Carlton Caves

David Zierler:

Okay, this is David Zierler, oral historian for the American Institute of Physics. It is July 16th, 2020. It is my great pleasure to be here with Professor Carlton Caves. Carlton, thank you so much for joining me today.

Carlton Caves:

It is great to be here, that is, at my house, but I would like to get back to a world where I can get out of the house.

Zierler:

Alright, so to start, can you tell me your title and institutional affiliation?

Caves:

Currently, I am a Distinguished Professor Emeritus and a Research Professor of Physics and Astronomy at the University of New Mexico. I am associated with UNM’s Center for Quantum Information and Control—the acronym CQuIC is pronounced “see-quick”—of which I was Director until I retired from teaching and administration at the end of June 2018.

Zierler:

And when did you attain both the Distinguished Professor and Research Professor titles?

Caves:

I was a Distinguished Professor—I became a Distinguished Professor in 2006—when I retired. Before that, I was a Professor at UNM, from the time I moved to New Mexico in 1992. When I retired, I was made a Research Professor, basically so I can be paid a tiny amount.

Zierler:

How exclusive is the Distinguished Professor designation? Roughly, how many professors on campus have that title?

Caves:

Oh, roughly 40 on the Albuquerque main campus. It is more or less the equivalent of having an endowed chair at a university that has enough money to endow those chairs.

Zierler:

Okay. Good. Well, let’s take it back to the beginning. Let’s start, Carlton, with your parents. Tell me a little bit about your parents and where they are from.

Caves:

My father is Morris Caves (1923–2006), and my mother is Mary (Buddrus) Caves (1924–2007). Dad grew up on a farm in southwestern Oklahoma, about a mile west of a little town called Blair. Farm is still there, but no longer in the family. Dad went to college at Oklahoma State University (then Oklahoma A&M) and then to the University of Oklahoma. Toward the end of World War II, he was commissioned in the Navy and was for a short time in the Philippines, but actually after the conflict had ended. Dad met my mother at A&M, and they got married in November 1945, just before Dad was sent to the Philippines. Mother was born in Muskogee, where I was born and grew up. Her father, Curt Buddrus (1894–1981), started a company making fans in Muskogee in 1939. After the war my father threw in with Curt and his son, my Uncle Ed, and several other people to run that firm. It’s still there on the east side of Muskogee, churning out fans, run by my cousin.

Zierler:

Wow, fans still made in the U.S.A.?

Caves:

Yeah, I don’t keep up with it at all, but I think at one point they had the Google contract for server farms—maybe still do.

Zierler:

Oh, wow.

Caves:

What they did when I was a kid was that they pioneered the market for making fans for large chicken houses. Every time it got to 100 degrees in Arkansas and Oklahoma, the chickens had problems, and they sold fans like crazy.

Zierler:

Right.

Caves:

For me it was a big deal that the family ended up in eastern Oklahoma. It might look to you—really, to anyone—as though there’s no difference between eastern and western Oklahoma. But at the time eastern Oklahoma was the better place to be. So I call it my dad’s greatest decision to move to eastern Oklahoma and throw in with my maternal grandfather. More open to the world and less susceptible to a fundamentalist brand of Christianity than western Oklahoma. I call it my dad’s greatest decision, but I am thinking now that maybe my mother didn’t want to subject herself to western Oklahoma.

My parents—it sounds self-serving, since they loved all their kids—but particularly my mother knew that she had something special in me. She was dedicated to my doing whatever I needed to do to fulfill my potential, not that she or Dad really understood what that needed to be, but they were willing to encourage me in whatever I wanted to do. My dad loved me, too, of course, but I was closer to my mom. The main things I took away from Dad were, “Life is hard work—work it out!” and “If a job is worth doing, it’s worth doing right.” That latter has informed my entire career.

Zierler:

Sure. Carlton, how did your parents fare during their childhood during the Great Depression?

Caves:

You know, it was tough out there in western Oklahoma where my dad was.

Zierler:

Right.

Caves:

That was not only the time of the Depression, but also the Dust Bowl.

Zierler:

Right.

Caves:

The heart of the Dust Bowl was actually in the Texas and Oklahoma panhandles, but certainly they were affected by it in all of Oklahoma and, especially, western Oklahoma.

Zierler:

Do you know if they ever considered, like so many other families, moving out West during that time?

