David Hitlin

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.

ORAL HISTORIES

Credit: Bob Paz

Interviewed by

David Zierler

Interview dates

May 6 & 25, 2021

Location

Video conference

See catalog record for this interview.

Usage Information and Disclaimer

Disclaimer text

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.

Preferred citation

In footnotes or endnotes please cite AIP interviews like this:

Interview of David Hitlin by David Zierler on May 6 & 25, 2021,
Niels Bohr Library & Archives, American Institute of Physics,
College Park, MD USA,
www.aip.org/history-programs/niels-bohr-library/oral-histories/47448

For multiple citations, "AIP" is the preferred abbreviation for the location.

Abstract

That’s right. I'll tell you a story. I worked at SLAC as a postdoc in the group of Mel Schwartz, who won the Nobel Prize in 1988 for discovering that there were two neutrinos. He was also my undergraduate advisor at Columbia, and then I was a postdoc in his group at SLAC. He was a very interesting person; he was always thinking about the next thing. The experimental thing that was going on now no longer interested him, even though everybody else in the group was engaged in it; he was on to the next thing. Mel liked to bounce ideas off me, and we actually had, in the corner of my blackboard, a list of things that we would be working on if we were only a little smarter. For example, we really didn't know how to find gravitational waves in 1970, but we knew it was important, so we had it on the list, along with other important questions. This was a conversation we had all the time. It was extremely interesting, and I learned a lot from it.

One day Mel came into my office, and he said, "What if you thought of an experiment that was for something really important, but it would take a hundred years to get the answer? Would you do the experiment?" That was amusing, so we discussed it for a while. My point is, we're actually now there. It doesn't have to be a hundred years; it just has to be a career's worth, or if you're 40, it has to be 25- or 30-years’ worth. And that's what major experiments take now. So, we are in a situation where you can say, "I'm going to start this experiment, and I may or may not retire before there's an answer." But I think the ethos of the field is, well, if it's interesting, you do it. You start it, and you hope you live long enough to get the answer, or somebody else will get the answer. But that was quite a prescient concept that Mel had back in those days.

Zierler:

Dave, let's take it all the way back to the beginning. Let's start first with your parents. Tell me a little bit about them and where they're from.

Hitlin:

My father, whose name was Maxwell, was born in Brooklyn, and lived in New York his entire life. He had a master's degree in business administration from NYU, but got it during the Depression, in 1932, and never actually worked in that field. His hobby was amateur radio (W2GEN: he was proud of the early call sign), that he pursued his whole life. He managed to turn that into a career. He became an electronic engineer and worked for various electronic companies in the New York area for many years, and always had a little workshop in the basement with things that he would be building. He built our first three television sets. We had a five-inch television, which was interesting because it had electrostatic deflection; The screen was based on an oscilloscope, so it had electrostatic scanning of the raster, as opposed to magnetic scanning, which was the way most TV cathode-ray tubes work. He built that, and then he built a 7-inch version, and then he built an 11-inch one.

Interesting. She had a rather fierce reputation, and over the years, I came to know some of her older students. She was rather demanding of them. They always kept in contact with her, but several of them felt that they had had a harrowing experience. I did not. I had just a wonderful experience with her. She was demanding; that was fine as far as I was concerned. She was a real model of how you do experimental physics. She was in her 50s when I was her student, so she was not in the lab day-to-day tinkering with oscilloscopes and that sort of business, but she was very incisive about what you should do, what you should check, what could be wrong, and very interested in the calculations that had to go into these things.

Muonic X-rays were actually discovered at Columbia, at the Nevis cyclotron, by Val Fitch and James Rainwater in 1952. They were predicted in 1947 by John Wheeler. Both Fitch and Rainwater went on to win Nobel Prizes for other things; Rainwater for a theory paper, which is actually worth noting, and Fitch for the discovery of CP violation in kaon decay. Neither Nobel award had anything to do with muonic X-rays. The case of Rainwater was extremely interesting because the Nobel Prize in 1975 went to Bohr—that's Aage Bohr, Niels Bohr's son—Ben Mottelson, and Jim Rainwater. I would guess that Rainwater was the most surprised person on the planet when he got the Nobel Prize. But the reason is clear if you look in the literature. Bohr and Mottelson made a career out of working on the collective model of deformed nuclei, which was basically an adaptation of Maria Goeppert Mayer's shell model to a deformed potential. A lot of nuclei have permanent quadruple deformations, which determines many of their properties. They wrote two books about this and worked on it for their entire career. That's what the Nobel Prize was for.

However, in the early '50s, Aage Bohr was a postdoc at Columbia, and he shared an office with Rainwater. Rainwater wrote a short paper, about a page long. This was before there was Phys Rev Letters; there was Physical Review Comments, which were short letters and comments in each issue of Phys. Rev. Rainwater’s paper said if you took the shell model and put it in a deformed potential, then you could learn a lot of interesting things. He showed a simple example calculation, and that was the end of it. He never did anything else in this area, because he was actually an experimentalist. But he had shared an office with Bohr, so Bohr knew about this. Bohr then took this idea and made a career out of it. The Nobel Prize committee did its homework and decided to give Rainwater a piece of the Nobel Prize, which was, I think, fair. He had the idea, and ideas are what they really give Nobel Prizes for.

Zierler:

For your own research, how did you know you had enough? When were you ready to defend?

Hitlin:

My thesis? That was actually unusual. Let's go on and then I'll get there. The muonic X-rays found by Fitch and Rainwater are gamma rays in the range of, typically, 1 to 6 or 8 MeV. The detector of choice for doing this kind of experiment was thallium-doped sodium iodide, a scintillating crystal; you view the scintillation light with a photomultiplier tube. It was the best way to do these experiments, but its resolution was not very good. If you were looking at something that's a narrow line, because of the poor resolution you saw a gaussian distributed over a large range of energy due to the poor resolution.

What was interesting about these germanium lithium-drifted detectors was that even though they were much smaller, they had vastly better resolution, by a factor of 100. The idea then was, and several people decided to do this at roughly the same time, that with germanium detectors you could see the muonic X-ray fine structure, or even the hyperfine structure due to quadruple deformations, and that's what I did my thesis on. Everyone with a cyclotron did this: in Switzerland, in Chicago, in Virginia, but it was first done at Columbia by Sam Devons, who had recently come to Columbia from Manchester. In fact, I'm in the middle of writing an article for the Royal Society. When RS Fellows die, they have an article written about their career. I was asked to write such an article about Sam, because he worked on my thesis experiment. I knew him quite well. Together with some people from Brookhaven he used a small lithium-drifted detector and did an experiment at Nevis on two isotopes of molybdenum and was able to resolve the fine structure and measure the isotope shift. Wu then got interested in doing this and they formed a collaboration. Devons' group, which was just one postdoc and a student, Wu's group, which was larger, and Rainwater. Wu thought it would be politically useful to have Rainwater, the discoverer of muonic X-rays, on the experiment. He didn't do much, but he was interesting to talk to. He lived right near Nevis. He had been a major contributor to the building of the cyclotron and knew everything about its operation. In the middle of the night, 2 o'clock in the morning sometimes, you would see the beam get better. The reason was that Rainwater couldn't sleep, so he would come in, shove the cyclotron operator out of the way and tune up the beam; he could do it better than they could. That was actually his main contribution to the experiment.

In any case, we had this powerful contingent of Devons, Wu, and Rainwater on the experiment. We embarked on a quite ambitious survey of many isotopes: basically, everything Wu could get her hands on. The way that worked was quite interesting, too. These separated isotopes are produced at Oak Ridge, where there's a repository of separated isotopes. They make them or made them—I don't know if they still do—using what were called calutrons. These were small mass spectrometers, basically miniature cyclotrons, that were originally used to separate uranium isotopes, along with gaseous diffusion, which proved to be a better method in the end. But they kept these calutrons in operation and then used them to separate other isotopes. They built up a store of isotopes which ranged from fractions of a gram to hundreds of grams of various isotopes. You could get neodymium-142, 143, 144, 145, 146—I think there is no 147, 148 and 150, just as an example. And then, tungsten-182, 184, 186; all kinds of separated isotopes. The kicker was that they're horrendously expensive. Most people who needed to use them used milligram amounts, but we needed tens of grams, or hundreds of grams, of each of the isotopes. You couldn't afford to buy these quantities; it would have been literally many millions of dollars. Wu made a deal with Gregory Ragosa, the head of the nuclear physics program office in the DOE, which was the AEC at the time; they would lend us all of the material we needed. They came in standard chemical reagent bottles. We would make a funnel out of filter paper and pour the samples into lucite boxes that we made in the appropriate size and use them as targets for the muons. We would then put them back into the reagent bottles. They would weigh the bottles before and after and they would charge us only for any loss; a small amount might stick to the filter paper, for example. That was the deal, so it cost us only a small amount of money to use this unimaginable wealth of separated isotopes.

There was an interesting incident along the way. We would take data at the cyclotron for several weeks every several months, making improvements and obtaining new isotopes in between. In the interim we would take the germanium detectors and isotopes back to Columbia. The Nevis cyclotron was in Ardsley, up on the Hudson River, about 25 miles north of the campus. We would drive the detectors and the isotopes back and forth. We kept the isotopes in a small cardboard box. At the start of one run, I had them in the back of my car, and was driving to Nevis when I was stopped by the state police; my car registration had expired. The reason for that was that New York was making a switch between all licenses renewing on January 1, to spacing renewals over a year. Naturally, because of the bureaucracy, or the mail—this was mid-January, or so—my new tag had not arrived. I had a choice of not driving or taking my chances, and I drove. The trooper gave me a ticket, but then wouldn’t let me go on; he impounded the car. They towed it to a garage—I think he had a deal with the garage; a kickback for each car he impounded. I called Nevis and had them send a car to pick me up. I insisted on riding in my car while they towed it, as they wouldn't let me drive; they had to tow it, of course. Nevis sent a car, and we brought the germanium detector and the isotopes to Nevis. It took me about a week to get my car back. That was a close call, but it's one of the vagaries of doing an experiment. We did these experiments for several years, improving the detectors as we went, making them larger, more efficient and with improved resolution.

Zierler:

Dave, is this a Columbia-only endeavor? Are you in contact with people at other institutions?

Hitlin:

In that experiment, we all were from Columbia. There were often collaborations between institutions, but that experiment was just Columbia. You were asking about my thesis. I did a very unusual thing. We had data, and I had analyzed the data, and I was going to stay a little bit to write things up for publication, so they gave me a job as an instructor, an academic rank below assistant professor; you teach, but there's no tenure at the end of the rainbow. By that time, commercial detectors had outpaced our homemade ones. They were much larger, more efficient, with higher resolution. I decided to go back to Nevis and do all my thesis runs over again with the new detectors. I didn't write my thesis in 1967 as I was supposed to, but I was still an instructor, and I was teaching an introductory physics course in the School of General Studies, Columbia’s night school, which was actually a lot of fun. The students were a quite different group than Columbia College undergraduates. Older, more mature, in some ways more serious. So, I had a good time doing that; the students kept calling me "Doctor," but I wasn't a doctor, because I hadn't filed my thesis. We redid all the experiments, and then I wrote my thesis in 1968. I had done a good job in General Studies, so they asked me to teach the same introductory physics course in the College. That was not nearly as much fun, because the students were, of course, much brighter, but a lot of them were premeds, and they argued for two more points on a homework problem, and that kind of stuff, which I don't have a lot of patience for. So, it wasn't nearly as good an experience as teaching in the School of General Studies.

I had a graduate particle physics course with him. He later wrote a book based on his lectures. The course was marvelous; he was as clear as day. You walked out of the room thinking you knew everything, until you went back and looked at your notes and saw that there was more to it than that. He was a marvelous teacher; the kind of teacher who would write on the board, and turn around and keep talking, while watching to see when people stopped writing, and then he would write some more. He was concerned that you were gathering what he was saying. He was great about that.

