Sam M. Austin

Kao:

It’s September 21, 2010. This is Philip Kao, and I’m having a conversation/interview with Professor Sam Austin of Michigan State University. We’re here to talk a little bit about the history of the MSU Cyclotron Laboratory and some of his reflections. I guess I’ll begin by asking how you came to the NSCL, the National Superconducting Cyclotron Laboratory.

Austin:

I got my degree in nuclear physics at the University of Wisconsin in 1960, went as a postdoc for a year at Oxford University in the UK, and then was at Stanford as an assistant professor for four years. Stanford rarely kept its junior faculty, so I had to look for a more permanent job; SUNY in Stony Brook, Washington University in Saint Louis, the University of Pittsburgh, Penn State and MSU made offers. Most of these had established nuclear physics programs, but MSU was a relatively new institution and had no reputation in nuclear physics. My choice to come to MSU was just a gut reaction. The MSU lab had energy and was going to improve, so with hard work, you could make something you could call your own, and not just add to somebody else’s accomplishment. It was probably not a particularly rational decision, or anyway not arrived at rationally, but for me it ended up being the right decision. Of course, at that time it was not the NSCL, it was the MSU Cyclotron Laboratory — NSCL came later, a name change from the Department of Energy.

Kao:

What was your first project when you first arrived at the cyclotron? I mean, you’re working with that 50 MeV cyclotron — it’s now called the K-50. Were you an experimenter there, or were you doing theoretical work?

Austin:

I was an experimenter, and we did experiments with the K-50. Originally, the K-50 accelerated negative hydrogen ions, each a proton with two electrons attached. The ion was then extracted from the machine by running it through a thin carbon foil that knocked off the electrons. That was easy technically, but it didn’t result in the highest quality beams. We first did experiments that could use these low quality beams. And then, later, we used special apparatus to extract a beam of much higher quality. The prettier experiments we did used the special characteristics of the machine that allowed it to produce beams of particles that had very well defined energy and time. One of the first was done by a student of mine, Helmut Laumer, and a postdoc, Cary Davids, who came to MSU from Caltech. It had to do with how certain light elements, lithium, beryllium, and boron, were produced in the cosmos. They are very rare and were too fragile to be made in stars, so had to be made in cosmic rays. At least that was the guess. And to figure out if that was possible we had to know the probability of a cosmic ray proton hitting, for example, a carbon atom and making one of these light elements. We measured those probabilities, the so-called cross-sections.

Kao:

Wow. And so this early period in which you were at the Cyclotron, and you were witnessing also the building of the K-500 and the KF-1200. Was there a change in the types of experiments that you were conducting when the facility was becoming the NSCL? Was it just more of a quote-unquote powerful beam or…?

Austin:

Well, you’re skipping over quite a lot of years. The K-50 Cyclotron began running in 1965, and ran for 14 years as a proton cyclotron. Around 1975, Henry Blosser at MSU built a magnet to prove that you could produce a high quality, cyclotron size, super-conducting magnet. It worked and we received a very small amount of funds from NSF to convert that K-500 magnet to a cyclotron. We thought it would be completed around 1979 or ‘80, and so the K-50 cyclotron was turned off in mid-1979. But problems arose and the K-500 didn’t actually operate until almost three years later. That was a big problem for MSU researchers. We were extremely disappointed at that time not to have an accelerator, but it ended up being good for us. We were a little known laboratory, and, almost all of us were, personally, little known in the outside world — we hadn’t been around for long and the laboratory was just beginning to build a reputation. To continue research we had to obtain beam time at other accelerators, and do experiments on them. We did this very well. Having this time gap when there was no accelerator at MSU ended up being very good for our reputation in the long term.

Kao:

It made you work harder?

Austin:

Well, not so much that. It forced us to go away from MSU, meet new people and collaborate with them, and do different sorts of experiments. The laboratory faculty made their initial reputation to a large extent during the time the laboratory had no accelerator.

Kao:

And, on a personal note, during this year — I mean I’d skip over some years, but when you first got here and you were conducting experiments, were there any projects or things that you were doing that made you grow in new directions that you didn’t foresee doing?

Austin:

I’ve always preferred to do several different things simultaneously, partly to continue progress when one direction had one of those inevitable holdups. At first, as I mentioned, we did experiments having to do with production of light elements in the cosmos. In addition, I worked with other people, especially Gary Crawley, trying to figure out how two protons interact inside a nucleus. One learns how they interact in free space by scattering protons from protons. But if you want to understand a nuclear reaction, you need to know how they interact when they’re surrounded by other protons and neutrons, so we had to figure out a way to find out. We did a lot of experiments and really pioneered the field of determining these effective interactions in nuclear matter.