Caves:

I don’t think so. My paternal grandfather, Fred Caves (1893–1987), had moved with his family from Iowa to western Oklahoma before World War I. Fred always said it was a pretty dumb move for a farmer to go from Iowa to western Oklahoma, but there they were. They had a substantial farm, roughly half a square mile … no, 440 acres, so more than half a square mile. My grandfather Fred was … a distant person, but my grandmother, Grace (Morris) Caves (1893–1967), originally from Kentucky, was the best grandmother you could ever have, wise and warm and enveloping in a way that touches me even now. My grandfather, well, was a very—he was a person totally inside himself. Didn’t really relate to other people. It took me years to get straight that Fred was my grandfather, since Grace always called him Daddy. But, you know, I’m sure it was tough out there on the plains. Fred occasionally let slip that he had aspired to be a historian, but life presented no such opportunity, and he ended up running the farm. I’m guessing Dad definitely wanted to get out—he grew a large vegetable garden his entire adult life, as a green affirmation of life, a stark contrast with the brown plains where he grew up (eastern Oklahoma has more than twice the annual rainfall of western Oklahoma)—and to get out, he went to college. When the chance came not to be stuck running the farm, but rather to take advantage of his engineering degree and become the chief engineer at the fan company, he took it.

Zierler:

Was your mom able to go to college as well?

Caves:

My mother went to Oklahoma State University, where she met Dad. It was not common for anyone, especially women, to go to college then. It shows her parents had expectations and wanted her to go to college. So, she did. I think the Depression was a tough time for my maternal grandparents—my grandfather Curt Buddrus and my grandmother Doris (Devalon) Buddrus (1894–1990). But they got through it, and then Curt started the fan company in 1939. And, really, that was quite a successful firm. They weren’t rich, but after a while, they made good money. Ultimately the company had hundreds of employees, and at some point—let’s say the ‘80s—its operating revenues were about 50 million dollars a year. So, it was a big deal.

Zierler:

Carlton, anything further you want to say about your maternal grandparents?

Caves:

Curt and Doris were both Muskogeeans. Curt was the wild card of the family, irreligious, blasphemous, oblivious of society’s pressures, living life exactly the way he wanted, traveling to northwestern Arkansas to buy liquor, Oklahoma then being a dry state. To the extent that independence survives in my siblings and me and our descendants, Curt has to be a prime source. Curt always expressed contempt for the Okies who left for California during the Depression. Doris seemed pretty conventional to me and was one of the sweetest persons you could ever meet, but she participated in all of Curt’s schemes, wild and otherwise, including founding the fan company and accompanying him on long auto trips all over the place.

Zierler:

Carlton, did you live in Oklahoma through high school?

Caves:

I was born in Muskogee, Oklahoma, in 1950 and lived there ‘til I went to Rice University as an undergraduate in 1968.

Zierler:

And it was public schools throughout, for you?

Caves:

I went to public schools throughout. There really weren’t any private alternatives in Muskogee, and my parents wouldn’t have considered it anyway. I think they were pretty committed to—it just wouldn’t have occurred to them to send us to private school.

Zierler:

Was church part of your upbringing at all?

Caves:

Yeah, my parents were pretty strong Christians. They were prominent members of the First United Methodist Church on the east side of Muskogee. Church was a big part of our life. To a considerable extent, my social group in high school was the people I knew from church. Having said that, I converted to Judaism in 1989, and now I’m a Jewish atheist.

Zierler:

[laugh]

Caves:

Easier to do, the atheism, that is, if you’re a Jew than if you’re a Christian.

Zierler:

[laugh] Well, I’ll make a note of that. We’ll have to get to that when we work up to 1989. Carlton, I’m curious, did you start expressing an interest in the natural world and mathematical concepts, maybe even before you were exposed to these things formally in your education?

