the late 1960s; research on the Bell inequalities as a post-doc at University of California, Berkeley with C. H. Townes in the 1970s; collaboration with Abner Shimony and Michael H. Horne; atom interferometry and its possible applications; quantum mechanics and its conceptual problems.
Prandtl created a revolution in fluid mechanics that was very similar to the revolution in quantum mechanics. He developed, just after the turn of the century, a mathematical ability to understand fluid mechanics. And once he did this, man, he and his students just solved all the outstanding problems. The same thing happened in quantum mechanics. Once the mathematical structure was developed, very quickly all the problems got solved. Well, Bob von Kaman was one of his brightest students, and was now at Caltech. And Dad and his brother were trying to figure out what they wanted to do for grad school, and this was very exciting stuff. Aircraft were just starting to be built, high-speed aircraft. Actually, the concept of supersonic aircraft was in sight at this point. And it was a very exciting time. So they decided rather than going into physics, they would work for Bob von Karman. [Dad and his brother were both undergrads in physics at Caltech and were then deciding what to do for their PhD's.] Anyway, so he went on and designed aircraft for the war. Dad was Chairman of the Aeronautics Department at Johns Hopkins, and then he became Vice Chancellor at UC Santa Cruz, and then went back to Caltech and got an endowed chair in engineering. But all along the way, he always was trying to understand physics, and there were very strong similarities between the mathematics of fluid flow and the mathematics of quantum mechanics, and he didn't understand quantum mechanics. And he kind of pre-programmed me as the guy who might help try to solve the problem that he couldn't solve.
At what point do you appear in this trajectory?
Well, I was born a year after Pearl Harbor. And so the first thing that I can remember really was at the close of the war, we moved back to Baltimore. My dad took a job at Johns Hopkins to create an aeronautics department there. And so as I grew up, basically as a kid, I just would come in after school to his lab. We lived in the suburbs, and so I would do homework— I was supposed to be doing homework, but mostly what I would do is just sort of wander around the lab and gawk at all of the nifty laboratory equipment. And I kept thinking, "Wow, boy do these guys have fun toys. When I grow up I want to be a scientist so that I can play with neat toys like this." And I did a lot of science projects in high school and the like. My dad was absolutely a marvelous teacher, my whole formative years. Every time I asked a question, he knew the answer and would answer it in gory detail so that I would understand it. I mean, he didn't force feed me, but he did it in such a way that I continuously hungered for more. I was an electronics whiz kid. He taught me some of the basics of electronics, and I just went off and built some of the earliest computers and the like. I built the world's first video games, and I actually won a whole bunch of prizes in the National Science Fair for doing this.
Were there other children in the family?
My sister, one sister. She basically went into the humanities rather than into science. She's a freelance editor in Manhattan now. Although interestingly, she is unique among women that she has probably dated more men who have gone on to win Nobel Prizes than any other woman in history, and/or had affairs with some.
So you chose physics at some point.
Originally I was thinking since I was such an electronics buff, I originally thought I would go into electrical engineering and computer science, even though computer science didn't exist yet as a field. But once I got to Caltech, I very quickly— My dad had advised me, "Look, get the most basic, broad education you can." And being an engineer, there was always kind of a snobbery between engineers and scientists, and being an engineer he was on the lower echelon of the snobbery, and he kind of wanted me to be on the upper echelon of the snobbery. I guess. But the bottom line was that he urged me to get just as broad an education I could, and not be stuck in any one narrow field, and with the promise that if you really understand everything from broad basic principles, you can do anything you want. And that has proven the case, I think.
Is there going to be anybody else in the picture who is influential, aside from your father?
Certainly the two most influential people in my life were my father and my thesis advisor. That was Pat Thaddeus at Columbia. But in truth, probably my father still is far more influential than anyone else. There are any number of other people who I have kind of worked with and interacted with who have had relatively powerful influences. I mean, clearly John Bell for one, although I never really knew him personally, clearly he and I interacted very powerfully in our work. [Note added: See also Townes, below.]
What about at Caltech?
At Caltech? Not really. I can't think of anybody there that kind of blew my socks off as far as a powerful influence.
That would have been in the '60s that you were there.
Yes. I was in the Class of '64, high school class of '60.
So who was in physics then?
Well, let's see. Well, probably the two great luminaries on the physics faculty were Feynman and Gell-Mann among the theorists. There were frankly no really world-class experimentalists on the faculty. In fact, there was one guy, and I forget what his name was...?
He was in low temperature physics.
Because I know David Goodstein is the husband of the archivist there who was the person that I know.
It wasn't Goodstein. I forgot who it was. When I went to Columbia, that was a dramatic piece of culture shock.
Well, Columbia was dramatically bigger. But Columbia not only had a counterpart to Feynman, i.e. T.D. Lee who certainly— Frankly, I think he was a much brighter guy. I didn't particularly care for Feynman. I didn't really like him very much. And frankly a lot of the stuff that he did and said was wrong, especially in foundations of quantum mechanics. But T.D. Lee was one of the brightest guys I'd ever run into. But also this was a world class experimental physics school.
Who was at Columbia?
In atomic physics, the luminaries were— Well, this is really where atomic physics started with Kusch and Rabi, Ramsey was a student of Rabi came out of there. Lamb was an opponent of Rabi's. Lamb and Rabi did not get along at all. Novick was still there, Kusch was still there. And plus they had a very powerful set of high energy experimentalists: Lederman, Sam Ting, who I routinely came to blows with. Who else? There was Foley, Serber were very prominent. Madame Wu was there, had her whole nuclear physics group. Sven Hartman was doing solid state and laser interactions in matter. And then there was also another group there that was associated with electrical engineering and plasma physics, a guy named Bobby Gross running that. [Other important players included Will Happer and Bob Novick.] It was a very dynamic, active place.
So you came in and whom did you link up with?
Most of the influence I had actually was through— I peripherally interacted with all of these guys, but dominantly with my thesis advisor.
How did you come to pick him?
Actually he had a marvelous idea, which we ended up never doing. I asked him what he was going to do and he says, "Well, I'm going to put a radio telescope in a U2." And I thought, 'Wow, that sounds like it's going to be great. That sounds like fun." Well, it turns out we never did that. I was interested a little bit in general relativity at the time, and this was the era of the first measurements of the cosmic microwave background. and so my thesis was actually the third measurement of that. And in fact, the presently accepted value still fits nicely within its error bars, 2.73 degrees. My error bars were very conservative.
Why was he listed on that paper at the Goddard Space Institute instead of at Columbia?
Basically, Goddard Institute, there was a branch of Goddard Space and Flight Center, which is in Houston, that was an academic branch that was set up to be in New York and have scientists in associated fields from NYU, Columbia, Brooklyn Poly, Yeshiva. I think there was one or two others. But essentially, the major universities in New York City could bring in staff scientists who then would be adjunct professors at these various universities and then put them on the staff. That was one of these. Pat was a student of Charlie Townes. One of the other people who also probably had a very powerful influence over me, has always been Charlie Townes.
Even before you went down to Berkeley'
Even before I even knew him. His legacy was still there. Lamb had left, Townes had left, and Townes and Ray Chiao had left. Actually, all of these guys had been a powerful influence there at Columbia, and their legacy was really still left behind. All of the people who created lasers and astrophysics, were all Townes' protégées. Pat Thaddeus was a Townes student. Just all the names in astrophysics and lasers were Towns' students.
I want to pursue that, but also at the same time, you were taking courses in quantum mechanics at that point, or you were done with quantum mechanics?
Yes. No. There's a little known piece of dirt, and I'm not sure you want to publish this. And in fact I found that I was not alone. Columbia had a requirement that you had to get a B or better in four courses. One of them was in Advanced Quantum Mechanics. One of them was in Physical Mechanics. One of them was in Electricity and Magnetism, and Nuclear Physics or something like that was the other one. And these were really tough courses. And so if you didn't get a B or better, you had to take it again. And you had to keep taking it until you got a B at least. Well, one of the semesters I got a C in the quantum mechanics class. So I took it again, and I got another C. So I took it again. Well, it turns out, I am told, that Charlie Townes also had to retake that course at least twice. So I at least had good company.
Who was teaching it?
This was Gerry Feinberg when I took it the first time, and then Gerald Wick the second time, I think. Feinberg seemed to be very bright and whatever, but he really didn't explain it to me very well. One of the problems I have, I'm very different from many physicists, which is both a blessing as well as a major impediment. I am not really a very good abstract mathematician or abstract thinker. Yes, I can conceptualize a Hilbert's Space, etc., I can work with it, I can sort of know what it is. But I can't really get intimate with it. I am really very much of a concrete thinker, and I really kind of need a model, or some way of visualizing something in physics. Once I can do that, I am very good at using that conceptual model to do calculations, to innovate new ideas, answer a problem, etc.
Concrete model like a piece of equipment?
It's the difference between abstract pure mathematics [and applied mathematics]. There exists a set of numbers with algebraic structure of such and such, and we will define a particle as being something for which this operator commutes with that operator, etc. I haven't the foggiest idea what any of that means. But an electron is a charge density which may be Gaussian in shape and its shape, and it's about this big, and it's held together by various forces, and this is how the forces work that kind of hold it together. The difference between those two [concepts] are very dramatic differences of thinking. Now there's a whole class of physicists who can only think in the former method. I can only think in the latter mode. And many theorists will write papers and write down, "Here are the equations of general relativity in tensor form." Well, it's beautiful that way. Yeah, but what does it really mean? Yes, you can make Maxwell's equations in as compact a form as possible, but you haven't the foggiest idea of what they mean by looking at them like that. When you break them down into Faraday's Law and things like that. then all of a sudden you get a real feel. Then all of a sudden, "Oh, of course. The electric field does go this way, and the magnetic field goes that way, and they are all tied up in these nice, beautiful patterns," and you can draw pictures of them and see what's actually happening. But if you just have some four vector notation for Maxwell equations, it really is very difficult for me to understand.
Now, Einstein had to actually make that transition somehow. He started out a very concrete thinker and looked at very simple conceptual problems. Apparently, he had a big problem learning tensor theory. He apparently claimed he wasn't a very good mathematician either. I [similarly,] am no where near as good of a mathematician as I would like to be. There was a question that came up, and I still remember it, and it was just kind of characteristic. I had great problems all my life understanding the square root of minus one. In high school, I learned that the square root of minus one was called this little symbol i. Well, there is no such actual number. Well, so it's an imaginary number. If you multiply i times i you get minus one. All right. But suppose I go to the store to buy 1+i candy bars. I could buy one candy bar, or one and a half candy bars, but I couldn't buy 1+i candy bars. But it's useful because it makes the equations work out, and once you play with it, the equations work out better that way. So then when I get to Caltech, the first thing we learned was that we were going to talk about waves. And so I could go, sine (kx-wt). That's obviously a wave. But e, that's a complex number. "Well, why don't we use sine rather than this complex stuff?" "Well, it's a mathematical artifact. Don't worry about it. It just makes the equations look nicer."
So you can take the real part, or you can take the imaginary part, or the modulus squared, etc., and get powers and things like that, and then the equations work out very nicely. And then you can do Fourier analysis and you can have complex frequencies... damped oscillations. Okay, then we do electromagnetism. I thought, "No problem. The electric field is really always real, and the magnetic field is always real. We can understand that. Okay." Then, we get to quantum mechanics. And in quantum mechanics— In fact, Schrodinger wrote a paper about the role of it in his equation, and it turns out that you can't get rid of it. You can't make the equations always real.
Then you go over into another branch of physics that you're learning simultaneously —special relativity. You have the spatial dimensions. Typically this is Minkowiski or Lorentz, but supposedly spatial dimensions are sort of real things. And time is not really a spatial dimension. And so they put them all together and make time imaginary. And the equations again work out nicely, but you can always calculate [real quantities]. You never end up with a complex spatial dimension or an imaginary time dimension, etc. Time is still stuff you measure with a clock, and positions are stuff you measure with a ruler. But the equations worked out better there. But in quantum mechanics, you couldn't get rid of the i. You now had these two sources of i that were coming in here. Then when you finally get into relativistic quantum mechanics, the equations have complex numbers in them. And one possible source of i was from relativity. Another possible source was from a phase shift of a wave function. In relativistic quantum mechanics, it was not obvious where this i was coming from. Was it the relativistic i, or was it the phase shift i. And so I remember actually Gerald Feinberg, someone raised his hand in class, and I assumed with all of his background in mind, asked Feinberg, "Well, what does i mean in this equation here?" And Feinberg smiled and said, "The square root of minus one." And he did the rest of his lecture and never explained, probably because he didn't know.
Were you already sort of worrying over...?
Hell yes. That was one of the reasons I kept not getting B's in it because a lot of it was just mathematical manipulations, and I was not very good at it; and didn't understand, didn't know why I was doing it. And I felt very uncomfortable with it. And once I felt uncomfortable with it, my brain kind of refused to do it.
...Tell me about Thaddeus, how you worked with him and what makes you say he was such an important influence on your physics?
He taught me a lot of quantum mechanics. He taught me really how to use matrix mechanics.
Did he share some of your mistrust of it?
Hardly. When I started reading all this stuff, it was very much the opposite. He thought this was a terrible waste of [my] time, and I was ruining my career. And in fact, at one point later on, even after I finished the experiment,* he had written a recommendation for me, where basically someone said [to me], "Look, you can't send his recommendation in for a job application because he's totally panning you because he figures you're going to go off and do some more quantum mechanics experiments." [In the letter, Pat says] "Don't hire this guy if there's any chance that's what he's going to do, because it's all junk science." Since then, he's apologized for that, but you know, he has since then decided that there really was some interesting physics there. But even still, after I'd just finished the experiment with Stu Freedman, he thought this was all [junk]— "Well, of course you got exactly what you expected, what was the point?" And in fact, most of the physics faculty at Berkeley all said [the same thing]— Gene Commins had kind of a similar attitude. "What a pointless waste of time all of that was."
(*Stuart J. Freidman and John F. Clauser, "Experimental Test of Local Hidden-Variable Theories," Physical Review Letters 28 #4, (3 April 1972), 938-941)
What about Townes?
Townes was just the opposite. Without Townes, I could never have done that experiment. Townes was the guy who actually twisted Commins' arm to put Stu Freedman on the experiment and to steer Atomic Beam Group funds into doing the experiment. And if I had not convinced Townes early on [it never would have happened]— In fact, the first thing I did when I arrived at Berkeley was that I gave a seminar describing this to Townes' group, and Gene Commins was there, as he had done this previous experiment with Carl Kocher. And at the end of this, Townes kind of puts his arm around Gene Commins and says, "Well, what do you think of this, Gene? It looks like a very interesting experiment to me." So if Townes puts his arm around your shoulder and says, "Looks like a very interesting experiment," I mean, Commins thought it was a total crock. But when Townes says he thinks it's interesting, it's interesting. It becomes interesting at that point.
It's surprising to me because Cummins' experiment was precisely on the Einstein-Podolsky-Rosen paradox.
They did not understand Bell's Theorem. Nor did they understand just the significance of what it all meant at that point, absolutely not. But it was Eyvind Wichmann, a theoretician at Berkeley, a very bright guy, I think, who actually suggested that they do the experiment. He's a theoretician, a very bright guy. While I was at Berkeley, I actually audited a course he was giving. He apparently suggested to Commins that they can actually do this, and apparently they originally planned on doing that as a lecture demonstration. Originally it was built on this rolling table with wheels on it so they could roll it into a lecture. Kocher was going to do this kind of as just a little project, and polish it off, and then get on with a real thesis project. Well, it was much more difficult than they originally anticipated, and it became his thesis project. No, you could not have done it as a lecture demonstration. It took weeks of counting to get any decent statistics.
It's interesting. The other experiment, which was related to this point was the Wu-Shaknov experiment where they did this with positronium decays. And the obvious conceptual model that people would have is what I refer to as the Furry hypothesis, where you assume that each photon is a real photon, it looks like every other photon, etc., and there's no interference between two separated photons...
So the wave functions kind of dissolves in the...?