Sure. I wasn't worried about that, although it was clearly a concern for potential employers. I mean, the one thing physicists learn is how to learn. I've always said that. But the problem is, how do you get a job? So, I applied to lots of places, labs and universities, and I wasn't getting any positive responses. I got one offer for a nuclear physics job in Israel at the Weizmann Institute, with Gabi Goldring, who had been a collaborator of Sam Devons' at Imperial College. I guess Sam told him I was worth looking at, and he offered me a job. I only met Goldring once, twenty years later, at the home of a mutual friend at the Weizmann.

That was the only job offer I had, and then one day Wu said, "How's your job search going?" And I told her what I just told you. She had a mannerism: when she was thinking about something, or she didn't like something, she would shake her head from side-to-side. She did that, and that was the end of the conversation. A couple of days later, I was on the experiment floor at the cyclotron, which was a very noisy place, because it had motor-generator sets to provide clean power to the cyclotron magnet. The telephone ringer was a big fire bell. So, the fire bell rang, and it was Joe Ballam the Director of Research at SLAC, who said, "Why don't you come out and give a seminar?" A couple weeks later I went to SLAC for a seminar and interview. What had happened, I surmise—I don't know, but I surmise—was that Wu was a good friend of Pief Panofsky (W.K.H Panofsky, but known to all as Pief), the director of SLAC. They had known each other at Berkeley in the '30s, and they had remained friends. I'm guessing that she called Pief, and Pief then went to Joe Ballam, who called me.

So, I went out to SLAC in February 1969. There was a tradition at SLAC, I don't know if they have it anymore, of the brown bag lunchtime seminar in the Orange Room. At SLAC all the seminar rooms were named after colors. The chairs in the Orange Room were orange. The chairs in the Blue Room were blue. The chairs in the Yellow Room were yellow, etc. So, people would munch away at their sandwiches in the Orange Room while you gave your seminar. I talked about muonic X-rays, which was not something that the people at SLAC knew very much about, or probably even cared that much about, but that's what I had to talk about. And then you went around for interviews. They had a rigorous system of interviewing with each of the group leaders—there were groups A, B, C, D, E, F, G, and you would spend about an hour with each of them. Four of those people, if you count Mel Schwartz (who was at Stanford), wound up winning Nobel Prizes. Amazing. Each one was quite different, but extraordinarily interesting. Some of them probed you; some of them just told you what they were doing.

That's right. I've always lived a little dangerously, so I decided I would take my chances. I got the Stanford job, and I worked on the stopping muon experiments. I had an office on campus and was teaching at Stanford, but I spent almost all my time at SLAC, continuing to work on the K⁰ experiments.

Nobody ever gave me direction. I just did my thing. I taught one course each semester, that's standard, and continued whatever research I was doing. Most of the work I was doing was at SLAC. I did more work on the campus HEPL muon channel and a pion channel, but that never came to anything because the superconducting linac never worked. They changed the RF to a higher frequency, using a smaller accelerating cavity, to make the machine less expensive. That was a disastrous decision, because it brought them into a whole new realm of breakdowns and instabilities. They struggled for ten years or so, but they never made an accelerator that worked. They made one small section—I think it was 80 MeV; the original goal was 2 GeV. Years later, NSF closed it down. I designed muon and pion channels that could be built if there was an accelerator, but there was never an accelerator. Most of my research was at SLAC, but I was fully part of the Stanford community, and did my share of committees, arranging seminars and things of this sort. I'll tell you a colloquium story. One of the colloquia I arranged was for Fred Reines, from Irvine, who discovered the neutrino and did a lot of work with reactor neutrinos. I was taking him around to talk to various people, as is the custom. We opened a hallway door, and coming the other way was Joe Weber, who was famous for his gravitational wave antenna bar that claimed to see gravitational waves but didn't. He was visiting Stanford because Bill Fairbanks was working on a superconducting version with squid readout of the aluminum bar antenna. They saw each other, and it was amazing: Hugs all around; "Freddy!" "Joey!" It turned out that they grew up together in Hoboken.

I can also tell you another story that involves Fred Reines. When weak neutral currents were discovered around '72-'73 in the Gargamelle bubble chamber at CERN, and then later at Fermilab, some of us had the idea that you could use what we knew about nuclear physics to measure something new about weak neutral currents. That is, there is a process called Coulomb excitation, whereby you bring a charged particle, typically a proton or heavy ion, not to collide with a nucleus, but to pass close to it. The passing electric field can excite the nucleus, and you can measure the de-excitation gamma rays. Our idea was if there were weak neutral currents, you could have the analog of Coulomb excitation, where not a photon, but a Z boson was responsible for exciting a nucleus when a neutrino passed by. You could then measure the de-excitation photon. The reason that's interesting was there were certain selection rules involving the quantum numbers of the excited states of these nuclei, so you could learn about the spin and isospin characteristics of the weak neutral current.

We got two theorists from Stanford, Dirk Walecka and Tom Donnelly, to do some calculations with us, and we figured out how to do these experiments. There were several options. One uses higher energy neutrinos with carbon-12 as a target. Another case, with lower energy reactor neutrinos, was fluorine-19. It turned out that there was a scintillator, barium fluoride, that had been recently discovered. It was similar to the sodium iodide that I mentioned earlier as the first muonic X-ray detectors. Traversing particles produce light, you look at the light with a phototube. The idea was to make a target out of barium fluoride, and then if any of the fluorine atoms were excited by the reactor neutrinos you would see the de-excitation gamma rays. The target and the detector were the same, which is an elegant idea.

Zierler:

And where does Mark III enter the picture?

Hitlin:

Let's do Mark II before we do Mark III.

The Mark I detector was a very forward-looking device. It was the first large 4p detector, and it did great physics. But there were parts where it didn't work very well, in particular the electromagnetic shower counters, which were the responsibility of Berkeley. They used alternating levels of lead and plastic scintillator, viewed by phototubes. As they were building them, they cleaned the plastic with KimWipes, basically tissues, except those tissues have glass fibers embedded in them for strength. This produced microscopic scratches on the plastic scintillator, such that the decay length of the light became very short. If light was produced in the middle, it was very greatly reduced by the time it reached the photodetector on the end. So, there wasn't very much light out of these things, and they performed poorly. That was the motivation for doing the Lead Glass Wall, to take off a side of the MD and put in good photon detection.

Now Burt was going to build the Mark II, the bigger, better detector for SPEAR, but really for PEP. I joined Burt’s group right about when people were thinking about getting started on this. So, I looked around for someplace where I could contribute, and I became aware of a new type of detector that was being built at CERN by a group led by Bill Willis. He was at Yale, then went to CERN, then to Columbia. The accelerator they were using was the Intersecting Storage Rings (ISR) at that time, and he had built a lead-liquid argon detector, which covered only an octant. It had radiators of lead, but the sensitive material was not plastic scintillator. It was cryogenic liquid argon. This detector worked really well; it was a good idea. The device they made was really just jerry-rigged. It didn't have any good insulation when it was brought down to 79 degrees Kelvin; it was just covered with blue Styrofoam. The Styrofoam would crack because of the expansion and contraction with temperature cycling. So, the detector had icicles hanging from it, but it worked. I mentioned this to Burt as an interesting idea for Mark II. He said, "Get on a plane, go to CERN, and learn about it."

So, literally, the next week I went to CERN. Bill was very helpful. They showed me everything about it. I went into enormous detail about all the circuitry for the electronics, and all the ins and outs of what you need to do to keep the argon clean, etc. When I came back, I was very enthusiastic about it, I made a preliminary design, and I ran into some opposition. Not for the design, but for the fact that the Berkeley guys were the shower counter guys. That was their domain, and here I was poaching on it. We fought that out and made an arrangement where we would do a good fraction of the work at Berkeley and some fraction of the work at SLAC, but I would be the system leader. That worked fine. We had our differences, but it worked out nicely, and it produced something quite good; we built this large device that performed fabulously. We had to build a prototype, of course, to prove to people we knew what we were doing. At SLAC we built a cubic foot scale prototype using my design. The LBL people decided they needed their own design, so they made another version. We designed them so that both would fit into the same cryogenic apparatus. We would put one in, run, take it out, and put the other one in, running with the same electronics but a different mechanical design. Mine worked a lot better than theirs, so we adopted that design, and we built it.

Another aside there. I got in real trouble with Panofsky over the prototype. We had a very small group of people, and we were working day and night for weeks in this beam test. We had a very strong cobalt-60 gamma source that we could put beside the dewar to see a signal when there was no beam, just to tune things up. This source had a radioactivity label on it; we kept it in a lead container on the floor against the wall in the area. When the run was over, and we were completely exhausted, we all just went home and went to sleep. One of the janitors found this bright shiny object, which was not in the lead container, and put it in his pocket. He then, after a few hours, thought maybe that wasn't a good idea, and he turned it in. So, I was called on the carpet by Panofsky for letting a radioactive source get out into the wild, and really raked over the coals. That's the only time he ever did that, but I deserved it.

We then built the full liquid argon calorimeter and put it in the beam. It worked right away. I had bought a bottle of champagne that I put in the kitchen in the control room at SPEAR with a tag on it that said, "Open when we see a Bhabha peak,”—that is when we could see a full electron peak in the detector. We ran overnight, and at 10 o'clock the next morning, I was drinking champagne, because we could see the peak that quickly. That's how well it worked. It really was good. It was a workhorse. It was used at SPEAR, it then went to PEP intact and worked there, and finally it went to the SLC. For a long time, it held the record for the best energy resolution for a sampling electromagnetic calorimeter in the Particle Data Group review. I think somebody's done better now, but that's 30 years later, so it was very successful. So that was the Mark II, and now we’re up to the Mark III.

Zierler:

And a decision that you faced at this point.

Hitlin:

Yes, I had a decision to make. What happened was that Jasper Kirkby, a guy I knew very well because he had also been in Mel's group, put in a proposal. He had already had an experiment called DELCO at SPEAR. DELCO sacrificed other capabilities to have good electron recognition, as a strategy to find charm using the high p_T electron signature. He had put in a similar proposal for PEP, with Taylor, which was turned down. So, he revised it, and put it in again for SPEAR. They wanted to build a new version of DELCO at SPEAR. I saw the proposal, and remarked to Burt, "This doesn't really look like the right thing to do at SPEAR going forward." If you're going to run SPEAR, you need to emphasize exclusive state reconstruction. That is, you needed to have a large solid angle and the capability of good tracking of charge particles, good energy resolution for electromagnetic radiation, good particle ID. All the things the Mark II had, except it couldn’t be as Cadillac as the Mark II. I said, "I think there's a way that this can be done much cheaper and still do as good a job. This is the direction to go, not another DELCO-type experiment." Burt read Jasper’s proposal, thought about it, and said, "Why don't you put in a proposal?" It was just me and my idea; it wasn't going to involve anybody in Burt's group, because they were all getting ready for PEP. So, I involved another group at SLAC, Bob Mosley's Group D, which had built a streamer chamber facility that had done some good physics but was kind of at the end of their string of being able to do interesting physics with it.

No. I didn't think about it; I had maintained good relations with everybody. I worked with the same people, Mosley’s group, and expanded the collaboration to several other universities. We just picked up and carried on. I'm certainly not at all unhappy that I wound up at Caltech and not SLAC, especially given the way things have evolved at SLAC over time. A university appointment gave me more freedom and the ability to teach, which is fun, at a really great place.

The Mark III had to be done on a shoestring, but had to have really good capabilities, which made for an interesting design challenge. The first thing we decided to do to save money was to reuse the magnet of the original Magnetic Detector. It had steel plates for a flux return, and a normal-conducting aluminum coil. The coil had been taken out and stored in one of the end stations, and the flux return had been disassembled. We were going to test it and reuse it, because that saved a lot of money. I remember I was teaching at a summer school in Hawaii in the summer of '79; I already knew I was going to Caltech in the Fall, but I was still at SLAC. I got a phone call early in the morning from Bob Mosley. I guess he waited until I had some chance of being awake, because of the time difference. He called me with the news that the coil, which had been supported on wooden 4x4 cribbing, had fallen off the cribbing on one side and now had a dent in it. That was a crisis; we didn't know what to do. We decided to measure electrical and cooling integrity of the magnet coil. We did a ball test; the aluminum conductor has a central round hole for cooling water: you drop in a steel ball and make sure it comes out the other end with no obstructions. That was a way to see whether the dent in the conductor had impinged on the hole. It was fine, but in the end, Pief decided it was too big a risk, and that we should build a new coil.