Caves:

You know, I clearly was interested in science, but I wouldn’t say that interest was so much in the hard sciences. That is, the mathematical sciences. I can remember reading a Time-Life book about cosmology sometime in the early ‘60s and just being blown away by the whole idea that scientists are addressing questions that I read about in science fiction and this is what I’m interested in. For kids who grew up, for scientists who grew up in the ‘50s and ‘60s, you can’t avoid the influence of the Space Age and science fiction as a big influence in where your interests went. Because this seemed to be looking at big questions that were profoundly interesting, and science seemed to be addressing them. That was a big influence on me, and I think the way you can tell a good scientist from those just interested in the science fiction—the ones who succeeded from the ones that didn’t—is that when those just interested in science fiction started doing science, they found out there is a lot of tedium underneath addressing those big questions. And they didn’t want to do that tedium. Those were the people who dropped out. And those of us who found out that the tedium’s part of it, so [laugh] you really have to be interested in doing that, too, well, we’re the ones who ended up being real scientists. And that’s tedium either in the lab for experimental physicists or doing tedious, thorough analyses for theorists. If you can’t get into that tedium, if you can’t find that rewarding, then you’re not going to be a scientist.

But what I actually did as a kid was I collected insects. I can look at that now and say, “That’s the same thing.” The urge to categorize and find patterns, but it’s more than that. I was really interested in the natural world. I collected tadpoles and had them grow into frogs in the house, and I collected and displayed insects. I got into collecting and displaying insects because butterflies are flashy, but if you’re really into this, you’re going to end up collecting all kinds of insects, including the ones that sting you and bite you and the ones that aren’t pretty at all. So really, it’s the same thing. If you’re only interested in flashy stuff, you’re not really going to do it. But I was pretty serious about it, you know, serious in a kid’s way. So, that’s what I did. I wouldn’t say I was any good at conventional ideas of kids’ science … you know, I didn’t do big science fair projects.

Actually, the thing I was most interested in in high school was speech and debate, and that was how I … so, there was a cohort of people in my class. My debate partner was generally Mike Synar, who was later the Congressperson from that district in Oklahoma. And the others were Elizabeth Williams, Joe Lunn, and Shari Temple. And that was just a great crew. I said before that my social group was the church group, but that’s only half the story. The other social group I had was the high school debate students. This just opened your mind, first, to being logical, to making arguments and to being serious about learning about things and, second, to there being a much bigger world out there beyond Muskogee, Oklahoma, to things that are outside the experience you would generally have in Oklahoma. We had a great debate coach, Orville Eaton, later mayor of Muskogee. Orville was an odd duck. Didn’t really teach us much of anything, but somehow inspired us to teach ourselves, to learn how to be high school debaters and extemporaneous speakers on our own. Not actually teaching us [laugh], instead inspiring us in an undefinable way.

We might have learned more with someone else, but this way we owned it. I think all my debate compatriots would say that. When I was in high school, Mike and I—so I should say Mike Synar is the person who got Elizabeth Warren into politics. And how did he do that? Because he knew her from high school debate. We debated Liz (then Liz Herring) and Karl Johnson (both two years ahead of us, so I’m sure we lost), who were from Northwest Classen High School in Oklahoma City. And for us, Oklahoma City was, “Oh, God! That’s the big city!”—even though by current standards, OKC was then just a big town. Just coming into contact with these obviously serious, capable, intelligent kids from all over Oklahoma—such a huge influence. That you can set your sights and do something, and you can get beyond Oklahoma. I really needed to be pushed on that, because I hated going anywhere when I was a kid and I hated being away from home. I was a real homebody. It was just a fortunate thing that I decided to go to Rice University. And I think high school debate had a big part of that: “You can do it, Carl.” These days my parents would be pushing me—should have been pushing me—to go to Harvard or Caltech or MIT. But, actually, I’m thankful I went to Rice.

Zierler:

Carlton, when you were thinking about colleges, were you considering physics, specifically? Were you thinking about being a physics major right from the beginning, or did that come later on?

Caves:

I’ll tell you that in a minute when we get to Rice. No, I knew I wanted to be a scientist and I wanted to go to a place where I could get a good science education. And as I said, I didn’t have parents who would say, “Well, what you should do is go to Caltech or MIT or Harvard.” Rice, in Houston, was the safe alternative, both for me and my parents, not so far from home—you could drive there in a (long) day and, culturally, not so different from Oklahoma—a place to go that had a reputation as an institution where you could get a good science education. And my cousin, Lee Buddrus, had gone there two years before, so I knew about it and felt comfortable going there. So I really worked hard to get into Rice, and I did. That’s how I got out of Oklahoma.

Zierler:

[laugh]

Caves:

[laugh] I had more ambitions by the time I left Rice.