Well, it becomes localized around the propagating particle. Well, it turns out that for the positronium annihilation, the prediction for this polarization correlation was originally calculated by John Wheeler. And he got it wrong because he assumed the Furry hypothesis. He didn't realize he was doing it at the time. And it was not until Pasternack came along and redid the calculation, including the EPR interference terms, that the calculation was done correctly. And so Madame Wu now said, "Gee, I have two well known theorists giving me different predictions. Maybe I'd better measure it." And so she did. But no one at that point realized that [it was an example of EPR]- I don't think in any of those papers, Wheeler's nor Pasternack's nor Wu's, do the names Einstein, Podolsky, Rosen ever get mentioned. And it wasn't until Aharonov and Bohm looked at it and realized, "A-ha! This is exactly what we're looking for."
When you were in these quantum mechanics classes, were you reading stuff like Von Neumann or Gleason or any of these?
I did, yes, but not as part of the course work, no. Only because I couldn't understand it. And a lot of it was when I finally got a book by de Broglie called Nonlinear Wave Mechanics. I read that. And then I got on to read Bohms' papers and then read Einstein-Podolsky-Rosen, and all of the various things that Aharonov and Bohm had written on all of this. And so I had all of that read. I was familiar with all of that when I first ran into Bell's paper. It was all because I couldn't understand quantum mechanics. I kept looking for a conceptual model, "What's going on here? How do you understand?" That's the problem. Here's the conceptual problem that has bothered me for years, and it's kind of related to my description of the electron if you will. In quantum mechanics, the books all make this seem like simple wave mechanics, i.e. what you would see— a direct analogy with waves on the surface of a pond. And they show pictures, "Well, you throw the rock in here in the pond, and you get these nice circular ripples propagating out, and I can put a barrier with slits in it, and then defract the waves through the slits". These are very simple things you can actually do in wave tanks, and even in undergraduate laboratories you have ripples where you have a ripple generator and you can make interference patterns this way. And then even worse, they say, "Okay. A particle, we can represent kind of as a wave packet," whatever that means. Well, okay. Sort of imagine this little cluster of waves, and they are all propagating in real space. Now, the description of ten particles, nine particles, eight, three, two, one, the meanings of the terms should all be the same (depending on how we do this) as that of the one particle case. "We talk in terms of x, y, z, and t, just as if they were in this wave tank. So this wave function, I guess we could say at any time, "I could go along in the wave tank, I could freeze the motion, and with the ruler measure the depth of the water at every point in x, y, z, and t". That's what x, y, z, and t mean. So I have as a function of x, y, z, and t, which would be what I would get from my ruler.
Now consider a two particle case. 4r is no longer a function of x, y, z, and t. It's a function of x1, y1, z1, x2, y2, Z2, . Has space and time grown? So far I thought we only had three spatial dimensions and one temporal dimension. But now I see seven numbers as the arguments. So how do I now go along with my ruler and measure at every point? I want to measure in x. What does x mean for two particles? Let's say we have two particles in that wave tank. I can't do this anymore. And the description of what x means in the argument of y , it's got to be the same for a one particle case as it does for a two particle case. So if I can't do it for a two particle case or a three particle case where I have now ten arguments, or a four particle case, there are not that many spatial dimensions. So if I couldn't do it for four, three, two particles, I shouldn't have done it for one particle either. So that means going through and measuring the depth of the water at each point in the tank is not what x, y, x, and t means.
When you guys were graduate students, did you actually talk about these things among each other, or you just sat in the corner?
I sat in my corner and tried to understand it myself. Nobody else talked my language. So how do I build a wave tank, for example, of quantum mechanics for two particles? I can't. So that means that I can't even do it for one particle. Which means this whole idea of wave packets that all of the books put in there is to try and make you feel comfortable with it, all of those chapters, you might as well rip up and throw them away because they are wrong because that's not the correct conceptual model. But I can show you dozens of quantum mechanics textbooks where they will show you this and try to make you feel comfortable with it as a conceptual model. But it really can't be. That is not right. [Note: Almost every such book describes a "probability current". But that concept is muddled (wrong) if there is more than one particle. The books don't recognize this part].
Now, imagine instead, if you will,— suppose I have a roller coaster that has a wavy track. I could go through with a ruler and measure the height of the track at various points in space. So I could generate (let's not call it y) let's call it 9 of x, etc. And if this roller coaster track moves in time, I can get 9 of x and t, if the track is moving up and down, like an earthquake is going through it or whatever. I now put a car on the track. And in a given time, that car has a position. I put a second car on the track. So I can say, "Okay. That first car has a position x1 at time t, the second has a position x2 at time t, ... So now say I put five cars on the track, and I have this magic complex number down on the hotdog stand at the base of the roller coaster that just flashes up and displays a complex number which depends on the positions of all of the cars on the track, assuming that the cars are on the track. How do we know where the cars on the track are and what is the relationship between the cars and the track and this complex number. It [y] has absolutely nothing to do with a wave. [The "real" physical-space wave, ???(x,t), never appears in the equation]. That's what the x1, x2, y1, y2, etc., mean. They are assuming that there are cars. This whole idea of having a wave packet was trying to convince you that this blurred thing was a particle. No, it's not a particle. The x really is a configuration space dimension, and not a dimension of the wave. And why a wave equation [for ???] is anybody's guess. It's totally obscure to me why Schrodinger's equation should be in fact a wave equation. It's not, really, because it's got complex numbers in it.
Have you written that up, by the way?
No. Maybe I should. That's the difference that I was describing to you earlier when I say I am not an abstract thinker. The abstract thinkers just have this thing and they say, "Oh, well, blah, blah... We have N particles. We do this, and then we manipulate the algebraic structure..." Whereas in my case, I have to see it. And once you see it, you say, "Oh, we're looking that the x's, y's, and z's have very different meanings. One is a configuration space variable, and one is an actual real spatial variable." And only a handful of physicists really seemed to have paid much attention to this. DeBroglie, in his book, was acutely aware of these differences, although he never explained it the way I just did with the hot dog stand and the roller coaster. And I always kept trying to think...
Now, in Bohm's theory, he uses the wave tank model, and he has the waves propagating in real space, and the particles are like surfers on the wave, that are kind of surfing around on very real waves. Again, now we have ??? as a function of a background wave pattern, which is still something we can measure with a ruler in a wave tank, plus we also have the particle itself, which has a very well defined position at any given time. Now, both Bohm and deBroglie, when they wanted to try to generalize that simple model, they could do it for one particle. When they wanted to try generalizing to more than one, they came up against a brick wall. They didn't know how to do that because it was all linked into the fact that these x's didn't mean the same things that the quantum mechanics, which was trying to pull wool over your eyes and make you believe.
I had a question about that period. How did you get to go up to MIT for this seminar that you gave?
That was interesting. I had one of my buddies at Caltech that went to Harvard and was a student of Ramsey's. And Ramsey had Dan Kleppner as a junior faculty working for him. Well, Dan Kleppner curiously — much to Harvard's obvious mistake, Harvard did not give him tenure. And so MIT went down there and snapped him up in an instant and said, "How would you like to move down the street and come down to the other end of Mass Avenue because we can do just as good physics." And so my buddy was working actually in Ramsey's group but on Kleppner's experiments. And so when Kleppner moved to MIT, (this was Frank Winkler), he moved down with Kleppner, although Ramsey was still his nominal thesis advisor at Harvard. He was still a student at Harvard, but he was doing work down the road at MIT. And so we [my wife and I and another Caltech classmate at Columbia, Jim Whitney] would frequently go to parties together, commute up and down from New York, and he would come down and visit New York and stay with me, or I would go up to Boston and stay with him. And we were good buddies and went to a lot of parties together and the like. And so we talked about what we were doing in physics, we [Frank and I] were both in physics. And so he said, "Well, here is a student of Kleppner's named Dave Pritchard who is doing his crossed-beam scattering experiments." And I though, "A-ha! Here's a way of doing a Bell's Theorem experiment." And so Frank interceded and said, "Yeah, I talked to Kleppner, and he would be very happy to have you come to MIT and talk about it." Well, I was still also looking for a job at that point, so a lot of my effort was involved in searching for jobs.
It was not a very great time.
Yes. Einstein had similar problems. So, I'm looking for a job, and a lot of the talks I gave were kind of interviews for jobs, none of which I ever got.
But this one, if you gave a talk there on Bell's Theorem...
It was just a group seminar.
That wouldn't be the most propitious talk for a job, would it?
I was sort of young, naive, and oblivious to all of this. I thought it was interesting physics. I had yet to recognize just how much of a stigma there was, and I just chose to ignore it. I was just having fun, and I thought it was interesting physics. I was just trying to understand what was going on. And since I really couldn't understand it, I also had difficulty believing it. It was also the reason I had problems with quantum mechanics. I just didn't really believe it all. I was convinced that there were things that were wrong. My Dad, had always taught me, "Son, look at the data. People will have lots of fancy theories, but always go back to the original data and see if you come to the same conclusions." When ever I do that, I come up with very different conclusions — every time I've done this, "Boy, these guys built a house of cards based on very slim evidence."
I'm going to ask you one more question, and then I suggest that we take a break. There's this guy at Harvard, Costas Papaliolios.
Oh yes, a charming fellow.
Does he come into the story this early?
Oh, yes. Yes. What he did— In fact, he was contemporary at that time. In fact, Frank Winkler knew him very well [and introduced us]. His nickname was Cos. And he was one of the few people who had actually read any of Bohm's work. And Bohm had come up with a second theory, and it was just an abstract mathematical formulation. Very much unlike his original 1950s paper, he had a student named Jeffrey Bub, and this became the Bohm-Bub theory. And the Bohm Bub theory totally got away from this nice physical model that Bohm had proposed, and now there was just some sort of an abstract Hilbert space kind of thing with mathematics. But this was one of the first so called spontaneous localization theories. And Phillip Pearle has been studying these a lot recently. A whole bunch of papers— In fact, there's this paper I sent you which was called Interference of Small Rocks and Live Viruses.* And there is a list of at least half a dozen or more theories of spontaneous localization. In fact, some of it even involved people like Steven Hawking. He had one of them, among others. So in the beginning of that paper is a very sacrilegious description of why we see dots. It's not at all obvious as to why we see dots. Schrodinger's equation describes continuous waves, but we look at the screen and we see dots. What I suspect may be happening in the soap opera behind this was Schrodinger is sitting there at the blackboard saying, (???), absolute value of (???). Yeah. No... no. The real part of (???) ... So this is when one of his friends comes in and says, "What are you doing Erwin?" And he says, "Oh, well I have this theory and I'm calling it quantum mechanics. It describes the motion of particles." The friend says, "Well, you know, the guys down the hall, Davisson and Germer are doing some experiments like this. Why don't I go down and find out if it [your theory] has anything to do with reality." So his friend goes down and comes back. "Well, I've got good news and bad news." "Well, what's the good news?" "Well, the good news is that your equation has something to do with what their experiment shows." "Well, what's the bad news?" "Well, you have this wave, right? They don't see that." "Oh, what do they see?" "They see dots." "Damn! Well, how does this good news work?" "Well, the probability of the dots is proportional to your (???) " "Damn!" "Don't worry Erwin, we can fudge it. Good PR. We can sell this." "Oh, yeah? Who's going to buy that?" "Look, we've got this really slick PR man who can confuse anybody, Niels Bohr. He'll do it. And we've got this slick mathematician, Johnny von Neumann, who can build this neat theory that will confuse anybody. Look, don't worry about it. The fact that it doesn't agree with what they see there in the lab, we'll sell it anyway."
(*John F. Clauser, "DeBroglie-Wave Interference of Small Rocks and Live Viruses" in Robert S. Cohen, Michael Horne and John Stachel, Experimental Metaphysics (Kluwer Academic Publishers, 1977), 1-11)
So Bohm-Bub was one of the first theories that described spontaneous localization. That is, is there a process to describe the on Neumann collapse, which is essentially the same thing, which Wigner and Shimony and d'Espagnat have shown cannot be described by Schrodinger's equation. It's a non-unitary process. And effectively, how do you get from these waves to the dots? And Born just says, "Well, it's just a probability. We interpret it." Some of us would call it fudge, but they get away with it. It was part of the slick PR in calling it an interpretation. So the first such [spontaneous localization] theory was Bohm-Bub, where they introduced so called
collapse equations, and this introduced a new time scale. How long does this collapse take to go from a wave to a dot? It's a non-local process because the wave can be distributed all over space and now has to collapse down to this one little point. The most recent of these is the so called Ghiradi-Rhimini-Weber Theory, which Bell. in his latter years, was championing. Phillip Pearle has such a theory. Diosi and a bunch of Hungarians have looked at this. The idea kept reappearing, and one of the things that I pointed out was... They admitted that the two slit experiment was a good test case for such theories because the question is if you get the two slits farther and farther apart, what is the chances that this thing is going to collapse while it's trying to go through the slits. If it does that, the interference effects go away.
Now, I'm confused. This argument about the two slits comes at a much later date than the...?
Well, what Papaliolios did was the following: Okay. There is an extra free parameter, which is how long the collapse takes. So what did Cos do? Cos read Bohm and Bub's paper, and Bohm didn't know how to figure what the time scale was set by. So he guessed. He said, "We need a time here. Where can we pull a time from? How about thermal fluctuation times?" Which means it's going to be on the order of microseconds or things like that. And what Cos realized was— And Bohm and Bub did this for a simple two state system, and what Cos Papaliolios did was he just sat down and said that you could consider a photon a two state system, the polarization states. "Suppose I put two polarizers very close together. It has to collapse into a definite polarization state as it goes through each polarizer, and if I put them really close together, I should see a breakdown of the Malus Law. That is, that I won't get the cosine square distribution in the intensity as I rotate the polarizers." So he went down to the undergraduate laboratory, found a pair of polarizers, put them smack on top of each other, and tried this. And I talked to him later on and he says, "Well, you can do this better by taking the individual Polaroid grains and looking under a microscope and looking for various orientations of the individual grains, and then you can actually get the time for the light traveling from one grain to the next." Now, they are extremely short times, which would throw out Bohm's imagined or guessed at time. But every one of these theories all has an additional set of collapse equations as a real process within it. In von Neumann's theory is just sort of an on Ansatz. It's going to be there to explain measurements, and it's just some other process.
Did you know Papaliolios?
Well, Frank Winkler actually introduced me, it turns out.
Did he come to the seminar that you gave?
I don't remember if he did or not. Frank introduced me to him. I don't remember when I met him, but he's a charming fellow.
That's interesting. He's practically the only other person around...
To have read the paper, and comprehended what it said and said, "Hey, look. I can test this." So he wrote a great quickie letter or something, where he says, "Hey, I can demolish Bohm-Bub's collapse time."
Had you been in touch with Horne and Shimony or that...?
No. This is long before. At that point, I didn't know of their existence. And in fact, it was at that seminar where I met Kocher and Pritchard said, "You know, Carl, wasn't your experiment to test this?"
Then Ghirardi was at this Bell's Unspeakable Conference, so I went up to Ghirardi and I said— Because Zeilinger had actually been doing these experiments that I proposed in that paper [Interference of small Rocks and Live Viruses], where I suggested that, "Hey, look. If you use this Talbot Lau interferometry," that I was using for doing deBroglie wave interference experiments, you could do this with extremely large molecules. In Ghirardi's theory, it depends upon the mass of the particles, (the collapse time). He actually has equations, and I forget how they work, but if you basically— If some particles are big enough, then quantum mechanical interference effects disappear in his theory. So I dragged him up to the blackboard in Vienna and I said, "All right. Give me a number. You can't play like David Bohm and back away from once the experiment is done." I then got Anton Zeilinger up [to the board], but I said [to Anton], "Okay, you're witnessing this, and you're doing the experiment. So these are the targets in order to shoot down GRW." So Anton has now has done this with bigger and bigger particles, and in that paper, I pointed out that you could do this with things up to, I don't know, 106 or 107 nucleons, which is comparable to a virus. You could actually put a virus through. That idea kind of came up...that was a humor session after an atom interferometry conference. This was in France, I guess. No, it was in Germany. This was at the Jurgen Mlynek [Mlynek's conference] at Constance.