We kept the Mark I steel, but we built a new coil. Everything else had to be built from scratch. We built a very nice drift chamber, a big cylindrical thing, perhaps 6 feet in diameter and 8 feet long. We reused the tooling that had been built for the Mark II drift chamber, which consisted of a cradle that you turn as you go to make the stringing of wires more convenient. The support is a wide leather belt, held together by giant staples which go through both sides of the leather and hold it up. This cradle had been used for the Mark II and then stored. We decided, again to save money, that we would use it for our drift chamber. That was a good decision until very near the end. We had strung all the wires and were testing the chamber when some of the staples pulled out. The belt broke, and one end of the cylindrical detector fell to the ground. We fixed the cradle and put it back, tested it, and it turned out there were a few wires broken, which we were, fortunately, able to repair.

The lesson I take from these episodes is that it's attractive to try to save money by reusing things, but you may or may not save money in the end. But this time we had formed a collaboration which involved Washington, Illinois, Santa Cruz, SLAC and Caltech—a reasonable sized collaboration at the time. Eventually, we added Texas when Joe Izen moved from Illinois to the University of Texas in Arlington. It was good collaboration; it held together for many years. We did a lot of interesting physics. We made, among other things, the standard absolute measurements of charmed meson decay fractions and the first measurement of the ratio of the lifetimes of the charged and neutral D mesons. We never got much credit for the latter because we did it by an indirect method that measured the semileptonic branching ratios and used isospin, but we got the right answer. People really wanted to see the lifetime difference with their own eyes, so the experiment that got the most press measured the actual decay length of the D’s by using silicon detectors. But we had the answer first. We continued with the Mark III for a long time and did a lot of good science. At the start I had, of course, to form a group at Caltech. It consisted of two or three grad students and two or three postdocs. Six or seven or eight people at a given time. Several of these people went on to do quite well in the field. That's actually been an interesting phenomenon. Apparently, the number of opportunities now in fields other than high energy physics, and/or the overall attractiveness of high energy physics have both changed over time. We used to have a very high fraction of the postdocs and graduate students go on in the field. That number of retentions has reduced substantially now. They can get jobs for twice or three times the amount of money in artificial intelligence or big data, or things of that sort, so a lot of them do.

Zierler:

When in the chronology does SLD get going?

Hitlin:

That was the next thing I was going to talk about. PEP was the big game in town, but I had nothing to with PEP. I had nothing to do with the Mark II moving there; I was fully involved with the Mark III. Then, Burt got approval to build the SLC, which was a clever version of a linear collider, where instead of clashing head-on beams, you used the time structure and some interesting manipulations to make one beam, hold on to it for a while, make the other beam, let it out, and have them collide. That required new tunnels and a new detector hall. They decided to build a single detector hall. Then, the question is, what detector went in there? Burt, of course, said, "Well, I have the Mark II, and we know how to move it, so why don't we just do that?" With some upgrades, that in fact happened. But some of us thought that the experimental situation at the SLC energies was quite a bit different from that at PEP and SPEAR energies. You could envision a much better detector than Mark II to confront the physics of SLC. So, we got together people at SLAC—Marty Breidenbach, Harvey Lynch, David Leith, Charlie Prescott, and me. A bunch of people from MIT—Jerry Freidman, Henry Kendall, Wit Busza, Lou Osborne and others. We decided to design what we called a grounds-up detector. In other words, if we had a clean sheet of paper, could we do better? Initially, the detector was called GUCCI. That was my coinage, that stood for Grounds Up Correctly Conceived Instrument. My wife actually bought me a Gucci handkerchief, because it was the only thing she could afford in the store, I think. So, it was named GUCCI, and we had GUCCI notes, and all these things. And then people got cold feet; they decided that wasn't a good look, that the name was too upscale. We didn't want to do that, so we thought about another name, and Marty Breidenbach came up with SLD. If you asked what it stood for, he would say, "It's what naturally comes after SLC." If you asked other people, they would say it stood for SLAC Large Detector. When there were a lot of problems with the accelerator, people would say it's SLAC's Last Detector. So, you can take your pick of what it stood for.

Zierler:

All plausible.

Hitlin:

Yes. We had a good time building SLD. I was the system manager for the electromagnetic and hadronic shower detectors, again because of my expertise on calorimetry. We built another liquid argon calorimeter. Marty Breidenbach and Charlie Baltay of Columbia were the spokespeople. They didn't have the concept of rotating spokespeople. They were, and I think still are, the spokespeople. We didn't have an election; that's a modern concept which didn't exist back then. There was a problem, of course, that the SLC barely worked for a very long time. There was a rush to get some physics out before CERN swept the table, which it eventually did. So, Mark II stayed in there and did some physics, and in fact, first showed that there were only three light neutrino species. That made the CERN folks angry, because the Mark II scooped them, but it was perfectly legitimate. Eventually, the overwhelming statistical advantage, because of the better performance of LEP, just overwhelmed what Mark II could do. Over time SLC started working somewhat better, and SLD had its turn. There happened to be an ace in the hole for SLD, which was not a central idea in the beginning. It started as peripheral but became the saving grace. A group of people at SLAC had been working on a polarized electron source. SLAC had expertise on polarized electron sources, so we incorporated this into SLD. That's what Charlie Prescott used to measure sin²θ_W back in the late '70s. They built a new and improved version of the polarized electron source, and we were able then to take data with the SLD using it, reversing the spin of the electron to limit systematics, and measuring a set of quantities, most importantly the forward-backward asymmetry, which was something that LEP couldn't do. So, that came to the rescue, allowing us to get some unique physics out of SLD. Otherwise, SLD was somewhat disappointing in that regard. We built an excellent detector but got only a minimal amount of physics out of it because the accelerator didn't perform very well.

It does emit very low-level alpha radiation, but it's with a many thousand-year half-life. It's not a very intense source of radiation. You do have to handle it carefully; you need a permit, but it's not dangerous if properly handled. You could just walk around with it and work with it. The pyrophoric nature is quite interesting. It means that it actually oxidizes very quickly, so if you want to make an electrical connection, you have to grind a clean spot so there's no oxide, and then quickly make the connection. When you're doing that grinding, it's quite spectacular, because the sparks that come off catch fire in the air.

Willis had only a small amount of depleted uranium to make his test calorimeter. He didn't have very much of it, so he made a uranium hexagonal section in the center, and then surrounded it by six other hexagons made of steel, and then put a layer behind that which had all seven hexagons made of steel. So, there was only one central section made of uranium. He put it in a test beam, and he analyzed the data, and he found, lo and behold, he got this energy back into the argon and he got compensation. He got very good e to p ratio. This had been in the literature for a long time, and I knew about it. So, we decided when we built the SLD calorimeter, that we would use uranium.

Many other people had the same idea; it became fashionable to build uranium calorimeters. Five or six large experiments at the same time in the mid '80s wanted to make uranium compensated calorimeters, using this idea. There was our experiment, SLD; there was the D0 experiment at Fermilab; there was ZEUS at DESY in Germany; there was the UA1 Upgrade at CERN, run by Carlo Rubbia who got a Nobel Prize around that time for discovering the W and the Z bosons. Everybody wanted depleted uranium. There's a lot of depleted uranium around. In fact, the Department of Energy actually had huge amounts of it that they were quite happy to get off their hands. The problem was that it was in the form of a powder—uranium sulphate, in big barrels, stored—I don't know where. Somewhere.

The DOE thereupon formed a committee. I was the SLD representative on this committee. We had a bunch of meetings. The idea was to find a way to turn the uranium sulphate into uranium metal, and then roll it into big sheets that could be used for our purposes. We obtained some depleted uranium sheets, and built a prototype calorimeter, largely at Caltech, and put it into the beam at SLAC. We took a lot of data, but we didn't see any compensation. I got a lot of flak for that. “Don't you know how to build a calorimeter? What's the matter with you? Willis could do it. Why couldn't you do it?” That went on for a while.

But it turned out that Willis had made a mistake. The answer to the problem was found by a colleague named Jim Brau of the University of Oregon, who I had previously known quite well because we were both assistant professors at SLAC, neither of whom got tenure. Jim did a Monte Carlo simulation and found the answer. The problem is that if the sensitive material that's going to gather up the fission neutron energy is a large nucleus like argon, it's very hard to put the energy back into the argon. If you have a sensitive material, like a plastic scintillator, which has a lot of hydrogen, a small nucleus, then you can put the energy back. It turned out the only one of all of these proposed large detectors that had a plan to use a plastic scintillator was ZEUS, the DESY one, and they built a very nice uranium calorimeter that worked beautifully and had compensation. But if you use liquid argon as a sensitive medium, it wouldn't work.

In fact, Jim also figured out what the problem was with Willis's experiment. I described to you the structure, that only the central section was uranium; all the rest was iron. So, when Willis’ group analyzed the data, they chose events in which there was a lot of energy in the uranium. By doing that, they biased the result so that what they were seeing was mostly electromagnetic interactions, not hadronic interactions, which fill up the entire device. So, he was choosing a subset of the data that was electromagnetic, and therefore had the desired e to p ratio; it was actually a physics error. When we understood this, we decided we wouldn't bother with uranium, and we made the whole thing out of lead. It was much cheaper, and it didn't work as well as Willis's erroneous theory said it would work, but it worked fine. The other groups that had liquid argon or other materials that didn't have a lot of hydrogen went on and used uranium, and, of course, didn't get compensation.

So, it was a very interesting little detour where we figured out the problem before we actually built it, but other people didn't. In fact, DOE asked me to run a special workshop on compensated calorimetry at Caltech, in which we had people from all over come to discuss this problem. I was very disappointed that Bill Willis himself did not show up to take his medicine; he sent somebody else from his group, which I thought—I have great respect for Bill Willis, but I thought he should have come to get shot at. Instead, he had one of the younger people get shot at. That's the uranium story. We built a very nice detector, but we didn't have a well-functioning accelerator to put it at. What saved our bacon a bit was there was a group that wanted to use polarized electrons to produce the Z bosons at the SLC that joined SLD rather early on in the game. They developed this polarized electron gun for the accelerator, which could give you linearly polarized electrons either forward or backward. You couldn't polarize the positrons, only the electrons, but it was good enough. They were aiming to do a set of subsidiary experiments which turned out to be the main thing that SLD actually was able to produce in physics. The reason was that it was a unique facility. The competitors at CERN, at LEP, who had four detectors, much more data, and basically did all the other physics far better than SLD could, did not have the ability to make a polarized electron beam.

So, we had this unique feature that we exploited to make a number of unique measurements. The main one was the forward-backward asymmetry, with respect to the linear polarization of the electron. We found an answer which was a bit different from the nearly equivalent, but not quite the same measurements that were made at LEP. A lot of effort went into trying to reconcile this, to see were we right, or were they right, or were we both right, and why was there this discrepancy? No problem was ever found with the experiments, so both experimental results stand, and they are not in agreement, to this day, 30-some years later. It remains a mystery in the Standard Model, which may someday be cleared up.

Zierler:

Dave, what theoretical guidance, if any, might help solve this mystery?