A related undertaking was good for MSU researchers and for the lab. At that time, in nuclear physics at least, very few workshops were held, whereas, in 2010, there is a workshop almost every week. Gary Crawley, Hugh McManus and I organized a workshop that was held at Gull Lake, an isolated spot in mid-Michigan. We managed to persuade all the important people in the field to come, including Hans Bethe a Nobel prize winner, and famous in nuclear physics and astrophysics. It was the first time we’d really exposed ourselves on the international scene. That was a big plus for me personally, and was good for the lab’s reputation. Prior to that time, if you went to a conference, and saw some work from MSU shown on a slide by somebody else, you’d say, “That’s really amazingly good.” It didn’t happen very often. Eventually we started to see our work quoted, something that’s crucial for any researcher. It’s a slow process. It takes a while before people know you exist and much longer to learn that you and your laboratory do high quality research.

Then, in 1975, I became interested in the Division of Nuclear Physics (DNP), a division of the American Physical Society. We arranged to have their annual DNP meeting at MSU in 1975, in the Kellogg Center. It had a large attendance and was another step in growing the reputation of the laboratory.

I also want to comment on different styles of research in nuclear physics. I know you’ve participated in some experiments here. They’ve all been group experiments, done by groups set up to work on one topic. That’s not how it was initially, because we had quite a large faculty, but only enough funding to support a few postdocs and students. It was impossible to have groups — If you wanted assistance to do experiments, you had to persuade others to work with you. This, for me, is still the ideal way to proceed: you put together a team and you do an experiment or maybe you do several — other team members are probably doing other experiments at the same time. But then, that team disperses. There were no groups, there were just floating collaborations.

Kao:

That’s very interesting what you just said. It sparks in my mind the issue — you were saying that one can prefer the style of arrangement with groups. I see positives and negatives with groups. There’s a certain cadre of people, experts in one area, and they can really develop. But, at the same time, maybe they are not as flexible, and they’re not sharing information as much or producing knowledge as much.

Austin:

It’s somewhat more complicated, I think. One problem is that it’s hard to administer a laboratory consisting of groups because each group tends to think they have a certain right to resources.

Kao:

To the equipment?

Austin:

To equipment, students, postdocs, etc. It’s against human nature for group leaders to cede resources to others, even if their group becomes less active. And that may happen for many reasons: a crucial person gets sick, or the group’s research runs into a dead end. From a lab-wide view, they’re not using their resources wisely. Consequently, it takes strong management to make sure that people stay active or to refocus resources on the most important experiments, especially, when there aren’t enough resources. Not having groups is advantageous from that point of view. On the other hand, some types of experiments just need a group. Certain experiments with heavy ions, for example, generate huge amounts of data. It might take a student or a postdoc two or three years just to analyze that data. You have to have a fixed group for such experiments and can’t afford to have people floating in and out leaving behind an incomplete analysis. So sometimes, groups are necessary. I, personally, don’t feel this is ideal but I also realize that at present it has to be, in most cases, because experiments are more complicated.

Kao:

Yes, they are. A little bit more on the earlier period. I know that getting funding, having the K-50 run, and building that first K-500, were a huge effort. For this laboratory, in this era, this was a big innovation in terms of the actual facilities. What other innovations were being, you’d say, produced in the early years, that were quite significant, that still could use reanalysis, could use a revisit? I’m sure lots of data were being generated, lots of procedures or processes put into place. What were some of the innovations you would chalk up from these earlier periods of time?

Austin:

The lab has always had a tendency to develop experimental equipment; it led to having niche areas you could concentrate on and have an edge on other laboratories. If you’re a new lab, having an edge lets you develop more quickly. Our’s has often been the development of spectrometers. We’ve pushed the advantage of being able to measure with very high resolution, so as to isolate different kinds of events accurately. We had a sequence of three of these devices. Two were built here and one, the first one, was purchased. Those built here relied mainly on the concepts of Jerry Nolen and Brad Sherrill. One of the spectrometers, no longer existing, the S-320, was the first spectrometer we built, mostly out of spare parts from other labs. The second, the S-800 is still being used and has become the workhorse of the laboratory.