Can you give me some dates?
This is recently. This is ten years ago. When I'm doing atom interferometry here at Berkeley. So at that point, Anton Zeilinger was the after dinner speaker, and his job was to tell jokes and the like. And he gets up and he describes, "Well, we've talked a lot here about atom interference, and it's all kind of related to the 2-slit experiment. Could we do this with really macroscopic particles? Suppose we did this with a living atom. Could the alive atom collapse its own wave function it was interfering. Suppose we did the experiment with graduate students and we threw them out of windows. How big would the windows have to be..." And so he just put in the numbers and Planck's constant and everything. He says, "Well, okay, we can make a projection of the graduate students all the way to the moon, and shoved them through a bunch of windows. You can see interference patterns, but would they [the graduate students] collapse their own wave function?" Well, my talk on the atom interferometry was the next day. And so that night, I put in the numbers and I said, "Well, maybe you couldn't do it..." The reason he [Anton] did this was that somebody in an earlier session, we were talking about Schrodinger's cat, and somebody had a slip of the tongue and said, "A Schrodinger's cat is anything you can put through a slit." That was the slip of the tongue. And so he [Anton] echoed this in his joke, that he was going to put graduate students through slits.
*(John F. Clauser and Shifang Li, "Generalized Talbot-Lau Atom Interferometry," in Atom Interferometry, P. Berman, ed., (San Diego: Academic Press, 1997) 121-151)
So my talk was the following day, and that night I just plugged in the numbers with the Talbot-Lau interferometry I was doing. And I said, "Well, maybe you couldn't put a cat through a slit and see interference, but could you put a cat's DNA through a slit and see interference? And since the DNA has all of the cat's blueprint built into it, this would be equivalent. Yes." So, I pointed out at that conference that you really could do this with a Talbot-Lau experiment with cat DNA. It would be tougher with DNA. You could do it with a virus. Maybe not with the DNA, but close enough. (I didn't know how big a DNA was at that point.) Anyway, so then I kind of put the numbers inside, and said, "Well, maybe I should publish this and put it into Shimony's Festschrift." And so now Antoine has now done Talbot-Lau interference with— He's been doing it now with carbon buckyballs and still seeing interference. And the hard part is to figure out how big you can go. Well, I was suggesting that maybe he could use alkaline metal clusters.
Was it Herschel Pilloff who funded you?
Yes. Yes. And then the government shut down and he ran out of money and then he got cut back. He said, "Hey, look. We're going to have to cut something." And I said, "That's all right. I've got other things to do."
I really want to know about him, so if you have anything to tell me about why he was funding this stuff...?
Hersch is an interesting guy. He in a certain sense is sort of like all of us. You know, where do you get a job? The real question is where can you get the pay to survive. So he was a physicist and then I guess he got a job with ONR being a contract monitor, and it was as good as any. And now, I think he's gotten back into doing some physics experiments.
Did you contact him?
He contacted me out of the blue. My original plan was to use this for commercial enterprise.
This atomic interferometry?
Yes. What I wanted to do, and in fact, I had a whole series of patents on this, in fact, I patented atom interferometry. You can do it. And the primary application I thought was useful was to use it as a gravity sensor, and you could use it for— Originally I was thinking of using it for inertial navigation. And that kind of came out of the Hunt for Red October. In the book, they had these gravity gradient sensors that they could use to sense the presence of sea mountains so they wouldn't run into them. In the book, it actually described how it worked. It never made it to the movie. It described how it worked, and I put in the numbers, and that device that they described couldn't work. But I figured, "A-ha! With atom interferometers, you could do that." And so you could use it for inertial navigation, for gravity sensing, and then all at once I realized, "Oh, another thing you could do with it is use it for looking for oil." That became a primary interest. My thought was you could build a gravity gradiometer, which you could then lower down a bore hole, and use to map out all of the surrounding dirt, rock, whatever. And it would be like sending down Superman with 3-D "gravity-ray vision". And you could find, "Okay. The oil is over there, the gas is over there, the water is over there." And in fact, Shifang Li was my post-doc. I had a short grant from the Office of Energy-Related Inventions, where I proposed that such a device be built. They gave us $100,000 to do a feasibility study on this, which we did. And at one point, I was trying to entice people into funding construction of one of these devices. We worked out all of the details on how you could actually do the tomography inversion program, so that from all of the gravity gradient data if you went down the bore hole, you could essentially invert that[the data] back to get the mass distribution surrounding it [the bore hole]. I worked out all the basic details of how to build it. We had a design on the books.
Then I started talking to some of the oil companies and the well-logging companies. I talked to Cheney's old company, Halliburton. They actually expressed the greatest interest, and they sponsored me into a meeting of a consortium of oil companies. And it was very different from the Gas Research Institute, where apparently Gas Research Institute represents all of the gas companies. Gas Research Institute decided, in their infinite wisdom, how to spend the money, and everyone [all of the gas companies] had to pay it. Whereas the oil group were sort of more like the Swiss federation of fiercely independent cantons. You would present an idea, and they [the oil company representatives] would all come to this meeting. And anybody who wanted to be part of that and sponsor research could buy into it. And then they [the oil companies] could share in the results of the research. But the guys who didn't buy into it couldn't. You had to be sponsored by somebody to get into that meeting, and Halliburton sponsored me. I think I got one other taker. Shell was sort of interested. And what I pointed out is you could really come out like gangbusters with this in secondary oil recovery because in so-called depleted fields, 80 percent of the oil is still there. You just need to know where to look.
So the [remaining] question is, why didn't Chevron buy it, because that's where most of their work is, (in secondary recovery and depleted fields). The bottom line was none of these guys were interested. And so...
Is this before Pilloff comes on the scene?
In parallel with that. As a follow-on, that was kind of one of our motivations for going back to Cal was to do the experiments. To see, (a), if you could do the interferometer— As soon as I proposed it, everybody else who had really well equipped atomic physics laboratories, I proposed in the conference that Zeilinger and Rauch ran in Vienna back 15 years ago or so. They called that conference Matter Wave Interferometry. And it was mostly neutron interferometry at that point. And I stood up and I said, "You guys are using the wrong particles." The leading lights were all there and I stood up and said, "It will never work."
And you were already talking about oil?
Well, at that point, I was thinking in kind of the idea of doing oil and/or inertial navigation. And I brought it out, and with neutrons, they had already built gyroscopes and accelerometers, gravity sensors. And I pointed out, "You haven't seen anything yet. This one goes up with the mass." And you could very easily build the world's most sensitive gyroscopes and gravity sensors and gravity gradiometers with atoms. And there was stunned silence in the room. I didn't get a single question. They had no idea what had hit them. But then very quickly, Jurgen Mlynek in Constance decided, "Hey, what a great idea." And he went out and did it. He built the first one, and in the meantime, I was just trying to find a job, and find some money, and get a laboratory. And then as soon as he did it— So then I was trying to suggest that you could also use these for doing relativity experiments in spacecraft. And so I went to a conference in Annapolis that was sponsored by NASA, and I proposed that yes you could do the measurement of the Lens-Thirring effect, etc.
Well, Francis Everitt, (who was running the Stanford gyroscope experiment), he said it would never work. And so I said, "Well, okay." He said, "Well, I tell you what. Why don't you come down to Stanford and give a seminar and describe the details of how this is all going to work." So I went down to Stanford, and he gave me the grand tour of the GPB [Gravity Probe B (a colossal NASA experiment in a satellite to measure the Lens Thirring effect from general relativity)]experiment they are running there. Before I went, I said, "For my seminar, be sure you invite Steve Chu. Steve Chu is the world's expert in atom cooling, and he's the one who could tell me whether or not I'm making a total idiot of myself, whether any of this is going to work. He would know more than anybody else whether this is feasible." Steve was a student of Gene Commins, and I knew him when he was a punk graduate student working down the hall from us from where Stu Freedman and I were working). So Steve said, "Oh, this sounds interesting. Clauser always has something new and crazy to talk about, but whatever it is, it's always interesting." So he cancelled his class and brought all of his students in. And so I gave the talk, and he and his students went back and said, "You know, yes. Of course it's going to work, but there's a much better way of doing this with lasers. Exactly what we're doing here with the atomic clocks, we already have this atomic fountain experiment. All you have to do is just modify the atomic fountain experiment and use the internal states, and we could build an atom interferometer with the hardware we've already got sitting there operating in the laboratory." So he did. And before I gave the talk, it had never occurred to him that he could build an atom interferometer out of it. But then once I gave him the spark, man, he just took off, and Steve's a very bright guy. In a matter of months, he came back and had one working. And then there was a whole bunch of guys in Germany in Bonn, and PTB Gei
Physikalisch-Technische Bundesanstalt. I can't pronounce it. It's a major German university there. And then, the guys in Tokyo jumped in. Shimizu building atom interferometers. So all over the world, all of a sudden, they jumped in. So very quickly that field got overpopulated, and so I got out. The final thing on the oil and gas was that I thought this was an important application, and couldn't get any of the oil companies to fund it — Even though I thought this was a neat way for finding oil and gas, none of them were interested. So I finally said, "This horse, I haven't heard a heartbeat in years (using an atom interferometer for well-logging). I've been listening closely, and keep whipping it, and there's nothing there." So it's time to get out of this field. But at that point, I realized that we'd developed this whole theory of Talbot-Lau interferometry that I had to develop because it [atom interferometry with slits] really didn't work very well trying to do just a simple 2-slit experiment. So we worked out this whole theory of the Talbot Effect. The first time anybody had ever done this, the complete theory of it, although Cowley and Moodie was close before, but they could only do it for infinite numbers [of slits]. So then all of a sudden, I realized that x-rays are exactly the same wavelength as the potassium atoms that we are trying to do atom interference experiments with. I wondered if you could use this to image refractive index with x-rays. So I went home, plugged in all the numbers, and I had done some work when I was at Livermore Lab with x-rays. (I was working on magnetic fusion, so I had some expertise in x-rays.) I've had to relearn and learn a lot more since then. So nobody was going to buy the oil stuff, the atom interferometers, and the atom interferometry field has just already gotten overpopulated. Plus these guys were not shooting square. There was a lot of backstabbing and theft of ideas.
In the atom interferomet
Yes. It was filthy. And I said, you know, "Let's get out of this. These guys are making me sick. I want to get out of here."
Is that something new in physics?
New to me.
Has physics become more commercial maybe?
Well, this wasn't really— I don't know. This is just a few bad apples, guys who are just inherently dishonest. It just kind of poisoned the whole thing for me. And at this point, my life was repeating. We did the first Bell's Theorem experiments, and who gets quoted? None of our work. Everybody— you know, we did the first experiments. "Oh, well, Aspect did all of the first experiments." Well, that was ten years later. Okay. And then the same thing was happening all over again in atom interferometry. I proposed the original atom interferometers. "Who? No. You couldn't possibly. All of these other guys did it first." "Oh, my God." And in fact, it was very clear simple plagiarism on a lot of parts. With the atom interferometer, that was quite overt. So, you know, "Let's just get out of here, and let's do some of this x-ray stuff, but don't talk about it. Just get it all done and working as a fait accompli, and then lower it on the world as a bombshell. And that's the intention. Just so I don't have the vultures coming in and snatching it out of my hands."
Did you go from Livermore to Berkeley because of the atom interferometer?
Sort of. Originally I was thinking about— I left Livermore because it was— I was stagnating. The experiments were dying. The government had cancelled the magnetic mirror program, but they hadn't told anybody yet. They were just kind of running out the contracts. It was kind of obvious. The only thing that was left there, they were trying to encourage people to leave the project and work on classified stuff. I refused to do anything classified, so I said, "You know, it's just time for me to get out of here." The resources were petering out. You were backsliding relative to inflation at that point, because the new hires were soaking up all of the money. It's a well known effect. And the groups had become so big that my productivity had totally gone to zero. I wasn't doing anything at all useful or productive. So I said. "Let's get the hell out of here."
That was about '86?
Yes. Then I went off to spend a year, actually less than a year, at SAIC, Science Applications International Corporation. And that was-
And was that in California, too?
It was in Emoryville, just right down the street from Berkeley. They had an interesting project that I first started working on, and that was to find explosives in checked baggage using neutron activation. The idea was you take the bags— They were building these things, and they still use them in airports. The trick is you take a suitcase on a conveyor, and you run it through this neutron bath, and anything that will detonate, almost by definition, has a copious amount of nitrogen in it. It has to do with the bomb structure. And it also tends to need to be clustered together to detonate. Well, a thermal neutron will react with a nitrogen atom and give off very promptly a gamma ray of very specific energy. So the trick is if you can detect gamma rays, you can use this to measure the amount of nitrogen present. You can tune the energy discriminators just to look for that gamma ray directly. So then the trick was if you put a lot of detectors around this, can you somehow determine if there is a cluster? And what I realized when I got there was that, okay, in fact, you can use this to build a crude image essentially what's done in what's called SPECT [Single Photon Emission Computed Tomography] now for medical imaging. This is before SPECT was invented. I effectively invented it for this neutron scan. And run up all of the software to figure out, given you see various counts in various detectors as you are scanning this suitcase through this thermal neutron bath. How much nitrogen, and where is it in the suitcase, and is it in a cluster, etc.? And do we want to set off an alarm that this thing has a bomb in it?
And this is really something that they give you when you get there?
Well, it was an ongoing project, and how I fit in was that I figured out how to do the tomography reconstruction.
But why did you want to spend only a year there?
Well, I also went and figured out how to improve it [using an image-gradient dual-energy x-ray imaging scheme, now marketed by Perkin Elmer.] But that was an unclassified project, and then they wanted me to work on some classified work, and I said, "I just don't want to do that." And they were twisting my arm, etc. And basically, I said, "No. That's not my ballpark." Even though my dad had done a lot of classified work through the war and during the Cold War. There was this very highly classified— In fact, I've just seen the papers again. He wrote the original papers that described how to put earth orbiting satellites in space, and what you could do with them, how high they would have to fly, how many stages you'd need on a rocket, why you want them, etc. This is the project that created the Rand Corporation (for the Air Force).
And did this information have anything to do with your being educated in the `60s?
That's a long different story. No, it's not [related to my education in the '60's]. I, in fact, was defending the government in the Vietnam War. I think it has a lot more to do with what has happened since then, and a lot of the way the government has handled classified material.
And around '76, nobody thought that the experiment you did with Freedman was all that important. When did they start to say that this is really important?
Once Aspect did his experiments. I set the groundwork, but people were sort of skeptical, and then Shimony and I published our review paper in '80, I guess.
'78. And a lot of people read that. But most of the references were back to our proposal for the experiment and then Aspect basically — In fact, he came to Cal after he finished his first series of experiments, and he described the whole thing in the very same auditorium where I had given almost an identical lecture ten years earlier. Never mentioned my experiment. Never mentioned that the first experiment had been done right there at Cal.
Except that he knew it, I'm sure.
Oh, indeed he knew it. In fact, in here, there's a personal letter where he thanks me for editing his paper. He sent me the first draft and had me rewrite it because his native language was French, and he liked my writing, and so I rewrote his paper for him. Of course he knew. In fact, there's another letter in here somewhere that I can find where he used the same interference filters, where he wrote, after I left Berkeley, he wrote to Howard Shugart and said, "Can I have Clauser and Freedman's interference filters?" And I think somewhere I've got his letter requesting that. So, of course, you know, in fact whenever I was in Paris, we would go out to dinner. In fact, he came out here and stayed with me. I took him sailing, and his wife, and his kids. So yes, of course he knew about this. In fact, at one point while he was doing this he said, "You know, while I was doing this, I really wish I had been John Clauser because you did all of this stuff first." He told me that directly. You know, and if somebody asked, "Yeah." In fact, he does reference the work in his publication. But, somehow, that kind of got left out.