Hitlin:

There are models that perhaps could make both results correct, if there were for example a Z prime, an additional neutral weak boson, which also comes up in the context of other discrepancies with the Standard Model. So, it's possible to resolve this discrepancy with something beyond the Standard Model. We don't know quite what the right answer is. There may be a solution in the next generation, because people are talking about either building linear colliders or circular colliders that could—they're not primarily meant to, but they could go back into this regime and explore this problem again. That would be 20-30 years from now, but it might still happen. Anyway, the important output of SLD wound up being this forward-backward asymmetry measurement, which was unique. I guess, that's probably about it for SLD, but there was another related matter. That is, while we were building SLD, for example, my group was still working on Mark III. While you're building an experiment, you like to be doing current physics with some other experiment. To get Ph.D. theses, etc., you need data. You need to do physics. So, we always try to overlap these kinds of efforts while you're in these long time scale construction projects, so that you have some physics to do. We were doing that with Mark III.

Then, because of the crisis at SLAC where the accelerator, the SLC, was not working well, they basically had to throw all the resources of the laboratory at the SLC. One of the things they did was temporarily shut down SPEAR until the SLC problems were solved. It was supposed to be shut down for six months, and then it got extended, and extended again. We used to have television monitors that were up in lots of rooms that would tell you the status of the various systems. One screen would tell you the status of accelerators, another would tell you if the central computing system was up or down, etc. They were in many rooms, so I got a friend of mine who had friends in the computer center to add a line to the display that said, for example “179 days since the temporary shutdown of SPEAR.” And then the next day, it would be 180 days, etc. The director of the lab, Burt Richter, at the time, didn't think that was funny, but we kept it up. In fact, SPEAR never came back on. They just didn't have the resources to turn it back on. They were too busy working on the SLC. The question was how our group, without SPEAR, was going to do physics while waiting for the SLD to go on the beamline.

At around that time, in China, at the Institute for High Energy Physics in Beijing, which at that time was a brand-new thing, they were building an accelerator. There was a very interesting high-level competition that reached up to Deng Xiaoping, in fact. American laboratories were giving the Chinese a lot of help in starting up science in various chosen areas. They had four thrusts: physics, biology, other things. In the physics area, and particularly in high energy physics, they were getting advice from Wolfgang Panofsky, the director of SLAC at that time, from people like Sam Ting at MIT, from T.D. Lee at Columbia, and there was a competition. What sort of accelerator should they build if they were going to build an accelerator? Panofsky and T.D. Lee were advocating that they build a storage ring very much like SPEAR, and a detector very much like Mark III. Sam Ting thought they should build a 70 GeV proton accelerator. Not a colliding beam machine, just a proton accelerator. That would have probably been very much a physics dead end. The Chinese chose to build a SPEAR-like machine, BEPC, the Beijing Electron-Positron Collider, and they built the BES detector. The SLAC accelerator people gave them a lot of help. I went there in 1987 to give some lectures on charm physics. We had a summer school to help bring them up to speed on the physics. We gave them a lot of advice on building the detector; they built one that closely resembled the Mark III.

Actually, there's another little story there. BES copied Mark III in almost every detail, and among the things they copied was the design of our electromagnetic calorimeter. I think I mentioned before that we needed for Mark III to have a calorimeter that was as capable, but much less elaborate and expensive, as the liquid argon system of Mark II. We devised a system using layers of lead and sensitive wires in a gas medium to measure the shower energy. We had a very ingenious mechanical engineer named Knut Skarpaas, who designed an elegant construction technique. There was a central aluminum cylinder, and then a set of spacers that went along the top ribs, with holes in them. Then we strung the wires, and then we took plates of lead that had been encased in thin sheets of aluminum, bent them, put them over the ribs, and then used steel banding around to hold it in place. The load was transferred by the bands through the ribs to the central aluminum cylinder. There was a total of twelve layers, built one on top of each another. They had the same design at BES, but they were having trouble mechanically when they tried to band it. The load was not properly being transferred to the central support.

Then papers by Harald Fritzsch, and by Ikaros Bigi and Tony Sanda, that showed that a particular B decay, into a J/Ψ meson, the Ψ particle that had been discovered at SPEAR, and that Sam Ting discovered and named J, and a K meson, could have very interesting properties. You could have a very large experimental CP asymmetry, given the lifetime and the mixing, and it could cleanly get at the underlying physical theoretical quantities. CP violation was first measured in the K system in 1964 by Fitch and Cronin, and they got the Nobel Prize for that. But relating their measurement back to the fundamental quark mixing parameters was very difficult, because of hadronic uncertainties; the theory doesn't work well with light mesons, but it works much better with heavier mesons, like B mesons. So, what these theorists showed was that if you made this measurement of CP violation in the B system, you'd be able to work your way back without uncertainties as to what was going on at the deepest theoretical level.

This got everybody excited. We needed to figure out how to do this measurement There's a very clever way to do it, which is in e⁺e^-, use the U(4S) resonance, which is a bound state of a b and b quark, analogous to the Ψ” in the charm sector, which is in the 4 GeV region. The U(4S) was in the 10 GeV region. But you need the right kind of accelerator. If you produced the ((4S), which decays to a B meson and a B meson, then you could use the simultaneous measurement of the decay of the one meson and the other to measure CP violation. The trick was you need to distinguish the relative time of the decay of the one versus the other, which turned out to be very difficult. In conventional accelerators, where the beams had equal energies, you were producing the ((4S) at rest, so that the two resulting B decays went only a very short distance on average before decay; the mean flight path is only about 20 microns. We didn't have the technology to measure that distance with any precision.

The answer to that problem came from Pier Oddone, who at the time was the Assistant Director at Berkeley Lab, at LBNL, a very old friend of mine. He said that if the two beams were not of equal energy but rather of unequal energy, and you tune them correctly—in fact, for our case it was 9 GeV and 3.1 GeV, that provides the appropriate center mass energy to make an U(4S), but you make it in motion, along the direction of the higher energy beam. Due to the Lorentz transformation, the length before decay in the laboratory of these mesons produced in motion is much longer. In fact, with that boost, because of that ratio of energies of about 3, the 20-micron distance becomes about 250 microns, a range that you can measure. Not easily, but you can measure it. We knew how to do that. There were lots of workshops about this, and in fact, Pier's idea was surfaced at a workshop that was held at UCLA under the auspices of David Cline from UCLA, who used to run workshops all the time about everything. He had a very good nose for what the hot new physics was, and he would get people together to think about it.

Pier had this idea. The question then was, could you build an accelerator which had unequal beams? Lots of people thought about it and came up with ways of doing it. I have a table in which I gathered all this information for presentations I made in the past. There were, depending on how you count, something over 20 different worldwide proposals about how to build an asymmetric collider. It required not just building a new detector, you also had to build a new accelerator, because you needed a collider with unequal beams, and they didn't exist. It turns out that the ISR at CERN had run for a short time with unequal beams to do a different piece of physics all together, so at least it was proven that that idea could be done, at least with protons, if not with electrons and positrons. In other words, it was not an off-the-wall idea; it was something that you would at least consider.

So, everybody who had access to an accelerator complex got into the game. At SLAC, people—Pier, who had access to accelerator resources at LBNL, put people to work on this. Then, together with Gary Feldman, who was still at SLAC at the time, he came up with the idea of building a small storage ring, and putting it next to PEP. So, the big storage ring would have a high energy beam, the small storage ring, APIARY, would have a low energy beam, and you would collide the two. APIARY was a pun, not an acronym, because it would produce B's/bees. Other people came up with other approaches. There was an Italian design for colliding linacs. There was a similar idea to the SLAC design at DESY, which also had a large storage ring. There was a proposal to build colliding beams in Switzerland. There were important proposals from Cornell, which had done a lot of the e⁺e^- B physics, and sort of owned that field; they wanted to do this too, so they had a proposal. There was a proposal at Jefferson Lab, which had an electron linac. There were proposals at Novosibirsk. There were several proposals from Japan. There were more than 20 different ways of doing this. There were starting to be workshops on the topic. That's basically when I got interested in this. Frank Porter and I had been doing totally different things at Caltech, but we got together to work on this because we knew it was going to be a big deal. The natural place we would have thought to do this would be SLAC. I had a long history of working there, Frank had a long history of working there, so did Pier for that matter, all kinds of other people. But SLAC wasn't interested. Burt Richter, the Director, wanted to build a large linear collider. The SLC was meant to be a prototype for a really big linear collider, a several TeV linear collider. That's where he was headed. It was not going to be on the SLAC site. It was going to have to be somewhere else because it was too big for the SLAC site, but it was going to be basically spearheaded by SLAC as a project. It would be elsewhere in California if they had their druthers. That was their focus, together with all the problems he was having with the SLC. So, SLAC was just not interested. In fact, Burt called it “boutique physics,” which was not a compliment. It was not higher and higher energy exploratory physics. It was looking at fine details of something, and that wasn't really his bag.

Zierler:

Did you make the case that it wasn't boutique physics, that there was bigger science?

Hitlin:

Of course. Burt came around eventually, obviously, but that was his initial reaction. So, what could we do? SLAC wasn't interested, but we wanted to move forward, so we put together the Southern California Consortium for a B Factory. It was mainly us and Dave Cline at UCLA, but we gathered Santa Barbara, Riverside and Irvine, a whole bunch of the Southern California universities, to work on this. We had two different ideas. Frank and I wanted to do an asymmetric storage ring design. Cline wanted to do colliding linacs. He was working on doing it on the UCLA site, not on the campus per se, but across Veteran Avenue, where UCLA had ideas about expansion, which has not even happened to this day, but that was his idea. And our idea was to build it either at Caltech, or at Edwards Air Force Base, where JPL had a sub-site where they had tested rocket motors, and explosives. We designed the CITAR collider at Caltech: California Institute of Technology Asymmetric Rings. The idea was it would be underground, basically under the Caltech campus, which really isn't very big as university campuses go, but was the right size. If it was at Edwards Air Force Base, we would have built it on the surface: SPEAR was not in a tunnel. Because SPEAR, as I told you, was never an approved DOE project, it was actually built in a parking lot and shielded by concrete blocks. You could, with larger concrete blocks, do the same thing for a B factory; that was the idea at Edwards Air Force Base.

We worked on this together for more than a year, having monthly meetings and presenting what we learned. These were open meetings; one day, a couple of people from SLAC appeared at the meeting. At one meeting, Jonathan Dorfan appeared. Jonathan was at the time the head of Group C (Burt’s former group) at SLAC. He had been the progenitor and the spokesperson of the Mark II moving to PEP and then to SLC. He eventually became the Director of SLAC succeeding Burt Richter. And he was an old friend of mine, having worked on our E92 experiment with the bubble chamber that was his thesis. After that we worked together a little bit at SPEAR, but after that, we had never collaborated, but always remained in touch. Anyway, he appeared at the meeting, and we talked. Building an accelerator from scratch on a green field site is not simple. We decided to make common cause to see if we can do this at SLAC. That was why Jonathan came to our meeting, because he thought that was a good idea and something we could accomplish. So that was kind of the end of the Southern California Consortium; we decided to see what resources we could gather up at SLAC to think about accelerator design and detector design. At about the same time, there was a Snowmass meeting. These occur every five years or so: the high energy physics community would gather for two or three weeks, typically in Aspen or Snowmass—that's why they were called Snowmass meetings—and break up into organized groups, work on what they thought the future of high energy physics should be and make comparisons. It is always a very interesting process. It's still called Snowmass, but it's not at Snowmass anymore. In the wake of the GSA Las Vegas debacle where they were having questionable parties, DOE started looking askance at those gatherings. So, we don't do Snowmass in the way we used to. It was a good experience because people were there for extended periods of time. You could always go out for a hike and enjoy the environment, but being the sort of serious, crazy bunch of people that high energy physicists are, a lot of work got done, and a lot of interspecies communication happened, because people were in the same place at the same time. It was a very good thing. So, what goes on now has some vestige of that, but it's not quite the same.