Those were the most important single developments. But generally, there’s been a lot of apparatus development, some of it straightforward and simple, some not. We were, for example, the first to use timing and energy detectors to measure the masses of particles that had very low energies. This was necessary for the cosmic ray experiments and has become a standard practice now. The 4pi detector was a state of the art detector, and was used for a large number of experiments with heavy ions at MSU — it could detect both very light and very heavy ions emitted in any direction from the target. Another complicated but portable detector called the mini-ball, was also widely used for studying heavy ion reactions. Those devices defined the state of the art at the time.

We were also quite advanced in using computers for data taking and analysis. We had one of the very first time-shared computers, which was developed by Jack Kane, Walter Benenson and two graduate students. We were usually at the state of the art of computer systems, and we still are to a major extent. In the early days what was, typically, needed to do an experiment was a set of one or two detectors plus an electronic multi-channel analyzer to sort events by their energy. To study more complicated phenomena, bigger arrays of detectors were necessary, and the means to deal with the signals from a hundred or several hundred detectors. The arrays produced large amounts of data in digital form. In 1965 data was recorded on punched paper tape as the standard medium. Then, came punch cards, then magnetic tape, and then hard discs. There’s also been a big progression in the way to deal with the data before it is recorded. If complicated data is arriving at a high rate, it’s impossible record all of it. One can’t write to tape quickly enough and even hard discs can’t spin fast enough. So you have to do sorts and throw away most of the data before you record it. Designing the filters, called triggers, to throw away most of the data, but not the good data, is a major part of the present day experiments.

Kao:

Yeah, I recall observing some experiments, and you’re looking at the coincidences for example. With the onset of computing when you first came and during the experiments with the 50 MeV; were computers already being used?

Austin:

They were being used but in very simple ways. They were probably much less capable than any smart phone these days.

Kao:

I remember our discussion the first time I met you in person. You are talking a little bit — and this may be switching subject matter a bit. You had talked about the black box model of development, where experimenters do not understand all equipment in detail. The equipment has been designed for a particular group and they know how to use it. Then a new group of researchers comes in, uses that same equipment yet may not have an intricate understanding of the equipment. So, we have this black box design that’s pushing experiments in certain way, and then, there is a need of paradigm shift or perhaps to create another black box. Am I capturing this correctly, or you can expand on this?

Austin:

Maybe I’ll say a little about the historical development at MSU to try to clarify this. When we first began here with the K-50, an experimenter had to do everything, including all maintenance that didn’t involve major changes. Experimenters had, for example, to change the extraction apparatus when it failed, had to change the sources of ions, and had to operate the machine, tune it and send the beam down the beam line. This was the standard practice, and every experimenter did it.

But, as things got more complicated, it became clear that very few people could do it all. And as soon as there was nobody who could handle the whole apparatus, one had to assign part of it to specialists and say. “Okay, you specialists produce the beam at the end of the cyclotron and we’ll handle the rest. We’ll design the beam lines and we’ll take the beam to our target, and we’ll use our individual apparatus.” Well, pretty soon the beam line part of the apparatus got complicated, especially after we started making secondary beams, radioactive beams. So, again, there was nobody who could handle that and the experiments themselves. We now have specialists who handle the beam production.

And there’s also a specialist for the spectrometer, so the experimenter only handles the detectors and the targets, designs the way the experiment is done, and maybe does some of the particle optics. But as more and more of the apparatus is handled by someone else, you have to regard what comes to the experiment as coming out of a black box. You don’t know what’s in the black box, all you care about is knowing what its product is. The people who run the device that sorts out the different radioactive ions, for example, only need to know the properties of the beam the cyclotron is providing. They don’t want or have to know how it works, they just have to know what it produces. And the experimenter just has to know what comes out of the particle separator to his experiment. That’s why it’s called a black box, you just want to know the output, and don’t care how it was done.

Kao:

Right. And, who usually decides when there’s a need for a new piece of equipment — I’ve observed an experiment using a device called a kicker. Who usually decides there needs to be a development of a certain kind of equipment? Are they theoreticians, are they the engineers or are they the experimenters?

Austin:

Almost always, the experimenters.

Kao:

Always the experimenters, they’re the one who drive the requirements.

Austin:

For example, you talked about the kicker. People wanted to do experiments with radioactive beams that were relatively rich in protons, but found there were large backgrounds that prevented one from doing them. The kicker was a way eliminate the backgrounds. It would allow experimenters to do the experiments they wanted to do. The kicker was a complicated radio frequency device and our electrical engineers had to design and build it. But, the interest came from the experimenters. I suppose you might find an exception, but I can’t think of one.