Now he had the advantage that he was not that long a drive from CERN. So he would commute from Paris to CERN and chat with Bell. And since he was there all the time chatting with Bell, Bell tended to forget the earlier experiments and got more interested in his experiments. And even worse, the guy who was really getting left out of all of this is poor old Ed Fry. He did the first confirmation of our results. I confirmed it, about the same time, with my second experiment on that, and about the same time he published his results, the Fry and Thompson results. And nobody remembers Fry and Thompson. Who the heck are those guys?
There are a lot of guys that don't get remembered.
Ed Fry is a really bright experimentalist. He is one of the brightest and best experimentalists that I know. A really bright guy. And charming. I mean, he and I are also really Good buddies.
I'm always assuming that that paper that you wrote with Shimony, that review, that he wrote the first part and you wrote the second. Is that true...?
No. All of our papers—
For example, the one that you wrote, the four man paper, the one with Shimony, Horne, and Holt, did you all have exactly the same ideas at the same time, or how did that paper get written?
The overall drafting of that I did. But every paragraph, the wording and what it said, we all discussed. And I'd say that there was no— If there was any dominance in the writing style, it was probably mine. In CHSH and CH and in the review paper. And even in the CH, what Mike [Horne] and I did was we used Abner as a sounding board, and he would read the various drafts and the like and he would comment upon them and criticize them. And then we'd rewrite them in response. I'm sure all of those papers must have gone through probably at least ten or twenty drafts to get them to what we finally published. Originally, I wrote the bulletin of the APS, and he and Mike Horne, and Gil Nussbaum, who was the student of Pipkin's, and Holt came down to Washington. All of them came down to Washington and heard my talk.
Then I submitted my thesis to Columbia, and I think there was like two weeks or so between submitting the thesis and the thesis defense, which was kind of a dead time. And so I just went up to Boston— actually it was to Wellesley and stayed in Abner's house with him. And Mike came over pretty much every day, and we just sat there and took two weeks to hash the whole thing out. And then I had the job out here, and I had a boat there. And originally, we were just going to sail the boat all the way to Galveston and put it on a truck there, and truck it across to LA and sail it up the coast to Berkeley. It turns out we ran into Hurricane Camille, so we got kind of stopped at Fort Lauderdale. We didn't save any extra mileage by doing this, but we had a lot of fun sailing down the coast. So every time we put into a port, I would get on the phone and Abner knew my schedule. And so basically he would send off his re-drafts to all of the various marinas in the next city where we put in, some of which I picked up, and some of which are probably still sitting there for all I know. While I was sailing, I would be writing furiously away and editing various things. And we'd get on the phone and chatter about various versions, and we'd keep swapping drafts. This continued all the way until I got to Berkeley, writing the paper, and then we finally submitted it, pretty much right as I arrived in Berkeley. So all of this is done while I was out sailing on the Atlantic Ocean at that point.
When you did the paper with Horne or Shimony, the later papers, is that all by telephone?
That was by telephone and mail. Yes. Pretty much on both of those. In fact, the one with Mike, we, in fact, offered Abner a joint authorship of that paper, and he said he didn't feel his contributions warranted co-authorship. And when he and I wrote the review article, we offered a slot in the writing to Mike, and he was busy and basically decided he couldn't be part of that. And so, basically, we would attempt to just alternate drafts, and then get on then and pretty much talk about what we were going to write. I think the sections where we were discussing EPR and Bohr, and von Neumann, I think Abner did much more of that. And also all of the stuff on Stapp, those were mostly his ideas. The re-derivation of Wigner that we put in there I did. The introduction and the historical outline, that was pretty much mine. I did most of that. He was the one that pointed out— We were kind of looking around for something to call this, in both cases, and realized that the word "hidden-variables" had a stigma associated with it, so Mike and I decided to go for "objective local theories", and then Abner pointed out that really this was the philosophy of realism, and it already had a name, so we might as well use it. So we did. And that's how the terminology varied.
I'd say it was pretty much a joint effort on all of this. A lot of the word picking— This is something I learned from Pat probably more— Besides various ways of manipulating quantum mechanics and the like, the thing that I learned that really stuck with me more than anything else from Pat Thaddeus was how to write scientific prose. He drilled me mercilessly on this. I mean, I had read various good writers, and when I wrote my own first draft of my thesis, it was just plain bloody awful. It was just plain terrible. And he didn't have to do much convincing for me to recognize the fact that he was right. This stank. It was merciless. And he was the guy that kept insisting, "Hey, look. One idea in one sentence." He was the guy who taught me, "If you have problems writing a sentence, you are trying to put too many ideas in one sentence. Split it up in nice, simple, clear, concise sentences. If you have problems writing a paragraph, you probably don't understand what it is you are trying to say." Those pieces of advice I have taken and parlayed... And the real question is that you just have to be honest with yourself and ask yourself the question, "Do you really know what it is that you are trying to say? Do you understand what you are doing?" And the red flag you have to learn to recognize. And this is what he taught me in spades — is that if you are having problems wording something, just go back and be honest with yourself and ask, "Do you really know what you are trying to say? Do you understand what you are saying?" And at times, just on one paragraph, I would go off and ask myself that question and the answer was, "Well, maybe not." I'd spend a whole week trying to figure out what I was trying to say, and in doing so, I learned a lot of new physics, which I didn't realize before.
Another trick I've learned since then came with the advent of word processors when I was trying to put together a paper from totally disorganized thoughts. I would just sort of randomly write out pieces of it. Then the nice thing about the word processor is, I would find sentences that just really didn't belong there. They were on the wrong subject. So I'd stick them off in a stash some place, and then later I'd find another sentence that really didn't belong there. And I'd stick it off and, "Gee, well, I have a stash here on that subject. Let's put those two sentences side by side," and then go through over, and over, and over again, and then you go back and you look at the stashes on a given subject matter, and you could kind of move your sentences around. And as soon as you did that, you would all of sudden realize, "Oh, if I add up this, and this, and this, I come up with a whole new concept that I hadn't even realized before." And gain dramatic new understanding of something just from these disparate ideas that I had just sort of set aside because they didn't really fit where I was trying to write them.
When you were at Berkeley, were you in contact with Henry Stapp? Was he here then?
Was anything interesting going on there?
Well, we commented in the review article about what Stapp was doing. He had this whole notion of counterfactual definiteness that he was pushing. He was a total theorist, and really didn't have the faintest idea what was going on experimentally. There were two classes of theories unfortunately. There was the first class, which is what Bell's original paper was in, and I think I pointed that out in these recent papers. The theorists all assume an ideal apparatus, which you can never generate. Moreover, in Bell's case, he uses the fact that there is at least one pair of orientations where there is a perfect correlation. And once he has that, he can (A), derive determinism from that result, and (B), he then used that to prove the original theorem in his original paper. And very quickly, as an experimentalist, I realized, "Well, that's very nice, but that never happens." So in a certain sense, that means that this is kind of a useless result because it assumes something that cannot be. Well, all of Stapp's papers basically used the same ideal apparatus assumption. And there are a whole bunch of others that do that. There are dozens of subsequent proofs, and you can pick out the theorists immediately by those who say, "Oh well, of course, let's assume an ideal apparatus, blah, blah, blah..." And then they use the assumption of the idealized apparatus elsewhere in their proof. In my point of view, that totally invalidates the derivation.
Was it interesting to talk with him?
Yes. We spent a reasonable amount of time chatting. And some of this, we tended to get together in this group that Elizabeth Rauscher had put together.
Is that that fundamental "Fysiks"* group?
(*See H.M. Collins and T.J. Pinch, Frames of Meaning: The Social Construction of Extraordinary Science (London, Routledge, Kegan and Paul, 1982)
Yes. Those guys were a bunch of nuts, really.
Yes. But we kind of used that as a forum. The real physicists were over here in one corner, and all the kooks are in the other corner. And it was kind of an open discussion forum. Phillip Eberhart. He was part of that. He was an experimentalist. He also was a good friend of John Bell's.
And he was on the physicist side?
Yes. He was a decent physicist. Occasionally, Jeff Chew came down. Jeff was a total theorist.
Was he a plasma physicist?
No, no, hydrogen particle physics. He was the guy who championed the S-Matrix theory. He is Mr. S-Matrix of Chew and Franchi. And he was trying to explain everything in terms of the S-Matrix. You know, given the ins, what do you get for the outs? And the S-Matrix converts the ins to the outs. And if you know that, you know everything, so he would claim.
The other thing that I have a theory in my head which you need to shoot down before I use it, and that is that talking to the quantum opticians got you interested in some of these problems like whether neoclassical radiation theory would...?
Well, I just realized immediately that this was old stuff. That what Jaynes was doing was independently redoing that which Furry and Schrodinger, etc, had done before. It was the same ideas, and that he was trying to build a real space model of what was going on. He didn't use the same words, but the concepts were obviously identically the same, and he just didn't realize the history of what had been done by the founders. Where he was different from the founders was that his work was firmly experimentally grounded. He knew what the state of experimental atomic physics was, which had advanced dramatically since the '30s, in quantum optics and just general atomic physics. And he was unique in that regard. He had one foot firmly on the ground in formulating his theories, where Schrodinger and Furry did not.
So he was trying to do for the electromagnetic field what they had been trying to do to the matter field.
Yes. Yes. Absolutely. Absolutely. And he had a bunch of very bright students, who are now major workers at— as a matter of fact, a lot of them are on the faculty at Rochester. Carl Stroud, Crisp, a very bright guy. And Ed Jaynes, I found him a very charming fellow.
His papers are charming, certainly.
Yes. And a real pioneer, but a very provocative iconoclast in a certain regard. So I think he did a major service for the quantum optics community. And in fact, one of the things that Eberly showed— The bombshell that he [Jaynes] dropped was hey look, he could derive a Lamb shift semi-classically, and boy, did that get peoples' attention. And so what Eberly did, subsequently— And in fact, it's in his book. What's it called? Optical Resonance in Two Level Atoms. Actually, I'm not even sure it's in there, or if it was in one of the papers. But basically, what he realized is he said, "Okay. The way Jaynes is doing this is by bringing in radiation reaction," which was basically an unsolved problem even for classical mechanics. The alternative solution to it is Wheeler-Feynman. But basically, you get these runaway electrons where the electron reacts to its own field, and some very strange things can happen. And so what Jaynes did was, he said, "Well, okay, if we couple the radiation reaction— You have an excited atom," which he assumed were very real fields. He was basically assuming the old Schrodinger interpretation where the function was now in real space and not in the configuration space. But there was an objectively real charge density, and the charge density was oscillating and radiating, and that the radiation then reacted back on the atom itself, and this changed and acted to first accelerate and then damp out the oscillation of the decay of the atom, and this would also produce frequency shifts and the like. And so what Eberly did was he said, "Well, suppose you try and take what Jaynes is doing, and you try to do exactly the same thing, but do it rigorously within quantum mechanics, and quantum electrodynamics." And what he found was that Jaynes didn't get the Lamb shift by accident. That in fact, it was effectively the same Lamb shift calculation that had been done by Bethe and others and Lamb etc., but just viewed in a different quantum mechanical representation of the fields. And the fact that Jaynes got it was not at all accidental, and it really was a valid way of looking at it, even quantum electrodynamically.
Now, what surprised me was here's Clauser going along working on Bell inequalities, and suddenly you're working on problems like the photoelectric effect.
Well, the reason for that was very simple. I was convinced that quantum mechanics had to be wrong. Unlike everyone else around me, I was off— I just couldn't believe that objectivity...I couldn't believe in a breakdown of realism. I couldn't believe that one could have super-luminal communication because you could get reversal of time order events, you could get multiply valued history, all kinds of— You've got these causal loops.... So something had to be wrong. Actually, it turns out that I have now changed my opinion on that. I think that may be part of the answer.
Causal loops and stuff?
Yes. They are probably very real. Bring me back to that and tell you how that works. But in any case, I kept saying, "Well, we did the experiment, what could be wrong? Obviously we got the "wrong" result." I had no choice but to report what we saw, you know, here's the result. But it contradicts what I believed in my gut has to be true. The result, I didn't expect. I hoped we would overthrow quantum mechanics. Everyone else thought, "John, you're totally nuts." But, so we didn't. So then I kept saying, "Well, okay. What could be wrong?" Well, the first question had to do with the CHSH assumption, and so part of that we worked on, and that's what lead to the CH paper, and we brought that down to the no enhancement assumption, which is much weaker and much harder to violate.
But then the other question came out about, "Okay, had anyone every really done the zero'th generation experiment?" We were doing the first generation experiment. And you see coincidences. Now, what can this possibly say? Well, if you believe in realism, that means that there's something moving from the source to this detector, and there's something moving from the source to this detector. You don't know what it is, but there's something there. And the question is how big is the something? Could the something be localized like a particle, or could it be in fact a wave front that's spread out, a very low intensity wave front? But the detector just has a very low threshold, and maybe this collapse whole thing doesn't work, and the rest of this, just sort of stuff goes on and just randomly detecting it. Maybe Maxwell was right. Maybe Maxwell's equations are strictly correct, and it's just an electromagnetic wave propagating out. That was what Jaynes had suggested. That's a simple example of the Schrodinger-Furry hypothesis. And then at that point, it was right after the Fermi Conference that Jauch sent me his paper, which was Are Quanta Real? I don't know that it ever got published.
Yes, it did. I've looked at it. I haven't read it.
And where he was pointing out that, "Well, of course, the experiments all show..." And he was addressing this question about the wave particle duality for photons. And there are a whole series of experiments that Janossy had done, and these were in fact, done because Schrodinger lived just down the street from him. I mean, Vienna and Budapest are just a freeway exit apart. And so Janossy had done this whole series of experiments at the urging of Schrodinger.
This was in the '50s, wasn't it? I think so.
Yes. But it was very clear, when I looked at that, that from the experience that Stu and I had about trying to get coincidences. We knew what the absolute detector efficiencies we had were. It was very obvious— And we had to work all of this stuff out just to make our experiment work. What are the detection efficiencies, what's the accidental rate, what's the true coincidence rate? And these were just calculations we were doing every day to evaluate the apparatus. It was very clear that these guys didn't have the faintest idea of how to see coincidences if they were there. So I thought, "You know, maybe that's the loophole." And then during this period was when this whole controversy within quantum optics had started, and I just started reading some of these papers...
You mean the Jaynes versus...
I can describe that experiment semi-classically again, but I think I described it in one of the papers. And there was some real confusion going on as to what this all meant. And some of the smugness permeated my graduate and undergraduate years, "Yes, we understand quantum mechanics. It's just simple wave mechanics, blah, blah, blah." That was starting to wear off.
I had the feeling that the quantum optics people were never that smug. Do you think I'm wrong?
Some were and some weren't.
I remember Carroll Alley writing in 1961, "We don't know what a photon is."
Right. And the opposite extreme, of course, is going to be Lamb. So, what can I say?
What about Mandel? Did you have anything much...?
Mandel, I'm not sure. Well, there was this— I don't know that he liked me? He was not an easy guy to get along with because he and Aspect pretty well took over the whole business after I left it? Initially, he was very skeptical. And in fact, when I proposed the photon splitting experiment, he and Lamb were there, and they were saying, "Well, how do you generate a single photon? You can't do that? Nobody knows how to do that?" But we'd been doing it all the time with these cascade experiments. I said, "Hey, look, we do this tag photon scheme that we use when we're calibrating our detectors and the like? When I wrote a paper for that quantum optics conference called The Localization of Photons? Afterwards, he said, "Well, that's not what we normally mean by the localization of a photon." I said, "Well, we got it in the box. I don't know what else you could mean by it."
So there wasn't really much interaction between you and him?
Not really. Not really? And I think to some extent— I mean, I've read a lot of his stuff. He and Wolf had done some major contributions to the coherence theory in optics. And another guy who was a player in that game was Don Scarle, who became a good friend. Scarle did some of the first Hanbury Brown Twiss effect experiments that he reported. In fact, one of the things I was trying to do is I had read a paper by Fano, and I convinced myself that there was supposed to be a backwards Hanbury Brown Twiss effect.
What does that mean?