In any case, there was such a Snowmass in 1988, and then there was another one in 1990. The SSC was being designed and built, so the thrust was largely to figure out what the rest of the high energy physics program would be. I ran the group that made the comparisons between the different methods of doing a CP violation experiment. It becomes very contentious because experiment A had to be compared with experiment B and experiment C, to see, with a given set of parameters, which one is more sensitive, etc. So, everybody wants to have that be to their advantage, and you have to be a referee so that you make fair comparisons. We did that, and we wrote a comprehensive comparison report. We decided—well, at least, some of us decided—that the optimal way to do the experiment was with an asymmetric colliding beam machine that had both rings in the same tunnel. Both the high energy and the low energy ring would be the same size. That's a slightly unusual idea. It had two thrusts. One was while we were doing this work at Caltech, before there was any SLAC involvement, we were not accelerator experts. Frank Porter had a friend from grad school back at Berkeley named Jeff Tennyson, who was what I would call an itinerant accelerator physicist. He was really good, but he didn't have a permanent job. He would go someplace for a couple of years and work, and when he was no longer interested, he would go somewhere else. The most recent place he had been was Novosibirsk, where they do the same kind of physics that SLAC did. We haven’t talked much about Novosibirsk, but I've had a very long and productive relationship with people there going back to 1977, the first time I went there.

In any case, Tennyson came to Caltech. We raised some private money from a Caltech benefactor named Millard Jacobs to do some of the design work on CITAR. We decided there were two main problems. One was, since this was going to be a very high current machine, there were going to be problems with heating of the beam pipe at the collision point because of the high current image on the beam pipe. So, we used some of the money to hire experts on cooling at JPL, and they produced a nice report which turned out to be quite relevant to this problem. The other was having Jeff Tennyson work with Frank on how to design the collider. They came up with the idea of using equal size rings having a very large number of individual bunches of electrons and positrons and having them collide one with another. If the rings were of equal size, then the bunches in the two rings would have equal spacing. So, bunch 17 always collides with bunch 17, etc. If you recall, the original idea at SLAC was a small ring colliding with the existing big ring. In that case there are more bunches in the big ring than the small ring because the bunch spacing would have to be the same. So, it's not always the same bunch colliding with the same bunch. That turns out to be a big problem. This was figured out in one of the workshops we held in at Caltech. We had people come from all over the world, all the people who were interested in doing this kind of physics. One of the CERN accelerator physicists, Eberhard Kiel, showed that if you had rings of unequal size, you would get a comb filter effect, which would cause instabilities. So, the original APIARY idea at SLAC would not have worked. Nor would the one at DESY. That was a definitive result; you had to have equal size rings.

So, that's the CITAR design, and that's the way we took the project up to SLAC. We would use the PEP tunnel, which already had one ring, that could be used as a high energy ring, and we would have to build another ring of the same size, but capable only of lower energy. It would have smaller magnets, and be cheaper to build, but would have the same diameter. The idea of having a large current made up of a large number of small bunches that didn't each have a lot of electrons in them was not initially well-received. In fact, when we took it up to Berkeley to present the idea, we had sent them a report beforehand, and I'll never forget this. Mike Zisman gave the Berkeley response—this was in the days when we used plastic foils. Mike took the front page of our report, made a foil from it, and then made a bunch of black marks on it that were meant to look like bullet holes; they were shooting holes in everything we had done. But in fact, we were right; it is the approach we used in the end. Their issue was the rapid growth of microwave instabilities, which we dealt with by an extensive feedback system. We worked together on PEP-II with the people from LBL and people from SLAC, and Jonathan even involved people from Livermore Lab who had expertise in RF structures.

So, the eventual proposal was a three-lab consortium, plus a lot of universities, of course, for BABAR. The three labs were responsible for the accelerator. We had to jump through many hoops; there were, of course, all these other proposals going on. Eventually, one by one, they all failed, except for the Japanese collider, which was approved and was our main competitor. Their accelerator was called KEK-B. KEK is the name of their laboratory, so it was called KEK-B for an accelerator that produces B mesons, and their detector was called Belle. So, PEP-II/BABAR and KEK-B/Belle became intense competitors for measuring CP violation in the B meson system. We were the two groups left standing in the end. But in the U.S., we were nowhere near DOE approval—I'm telling you the story up to the early '90s. We were not approved at that point. It still wasn't clear that we were part of the SLAC program at all. We had a bunch of committed people. For the detector, the experimental side, we had maybe 20-25 people. For the accelerator, we had people doing bootleg work on the design. Not many people were really tasked to do it, but people were interested, and they put time in. We were designing technically challenging asymmetric colliding rings, and a detector that could do this physics with asymmetric collisions. So was Cornell, because at that time Cornell was the home of B physics and e⁺e^- in the United States. Cornell is an NSF lab, SLAC is a DOE lab. So, this was not only an inter-laboratory, but an inter-agency competition.

I ran workshop after workshop to get people committed, because you have to build a collaboration; you need a lot of people to do this. The BABAR collaboration eventually grew to about 600 PhDs and grad students from all over the world. So, the competition went on. It wasn't at all clear how this was going to turn out. First, we had to make the B Factory part of the SLAC program. At that point, we made common cause with David Leith, who was the SLAC Director of Research. He controlled of a lot of resources on the experimental side, so that was extremely important. He was also very important in building worldwide interest in collaboration. We worked very closely together to try to make this a part of the SLAC program. I was elected as head of the SLAC User's Organization. Back in the '70s I had been chair of SLUO for two terms So this, with a long hiatus, was the third time. I used this SLUO position to build the support of the laboratory community for PEP-II/BABAR. We ran workshops not for doing B physics per se, but to compare other experiments that could be done at SLAC. One of these was to build what's called a tau/charm factory. That is a collider like SPEAR, or like the Beijing machine, to be built at SLAC. That's a smaller, cheaper thing to do. It would do charm physics and tau physics, but not B physics. There was another experiment to do charm photo production at SLAC. So, we ran SLUO workshops and had discussions about which of these should be the centerpiece of the SLAC program going forward. We worked through that, and eventually it became clear that it should be the B physics program.

When we finally had the lab and community on board, we then faced the external competition. How do you get this into the DOE program? At the same time, Cornell was trying to get its proposal into the NSF program. That went on for quite some time. In the meantime, I was offered the job as the head of the high energy physics group at Argonne National Lab. Argonne, which used to have an accelerator, is a large multipurpose DOE lab not far from Fermilab. It had the ZGS in the '60s and '70s but didn't anymore. It has a very large experimental high energy physics group. They offered me the job of running that division with the idea of positioning the group to be a prominent player in building detectors for the SSC, which was an interesting opportunity. I thought about it very seriously. I went out there, looked at houses, all the things you do. In the end, I made what was regarded at the time as probably a foolish decision, which was to stay at Caltech and to work on getting the B Factory approved. Had I let it go, that doesn't mean it would have failed, but I was certainly a prime mover in what was going on.

Anyway, I decided that I was committed to the B Factory, and I was going to just go for it from Caltech, so I turned down the Argonne job. We needed to find a way into the federal budget. Richter was a master of this sort of business, working with the various congressional offices, with the OMB, etc., [HDG1]to bring this to the fore as a part of the SLAC program. The scale was just under $300 million at that time, not a small number. We organized delegations through SLUO to go to Washington en masse to visit the congressional people in your own district, or around your district. This proved that politics is indeed local. For example, if I went to talk to a congressman from Massachusetts because we had MIT collaborators, their initial reaction would be "Well, why are you here? You're not from my district." I knew that would happen, so the ready answer was, "Digital Equipment Corporation is in your district, and we use Digital Equipment Corporation computers." You always had to have an answer ready, since a question like that was going to come. We walked the halls of Congress, and we talked to everybody. I remember talking to Nancy Pelosi at that time. She had just become a congressperson in the early '90s. Among all the various people, she was one of the most impressive. We also met with Dianne Feinstein in Burt Richter's office at SLAC. She was much less impressive but was helpful nonetheless in getting the approval. We also went to Sacramento seeking state aid, which never panned out. We worked with Carlos Moorhead, the congressman from Pasadena, who was the longest serving member of the California delegation at the time, which means something; seniority means a lot So, it fell to him, for example, to organize a delegation letter; the idea was you wanted to make sure you had the full support of the California House delegation. We did all of this kind of work in support of getting the funding.

It needs to be said that this was all in the context of real threats to the existence of SLAC as a high energy physics laboratory, because the SSC was being built; it was going to be the big game in town for DOE. They eventually made a horrible mess of it, but that's another story. They were becoming concerned with how to find operating funds for the SSC. New money was provided for the construction, but when it comes to operations, you always have to redirect the program to find the operating money. You may get something additional, but most of it is expected to come from the current program. So, how do you do that? You have to stop doing something else, and the favorite candidate was to stop SLAC hep. There was Fermilab, which was the flagship, and was running the Tevatron, and looking for the Higgs and the top quark. So, one idea was to stop doing high energy physics at SLAC. There were committees considering this and we were quite worried that they would recommend that. So, it was very important for SLAC to have a first-class flagship project that could do important physics, and to say, this is our program, this is important, we need to do it. We made some headway but were not getting all the way to approval. Then, the SSC was canceled. That was, as far as I considered, an unbelievably stupid decision. It was a purely political decision on the part of Clinton. They were trying hard to balance the budget, and they were being very stringent about outlays of this kind. They decided that one way to do this in the scientific realm was to kill something, and apparently the two candidates were the Houston Spaceflight Center and the SSC. It came down to the SSC, which I think is extraordinarily unfortunate for U.S. science and for basic knowledge. Everything that's being done now at the LHC, with major U.S. participation, would have been done years and years and years ago, better, had there been an SSC. So that cancellation was just dumb, and the U.S. program has never actually fully recovered. The physics community bears some responsibility, but the DOE bears a great deal of responsibility for mismanagement. It was unbelievably badly done. I was never part of SSC activities, but I was on review committees, so I could see what was happening. It was a bad show. In any case, when the SSC went down, it was a new ballgame in high energy physics. One way to think about it is that PEP-II and BABAR were something of a consolation prize. We'll never know, but it is what happened. So, soon after the SSC cancelation, we were approved.

Zierler:

Are you the main spokesman at the beginning? That's part of the approval, that you would be spokesman?

Hitlin:

No. We didn't have a spokesman, per se, in the detector organization, but I was running it. I didn't have a title; I was just doing it. The first thing we did when we were approved was to break out the Aberlour from David Leith's desk drawer. He was Scottish, and when he came to SLAC in the late '60s, he brought along the concept of single malt scotch. He was a great advocate and emissary for single malt scotch. Beltramo's was a large liquor store in Atherton. It's no longer in existence, I think, but it was a very good wine and liquor store. He convinced them to start carrying several varieties of single malt scotch. To this day, I still have many kinds of single malts at home. Later on, maybe, if there's time, I can tell you a story about single malt scotch in his retirement, but let's leave that. Jonathan and David Leith and I toasted our success. And then we went and had cake and champagne with everyone else.

And then there was a problem, whose name was Daniel Moynihan, the powerful Senator from New York, doing constituent service. He put his fingers into the pot, and then we had a new situation. Yes, the B factory was approved. DOE would build the B factory and associated detector, but it could be at SLAC in California, or it could be at Cornell in New York. The money would come from DOE, which would be straightforward if it were built at SLAC, but perhaps it would be built at an NSF lab in Ithaca. So, how did DOE resolve that one? They formed an inter-agency committee run by Stan Kowalski from MIT, a neutral party. This had to be done very quickly because of the exigencies of getting the project into the budget. So, with very short notice, we had a committee proceeding. There were two meetings. One was at SLAC in which the two accelerator concepts were compared, and the other one was at Cornell in which the detector concepts were compared. I basically didn't sleep for about three weeks, the time we had to put together a report and plan the presentations. We had a very small crew of 20-25 people, and everybody worked really hard. I had to put this all together into a coherent document. I thank Steve Jobs and Tom Rokicki, for that. The reason was the Next computer. At the time, Steve Jobs had left Apple, and he had formed another company called Next, which made a really great computer. I had access to one, not at Caltech, but at SLAC. It had a beautiful high-resolution screen, and it had an implementation of TeX, the typesetting language, written by Rokicki, which was far better than any implementation of TeX ever. It was as good as anything that exists even now, in that it allowed you to write text and see it rendered nearly instantaneously, so you would work on a very complicated manuscript very efficiently.