Kao:

Perhaps this is another question, shifting gears just a bit and still staying on target with our conversation on history, were there times when certain decisions were taken that might have led the laboratory in a different direction? We know that the laboratory got funding for the K-500 and the K-1200, and there was an instance quite recently about building the rare isotope accelerator RIA?

Austin:

Yes, that’s correct.

Kao:

These things are already in history and these decisions and these paths have already been taken but do you remember a time in the past when a certain decision could have led to a different history? Or perhaps developments in the physics world that have driven the cyclotron laboratory in a very different direction?

Austin:

For this lab, there have been two main paradigm shifts. When we began, the heaviest ion we dealt with was a lithium ion. It’s always true that when you have a new accelerator, you do the easier things first. Then, if these experiments delineate a fruitful direction, other people do similar experiments, and all of a sudden, you have competition. It becomes much harder to do forefront experiments because other people are also doing clever things. After the cream is skimmed, the experiments get harder and harder, and often less rewarding because they don’t produce unique new results. And there were starting to be competitive facilities. After we built the K-50 cyclotron, several copies of it were built: one at Princeton, one at a NASA Laboratory, and there were other similar devices.

So, soon we decided to go into physics with heavy ions. None of us knew much about heavy ions, but we saw that that was a direction of great future interest, and we thought we could be near the forefront if we hurried. That was the impetus for the K-500, and later the K-1200. It was a complicated history, but we ended up proposing a new accelerator system to the NSF and eventually got it funded.

I suppose we could have proposed a different kind of machine. But at the time that we were thinking about what would come after the K-50, the laboratory director Henry Blosser, an accelerator physicist, became interested in trying to build a heavy ion cyclotron, and that influenced the decision; you have to have somebody to build such accelerators, you can’t buy them off the shelf. I don’t recall exactly how it all came together, but we realized that the future for us was probably not in light ions and there was someone in the lab interested in building a machine that could make heavier ions. The experimenters agreed to change their research interests because it looked like the road to the future. We could have chosen not to change, but that would have, in hindsight, led to a slow decline in productivity and influence.

For the next ten years we did primary beam experiments in which we’d accelerate a beam and do physics with the beam ions. But then it appeared that experiments with radioactive beams were going to be an important part of the future of nuclear physics and we made another paradigm shift, to radioactive beam experiments. These are secondary beam experiments — the accelerator beam hits a target, the radioactive products produced are sorted and the interesting ones are used to do experiments. These products had different properties than the original beams and made it possible to expand the range of physics we could do. We made a conscious decision to go in that direction. It’s become a laboratory paradigm, to be aware of promising new physics opportunities, and figure out how to take advantage of them. Many other laboratories haven’t made these kind of decisions and have then lost funding and been closed.

Indiana University, for example, had a higher energy light ion machine that was of good quality, was in competition with MSU for limited NSF funds, and was closed because NSF couldn’t afford to fund both IUCF and MSU. It was unfortunate that Indiana was closed; they were more productive than some others that were still operating, but NSF had to decide between them and us. The decision go in the direction of radioactive beams was a strong positive for us then, and later led to an MSU decision to compete for RIA and FRIB.

Kao:

Can you comment now a little bit on the community of the laboratory when things were just getting started? You were growing, you were attracting more and more people and yet you had to build a sense of community, or at least the community evolved out of people working, committing themselves and seeing a lot of satisfaction from the work they’re doing. Can you talk about the early years of the community of this lab? And also, I think there are some other laboratories that may be affiliated with the university but are not necessarily typically located on campus. Were there any benefits to having it be on campus?

Austin:

Yes, but I’m not sure that being right on campus is crucial. For example, MIT had an accelerator called Bates that was not on campus, but was close enough. It’s very useful to be close to campus so students, and postdocs especially, stay in contact with academia. In particle physics, postdocs and students typically move to Fermilab near Chicago or CERN in Geneva, Switzerland, and seldom see their home campus. That is not an optimal situation, it’s not as intellectually rich for students, and, it’s less fun somehow. You have no choice if you are a particle physicist, and that’s becoming more and more the case in nuclear physics. Being located at a university tends to develop interests in a broader range of science, not interested only in their own narrow area, and in collaboration. As the lab got bigger, it became harder to maintain that approach. We’ve mostly done so, but it is more difficult when you don’t know everyone well. When we were small, we all met in the hall many times a day. To maintain enough connections to stay friendly, and cooperative — that’s the crucial part.

Many people come here as users. Being located at MSU is a big advantage for MSU researchers, but it also carries responsibilities: we have to insure that the facilities serves its outside users well and reliably.