Well, what it means is that I dropped the minus sign in reality. But I predicted that if you had two atoms excited simultaneously— Fano described the Hanbury Brown Twiss effect as you had a pair of atoms that were excited, and you had a pair of detectors, and if these two were close together, you could have a certain probability of coincidences similar to the Hanbury Brown Twiss effect. It was a quantum mechanical derivation of the Hanbury Brown Twiss. And I looked at Fano's calculations and I thought, "Suppose the two detectors are exactly diametrically opposite the two excited atoms. You should also see the same enhanced coincidence effect." That was, "Well, yes except for this missing minus sign in my derivation," and the whole thing was wrong. So based on this missing minus sign, I figured, "Okay. Let's set up this experiment on the photon-splitting, so we can do both experiments." So I tried that and I didn't see it. I went back and checked the derivation, "Oops. Sorry about that one. Well, let's move on to the photon-splitting." So that was an experiment with a result that never got published because it was just plain wrong. And it wasn't there, and there was no reason that it should have been there, as a matter of fact.
Did you go to the next quantum optics conference? See, that was '72?
No. I just went to this last one. I think these are the only two that I've been to. That one and this recent one. Now, what was that thing we wanted to get back to?
We wanted to get back to the fact that you've actually changed your opinion of realism.
I have several thoughts. and none of them are good enough yet to publish. Let me step back and let's talk about causal loops for a second. If you make the assumption that it is possible to propagate a signal faster than the speed of light — to some extent what you need to do if you want to keep realism and keep the experimental results, assuming that all of the other loopholes get plugged — you have to assume that nature somehow is capable of sending signals faster than the speed of light.
Doesn't that mean you are going to keep locality, and why do you need keep locality?
Well, I'm throwing out locality. Keeping realism and objectivity. The cornerstones are locality and realism. So chuck one, take your pick. So I'm still a realist, and what do I have to give up if I chuck locality. Well, I have to somehow propagate signals faster than the speed of light. As soon as I do that, I automatically create the possibility of causal loops. Now, in a causal loop, A sends a signal to B— And it's all in the back of Bohm's textbook on special relativity. He has a very nice appendix in there where he describes all of this. But A sends a signal to B, B to C, C to D, D sends a signal back to A. And all of the observers are moving relative to each other, and I think that at least two of the four transmissions have to be super-luminal. And then just applying standard special relativity, A gets the answer from D before he sends the signal to B. So he's reversed the time order of these events. So he doesn't like the answer he gets from D, so he doesn't send a signal to A. So it doesn't arrive at B, so it doesn't arrive at D. So he didn't get it, so therefore he can't dislike it, so he does send it. So what does this say. Well, it says, the naive question is, "Well, does he or doesn't he send the signal." He can make the decision, "I will send the signal if I don't receive a signal from D," since that occurs in the other order. So, yes, he does, and no, he doesn't. And naively, I want to say, "Well, this is clearly absurd and impossible. It cannot happen, therefore one of our assumptions must be wrong. The only new assumption was that we could propagate super-luminal signals, therefore that must be wrong." That's the standard logic. Now, let's look at this for a second. What do we have. We have yes, he did, and no, he didn't simultaneously true. History is multi-valued! Where else did we encounter a very similar dilemma. The particle could go through the first slit, or the particle could go through the second slit, but the two are mutually exclusive, but both do occur. Well, let's wake up and smell the physics for a second. Where did we get these. We got one from quantum mechanics. That was the fact that history could have gone both ways, and in fact, must have gone both ways. The other we got from special relativity, which we got without knowing a lick about quantum mechanics. These are very different sources of exactly the same dilemma.
Except that you've put in super-lumin—
And also special relativity?
Right. I put super-luminal, but I never introduced quantum mechanics. All of those books could be burned, or never invented yet, and yet I run into the same dilemma. This is a staggering thought to me. How can I have the same dilemma appearing twice with exactly the same results, from very different assumptions. One of these theories [QM], we don't at all understand the assumptions. We have no idea how we got from Schrodinger's waves to Born's dots on the screen. So I then asked myself the very simple question, "Are these the same phenomenon. Is this schizophrenia of a multi-valued history exactly the same phenomenon that one would get by assuming that the internal workings of nature require, ubiquitously super-luminal communication." Indeed, history is not single valued, and that in fact the multi-valued nature of history is exactly what you see in a quantum superposition.
Is this something you've been working on recently?
No. It's just sort of an idea I've had over time. Now, it turns out when you look at Wheeler-Feynman time symmetrical electrodynamics, classical theory, you run into very similar kinds of things because you have these things going backwards in time. And in fact, for example, the simple electron/photon scattering events. You have an electron coming in, moving backwards in time, scattering a photon, etc., and all of these involve reversals of time, and the electron existing in multiple histories all at the same time. And in fact, they occur as coherent super-positions. All of these events suggest to me very strongly that in fact, this is exactly what a quantum superposition is. It's just simply our inability to understand the arrow of time, and the lack of a unique self-consistent history. I mean, if you have things moving backwards in time, this in a certain sense, can be equivalent, I believe, to saying that history has multiple values. I'm not sure that I know how to explain that at the moment.
I was going to ask you, the physicists talk about realism and objectivity and so on, do they talk to each other about this? Do they think about it? I'm not talking about you, now? I'm talking about when you speak to other physicists.
No. Not that I know of. Whenever I talk like this to other physicists, they all think I'm nuts. They quietly leave the room, start up a conversation with someone else, change the subject.
So they don't like to talk about objective reality or?
Not so much. No. A few do, but rarely. Most of them feel very uncomfortable.
And what about all of this stuff that people like Raymond Chiao doing about super-luminal signals. Does that have any relation?
I don't think so. I worked with Ray fairly closely when I was working with Townes, because he was Townes' hot shot at the time.
Still a graduate student?
No. Townes left Columbia and left all of these students there, including Penzias among others, and went off to MIT. He took Chiao with him. Pat was one of his students, had already gotten his degrees, and I think Ali Javan may have gone with him. But most of the experimentalists all had their experiments glued to the floor. And then at that point, Townes saw the light and decided that lasers were passé, and astrophysics was the new thing. So naturally, everybody had to see a similar light, so everybody else in his cubby of associates all of a sudden dropped studying lasers and quantum optics and started studying astrophysics. So then Townes moves to Berkeley and he gets to bring Ray Chiao along with him. Ray Chiao pretty much did whatever Townes— During most of this period, Ray Chiao pretty much did Townes' bidding. Then Ray started having some ideas of his own. Then in the meantime, Townes is sort of in semi-retirement, although he's still very active, but he's now off doing his infrared astronomy actively. He has to be in his '80s now. He has grants, students, and post-does, and it's a major operation. The guy is absolutely amazing. He still climbs to the fifth floor of Birge Hall two steps at a time, and never takes the elevator. He's an amazing guy. But now Ray has gone back to nature and decided, "Well, astrophysics has petered out. I'm going to go back and start doing some quantum optics." I think to some extent, I think I may have inspired him to do some of this.
In the '70s already?
He and I knew each other quite well during this whole period. In fact, when I was working on radio astronomy, I worked with him for a while. We got to know each other pretty well. Mostly what Ray has done so far, there is the whole question of what do you mean by propagation of signals. In fact, there's this book that I have here by Brillouin. He has in there some calculations. Originally due to, not Rayleigh, but one of the grand old men of physics, Lord Kelvin or somebody. Sommerfeld. That was the name. Of propagation of signals, what do you mean. I have this entity sort of moving along in time, but it doesn't necessarily maintain a constant shape. It slowly morphs itself into different shapes. Well, you have a group, and then you have waves propagating through the group, or propagating backwards through the group. And you have a group velocity and a phase velocity. Well, but suppose the envelope changes shape in time. Or suppose I just open a shutter. Now I have a sharp little front going out. It's no longer this nice smooth shaped Gaussian wave packet. What are all the various velocities associated with it. Well, Sommerfeld solved the problem about opening a shutter, and it turns out you have these little forerunners, that if you actually solve the equations in a dispersive medium, you see there are forerunner steps which have a finite sharp cutoff, that propagates with the speed of light. There are a dozen different velocities you can attribute to propagation of signals in material medium, and Ray has been studying all of these, and then bringing in additional ideas where one has things like tunneling effects. How long do these things actually take to propagate signals through or phases through, etc. And he's been cataloguing them and trying to understand them and sort them all out into various orders very successfully, I believe. And also discovering new physics in the process.
But at any rate, not anything directly related to what you were just talking about?
Oh, no? I don't think so.
Do you go to meetings at this point?
Well, I occasionally get invited to come to meetings. Very recently I've been invited to give the same old talk about the history of Bell's Theorem. I haven't really done much on speaking since I was doing the work on the atom interferometry. I got invited to a lot of meetings then, when I was at Cal. But when I switched over to this x-ray stuff, I mean, I've been going to several meetings related to x-rays, but purely as a listener, learning the tricks of the field. I got invited to give one talk on some of the x-ray stuff at a winter quantum electronics meeting. Nobody was interested, nobody came. It was kind of a waste of everybody's time. But I haven't really published anything on this yet.
Sure and that's understandable.
It's commercially potentially very valuable, plus I prefer at the moment not to create competitors. And what I've found in the past is that when I come up with a brand new idea, all of a sudden I am creating my own competitors? So I would prefer not to do that until I've gotten so far ahead that there's no way anybody could catch me and overtake me.
Now, another thing I wanted to ask you was in the '70s, going back to that, this particular archive is about experimental metaphysics essentially, and in the '70s...
You don't want to call it metaphysics, but— Experimental natural philosophy.
That is the name of one of the books?
All right. Whatever. Call it what you wish?
A lot of things were going on in Europe. I mean, there's Bohm, he was at Birkbeck College then, and he had a bunch of people that were working on various experiments. And the Italians, apparently, were doing.
Well, there is Selleri's group, I think it was in Rome... [trying to figure out his name]
Anyway, were these people part of your conversation?
Hardly. They were in Italy? The only time I think I ever met Selleri was at the Erice Conference. I may have met him once since then. I don't remember.
Was that of any interest?
I wasn't very impressed with it, frankly. A lot of his stuff was just recasting the Schrodinger-Furry hypothesis in various other forms? I didn't find anything that were brand new ideas there, no. To be perfectly frank, and perhaps a little smug, through the '80s, most of the ideas were mine. And the primary outside idea, there was really only one, and that was by John Bell, and that was his original Bell's Theorem paper? But pretty much all of the work through the `80s, the truth of the matter was, I did it.
And the stuff that Bohm's group was doing [subsequently], was that of any interest?
No. In fact, he deteriorated terribly. The pure elegance of what Bohm had done was what he developed in his 1950s, theory, and de Broglie simultaneously did this in real space, which gives you a space/time model. And it had particles and waves and how they interacted, blah, blah, blah. And you could see how all this worked. And clearly John Bell had read and had been enamored similarly as I was by all of this. And in fact, this was very similar to the kind of things my dad had talked about, whereas the particles were effectively viewed as strong non-linear terms in the partial differential equations, that we are creating closed soliton-like objects that would move around. He had particles and waves, and if you knew what the correct equations were, you could probably calculate these and generate them. There is a whole new possible physics here. Then the Bell's Theorem stuff, that pretty much, to some extent, shot down most of at least Bohm's and de Broglie's ideas for the guiding wave. Although there are some cute ways of doing this. That's another one to come back to in a minute, which is related to this question of breakdown of space/time. But in any case, to me, those ideas were what was stimulating Bell. Bell came out with his original result and then Shimony and Horne and I kind of worked this all out and to do the theories, Freedman and I did the original experiments. And the only other new idea then was the whole question of do photons have reality, and that these are experimentally testable. And those were the primary new ideas from that era, and nothing much happened besides those.
Did the spontaneous down conversion?
That was Mandel's contribution.
And was that important, or just another technique for creating particles?
In my opinion, it was just another technique for creating particles. It made it a lot easier to demonstrate the effects, and Yanhua Shih at the University of Maryland has done some really exciting stuff with this. He's a very bright guy, I'm very impressed with what he's done. And then once you have this, then all of these various quantum cryptography things could come about. It was curious. I heard Artur. The guy who invented cryptography. Artur Ekert. And he, at the Vienna meeting, he realized that the CHSH inequality was the basis for doing cryptography. In fact, I realized that back in the '70s. I had no idea that was important, so I never published anything about it, but it was pretty clear.
Do you have notebooks with any of this stuff in them?
I don't think any of that I wrote down. It may have some of the long calculations. Try to do ideas on doing experiments. I had some ideas on doing experiments for these two pi spin rotations. I had some ideas on how to— Stuff that had long calculations. Spent a long pile of calculations just doing the QED of generating and calculating the two photon correlation, just for the simple experiments that we've done. Those were not trivial calculations.
And you've kept them around somewhere?
Some of them, yes. Although others have done those. Ed Fry actually published a very nice paper where he went through. He did that much better than I did. And I think the original calculation for the CHSH paper, I think Dick Holt did all the details, QED on that. And that is in his Ph.D. thesis. That was a very nice contribution to that paper. It's not a simple calculation to do, and you can fill gobs of books on that. I've got crates of notebooks. All of the stuff from Livermore, peach baskets full of notes, all on atom interferometry and various things. A lot of them are not worthy of publishing, but..... Questions about Caltech's physics curriculum. When I started at Caltech, the physics curriculum had been set up by Lauritsen and Leighton.
And then so Feynman took over [at Caltech], and Bob Christie gave the first course in quantum mechanics, which I didn't take, but I audited my junior year, or senior year, I guess. And that was out of Dicke and Wittke's book, a very well done course. So I really didn't have any formal and decent training [in QM] until I got to Columbia, of course from Feinberg and from Messiah's textbook. But all of that stuff [the Lauritsen-Leighton curriculum] is no longer taught. The so-called modern physics curricula went by the wayside. There is just too much to be swept into an undergraduate curriculum. And so I profited very heavily by being the last of the generation to be taught that. In fact, the very last class that went through Caltech that actually took the course that was set up by Leighton our senior year. And that's the stuff that I'm using right now more than anything else. Now, to some extent, there is a fairly similar course of stuff that's taught by Segre at Cal, but I don't know if that's being even taught anymore. It's kind of along similar lines, which included everything from nuclear physics, relativity, spectroscopy, and all of the basic survey of physics. You know, there's so many fields of physics now that there's no way that you can keep up with them all.
[Tape 3, Side B]
Matthias Reinsch who was a student at Cal recently, absolutely brilliant mathematician. In fact, I really couldn't have written this paper, the Talbot-Lau effect, without him. I'm just a terrible mathematician, and especially in things like this analytic number theory, stuff that only Gauss understands. And I would struggle with these problems and come in and he'd say, "Oh, a piece of cake." He'd come back in an hour or a day and go, "Here's the answer." And then I'd go off and do whatever else I was doing. Similarly I put him in the laboratory and said, "Okay, how would you like to do such and such." And he would spend weeks, and weeks, and weeks on the simplest hardware problems. He was all thumbs, there are just inclinations that people have. That was something that I could pop off in ten minutes. And he would be sitting there analyzing all of the various ways of doing it. "I don't care! There are lots of ways to do it! Pick one, and do it! Let's get on with this."
Here's the thought. This was my dad's idea, and I believe it could be related. I have no idea whether it is. I've never been able to work it out. It involves mathematics more than what I am capable of. In fluid mechanics, as described by partial differential equations, and linear partial differential equations, there are three kinds. There are elliptic, parabolic, and hyperbolic equations. Elliptic equations are like LaPlace's equation for boundary value problems, electric fields inside a cavity, etc. In an elliptic equation, there is no speed of light. Every point on the boundary, in a boundary value problem affects the solution everywhere. So if you have x and y, if I move something here, it affects x and y everywhere. The wave equation, on the other hand, is hyperbolic. And things propagate along characteristics that are— like electricity and magnetism, this all propagates at the speed of light. And you can not have influences in x and t, at least for one set of solutions that the forward moving solution is going forward in time. You cannot have influences outside of the light cones.