I had to get the various contributions edited, make it all make sense, and put this report together, and then go to these proceedings at Cornell. Both labs’ proposals were perfectly viable, both for the accelerator and the sector. But the approaches at SLAC and at Cornell were very different, and that was one of the arguments we made. The Cornell style was do everything cleverly and cheaply and make something that sort of worked, and then fix it. That had always been Cornell's style. originated by Bob Wilson, who built the Cornell accelerators. He eventually became the first Director of Fermilab, and he built the first Fermilab accelerator in that style, which was not a good idea on the scale of a 200 GeV accelerator, so it didn't work very well at the beginning. They had to spend a lot of money and a lot of time to eventually get it to work. So, that was Cornell's style, and that's the way they would have done the B Factory. And it would have worked eventually. SLAC's style was different. We always emphasized that we were building a factory. A factory runs 24 hours a day; it has to be reliable. It has to work. So, our designs were very carefully engineered, and had some redundancy. All electronics were burned in so that if anything that was going to fail did so before you installed it, etc. It wasn't quite the space program, but that was the approach we took. We said, "We’ll build this. It will work, and we'll get on the air right away." We had competition from Japan because they had also been approved, so we were in a race.

Our argument prevailed. The Kowalski Committee wrote a “on the one hand, on the other hand” report. It just made comparisons; it didn't make any judgments. They basically put it back in the lap of the DOE office of high energy physics, which perhaps not surprisingly decided to do the B Factory at SLAC. So, that's what we did. Senator Moynihan did his best constituent service, but it didn't work. Cornell actually kept running and did a lot of physics while we were building. The luminosity of their existing collider was nowhere near our proposed PEP-II luminosity, which actually led, a little bit later, to a story. In the late '90s, when we were about ready to come on, NSF decided that the Cornell collider would be turned off. They then were able to actually reconfigure their accelerator to run at a lower energy and do some nice charm physics, so they survived years longer. They did that quite successfully, but they went out of the B physics business. So, their group had a party when they were being turned off. I guess it became a sort of rowdy party. They had a total integrated luminosity of 9 inverse femtobarns. That's the unit of measure. So, they took a photo—this is before the iPhone, so somebody must have had a real camera—of two of their collaborators, who I can't identify, mooning us, with “9 inverse fb^-1” painted on their buttocks, and with the spokesman of their experiment at the time, holding up a sign that said, "To our friends at SLAC." I was sent a copy of this, which I still have. To motivate the troops at SLAC, as we were working quite intensely, I put this photo up on my office door, which got me into a lot of trouble, because one of the secretaries took great offense. She caused a ruckus and eventually quit over it. They were going out of business, and were saying, this is how much we’ve got, what are you going to do? Well, it turned out that we did very well, and we got triple that amount in our first year, on the way to a lot more and a huge data sample.

By the end of 1994 we were fully approved. The groups building the accelerator had been working together, the three-lab consortium doing design and R&D, for years now. On the experimental side, it was still a small group of people. We needed to expand it dramatically. David Leith and I had been working at that, and had made some sorties around Europe, finding groups that were interested, and which funding agencies would be interested in participating. So, we decided to have a workshop in December 1994, with just a few months' notice. I made a workshop announcement that looked like a formal wedding invitation. You're Invited, … to a workshop at SLAC, and sent it to everybody we knew. Hundreds of people showed up from all over. We used that workshop to start putting together a formal collaboration. We had people from the U.K., France, Italy, Germany, Russia, a few others, eventually Spain, Norway, Taiwan—it became a very large collaboration.

We had to do some rather quick organizational work. First each country's delegation had to organize, and the Europeans had a lot of experience with how to operate in the CERN environment, so they knew how to do this better than the Americans did. They chose national representatives, and we identified US representatives, and we formed an exploratory committee. We decided on an organizational structure. We formed a subcommittee to write a governance document with a set of rules. We had had those kinds of things in other experiments; we weren't starting from scratch. We made a structure that had a spokesman, a deputy, etc. We had to find a title for the collaboration. The way this all worked was to have a nominating committee of senior people with some balance between the U.S., which was about half the collaboration, and one representative from each of these other countries, except for France, which had two funding agencies. That group nominated me as spokesman. We also had a Collaboration Council, with one representative from each institution, to deal with some of the more formal institutional needs, not the administration of the project, or the physics. My nomination was voted on by the Collaboration Council, and I was selected for a three-year term. Three years later, I was renewed for another three-year term. We were still in the middle of construction at that time. That's the way it worked and works now in most collaborations. In the old days, including a whole bunch of things I did, if you founded the experiment, you were the spokesman, until you decided you wanted to do something else. It doesn't work that way anymore. As these things got larger and larger, they became somewhat more democratic. Basically, to gather these resources, you have to give people a voice. So, it's usually clear who the progenitor is, and who makes it go, etc. But eventually people want to have their turn at it, so it's now become an elective game, which I think makes a great deal of sense.

We then had to choose a collaboration name. I had started using the name BABAR back in the late '80s. I did that more or less to make fun of the fact that people made these somewhat tortured acronyms for their experiments. Sometimes it would be as straightforward as CMS for Compact Muon Spectrometer. That's not very elegant, but it actually makes some sense. Sometimes it would be ATLAS, which is another LHC experiment, which chooses letters from a bunch of words, not all of which are the first letter in a word in order to spell it out. I regard that kind of thing as silly. I thought let's have a name for the experiment that's not an acronym but has some relevance. I decided that Babar, the children's book character, sounds a bit like B and Bbar, but it's not an acronym, so let's call it BABAR. I started using clipart of various Babar images, and I started buying first editions of Babar books and Babar toys; I now have a collection. Some people liked the idea, some people absolutely hated it, thinking it wasn't serious. David Leith didn't like it at all; he thought that a new collaboration should choose its own name. I was not happy, but I decided to have an election to choose the name of the collaboration. We chose a committee chaired by John Fry from Liverpool. I was hands off; he ran the election. He asked people to submit names, then he and his daughters sat around on a Sunday afternoon, went through the long list of submitted names, and pared it down to about eight names, including, of course, BABAR. As soon as he told me that, I said to myself “No problem. BABAR will win." Everybody had heard of Babar; nobody had heard of the other choices. It wasn't any sort of sophisticated ranking voting. It was just choose one. So, naturally, a majority—I don't remember if it was a plurality or a majority—went for BABAR, and the rest of the votes were scattered around all the other random names. So, the collaboration name became BABAR. A coda: Klaus Schubert, the German representative on the Executive Board kept insisting, after the fact, that the name had to be an acronym. He came to a meeting with a list of quite tortured acronyms for BABAR but was unable to muster support for the idea.

Zierler:

There you go. Name recognition is valuable in physics as well as politics.

Hitlin:

Exactly. I thought it was an amusing name, but Babar is copyrighted. I had been using the logo long before we were an actual project. When we would have a workshop, I would make signs like, "This way to the workshop," and it would have a Babar character on it. One day I got a call from the business office at SLAC asking, "Do you have permission to use the logo?" I told them, "No. You're the business office. See if you can get me permission." They went away and near as I can tell, did exactly nothing. But then they came around again and somebody else said, "Do you have permission?" So, I decided I had to do something. It turned out that the Babar books were originally written in the '30s by Jean de Brunhoff. He wrote about six books before the war for his children, and his wife illustrated them. Then he died quite young, and sometime after the Second World War, his son Laurent wrote many more Babar books; there are now a lot of books, TV cartoons and licensed toys, and all that kind of stuff. It turned out that Laurent de Brunhoff wound up living in the United States and was married to a woman who was a professor of English at Middletown College in Connecticut. I don't remember how I knew that, but I knew that, so I tried to contact him. This was the old days when you could call Information. I called 411 and asked for the phone number and address of Laurent de Brunhoff in Connecticut, and they gave it to me. I wrote him a long letter explaining about B mesons and CP violation and how much we had enjoyed using the Babar logo, and how it would be a great honor if we were to be able to associate Babar with his science, and would it be possible to get his permission for this? I never received a reply, but apparently, he contacted his licensing company, which was called Nelvana—it's now gone on to some other company. They were the licensee for the books and the toys. Nelvana contacted the SLAC business office and gave us a five-year royalty-free license. So, we were all set. The license came with a bunch of, you can do this, and you can do that, with a style manual of how you could depict Babar, which was actually quite amusing. It had rules, like Babar never carries anything in his trunk. And then, you turn the page, and there's a picture of Babar with a tennis racket in his trunk. So, their rules were sort-of-rules, not really rules. In any case, we used Babar, writing the copyright information under any use of an image, and put it on our webpage, and then all the collaborators made their own institutional webpage the same way. We decided, for example, we wanted to make T-shirts, so I had to get permission to make T-shirts. Would that be allowed? They wanted to know how many T-shirts. I said 500, which was fine; it was such a small number that it didn't bother them. They said OK, so we made BABAR T-shirts and other things of that sort. We had a lot of fun using this logo, and the license was renewed several times. I don't know if it's still renewed, even though we're publishing papers, but it did work very nicely.

Our second collaboration meeting was at École Polytechnique outside of Paris in February ‘95. I got there a day early and I went to the Clignancourt flea market in the north of Paris and was browsing through the used books looking for Babar first editions. I found one, and opened it, and inside the cover was what I call a bluebook. The kind used to write exams in college; this cheap set of lined pages stapled together, with a light blue cover. It was all written in beautiful cursive and was about 20 pages long. It was just inside the book, so I opened it up, and it was the Napoleonic Code contract of marriage—France has Code Napoleon, not English law, and in Code Napoleon, to get a marriage license, you must have a contract that says the groom is bringing this to the marriage, the bride is bringing this—an inventory of your piano, and your stove, and your this and that. This was the Contrat de Mariage for Jean de Brunhoff, the original author of the Babar books. I needed this, of course. The owner of the store had put it in the book to just keep it flat and keep it safe. I bought the book and the Contrat de Mariage for, around $75. This was before Euros, so I don't remember the conversion rate, but around $75. We eventually scanned it in, and it's accessible from the BABAR website. The SLAC archivist gave me an archival envelope to store it, so I'm taking good care of it. We had a lot of fun with the Babar stuff, to this day. We should get onto Babar physics.

Zierler:

Let's talk specifically about the connection between CP asymmetries and the 2008 Nobel Prize.

Hitlin:

We’ve already discussed the need to have an asymmetric collider. It also had to have a much higher luminosity, intensity, than any other accelerator had ever had. The combination of those two generated quite a bit of skepticism in the community, including with Burt Richter, who was not only a particle physicist, but an accelerator designer. So, he was very skeptical, and part of convincing him that this should be part of the SLAC program was convincing him that we could in fact build A collider with this unprecedented luminosity, which eventually, we did. We had to build this robust B factory, as I said, that in an unprecedented configuration that reached unprecedented intensity. We did that. DOE cooperated by giving us a budget which was basically what we asked for, and giving it to us on a timely schedule, which is not often something that happens. They usually stretch out the funding profile. But they gave us the money we needed in a timely way, and we did our job in a timely way, and we were able to get on the beam line about a month ahead of schedule. We took data quickly; normally when accelerators start, they don't work very well, but they eventually learn how to run it and get better and better. It might even take years to get up to the design intensity. But PEP-II got up to the design intensity in less than a year. The whole thing went off very well. Not that we didn't have our problems. We had plenty of problems. It took a lot of technical ingenuity; it took a lot of financial ingenuity to get around problems and deal with shortfalls. We did all of that. We, in fact, won an award from the DOE for project management, which was interesting. Jonathan Dorfan and I went to Washington to accept this award for how well we had done the project management. We came very close to missing the award, because a big trailer truck that was backing down an incline got stuck and blocked the street. We couldn't go in; we couldn't go out. By the time they got the truck unstuck, we got to the place and ran in as they were announcing the award; we were almost late to our award for being on time. But we got there just on time to accept the award, at which I remember our closest competitor was the re-piping of the National Petroleum Reserve in Texas.