Kao:

Can you also talk a little bit about this PAC (Program Advisory Committee) process? I’m sure some of the earlier laboratories around the country weren’t what are called user facilities. Researchers and professors had their own equipment and they sort of had a monopoly. They could run whatever experiments they wanted at any time on them. It seems like with the PAC, there’s more of a process; you have to get in line. It’s now more of a democratic process.

Austin:

We were never under-subscribed, so it was never a possibility to do whenever you wanted to. We had a meeting every couple of weeks in the conference room, and we decided what would be done. It was much more flexible and less formal than it became later, but still, one rarely got all the beam time desired. Then, in the early 1980s, we became the NSCL, a real user facility. As a user facility, you’re expected to do more than just serve local researchers; you have to attract and serve outside users or you’re not doing your job. There are always two goals: to treat outside and inside users fairly and to do good physics. Once you have outside users, demand for beam time is usually much greater than the accelerator time available. For MSU, it was and is usually two or three times as much demand as time. A PAC is then needed to make sure the best physics is awarded beam time and all experimenters, inside and outside, are treated fairly.

This raises issues. If a group is productive, it’s necessary to make sure they get enough beam time to survive. Students won’t want to work in a group that doesn’t get to do experiments; if a group gets no time for two years, effectively they won’t have students. So the lab management has to try to ensure that competent experimenters are awarded time. As a result, nobody gets as much time as they want, ever. And some are not rewarded any beam time and drop out of the process here; they may go to other accelerators with less time pressure.

Proposals to the PAC must also be much more detailed. One might argue, “Well, I shouldn’t spend a month planning for this experiment because if I don’t get accelerator time awarded or the experiment doesn’t work, that’s a waste of time.” But, persuading a PAC that the physics is good greatly increases the probability of receiving time. A question arises: does this make it more difficult to do risky experiments. Ideally, PACs are wise enough to let experimenters take risks for important, if difficult, experiments and the management of the laboratory has to try to ensure that risks can be taken. PACs have a great influence. Although, PAC recommendations are only advisory, directors almost always follow what the PAC recommends. If they don’t do so, PAC members may not feel it’s worth the effort to make wise choices. And your users will think you are being unfair.

The laboratory management usually keeps some discretionary time, typically 10%, that the director can give out to encourage risky experiments, or for urgent needs that come up between PAC meetings. Almost every laboratory has this discretionary time process and the Director usually reports awarded time to the PAC. So the PAC has some secondary control but everybody realizes a laboratory has to have some flexibility. No PAC is going to object, unless a director is capricious.

Kao:

This has been a lot of this talk about process and change and PACs. Well, just right now, a little bit about the science of it. The experiments — we hear these days of this term called “research-led teaching” where people are conducting research. Maybe they’re professors or people from the physics department, they conduct research here. They learn something they may be able to institute in a textbook or in a graduate school course. Beyond just the publications, how else have you’ve seen this lab been sort of, not necessarily only at the forefront of science, but that it’s in the business of enhancing science, producing science and furthering scientific knowledge? What’s the afterlife of an experiment like? Were there a lot of publications even in the early years right after the experiments. Did all experiments lead to publications or some didn’t?

Austin:

As to research led-teaching, my feeling is that the most successful approach is to involve undergraduates in the research itself, either as part-time employees, or in special programs for the brightest. We do this a lot and many undergraduates appear as authors on research papers. As to your other point, yes some experiments failed. I’m not sure my guess of 20% or so is right, but some fraction of experiments, especially the more daring ones that could have the biggest influence, don’t necessarily work on a first try. The failure rate was larger in the earlier days of the lab because we could record only a small fraction of all events, and hoped our filters got rid of the right events. Now with the advance of computers we can record much more and do the filtering later. This avoids many errors.

One of the things that the management of this lab has always insisted on is that you publish your data, if it’s worth publishing, i.e., if the experiment worked sufficiently well. And, the PAC always requires that a proposal for a new experiment shows what happened to data from previous experiments. If it isn’t published, you’re unlikely to get more time. That’s the usual process. I don’t know if that’s what you meant.

Kao:

Yeah, that’s one aspect of the afterlife of an experiment because some people conduct an experiment, and they may just publish one paper of it. And, that paper leads to further experiments and people reference it. So, you are contributing in this dialogue.

Austin:

That usually happens. Of course, PACs are not going to give time for experiments they think are boring. One problem with the PAC competition is that sometimes there is data needed to understand other data. But, if PACs think it’s intrinsically boring, and won’t recommend beam time for it. Dealing with this issue is another use of director’s discretionary time.