Now, it turns out in fluids— This is actually very interesting, and there are two things you can think about. Let me go to the first one. The first one is my dad's idea, and then I'll bring you back to the second one in a minute. In fluid mechanics, you have very similar sets of equations, and the transition between them is between subsonic flow and supersonic flow. In subsonic flow, if an object is some place in the room, it can change the pressure everywhere. In supersonic flow that's not the case. Things have to propagate with the speed of sound. And you have a transition from subsonic to supersonic, and you can have in the same flow regime, you can have transition from subsonic to supersonic, at which place the basic equations have to change character from elliptic to hyperbolic. Classic examples of how this is done is in an expansion wave through a Delaval nozzle, and/or through a shock wave. Now, shock waves are interesting, and they are sort of one-sided solitons if you like.
I don't know what a soliton is.
A soliton is a wave that propagates without change of shape, which is due to strong non-linearities in the wave equation. This is essentially a dispersion free wave. I.e., it could be a particle for example, a highly localized region of strong amplitude. And these characteristically occur in non-linear equations. And in fact, the equations of fluid mechanics are non-linear, and that is what brings up this transition from subsonic to supersonic flow. Now, suppose, for example, that a particle, let's consider, in a realistic framework, and let's say we could really magnify a particle and put it on a table and look at it. Suppose it has a soliton-like character, and suppose it has a boundary. Moreover, suppose it is due to strong non-linear teems in some underlying four dimensional partial differential equation, relativistically invariant. But suppose further that the surface of that particle is very much like a shock wave, a spherical shock, such that the solutions to the equations inside the particle are elliptic, and outside the particle, are hyperbolic. So everywhere inside the particle, there is no speed of light. Now imagine this particle moving in time. Okay. Now in four dimensions, this particle is now a tube. The tube is now elliptic throughout its length. There is no speed of light inside that particle. That particle knows everything there is to know about the future and the past, inside that tube. Hence, rule number one, things in the future via the elliptic solution in that tube can influence the past. Hence, if any of this has any validity, we now have a solution to the problem of trying to understand \super-luminal communication. It's very easy to occur, because the speed of light is imaginary in Laplace's equation, is an elliptic equation. There is no speed of light. There are no characteristics. Whereas for the wave equation, which comes out from Maxwell's equations, there you do have characteristics, and these are all the low amplitude solutions, and only when the field amplitudes, whatever they are, get strong enough, you actually get the non-linear terms creating the soliton like characters, and it's conceivable that this is what a "particle" is.
Is that something that you've been...'?
That's my father's influence. Absolutely.
Is that something that's been with you?
Yes. But I'm not that good a mathematician to really work it all out, to be perfectly honest. Here's another thought, the one I wanted to come back to. We define coordinate systems for measuring things. Ideally, Nature doesn't need our coordinate systems, they are just useful tools for us for measuring things and the like. So how the observer characterizes— it is really quite irrelevant. The equations of Nature must be independent from our choices for coordinates. One of the coordinates we choose is time, and we assign, in a certain sense arbitrarily, an arrow to time. Now, there are two coordinate references, if you will, that one could use. One could be a predictor, which is certainly the sense what quantum mechanics is. I know the initial conditions, what can I say about the future. An alternative coordinate reference viewpoint could be that there is "one" solution. Say, for example, that we were to assume that Nature is deterministic. Nature just sort of propagates. And there are trajectories, and whatever, particles, and things like that. And I could walk in as a historian after the fact, then I don't need an arrow of time. If I'm a predictor, then I instill an arrow of time. But if I'm a historian, I have a very different viewpoint. I'm just looking at all of these wiggles and trajectories as things propagate through time. There is a solution.
Now, if I were trying to solve my undergraduate boundary value problems, I could view this in two different ways. One is suppose I were solving a hyperbolic equation. I would start with a set of initial conditions, and a set of derivatives. And these are called Neumann boundary conditions. Alternatively, if I were to solve a boundary value problem, like fields in a cavity, I would specify the values at all of the boundaries, and I would use Dirichlet boundary conditions. Now, suppose I am a historian. A historian could look at this set of wiggles of how particles are propagated and say, "Oh, that's just a boundary value problem. I can put a slice here at t=0 and here at t=1." And I know the conditions along both boundaries because I'm a historian. And I can give you a solution to these equations. And this is kind of related to, effectively, what Wheeler-Feynman's time-symmetric electrodynamics does, is just looking at it from that point of view. You have waves coming from the future and the past, and you are converting a prediction problem to a boundary value problem looking at the past.
Now, let's step back to our original set of ground rules. Our original set of ground rules was that Nature dare not allow us to make a difference from our imposition of a viewpoint or a coordinate system. That means on the face of it, is that in fact, that the historian or the predictor must always give exactly the same results. So that very well may mean that I should always be able to solve all of natures problems as a boundary value problem, i.e., as stuff coming in from the future.
But I think it was interesting that you were talking to him [Yakir Aharonov] during this period.
Yes. He was a junior faculty at Yeshiva, and the programmer that I was working with at the Institute for Space Studies was also a graduate student at Yeshiva in physics. And he mentioned that Aharonov was there and was giving a course in von Neumann Measurement Theory. And so I asked him if he would introduce me. The class must have had eight or ten people in it, max. I guess there aren't that many physics students at Yeshiva, being that it's really a rabbinical school, a seminary. And so I introduced myself to him, and had told him what I was doing and asked his opinion of all of this, and he thought it was really quite interesting and would be well worth doing. Other than that, I don't know if there is anything spectacular about it. He had a very different way of dealing with quantum mechanics than I was used to working almost solely in the Heisenberg picture, and did a lot with just looking at the expectation values of time-dependent operators.
Did any of that feed into what you were thinking?
Not really. It just sort of improved my overall understanding of what— I confess even to this day that I still don't understand quantum mechanics, and I'm not even sure I really know how to use it all that well. And a lot of this has to do with the fact that I still don't understand it. There are these various things within quantum mechanics where book after book, there are these chapters where they talk about probability currents and probability densities as if they were real things. Yet in fact, they only work for one particle systems. You can not generalize those to - particle systems because the number of dimensions becomes bigger very quickly, bigger than what you see in real space. And very quickly at that point, you realize that for more than one particle, you're in a configuration space. And this whole issue of mixing back and forth between one particle and - particle systems, and when you get into one particle systems, people lively talk as if this were physical x, y, z space, a real space, and not a one particle configuration space. And these things are very, very different in concept. So with the roller coaster of yesterday. The roller coaster track is the real space, and the position of the car is a hypothetical position of the car at any given time. And these refer to very different things. So anyway.
Do you use quantum mechanics when you do your own calculations, in which part of your physics?
As little as possible.
Like, if you are doing atomic interferometry or any of the...?
Yes. Yes. Very quickly there, that's all done effectively in the one particle case where you are effectively assuming that you're dealing with real space. All of the books are written that way, and that's the way people do it, so I do it that way. I haven't the faintest idea why I'm doing it that way. And in fact, you listen to a lot of people and they say, "Well, you don't have to understand it to use it. Just follow the rules." So those are the rules, and if I follow the rules and I publish the stuff within the rules, people nod and say, "Yeah. Okay. He seems to have done that right," whatever it was that I did.
Because certainly looking at your papers, it seems as if you know quantum mechanics as well as...?
Well, it's amazing what you can hide. It shows how far you can get in physics without really knowing anything. Actually it turns out it's quite the other way around. Every time I have looked through my old notebooks, and seen problems that I have solved in the past and whatever I sort of shock myself into realizing that I've forgotten far more physics than most people ever learned in the first place. But anyway, he [Aharonov] and I became really pretty good friends in the process, and he and I are still good friends. Every time we run into each other at a conference, we usually have dinner together. He was running around— There was a meeting in Spain. He had two passports. One was Israeli and one was U.S. He had lost or he was convinced that someone had stolen one of these two. I suppose all Israelis are paranoid, the way the international situation is working. But anyway, I don't know what else to say. Clearly, one of the brightest theorists around these days in the general foundations of the quantum theory. Okay, what else do you want to know. Some of the others mentioned. Narton I didn't know. That was a friend of Shimony's. Max Jammer I think I mentioned yesterday that he came to give that one talk at Columbia. In fact, Len Kasday and I kind of together went to hear his course. And it was kind of interesting. He had the various students present us "guest lectures". He would pick a topic and they would prepare a talk.
He must have talked about the 1920s and 1930s among other things.
I'm trying to remember what he had in the course. I guess some of the stuff in his book.
But you had already read all of those old papers of Schrodinger.
Well, there was lots of stuff to read. I learned a lot about the history of it from listening to him and reading his book.
I always think he was kind of a medieval encyclopedist. He's one of those people who knows everything?
Yes. He was very knowledgeable. In fact, that book, where he actually describes the history, he goes into great detail about the actual physics of what was going on. My own opinion is that it was like— Well, it's coming out in those last two papers on the history of this, was that the state of the experimental foundations for all of this were really in a terrible state. How they could come up with the conclusion that quantum mechanics meant what they thought it meant, or claimed it meant. Which turns out to be, probably, in retrospect, a correct interpretation. They really couldn't measure anything. That's why they all used these Gedanken experiments. And most of it was based, and in fact, if you read his book, a lot of it has to do with x-ray physics experiments by Compton and the like.
So really until the late '60s, the experimental situation was very bad.
Oh, it was terrible. It was terrible. And in fact, that was one of the things I pointed out. You would think that right after World War II, there was just this whole sea change of abilities with experimental physicists that came out of World War II. A massive amount of money went into the war effort, both the atomic bomb at Los Alamos, and the radar project at MIT. Pat Thaddeus was telling me about the guys that went to work on the radar project. I mean, microwave technology just didn't exist before the war. And after the war, they all came back and they knew everything there was to know about microwaves. And boom! Atomic physics just took off because here are all these microwave transitions that previously were inaccessible in atomic beam spectroscopy.
But could you have done some of the experimental work you did right after the War?
Well, there were a lot of things. Phototube technology has advanced quite a bit? You might have been able to do it, but it would have been a lot harder given the available photo multiplier tubes? But we were using whatever the best thing available off the shelf was, relatively only recently developed. That came about through pressure from the high energy physics community for particle physics experiments [that] needed good, high quality photo multipliers. There were two main companies. It was RCA and Amperex. I guess Amperex was a division of Philips, and they were the two primary vendors? And there were two schools of thought. One was the RCA school, one was the Philips school, and there was kind of a religious following of each, and each one felt the other was junk? Of course, if you went to another laboratory, they would say the same thing, only reverse the players.
Was that stuff that you used in your thesis work at all, or was that something you only became acquainted with as you started to struggle with the [Bell theorem] experiment?
Mostly stuff that— I had some experience with photo multipliers, but not really high quality ones until then?
So it was really here at...?
Berkeley. Yes. In (experimental) physics you just sort of learn as you go? It's really a lot of trial and error? One of the things you find is especially if you are doing something new that has never been done before, atom interferometry is a good example, it's never been done before. You'll sit around and think, "Okay. There's maybe half a dozen ways to try. Which way do you want to try?" Well, if you are more familiar with some techniques in some area, you will kind of have a preference for one of these methods. The reality is that probably there are at least half a dozen ways, and three or four actually will work, but you don't know. Or maybe a dozen possibilities and half a dozen will work. Somebody else at another laboratory will face the same situation, who is also pioneering in the field, and he will say, "Okay. Let's try this method because I happen to be more familiar with those tricks." And you get half a dozen laboratories and each one of them will try a different method. As in the case in atom interferometry, they all worked. Some of them had different applications in different areas, or were better for doing some things than others. And then typically what will happen is you will get these independent schools. And all of this will be done in a given laboratory mostly by trial and error? There's an enormous amount of trial and error that is involved in producing new techniques. Then what will happen is you will have these independent schools of thought, and each school is to you a black art. And typically, historically when this sort of thing happens, the way you learn the other school is you hire the other guy's graduate students as post-docs, and he comes and teaches you what it was that they were doing in that lab? And so most of the reality is the techniques of all of this are moved around simply by word of mouth and watching your peers.
But you seem to have no hesitation, when you're coming to a new field, of just going out and learning what you need to know. Like all of this optical interferometry?
Yeah. And the x-ray stuff. I built some x-ray spectrometers when I was at Livermore Lab. There we were measuring the x-ray spectrum coming out of the plasma. A lot of that I just picked up from Eric Silver who was an expert in all of this. He was one of the staff scientists there, and he knew lots about it. And so I worked with him and pumped his brains for all of the information on x-ray technology? But then still I didn't know very much about how you can generate it with real x-ray tubes and the like, how much power do you get, what would the spectra look like in reality, in any sort of detail, how do you calculate all of this from scratch? And so I just sort of sat down with books and my computer and just started calculating all of this stuff. The wave theory for that, this is all the Helmholtz equation. So the solutions are E" x, really pretty straight forward, spherical waves, and plain waves, etc? Once you know the wavelength, all the rest of the physics kind of disappears. I just do this personally with the Kirchhoff diffraction integral. Just pull that straight out of standard physical optics.
Is that pretty typical though, or do most people just learn certain techniques and spend the rest of their career with them?
It depends on the people. That's very much a people kind of thing. A lot of people spend their whole lives in one field.
Have you ever met Yuri Geller?
He actually came to the physics department one time. He was a total charlatan. And also part of that group were Puthoff and Taag, who, I think they believed their own bullshit. In my opinion, I believe these were all sophisticated magic tricks, as in illusions. I'm convinced that their extra sensory perception was all a crock. I met that Lilley guy who— I was very unimpressed by Lilley. Lilley was the guy, I believe, that was the pattern for the character played by George C. Scott in Day of the Dolphin.
I don't know that film.
It's worth seeing. Lilley and the character in Day of the Dolphin, he was a scientist who was trying to communicate with dolphins, and learn their speech and try to decipher pieces of it. When I talked to Lilley and I asked him just simple questions, he reportedly was the guy who had done this, but I asked him simple questions about their speech, and he didn't have the faintest idea what the answers were. For example, do dolphins have dialects, etc He had no idea. "Well, what do you know about the other phonyms and the rudiments of the language?" Nothing. George C. Scott, in his character, knew a lot more than Lilly did. Then there was another movie that was also based on Lilley, which had to do with the sensory deprivation experiments that NASA did, where they would take these tanks of water— They were worried about astronauts being in total isolation in space, and would they go nuts? And so Lilley apparently ran a series of experiments for NASA where they put these guys floating in tanks in padded suits, so there was virtually no sensory input. Apparently the brain can't handle that. It creates its own [sensations] if it doesn't have something coming in from the outside. There was a movie, and I forget the name of the movie. It was also based on those experiments. I thought it was kind of a lousy movie. Day of the Dolphin, I thought, was great.
You were, and are, as I understand it, a realist?
Very much. I wish I could justify it? Still wondering what am I doing wrong. How could all of my experiments have shown what they appeared to show, and how can I justify my position. Total gut feeling, I guess. I suppose it's a religion?
I don't know. Did I have any other questions?
Who reacted to Clauser/Horne paper and what were the reactions. I don't remember. Well, other than that big stack of requests for reprints. Other than that, I don't really remember anything. How did I come to publish in Nuovo Cimento. Oh, that was because the circular polarization experiments, the only theory that discussed them related to semi-classical theories, or the only people who had cared up until then were Franco Selleri's group? And so I figured why not publish in the Italian Physical Review? I don't know? Probably should have published in Physical Review.
Did that suddenly involve you in a whole dialog with those Italians?
No, not really. I really hadn't met them until I went to Erice.
Which was about the same time?
Right. It was. I had just finished that experiment when the Erice Conference was held [in 1976]. And I think pretty much shortly after that was when I left Berkeley. I think that was kind of the end of my career there, or plus or minus a few months, and then took off.
Well, that's all I can think of to ask you at this point.