We got the detector built on time, not without a lot of interesting problems that I won't go into. It's a story in itself, but we did it well. We put together a team that worked together very, very well. We were ready to do the physics analysis at t=0 in this blinded way, which I'll come to in a moment, knowing that we were in a race with the Japanese. To exhort the troops, I would always tell them that we're running a sprint and a marathon at the same time. In other words, we want to win the race in the short term, but we want to do things sufficiently well that we have longevity to be able to gather the enormous amount of data to take on the very large smorgasbord of physics that we were capable of doing. The centerpiece, of course, was the CP violation. Now, Kobayashi and Maskawa had given us a range of results that we would get for the CP asymmetry. You couldn't make a precise prediction because the value depended on the mass of the top quark, and the top quark had not been found when this work was all started. So, we used to show a range. It might be here if the top quark is this. It might be there. But eventually, the top quark mass will be known, and it will be a precise prediction. In the late '90s, the top quark mass was measured, and we could zero in on a target value of the Standard Model prediction. That would tell us with some precision about the underlying quantities of the Kobayashi-Maskawa matrix.

Then the question is, how do we do the analysis? The idea of blinded analysis was just starting. In fact, one of the first people to do it was my friend Bruce Winstein, who I'll come to in a little bit, who did a lot of the important work in the later stages for measuring CP violation in the K_L system, the other system that shows CP violation. Bruce made measurements, and then at CERN they had measurements, and there was a Standard Model prediction, and things kept going back and forth with experiments not agreeing, getting a different answer than the Standard Model, doing it better, everything changing around. So, for his last and best measurement, he decided to do a blinded analysis. The idea was, as I said, you do the analysis in such a way as you're getting a number, and you refine it, but you don't know the number because you're missing an important quantity which is kept in a sealed envelope someplace by a trusted person. Then, when you finally say, "This is it," you open the envelope, plug in this number, and now you know the answer, so you could not possibly have been biased. In our instance, since there was a Standard Model prediction which was reasonably precise, if the Standard Model was right, we knew the expected answer. The question is, how do we do the blinding, and should we do it at all?

So, we had a meeting, in Padua. I remember standing up in front of our group in a big auditorium and having a free-wheeling discussion about his. Do we do blinding? If we do blinding, how do we do it, because it could get rather complex to keep the number a secret? Do we do this at all? In the middle of the discussion, one of the younger people said, "Well, there's no reason to do a blinded analysis because we know the answer." That made up my mind to do a blinded analysis, because if you're going to come to it with that bias, I want to be sure you don't know the experimental answer, even though you think you know what the answer's going to be. We decided we would do blinded analysis, and we do blinded analyses to this day. There are instances where you can't apply it, but in almost all cases you can, and that's what we do. It's been a very, very successful idea. Our first result was not terribly precise; we refined it as we went along and got more and more data. But it agreed pretty well with the Standard Model. Our Japanese competitors got a similar answer at roughly the same time that didn't agree so well but had very large errors, which kind of covered the discrepancy. So, there was always the question of, who did it first? We did it first. I think even they agree, but so close that it doesn't make any difference. We're both recognized as having done it first, and effectively simultaneously.

So, there were two levels of the exposure of the number. One was the first one, in 2000, at a conference in Osaka, and another was when we had more data a year or so later; we had an effect which was fully statistically significant in 2001. I'll come back to that in a minute. Let me come back to Bruce Winstein for a second. He and I were old friends, and he was on a committee at SLAC called the Scientific Policy Committee, which back in 1989, had to give us an approval, had to say this should be part of the SLAC program going forward. We made a lot of presentations to that end, and he was on the committee. After the meetings there was a dinner associated with these committees. You take people out to dinner and discuss things that had happened during the day, etc. After the dinner, in the parking lot, Bruce and I were talking. He said, "We're going to approve this. We think it's a good idea, but it's going to take a really long time for you to actually make the thing work and get the data, because we really don't believe you're going to come up to speed as fast as you're claiming. The accelerator's not going to work that well." So, I said, "Okay, let’s make a bet." The bet was for a case of very good French wine—Château Lynch-Bages, which is a first-growth Bordeaux, Premier Grand Cru Classé—that we will have a three standard deviation measurement of CP violation in the B meson system by the turn of the century. That's the way we phrased it. I chose the wine. The reason I chose that wine is I had a case of this wine that I bought at a very good price from Trader Joe's because the 1977 vintage was regarded as not a particularly good vintage for Bordeaux wines. A lot of this excellent wine wound up at Trader Joe's, which is not like Trader Joe's now. It used to sell these odd lots of things. It originated in Pasadena, and I actually knew the owner, Joe Colombo. He was a friend of one of my postdocs, RoseMary Baltrusaitis. He used to be a member of the Athenaeum (the faculty club) at Caltech which has members from the surrounding community, not Caltech people. He used to drink in the late afternoon at the bar, as did RoseMary, so I met him through her.

I had this case of wine, which also had a physics connection because the director of Fermilab at that time, John Peoples, was an assistant professor at Columbia when I was there. In fact, when I became an instructor, I shared an office with him. One of his favorite wines was Château Lynch-Bages, the same wine. It was a wine that even a graduate student could afford in the mid '60s, amazingly enough. The b quark was discovered at Fermilab in 1977. So, I had this physics connection to the wine, and I had a case of the wine. I said, "Okay, we'll bet a case of this wine that we make this measurement by that time." I promptly forgot then about the bet. Then in 1999 I got an email from Bob Seimann, a SLAC accelerator physicist, who happened to be standing next to us when we made the bet. He had written it down, so he then, ten years later effectively, exhumed it and wrote an email to Bruce and I saying:

“Eons ago in the parking lot of the Stanford Park Hotel a bet was made between Professor David Hitlin of the California Institute of Technology and Professor Bruce Winstein of the University of Chicago. These esteemed scholars bet One Case of Chateau Lynch-Bages that a three standard deviation signal of CP violation would be seen in B meson decay by the end of the millennium. Professor Hitlin took the affirmative side.”

In 2001 we had a meeting in Ise, Japan, in western Japan. It was a meeting from a series which was called BCP, which was set up by the Japanese B Factory folks. We had a similar series of meetings we had been running since the late '80s, which was called Heavy Flavor Physics, the HF series, which had become largely about B physics. They were running one series; we were running another. At the end of the Ise meeting, we had a meeting of people there from the BCP series and people from my series, and I brokered putting the two series together. Rather than competing, why don't we all just do the same thing together? We'll give it a different name, which we called Flavor Physics and CP Violation, FPCP. We decided we would alternate it between Europe, North America, and Asia. Every year it moves around, and it's still going. In fact, next week, there's going to be an edition of it in China, which will be virtual, largely. At that Ise meeting, BABAR showed our first CP results. We were starting to have other results on the B meson lifetime, and several other things. The Japanese collaboration had its results; we had ours. People from Cornell still had things to say. Other people did. One of the sessions one tends to have at these meetings concerns the future. What are you going to do in the future? So, I listened to what people were saying, and typically they were talking about upgrading this a little bit and making the intensity of the accelerator a little bit more. I just looked at it and said, "Well, that's not enough." It's not ambitious enough. If you do a factor of two or three better, you know, statistics goes like the square root of the number of events you have. That's not going to be enough for a funding agency to give you $100 million to do it better; you need to be more ambitious. It turned out that at this meeting, there was another person there from SLAC named John Seeman, who had been the head accelerator person for PEP-II. He was there to give a report at this meeting about the status of PEP-II. I asked John what if we just said we're going to use as many components as we can, but we're going to really rip it up and start over. How well could you do? Could you do 10³⁶? I should give you some numbers. The luminosity goal of the PEP-II accelerator was 3 x 10³³ in unites of centimeter minus two seconds minus one. We got that within about a year, and we had plans to get up to, over time, 10³⁴, three times more. We had built that in as kind of a safety factor that we could do with small investments and experience. In fact, we reached that eventually. The Japanese had the same kind of plan, and they were behind us. They caught up to us, and eventually they surpassed the 10³⁴ by a bit, but that was years into the future. So, 10³⁴ was the standard that we were aiming to get to in a few years.

One of the theorists that I mentioned before, Tony Sanda, who had written one of the important original papers, while in the U.S. at Rockefeller University, but had then gone back to Japan, to Nagoya, gave a talk at this meeting, and said that you need more data to do more things and you really need to get to 10³⁴. But he made a typo on his slide, and instead of writing 10³⁴, he wrote 10⁴³. In other words, instead of three times more, seven orders of magnitude more, which was crazy. We all had a good laugh about that, and people often in the future would mention 10⁴³. So, when I went to John Seeman to talk about this, I said, "I won't ask you to do 10⁴³, but how about 10³⁶ instead of 10³⁴?" In other words, 50-100 times more than we're currently capable of. "Could you do this?" He was sitting in the back of this, listening to all the physics, but he was not a particle physicist. He was an accelerator physicist, so he didn't really need to pay full attention. He started thinking about it and making back-of-the-envelope calculations, scaling things. And he said, "I think you might be able to do 10³⁶."

So, when we got back home, we got serious about this. Let's get a couple people together and see what it would take to do 10³⁶, what it would take for a detector that's capable of doing the physics at that much higher intensity, which would mean you'd have to change a lot of things in the experimental apparatus, too. We set to work on this in spare time, basically, because we were all busy with BABAR data and making the accelerator run better. But we started, and eventually it started snowballing. We put people together and I did my thing of running workshops, and we started making progress. The technique was somewhat brute force. The idea was to do it the same way, in the same tunnels, but have much higher current, which would mean it would cost a great deal more to operate, and it would be really hard to provide the cooling. But we thought we could do it. So, we started pushing this, but weren't making a lot of headway. The Japanese started doing the same sort of thing, I think, not at the level of intensity that we were, but they started thinking about it, too. SLAC was not terribly receptive. This time, for different reasons, because they knew the handwriting was on the wall, and SLAC was not going to be in business to do high energy physics in the teens and twenties. They were going to last as long as they lasted, but they were not going to get more funding for a big high energy physics installation. So, SLAC wasn't terribly supportive, but they let us use some of the people to work on this. We wrote a good design report employing the brute force approach. The next interesting thing that happened was one of our Italian collaborators, Marcello Giorgi—who had been a prominent BABAR person, and in the later years had become one of the spokesmen, started talking about this among the Italian accelerator community—and Pantaleo Raimondi—who actually had been at SLAC for a long time, but at that point was back in the Frascati Laboratory outside of Rome—came up with an interesting idea.

A lot of work had been going on in the community on linear colliders, and it still is going on. SLAC worked on linear colliders for many years. There's a candidate site in Japan that they're trying to get support for. That would be a worldwide accelerator. It may happen; it may not. I don't know. But a lot of people had worked on linear colliders. The secret to linear colliders is you need very, very tiny beams with very small cross sections to get the needed luminosity. Instead of big currents, you have small currents, but you have very small beams. That was the original idea of the linear collider B factories that we talked about in the late '80s. That turned out to be an idea that wouldn't have worked. If you remember, I told you that the small ring on big ring turned out to be a dead end. The linear colliders for B factories turned out to be a dead end, too, because those small beams made a very disruptive collision and reduced the effective cross section. B meson production was reduced, so that wouldn't have worked either. But remembering that, Raimondi’s concept was to use techniques that had been developed to make very small beams and put them into circular accelerators. That idea had been used for synchrotron radiation light sources. The accelerator at SPEAR that I had worked on became a synchrotron light source, making the beams very small, so you had a very high brightness X-ray source. Other purpose-built accelerators of that kind were all over the world by then. So, people knew how to do that. Raimondi's idea was to build two small storage rings that used very, very small beams to make high luminosity without terribly high currents. In fact, the optimal way to do that was with rings that were much smaller diameter than the rings at SLAC or the rings at KEK. That seemed like a really interesting idea. We weren't getting very far with the U.S. It was a new idea, so we all made a turn and said, well, let's work on that idea and the detector that would be suitable for that regime, which in fact would be easier to build than the one for the brute force accelerator, and let's see how far we can go with it.