Kao:

With respect to risk-taking, sometimes, people get funding for experiments that are quite risky, and may not generate results. Over the course of the years, do you think it’s harder to pursue experiments that are risky or easier? In comparison to the past, was it easier to do something riskier in the past by the nature of the small size of the lab with the fact that it was still growing and things were more uncertain?

Austin:

It’s complicated. There is a tendency to feel that the training of graduate students, part of any laboratory’s task, is often best done with useful, but somewhat straightforward experiments that students can do all by themselves, and don’t need continual help during the experiment. When I did in my thesis, I didn’t see my professor, ever. I imagine that for the experiments you observed, most of them, the experiments were far from straightforward, and the professor was there a significant fraction of time helping students decide what to do. That’s usually the case at MSU.

If plenty of time is available, proposed experiments tend to be somewhat similar to what was done in the past because they will get beam time, the experiment is pretty sure to succeed, and the results are useful. But, perhaps we’ve done more risky experiments since we’ve had PACs; if time is oversubscribed, you can more easily get time for valuable, if risky, experiments. And, as we’ve grown, and had PACs, we were forced into closer contact with other people and could better identify the crucial experiments. I think many people will tell you that is the case. Although I’m not certain these arguments are right in detail, I am certain that PACs don’t knowingly approve trivial experiments. Most PACs prefer experiments that may really have a payoff, even if they are risky, and, I think, approve more such proposals if they can. And of course, there are different attitudes for risks. Even students, whose thesis time may grow if an experiment fails, sometimes prefer to have a more important if riskier experiment.

Kao:

Yeah, that’s understandable. Well, we’re drawing to a close on our great conversation here, and I just want to leave room for any other comments that you’d just like to add at this particular time. As you know, this is a part of an ongoing set of interviews on conversations on history of the Cyclotron. Are there any other comments?

Austin:

I’ll add one comment. I’ve learned over the years is that the most important function of a leader or a director at the laboratory is the hiring of people — and it’s the most important function by far. If you hire, and promote, highly competent people, they and the laboratory will do well on the average. If you hire less competent people, over the long term the laboratory will do poorly. We’ve always had the criterion that a new hire should improve the average capability of the group.

When you go into a hiring process, often, people say, yes, they want to hire the best, but sometimes, there’s a tendency to want to remain the big frog in the pond and worry that the new hire might outshine you. Or that it would help me to hire someone who might work in my area. That has to be resisted. The people at MSU, as a whole, are among the best in the country and world. We’ve hired pretty well on the average. And for me, that’s a crucial continuing issue.

Kao:

Wow, thank you. I think that’s a great learning lesson for all organizations actually, not just laboratories. I have one more quick question for you, my last question, unless you have something else. Definitely, I want to give you the last word but something came up to my mind. You talk about laboratory cultures that, if you were to have a postdoc come here, we do things differently here. This is East Lansing. This is Michigan State University. Can you talk a little bit about if there is a particular culture at this facility that might be different than if somebody wants to go off to Fermi or CERN or something like that? Have you seen any changes in this culture, if there is a particular collective mentality or mindset here? What sets this laboratory apart, besides its niche in detectors?

Austin:

I think that the Cyclotron Laboratory’s tradition of thinking ahead, trying to understand what is the best future physics, and making sure there are the resources to do it is a particularly crucial part of the culture. A close interaction of experiment and theory has helped to make these choices. Another is the tradition of putting the success of the laboratory ahead of one’s own success, knowing that the former helps insure the latter. We’ve made a real push to help our young people succeed. Most of the earlier faculty in the laboratory didn’t have this advantage because none of us were particularly well known, and there were no senior laboratory members to push us forward. Perhaps because of this, we’ve tried very hard to give our young people a leg up, to make sure they have the resources they need for their experiments, and are chosen to describe them in invited talks. It’s not that we want people to succeed who aren’t talented. But, when we’ve attracted somebody who is, we want them to succeed as quickly as possible. It’s also an advantage to the lab, since they can attract better students and, especially, postdocs. And more funding. We try to help make it happen. I doubt that we’re unique in this, but I think we do it more than most.

[Some of the material in this interview is described more formally in a book (Sam M. Austin, Up from Nothing: The Michigan State University Cyclotron Laboratory, Michigan State University Press (January 1, 2016), and in an article (The Michigan State University Cyclotron Laboratory: Its Early Years, Sam M. Austin, Physics in Perspective 17, 4, p. 298-333, January 1, 2016)]