Okay. You didn't ask very much about this whole stuff that I wrote up about the infamous stigma, or have you talked to others about that. Dick Holt felt that very acutely. It was rather interesting? When I was proposing the experiments, this was sacrilege to even think about this. But Physical Review Letters accepted it. And then after we published the results, there were a lot of people who were very interested who were kind of hovering and waiting for our results —the ones with Stu Freedman. These were the first experimental results? So a lot of people were kind of hanging around waiting for that? And then once we published them, the typical reaction was, "Oh well, yeah, we told you so. Of course quantum mechanics is all right? How silly of you to have thought otherwise. We're still not sure why you did the experiment."
Except that you and Shimony pointed out that these people were really in a mess because they had to give up locality?
Yes. But nobody believed their experiments anyway. Bohm and deBroglie's theories, nobody took them seriously at that point anyway. That was all quack physics. The common wisdom, of course, was that Einstein was senile and anything that deBroglie and Bohm wrote were just quack papers that nobody would take seriously.
I guess what we're trying to say is that once the '70s were over and all of these experiments had been done, people really got terribly involved, I think, with things like entanglement and?
Okay. I think in a certain sense, we laid the groundwork, and then when Aspect went through, that came later, and went around and did the same thing, now all of a sudden, once people had heard it before, it had kind of sort of settled in and they forgot where they heard it. But now all of a sudden, the second time around, it kind of hit them that, "Oh, this really is something new and different, and in fact, maybe we didn't understand what it was." And then it's like anything. There's kind of an exponential growth in understanding, the sort of spreading through the thinking. A few people would say, "No. No. You can't look at it that way, that's not true." "Well, why not?" "Well, because of Bell's Theorem." "Oh, well, what's that?" "Well, like in this week's Physical Review Letters. See, there's this paper by Aspect." "Oh, gee, that really is interesting." And certainly the previous stuff clear back to von Neumann and Bohm and deBroglie and all of that, that wall all long forgotten?
And then once people began to realize that this was useful, then a few— Well, the thing where it really took off was a lot of it had to do with Ekert and Bennett, those two guys recognizing that this was useful for cryptography. Once that was done, I was totally unaware of how much money and interest there was in cryptography. Heck, most of my computers didn't even require passwords. The only reason I have them on now is because we have all of the ones in the house all networked, and you can't put it on a network without putting passwords on them. Other than that, we just leave them all off. All of the computers that I was in charge of at Livermore, there was no password to get it. I thought it would be silly that anyone would want to try to hack into or steal. So they want to steal it. They can just come and ask me and I will give them the results anyway. There's nothing to hide here. But then I started to realize that this is a huge business, and an enormous amount of money, especially from the National Security Agency and the CIA. They are pumping billions of dollars, and once they started pumping money because they realized and got wind that this could be useful for cryptography, and they are in the spook business, and so the National Security Agency had started pumping massive amounts of money into research grants. And anybody and his brother that came up with any idea they wanted could get money on it.
So this set people back looking at Bill's theorem?
Oh, yes. And also on the single photon stuff. You can't detect a single photon twice. Once you destroy it, it's gone. And so a lot of it had to do with the money givers changing their opinion, and someone, particularly Ekert and Bennett, discovering that this stuff actually had some practical applications. Because once it was discovered there was some practical applications, man, the stuff just took off and skyrocketed.
It sounds as if it's a very interesting story here about the sociology of scientific interest or whatever.
Well, it's like the old song that I'm sure you've heard lots of times, but at least my own experience with it has been that it is so true? I don't know who came up with it, but the standard scenario is at first, it's preposterous. The second generation is, "Oh, well it's obvious, but totally unimportant." "Well, it may be useful." And the third generation is, "That's actually interesting?" And then the fourth generation is, "That's so important and so fundamental that I discovered it first?" I'm sure you've heard this scenario described?
Okay, the post-doc period, did the administration of the LBL take any stand pro or con on what you were doing. The AEC funders. The way that worked was, (because it has definitely changed), in those days, in the early '70s, there was a much closer relationship between LBL and the U of C Physics Department. It was highly blurred, and in fact, at one point, I guess I was teaching a course as well as working as a post-doc, and so they decided, "Okay, well, this money for the teaching needs to come from UC, but the work you're doing with the post-doc needs to come from LBL, and so I ended up getting two paychecks during that period." Now, the atomic beam group was an LBL group, which actually had some people resident on campus, some people resident on the Hill. And originally it had been much, much bigger. That was developed by— What's his name. [Bill Nauenberg] I think he still is the Director of Scripps at the Oceanographic Institute. So it was an LBL group. And the way the money worked was actually very nice. Howard Shugart was now in charge of the group. It had gotten much smaller, but apparently there was just kind of a blanket funding of all the groups that came through, I guess, on some major AEC contract. Actually, during the whole period it changed. Because it started out with AEC and then it became ERDA, and then it became DOE on the period of about two or three years.
It became ERDA in sort of the middle '70s as I remember?
Right. And then DOE took over, and if Reagan would have had his way, it would have become the Department of Commerce, but that didn't happen. And so these groups...
Was Commins under Schugart?
No. They were both full professors. At Berkeley, Berkeley is a collection of small empires, the Physics Department. There are so-called "groups". The groups had a relatively formal structure, and frequently you would get people collaborating on experiments within a group, they tended to get funding as a group. LBL just sort of funded the whole group, at least the atomic beam group? But, you see, on campus, Physics had at least a dozen so called groups. Townes had a group, it was the astrophysics group. And each one had its own set of floor space, and usually, groups consisted of one or two senior professors, and a few junior faculty, and a handful of post-docs and graduate students, all of which just kind of worked as a community, family... a family with sort of common interests. And once the group had a threshold size, it would have its own so called group seminars they could meet once a week where people could report on what they were doing and the like. Commins was sort of part of the atomic beam group, when he needed stuff from the group and particularly wanted to share some of the beam group's hardware, and then sort of wasn't a part of the group when they wanted to share his hardware. He kind of effectively was part of the group, but sort of kept almost an arm's length relationship. I don't know, but I think his funding came also through the atomic beam group? But some of the guys in the atomic beam group worked on experiments up on the Hill, some of them worked on experiments in the Physics Department. There was in fact one theorist in the group who was actually a nuclear physicist, which had to do with back when it was mostly measuring nuclear moments, which is Auriel Bonney just sort of put on there.
So I never had to worry about money. On the other hand, I never spent very much. I kind of used whatever shops there were on the Hill, and somehow the stuff got paid for. I just sort of got used to pulling stuff out of storage closets and using junk sitting around in communal storage rooms. And very rarely have I ever bought anything? I mostly just built whatever I needed.
I just wanted to correct this impression that I had. So I somehow assumed that Commins would be part of some quantum optics group, but they weren't. They were doing atomic physics.
He was doing pure atomic physics? Right.
And whatever quantum optics was?
He actually really wasn't doing atomic physics? In fact, I had a conversation with Steve Chu a few years ago. Because nobody really does atomic physics anymore. Very few people actually do atomic physics, as in study atoms? Everything is known about, virtually all of the atoms. There isn't anything left to learn. Everything that you could measure has been measured in great gory detail. Where atomic physics is particularly useful is that the techniques used in it, all of these bags of tricks, that you can use with it, and especially when you can combine this with laser and microwave spectroscopy tricks, and optical pumping tricks and the like, this has all now effectively become applied atomic physics. You use these tricks to measure other things, like build atom interferometers, like do Bell's Theorem experiments, etc. And Commins was doing the same thing. He was using atomic physics technology for studying weak interactions and how these bring about violations of fundamental symmetries, parity violations and the like.
So should I even be using words like quantum optics...?
Quantum optics in general has to do more with the actual dealing with photons. Well, historically, it had to do with the quantization of the radiation field into photons, whatever they were. It had to do with the effects of this quantization. It had to do with how this breaks down, and the transition to classical optics and the like. Again, it's one of the fields, quantum optics, quantum electronics, I've never really quite understood the definition of quantum electronics. It seems to involve a little bit more in the way of solid state. They all involve laser technology nowadays. They didn't then, in the old days. We didn't have lasers in the old days. These are all gray areas blurred together.
And it's not a very good question anyway?
Yeah. And I don't have a very good answer for it either because it's not clear that all of these things are truly well defined. Okay. [Reads from list of questions] "You used Kocher's equipment, but did you need new polarizers or detectors?" Ah, we ended up using one piece of Kocher's equipment, which was a stainless steel insert at a collimator. We originally looked at Kocher's hardware, and that was my original intention. But once we started the design on the new experiment, each time we'd try to think, "Okay. Well, how much of this stuff can we use?" And as the design progressed, more and more pieces became inappropriate. So they got pitched out, and we kind of joked about it. At the end, we looked back and we went, "Well, what are we really using out of all of Kocher's hardware?" And it was almost nothing. I'm not even sure whether we used the same vacuum pumps. We built the polarizers from scratch. These are these pile of plate polarizers with sided plates. The phototubes, we had much better phototubes. He had 8575 for the blue tube, and a noisy red sensitive tube. 7265 was the tube he used for the red. They were both RCA tubes. And neither was particularly good. The dark current in the 7265 was really terrible. Its quantum efficiency was pretty low. And RCA had just come out with a new line of tubes. They had gallium phosphide on the first dynode, I think was the trick. And it made us an order of magnitude improvement in the secondary emission of the first dynode. And they were far superior tubes. There was the 8850, and the C31000E, which eventually got renamed as the 8852. It was a red sensitive, and the 8850 was the bi-alkali blue sensitive tube. And red means anything redder than blue light, i.e., green is red in phototube-speak? Anything that isn't violet is red as well, or isn't at least deep blue. And those are about the only things we ever bought, were the new phototubes. At one point you refer to these as cheap experiments. It depends on what your lab equipment costs in this period? It didn't cost us anything. Like I said, these were mostly closet-fulls of junk.
I think it was at one of your talks, I think at the '72 Conference, where you said, "We've really got to work on this because it doesn't cost a bunch of money?
In all of this, the only cost is effectively the people cost. It's not like a big facility, like an accelerator facility, or the like. The hardware costs are really quite cheap. And the way I did it was I would always tend to resurrect old junk rather than going out and buying new stuff. There are two kinds of people, really. Those who kind of like to use old junk and/or build it themselves from scratch. And those who go out and buy shiny new boxes. Carl Weiman (now a Nobel Laureate) is very good at using old junk and/or using consumer grade equipment. For example, you buy a scientific grade videotape recorder and you buy an off the shelf VCR, you will find in fact you have a factor of ten in the difference of price. You go out and you buy a commercial Loran set, (I don't know if you can still buy them anymore), or you go out and buy a Loran C receiver for a time standard. If you tear into the Loran C receiver that you use for navigating sailboats around, you get out this precision 100 kilohertz signal. And again, it's an order of magnitude or more in price difference.
The same thing is true in laser diodes. This is where Carl Weiman made a killing. They have these laser diodes in every Sony Discman. Getting any literature on them early on was a bear. Sony wasn't going to tell you how these things worked. You might be a competitor. And a lot of the things you needed to know, "What's the line width look like, how many modes does it have and the like," they don't know, nor do they care. And so people would just sort of pull these things out and try them and see what happened, and it's amazing how you can build a diode laser system for atomic physics. Now, it turns out just by pure chance, that the rubidium lines happened to fall within the range that Sony and Sharp happened to make diodes for, and that was purely fortuitous, and Carl Weiman went a long way with that. So anyway, I've gotten pretty good at dumpster diving.
[Tape 4, Side B]
I think there was an article about a year ago in Physics Today about people who do this, even on grand scales where they recycle whole accelerators and the like. That's some guys who grabbed Panofsky's old accelerator when SLAC was first built, and there were a couple of guys, this is at the Naval Post-Graduate school, and they just simply asked Panofsky, "Well, what will you do with it." "I don't know. Junk it." And they said, "Gee, can we have it." And Panofsky, (with that linear accelerator), had built a whole empire on measuring cross sections, and more than anybody else. And so they just went out and rented a semi and piled all this stuff in and carted it down from Stanford to Monterey and set it up inside a stadium, under the bleachers in a stadium, and bang! They started doing meson physics, with virtually zero budget. They just put the whole thing back together again.
If you are innovative and clever, it's amazing what you can rework things into that the original builder never intended, but as long as you know what you need and you can somehow retrofit a lot of this junk to still work. In my case, you know, a power supply is a power supply. As long as it puts out a certain number of volts and a certain amount of current, it doesn't matter if it's a big old clunker, or something brand new off of the shelf. Vacuum pumps are the same thing, as long as it sucks a vacuum.
Didn't you tell me that Aspect turned out to use your polarizers?
No. The filters. The interference filters. We had to buy those. Those were custom made at an optical coatings lab. And he used exactly the same transitions in calcium that we did for his experiment. In fact, those are the same transitions that Carl Kocher used originally. And it turns out that the first one was not a Rydberg transition, which we discovered halfway through. It was sort of amazing that it worked, but it did. You did a Bates-Damgard calculation to estimate what the branching ratios were, and it should not have worked, but it was a highly mixed upper level— But in any case, it had worked on his experiment, we figured that it would work on ours too, and it did. And so since it was a known useful cascade that people could see coincidences using it, it got recycled through three different experimental groups.
The story I've heard is that d'Espagnat is the one who suggested to Aspect that he do that experiment, that Aspect was a bright graduate student and they thought (???).
That's new to me. I hadn't heard that. Well, they were both in Paris. For Aspect, this was a so-called second thesis. We do that here, we just don't get degrees for them.
In Germany, they do that, too, don't they?
I don't know? I don't know. Germany, historically, things were somewhat different. You tended to be a dozent, and there was really only one professor, and you kind of did his bidding. At American universities, it's all over the map, and it depends a lot on the people. Even at Berkeley, I was one of the few post-docs there who actually had his own experimental program, whereas other guys tended to be working with full professors, jointly as part of their research programs. Whereas in my own case, I was pretty much alone. And there were some that were really not innovative at all. They just did whatever the senior professor said, "Here, do this." So they were sort of like still overgrown graduate students. It's all over the map, and it depended upon the people involved, and how much initiative each guy had, as to how people do this. To some extent, there was also a question of opportunity. There were a lot of people who I suspect probably had ideas and things that they wanted to do, but the guys who were running the group wouldn't hear of it and would insist that they were the only allowed ego in the group. Fortunately, in my own case, Townes and Shugart were willing to allow independent ideas, but there are a lot of other guys I know that wouldn't tolerate that. "It's not part of my program. What's he doing using my money?" Because usually it had to do with the money. If a guy went out and he wrote the proposal and got the money, well, he would spend it. Now, the LBL groups that I was describing earlier, were rather loosely funded, leaving a fair amount of latitude. They just had blanket contracts that supported the groups, and the groups really didn't have to justify it that much in those days? I suspect that today that especially with the era of accountability that the Congress has been pushing on a lot of these laboratories, that that is tougher and tougher to do.
Now, am I correct that in your atomic interferometry phase that you were entirely supported by ONR?
Yes. The way that worked was that initially for the first year of that, I supported myself out of my pocket. I just went back to Berkeley and I said, "Hey, I want to build an atom interferometer. And is it okay if I use the laboratories and the junk piles?" And Shugart said, "Sure. Go ahead. As long as it doesn't cost me anything," because he didn't have any money. And I said, "Well, I'm trying to get some money, and if I do, then I can support this and take on some graduate students." And he said, "Fine." And so the department chair— It's interesting how you can play the politics. Harry Bingham was also a member of the Berkeley Yacht Club. We soaked up a lot of beer together. He liked drinking beer, and I think he tended to drink too much of it. He ended up dying of it eventually. But one summer, he was Interim Department Chair. And so I talked to him and we got along quite well and he basically respected my opinion on a lot of things, and I said, "Hey, can I do this?" And Harry said, "Sure. why not, John?" So he was a temporary appointment. So then I went through and I said, "Well, previously your Interim Department Chair thought this is a good idea." Once you start saying that these other guys think it's a good idea, they don't want to cross their peers, and so they of course, by definition, also have to think it's a good idea, even though they haven't the faintest idea what it's all about. And so as long as you can kind of make people believe this is the common wisdom and of course, they would be the stand out as a "hold-out" if they said no, you can get amazing things through. And so I just played the politics reasonably well? So then once I got some money in, they said, "Okay. Well, we'll give you an appointment."