So, we worked on it. This was the concept that we called SuperB. That was Marcello's idea, I think. It was a fine name and we liked it. It has a problem, however. When you Google it, you get the word superb. But other than that, it was a perfectly good name. We worked hard on SuperB. We got ourselves organized; we got involved with the Italian funding agency INFN, which at the time was run by Roberto Petronzio. He was very much on board about this. The idea was we would do this at the Frascati Laboratory, the INFN laboratory at Frascati, which had a small accelerator that ran at 1 GeV, producing pairs of kaons. We would do it there. There was a question of whether you could do it on the site or do it on a neighboring site that belonged to the second Rome university, which was very close to the laboratory. We went back and forth on that, making two designs. The staff of the laboratory, interestingly enough, was not terribly supportive of doing this, for their own reasons. And the director wasn't terribly supportive. But the head of the funding agency, Petronzio, was. So, they did a lot of the requisite Italian science funding policy work, and in 2009, in the wake of the 2008 downturn, there was a stimulus funding exercise in Italy, as there was in the U.S. Included in that stimulus funding was support for SuperB, to the tune of €250 million. That wasn't enough to do the job. We knew that and they knew that, but the idea was to get started and then come back when we had a full cost estimate and talk about the rest.

So, we set to work on doing that, forming a real experimental collaboration, designing the detector, optimizing the accelerator, improving it and improving it. Raimondi came up with good ideas, including something that technically was very innovative, called a crab waist, which was a remarkable improvement in accelerator dynamics, and allowed us really to really be quite sure we could do this without having very high current beams. We were all set to go, we thought. We started trying to form collaborations; we tried to get money from the U.S. for what was now an Italian project, but we would get what we could for support of the U.S. participation in it. In the end, we did not get that. As we made progress over a couple years, we then had a better cost estimate which was more like €500 million. So, the funding agency went back to the government for the rest of the money. And at that point, the government had changed, and they said, "Well, you can't have any more money. You can have the €250 million if you can do it for that, but you can't have a penny more." So, we went back and seriously looked at that. What could you do? What corners could you cut? But we just couldn't do it. You just couldn't squeeze things by a factor of two. So, we adopted a fallback position. We said, "Well, with €250 million, using the same techniques, we can build a really good tau charm factory." A lot of us weren't that thrilled because we didn't think that physics was as interesting as what we wanted to do, but we gave it a shot and then we decided that wasn't really viable either. So, after a couple years of trying, we just gave up.

So, we basically disbanded the collaboration and went on to other things. The Japanese adopted some of these ideas. They were stuck with the big ring. We had optimized the size of the ring at something much smaller. The PEP ring was about 2.2 km in circumference. The Japanese one is about 3.1, much larger than optimal. The design we had optimized for SuperB was something about 0.9 km in diameter. That has a whole bunch of advantages, including making it cheaper because it's smaller, but also a lot of advantages in beam dynamics. In Japan, getting civil construction is much more difficult because of the way they're organized than building apparatus that fits into an existing piece of civil construction. So, they decided they had to do this within this big tunnel. And using some of Raimondi's ideas, they designed an accelerator to do that. They were working simultaneously with us. We're very competitive but we're also very collegial, so we exchanged information. We had joint workshops. Some of them we made in Hawaii because they were halfway between, and everybody likes going to Hawaii. We had workshops in the U.S. on the west coast, we had them in Italy, we had them in Japan. So, we were very much in contact.

We had a lot of qualms about their design, largely because of some of the inevitable consequences of trying to do this technique in a very large ring. They didn't agree with us, but we made our concerns known and they took some of it into consideration. When SuperB failed, our group basically scattered to the four winds. Frank and I, because of these qualms, decided that we would—typically, when these things happen, the losers are free under certain circumstances, assuming they haven't been fighting, to join the winners. So, some people from the BABAR collaboration and the SuperB collaboration actually went to work on the Japanese experiment, which is called SuperKEKB and Belle II. They are running now, but we didn't join that. Frank and I decided to join an experiment called Mu2E at Fermilab, which I can talk about briefly in a little bit. Others joined an experiment at CERN, and some number of groups, including a few U.S. groups, joined the Japanese experiment. DOE was willing to fund a small number of groups to do a little bit there. They've been running at various levels for a couple of years now. They have now reached the same luminosity, the same intensity, that they had when they shut off KEKB, back in 2011, or something like that. Their goal was to get to not quite 10³⁶, but something like 5 x10³⁵. They're not there yet; they have a long way to go. Their technique, in order to get to higher luminosities, they have to shut down the machine and add RF stations, which are quite expensive things and require you to rebuild a lot. So, their plan to get to higher luminosity requires run, stop and build, run, stop and build, which makes it very hard to accumulate a lot of data in a short period of time. I have always regarded that as something of a flawed plan, which is one reason I wasn't enthusiastic. The other thing that's hurting them is they have a lot of backgrounds, particularly one background called Touschek scattering background. Our work on SuperB was largely concerned with working on controlling Touschek backgrounds. That was one of the main things that we kept saying in your design in the big ring is going to be a big problem. And in fact, it's a big problem. It's those kinds of backgrounds which are making it difficult for them to get up to higher luminosities. The accelerator is capable of doing so, but only while producing so much background that the experiment can't run. So, it's been a real problem for them. It doesn't mean they can't solve the problem. I hope they will solve the problem. But they haven't really yet. They are working diligently to bring things up to speed to get to accumulating this much larger dataset, but they've made rather slow progress at it. So, the dataset they have accumulated now is about the equivalent of what we got in the first year of running PEP-II and BABAR in 1999 and 2000. That doesn't do anything new, since both their older experiment and BABAR have 15-20 times more data than that now in the can. The assumption is that over time they will get to where they want to go, but I decided I didn't want to go on that journey, and that it would be best to turn to something different.

That something is Mu2e. As I've told you, the name of the game is to find physics beyond the Standard Model. We have measured CP violation nine ways from Sunday in the B system, and we have found no evidence of physics beyond the Standard Model. There are modes of B decay that are more sensitive to physics beyond the Standard Model, but they require a great deal more luminosity to study; they require a SuperB factory level of luminosity. So, that's what SuperB and SuperKEKB are about. These are rather small effects; the effects that they produce are governed by the scale of the new physics. At the time we were working on all this stuff, from 2000 to 2011 or so, we didn't know anything about the scale of new physics. Now, from the LHC results, we do. We don't know in detail anything about the new physics, but we do know if it exists, that it has a mass scale which is higher than the SUSY proponents, for example, had hoped. Typically, the mass limits on various new particles which could be involved is up above a TeV, or so. If you take a TeV scale, and ask how big could these effects be in B physics decays? The answer is, they're really small. So, at least in my book, the situation has changed. Being able to make those kinds of CP violation measurements at a Super B factory, knowing what we know now about the mass scale of new physics, is now very problematic. Assuming they get the accelerator working, and they take all the promised data, and they make these measurements of a very small asymmetry, it's very unlikely to be affected by new physics at a scale that can be measured. I could be wrong about that; I hope I'm wrong about that, but my best physics judgment is that CP violation effects due to new physics in the B sector are out of the experimental reach of Super B factories. It was that kind of consideration which led me to think that maybe I don't want to do that anymore, that I would like to find a sharper tool to look for new physics. Now, I should put a caveat—there are, in the BABAR data, in data from the LHCb experiment at CERN, some indications of perhaps violations of the standard model in B physics. Not in CP violation in B physics, which is what I was talking about, but in other areas. In results involving lepton universality. In other words, the decay of B meson to tau, mu, or electron in a final state, along with other things, should obey the Standard Model rules called lepton universality. We should know exactly what those ratios should be. In some of their measurements, that's not the case. But the effects are small. They're 2 sigma kind of effects; that is, they're not fully established. But they persist as they get better and better data, but they're still never really statistically significant. And the BABAR and Belle data tends to be more in agreement with the Standard Model, but not very good either. So, whether, in fact, that is evidence of new physics in the non-CP area, remains to be seen. Jury's still out. But again, I thought it would be important to look for a sharper tool.

Zierler:

Dave, what does it mean, a sharper tool? What specifically are you thinking about?

Hitlin:

An experiment that has sensitivity to effects due to even higher mass scales, like Mu2e. Mu2e is basically a single purpose experiment. It's aimed at finding charged lepton flavor violation, not at a wide spectrum of areas, as with BABAR. When you look at how new physics could manifest itself in this area, the scales to which it's sensitive are much higher than those that B physics and charm physics and the things we've been doing are sensitive to. It can find effects if the scale of new physics is 1000 TeV. So, you have this much greater reach in the scale of new physics if you look for this non-standard model effect in muon decays. The reason for that is not magic. It has to do with the fact that it's easy to make a muon, which are copiously produced in pion decay. It's very easy to make lots of pions if you have intense proton accelerators. You can therefore study a much larger sample. You can accumulate so many muons that you can look at very, very, very tiny effects. The effects we're looking for are a part in 10¹⁶. In fact, in order to get to this 10¹⁶ level, you need to look at least at 10¹⁶ muons to find a single example. We stop 10¹⁰ muons in our experiment every second. A year has p x10⁷ seconds. So, if a single experiment were to run for a year, it can have more than 10¹⁷ muons to measure. Accounting for efficiency effects, it can have a sensitivity in this range, 10^-16 to 10^-17. The magic is due to the fact that you can look at so many more muons. We make and look at more muons in a day than SuperB can study in a decade, for example. It's as simple as that. So, Frank and I decided to go in that direction and join a just forming Fermilab collaboration.

There are different ways of looking at these effects in muon physics. We have chosen the one that's called muon to electron conversion. That is, you bring a muon into the field of a nucleus, and it turns into an electron. Now, muons turn into electrons all the time; they decay to an electron and two neutrinos. I talked a little bit about that early on. That's allowed by the Standard Model, but for a muon to become an electron with no associated neutrinos violates flavor lepton conservation. So, that's the experiment. People have been doing these searches for many years, even back in the days of cosmic rays, when you only found muons in cosmic rays. The current experiments have a level of sensitivity for muon to electron conversion of about 10^-12. In order to do better, some Russians had an idea back in the mid '80s and designed an experiment that was supposed to run at a Russian accelerator at the time that was never built. The idea was resurrected in the United States and the host for it was to be Brookhaven National Lab, which has a 30 GeV proton accelerator, which goes back to the '60s, but is still used as an injector to the heavy ion collider. The idea then was to build this in the style of this Russian experiment, a very sensitive experiment for Μ to e conversion, and it became part of a package of rare decay experiments which was called RSVP which was to happen at Brookhaven. Eventually RSVP became too expensive for something to do at Brookhaven, and it was cut off. The whole program was just stopped. Again, it was part of this business of where the operating funds for the LHC are going to come from. And the other place they were going to come from was residual physics at Brookhaven.

So, a new collaboration formed, built around exactly the same idea, but to be done at Fermilab. That's what became the Mu2e collaboration. So, it was forming with a nucleus of people who had been part of the RSVP work, in the late aughts and the early teens of this century. Frank and I started looking at this, and we decided it was interesting physics, number one, and number two, something we could contribute expertise to. I don't like to just walk into an experiment and say, "Here I am. Give me some data to analyze." I like to build the things I then work on, and then do the physics. That used to be the standard method of operation. It no longer is for a lot of people, but it is for me. So, we wanted to see what we could contribute to this experiment in early phases, and we have a lot of expertise in calorimetry in general. Not only liquid argon calorimetry, but also measurements using scintillating crystal. BABAR's calorimeter was a set of scintillating crystals which were made using thallium-doped cesium iodide. A good deal of the early development work for the BABAR calorimeter was done at Caltech. We actually didn't then participate in the building of that in BABAR, because I had my hands full as the spokesperson, directing the project. But we did a lot of the early development work. We also needed a calorimeter like that for SuperB, but it needed to have much higher speed.