And so when you were reporting to Pilloff I assume you didn't really have to say a lot.
Well, really I was supposed to start writing quarterly reports, but then he seemed to only care about annual reports. And he had various forms that he wanted filled out? And I went to all of the meetings, and he went to all of the meetings, and he saw me standing up and talking, and he kind of thought what I was doing was fun? He was pretty much the only guy I really needed to please at that point.
Because when I went to see him, I talked to him briefly, and he was very dubious about what I was intending to do? But he was all excited about atom interferometry.
Oh, yes? Well, I was the one who invented it, and then he actually called me up and said, "Would you be interested in taking some of my money?" And I said, "Sure." Now, my problem was that by the time I finally got the lab set up, all of these other labs had pretty much passed me. But all of these indeed were very cheap experiments. The big cost is covering payroll. That's the nice thing about working here [in his laboratory at home]? Not having any overhead, all of the extra money that UC would just simply take, and in return would hassle me, I'd be having to pay to be hassled with bureaucratic overhead. Now, Bobbi does the accounting, and we just write checks and put things on credit cards, and the hardware gets delivered by UPS and the like. And all of the extra money that would go into the overhead, I just buy additional hardware with it. And so it's amazing how much you can accumulate with that extra third of the money. The next question, "for the '70s experiments, were there new skills that you had to learn?" Lots. "What skills did you bring from astrophysics or other areas?"
We sort of talked about that a little.
Not that much. A whole range of computer programming skills among other things. When I was a student, I actually did sort of a mix of stuff. A lot of it never showed up in my thesis. We did some microwave spectroscopy, and I worked in a lab, but then as part of the group effort. Then as new students came in, I kind of willed my lab over to the other students. In particular, Dick Nerf took over the microwave spectroscopy project.
Is this still with Thaddeus you were doing this?
Yes. In fact, it's interesting. I went back to Pat's 65th birthday celebration at Harvard. He's now at Harvard Smithsonian. The curiosity is, he is still doing the same sort of things and hasn't really changed fields that much. He's doing some marvelous radio astronomy right from the roof of the Harvard Smithsonian. You walk in there and the whole lobby is surrounded by a mural which is his map of the Milky Way Galaxy with CO microwave lines. It turns out carbon monoxide is the preferred choice for mapping out the galaxy. Originally, everyone was trying to use a hydrogen 21 centimeter line, but there is just way too much of it. You can't get any decent resolution or detail. There is just about the right amount of CO to make really nice detailed maps of the density of this stuff. But he also has microwave spectrometers that look pretty much the same as the ones we built there, at Goddard Institute for Space Studies. So yes. We played with the millimeter microwave plumbing, and we had some vacuum systems, and some electronics, and all kinds of stuff. So we had some fun toys there, but none of that really got into my thesis. And I willed most of that over to Dick, who carried on. We were trying to measure the dipole moment of the cyanide and radical. And we tried to measure the lambda doubling in CH radicals, which we never could actually produce. I don't know whether Dick has ever successfully produced them either. You could never actually see the transitions, at least when I was playing with it. And I lost track of how his stuff actually worked out.
So a lot of it had to do also with being able to program computers. Actually, that was stuff I'd brought back from summer jobs at Caltech. That's where I learned it. I learned scientific programming, or actually taught myself, stuff that I still do lots of, just an enormous amount. There's an old line. In the '50s and '60s, if you wanted to be an experimental physicist, you had to be a glass blower. If you wanted to do experimental physics in the '70s, you had to be an electronics expert. If you wanted to do it in the '80s and '90s, you had to be a computer programmer. What you need now, who knows.
But you really, as far as what you are saying, you had to be a computer programmer at least in the '60s?
Yes. And in fact, for this experiment, the second one, you also had to be a glass blower for that one, too. That was an all glass vacuum system, so I had to teach myself glass blowing.
I'm highly impressed.
Actually, there was a professional glass blowing shop there, when I was still a post-doc. It was gone by the time I returned. I think they just sort of waited until these guys retired, and have never rehired. I think I needed some glass blowing when I was back there, and I found there was still a glass blower in the chemistry shop, and he would do it for me? Physics really didn't need much glass blowing anymore. And electronics that all came from my high school years.
Did you interact with Wigner or Wheeler?
I did not interact with Wigner in the Bell Inequalities work, except for that paper I wrote on von Neumann's informal hidden variable argument. That was my only interaction with Wigner, and not at all with Wheeler.
Is there anyone interesting that I've....
Left out of here?
Is Carroll Alley a big player in any of this?
I don't really remember him very much. I know the name, but I don't really remember his interaction with this. Okay, "you brought Kocher/Commins type correlation experiments to bear on the semi-classical quantum electro dynamics debate. It struck me as ingenious. Did others follow up?" Yes. The famous Grangier, Roger and Aspect experiment, in which they claim to have done it first. For which they get the credit for the first sub-Poissonian statistics that were measured. But in fact, when I first met Grangier when I visited a New York meeting of the National Academy of Sciences, where I was one of the session chairs? And they just described that they had done this experiment and they were excited about it. I said, "Oh, what is it?" He said, "Oh, it was your experiment and we just redid it." I said, "Oh, okay." To my knowledge, they didn't reference my work in that paper, but they mentioned it to my face that that was exactly what they had done.
How did you go to the '72 Coherence Rochester Conference? Invitation, your interest, what?
Joe Eberly invited me. And I don't know how he heard about what I was doing.
It wasn't published yet at that point, if I remember? That was June of '72.
What do you mean published?
Your experiment with Freedman...
No. It wasn't published yet. And how he heard of it, I don't remember? And in fact, it was Joe Eberly who invited me to this last one, the Rochester Conference [of 2001].
So he just called you and said (???).
Out of the blue. And said, "I hear you are doing this. How would you like to come and give a talk?" And the same thing like Bill Phillips. He sort of kept calling me out of the blue at one point for a Gordon Conference. He said, "Hey, you seem to have done a lot of the pioneering work in this field, and I'd like you to come give a talk on it." On all of these, I have no idea. It's just the grapevine.
It's kind of interesting because you said that the people at this  conference weren't even thinking along these lines?
And how Joe Eberly knew about it, I don't really know. Let's see, had I published (???). There was another paper which I did publish, and I don't know whether it had been published or not, which was basically saying the same thing.* That was a Physical Review article. Who knows? Maybe he had been the referee or something. I don't know? And I don't remember if that had yet been published or not. The work with Freedman, we had not gotten any results from. But I published that one before, at least submitted it before. *Experimental Limitations to the Validity of Semiclassical Radiation Theories," Physical Review A 6 #1, July 1972, 49-54
Did you meet anybody interesting at that conference?
Yes? All of those guys. Mandel, Lamb, Mandel's students, Don Scarle, who else was there? Oh, and all of the guys that I had only heard about, from Mel Lax, John Klauder, Roy Glauber. I think they were all there? Glauber I've run into a lot frequently at various conferences on atom interferometry where he came. And he and I are reasonably good friends? And also Pierre Meystre and I have become good friends at the more recent conferences.
Was he strictly a theorist?
Pretty much. He's a charming guy. He and I are really good friends.
Anybody or anything that influenced what you were working on at that point?
Well, I think the only other thing, I think I mentioned it, was that one paper by Jauch on Are Quanta Real. That was a fascinating paper. At that point, I was actually wondering on the question: what are the experiments that actually show that photons exist as a particle-like character, and that was really his paper, was what got me thinking about that. That and Bell's original paper were probably the two most influential. "Bell's inequalities were studied by a host of philosophers, in addition to Shimony. Cushing, Teller, Stein, Fine, Jarrett." Well, Jarrett effectively redid the same thing that we did in this interaction with John Bell, that was published in these Epistemological Letters? I have a set of them if you are interested in them?
You gave me one.
Well, I have the whole set here. The original set. This is the Association Ferdinand Gonseth. That's what they looked like. Those may be impossible to get elsewhere.
I will tell the archivist at the Bancroft Library about your papers. I think what they do is they send you a polite letter at some point, say when you're 70, saying, "We understand you have some interesting stuff. Don't throw it away. Don't put it in a landfill," or whatever.
Well, if they find it interesting, they are welcome to it. Anyway, the stuff that Jarrett did, that was in that interchange with John Bell.* We didn't use the same words that he did, although Shimony has redone that. At the same time, he has a couple of books that he recently wrote, which discussed it in greater detail. But in essence. those concepts were fleshed out pretty much in that interchange with John Bell. And Jarrett effectively kind of rediscovered that. Was he a philosopher or a physicist? I didn't know. Did he call himself a philosopher? *"An Exchange on Local Beables", by J.S. Bell, A. Shimony, M.A. Horne, and J.F?, Clauser, Epistemological Letters No. 9 (1976), No. 13 (1976), No 15 (1977), No? 18 (1978)
I think he's a philosopher, but I may be wrong. One of the things that I never really understood in that whole period is did the philosophers and physicists talk to each other, were they each just doing their own things in two parallel streams or— Because there seems to be quite an industry of philosophical articles.
Well, there is an industry of popular literature. Capra is one of the guys. And they make money. People buy that stuff. I think Shimony had some kind of disparaging words about them and the things that have been done with Bell's theory as a result, and essentially the mysticism that has been drawn in and justified by Bell's theorem. I'm not sure I agree with much of it. There was also at the same time, during this early era, that group had also meetings with the group at the Esalen Institute. There was this place down on, and I don't know if it's still there, at Big Sur, on this marvelous beautiful old ranch, I think the original homestead? And the guy who owned it, Mike Murphy, he was kind of the darling of the local sophisticate set. Basically the Esalen Institute was kind of the heart of the California sensitivity movement. The Est Group was the one that came out of this. Est. And who was the guy that was in charge of that?
I can't remember? [Note from transcriber: Warner Erhart created Est]
Anyway, that was all kind of centered around Esalen Institute? Murphy?
Did you ever go down there?
Oh, yeah. They would routinely drag me down there once a year.
What was it like?
Well, one of the draws was that it was right on this rocky cliff overlooking the Pacific, a gorgeous estate. And they had a whole bunch of resident worker bees there, that ran an all organic garden and the like, and they would offer courses for mostly wealthy Los Angelinos and Bay Area folks who wanted to have their consciousness expanded.
So did you give a course?
Well, the guy who was running this [Mike Murphy] decided that quantum mechanics was related to this consciousness expansion, and would bring us down there. It was free for us, and there were hot sulfur baths that were there, and the rocks, and you'd go into these hot baths, all communal, with everybody naked, which I guess was part of the grand excitement. And then the hot water would sort of overflow the tubs and go cascading down the cliff into the Pacific Ocean. And so part of the highlight of every evening was a trip to the baths. And then during the day, we would sit around and talk about new aspects of quantum mechanics and the like, and how it was related to the great cosmic cockroach, or whatever. None of which I thought very much of, but what the heck?
[A letter to Clauser, February] 1969, Dan Kleppner. Basically, this is the bottom line for where he's going to pay my expenses, saying that the check is in the mail. But at the beginning— the way it was left at the end of that meeting after I had given the seminar, Pritchard said, "Okay, wouldn't have your experiment have done this, Carl?" And Carl said, "Oh, of course. That's why we did it. It's already done, basically." And in fact, in here [in Clauser's correspondence files] I think I ran across it some place. It was a letter from Carl Kocher showing that "he has done it." And his letter is full of bullshit. He hasn't done it. But he in fact insists that he has at that point. According to David Wick's book, David Wick says, "Oh, well, these guys were just talking past each other. They really were meaning different things." No? Carl Kocher, that letter in here proves quite the opposite. Carl Kocher was convinced that somehow his experiment was already disproving or confirming Bell's Theorem predictions.
Now, what is this atomic scattering?
Now, what I had— In fact, I mentioned it in that stuff that I wrote. You probably don't remember it. My friend, Frank Winkler, said, "Hey look. Pritchard is doing these crossed beam scattering experiments." In fact, I have his Ph.D. thesis, it's great, still here. It's on the shelf somewhere. Here it is. Pritchard was just finishing up as a graduate student, and here's his thesis. And in this, what's the title of it. It's faded. I can't read the title. In any case, basically what he did is he had two atomic beams of alkaline metals, and then he also had some Stern-Gerlach magnets on them, either ingoing or out-coming or whatever, beams. And so he had all of the basic stuff there. And the thought was, "Well, okay. If you could do scattering at 90 degrees in the center of mass, then you get only the even partial waves, all of which are effectively in spin singlet states." And so the trick then would be if you could select that out, you might be able to do a two atom example of Bohm's Gedanken experiment, where you are actually using real atoms. And that's kind of what Fry— I guess I never finished on that before. That's what Fry [recently] was setting out to do at Texas A&M. And Thomas Walther was working on that. It looks like they've kind of backburnered that experiment, by the way. So in any case, I was wondering whether or not it would be possible to do that. So I went up to MIT, and I don't remember how I got myself invited or whatever. I basically gave a seminar at Dan Kleppner's group.
So that was the first time you heard about the Kocher/Commins?
Right. That was the first time I had ever heard of that. And Carl actually didn't have any reprints available, but he wrote down the reference to it, and as soon as I got back to my office in New York, I went to the library and picked it up and read it. As soon as I read it, I realized, "No. This is a crock. He hasn't done it at all." But Kocher was convinced that he had already done it. I said, "Well, if you've done this then this is a very important result." And so that's apparently what Kleppner is echoing in here because he believed that what Kocher had said, and that my contribution would now be nothing more than just interpretation of the Kocher/Commins experiment, as having done all of this. Unfortunately, it wasn't so simple. I actually had to do the real experiment. But indeed, what was clear as soon as I saw what Kocher and Commins had done, it was clear that using two photons was going to be dramatically simpler than trying to use two atoms. We're talking about an order of magnitude simpler experiment. That in fact, the two-atom experiment still hasn't been done. They finally put it on the backburner. It was a shame, because Thomas Walther was a very bright young experimentalist, and he was going to be kind of— He was joining working on this with Ed Fry. Actually Shifang Li was—So the paper finally came out jointly by Ed Fry, Thomas Walther, and Shifang Li proposing the experiment, and Ed has given that talk any number of times, describing all of molecular physics.
I'm not surprised I was unable to finish it all because the stuff he had to go through to get that done, it was really complicated. It took him a long time-he had much more time on it than I ever did, trying to figure out how to actually do that experiment. And his idea for that was kind of stimulated by conversations he and I had had over the years about, "Well, how do we actually go about doing this." And so we'd bounce a lot of ideas back and forth as to how we might actually do this. At one point, I was suggesting molecular dimers with cesium2 molecules and the like. And then he bounced back and suggested that we might be able to do it with hydrogen molecules, and then finally for a while, he was on silver. Then the final result, now, I think is to do it with mercury. I think to do that experiment, he's going to require something like six lasers all running simultaneously. It's just a nightmare of experimental problems. And he still has problems. The big problem is you put a lot of laser power inside a vacuum chamber and you've got electrons. And he was trying to do coincidences with ions and electrons, and you just get way too many electrons, which I predicted, and apparently, that's what they apparently found when they tried some of this.
This is an interesting letter too? It's before you get to Berkeley, it's a letter to Townes? I think it's before you get to Berkeley?
Oh, my God. I misspelled it? I'm embarrassed. I wrote a letter to Townes...
What was misspelled?
I should have said related, and I spelled it r-e-a-l-t-e-d? [reading letter]
Yes. This was a conversation I had had with him on the phone. I was worried that Commins was going to go out and try and scoop me on it at one point.
But clearly Townes already knew that you were interested in it long before you get out there.
Yes. And like I said, he basically said, "Well, when you get here, we'll talk about it." So I didn't go into my job with him under any significant false pretenses. Maybe a little bit, I didn't tell him that I really didn't give a damn about radio astronomy.
Did you give a damn about radio astronomy at that point?
To tell you the truth, not really.
I see. Despite the thesis, despite the background radiation thesis?
Well, once I got involved in the quantum mechanics stuff, everything else paled in comparison.
I can understand that.