Zierler:

Had you traveled beyond Oklahoma before you got to Rice?

Caves:

You know, when I was a kid, my dad thought he was indispensable at the fan company. It was extremely difficult to get him to—

Zierler:

Vacation.

Caves:

—take a vacation. Our typical vacation was—my family, together with my maternal grandparents and my uncle’s family, owned a house on a reservoir, Ft. Gibson Lake, near Muskogee. These days you would say it was suburban because you could drive to Muskogee in 20 minutes. We’d go out there for two to three weeks with my parents’ best friends, the Vaughts. Dad would commute to work, and we—my brother Doug, sister Linda, and I, and the two Vaught kids, Larry and Donna—would spend the entire time swimming in the lake … well, I did a little insect collecting in the breaks between swimming sessions. The swimming was the whole point of it. Oklahoma summers are hot; if you’re not swimming, you’re boiling. [laugh]

I think we only made two trips, two real summer vacations, when I was a kid, again with the Vaughts, both to the Texas Gulf coast—we would stay in a beach house on Padre Island near Corpus Christi. Again, swimming, this time in the surf, was the theme, but my passion for insect collecting morphed there into collecting shells.

And then, one very important family trip—my grandfather Curt was a tremendous traveler—beginning in the ‘50s, when he didn’t really have to work much anymore, he just traveled with Doris. Almost entirely by car, an exception being that Curt was an early discoverer of Cozumel as a fishing destination. He and Doris would get into their Lincoln Continental—Curt liked to buy a new one every year—and, well, drive to anywhere in North America, mainly the southern part. I remember Curt coming back from Southern California and telling us, “They have these freeways out there that go on forever, and, dammit, you can drive as fast as you want”—a big thing for Curt—"and you can pass people without having to worry about running into oncoming traffic.” And we all thought, “Well, now, that sounds pretty exciting.” Probably how I got my first hankering to be a western American.

So Curt insisted, once, that we all, including Uncle Ed’s family, go on a trip with him and Doris. We drove to Mazatlán on the Pacific coast of Mexico—took four days—over Christmas and New Year’s at the end of 1962. And so he and everybody else who wanted to could go deep sea fishing for marlin and swordfish. Now, I had zero interest in deep sea fishing, so I was down there … well, two things came out of that trip for me. One, you can swim in the ocean at Christmas. [laugh] Wouldn’t I prefer to be living in a place like that? And, two, you catch edible ocean fish accidentally—at least, those who went fishing did—a restaurant cooks them for you, and you eat them. And, gee, these are actually good to eat—called dolphin locally, but as I didn’t discover for many years, mahi mahi everywhere else—not at all like the freshwater fish we’d catch in the lakes in Oklahoma. Those were the two things. [laugh]

Zierler:

[laugh]

Caves:

I was 12 years old when we drove down there. It was an eye opener for me to see something completely different … a place where people lived a very different way … yeah.

Zierler:

So, when you got to Rice, when did you decide to focus on physics?

Caves:

Well, that’s an easy story. I really didn’t know what I wanted to do when I went to Rice. I thought I wanted to be a scientist, but I didn’t really know what it meant to be a scientist. Being at Rice was the first time that I encountered actual scientists. I decided to become a physicist because of the introductory course in physics, the first semester, for which the lecturer was Harold Rorschach. Rorschach was a genius at lecturing. He was a great scientist, a great physicist, but also a great lecturer. For me, his ability to stand at the board and extract from equations things that told you the way the world worked … just standing there. You didn’t have to go out and do an experiment yourself. These equations described the world, and you could do things with them and extract the way the world worked. It was just magical. And I thought, “That’s what I want to do! I want to be somebody who thinks about the world and figures out how it works and has the skill set to do this.”

A big thing was that Rorschach knew how to do approximations—of course, because he was a theoretical physicist. He could take an equation and say, “Okay, so what’s the answer here?” I’d look at that and think, “Geez, it’d take me half—this is just before electronic calculators—it’d take me the next century to calculate the answer. How do I do all these square roots and logarithms and so on?” But Harold would say, “I don’t really need the exact answer. All I need is an approximate answer. With the right approximation, the formula simplifies, and the answer is easy.” For me, that was magic, that he knew how to take these things and extract information from them. In a way that, geez, I would never have imagined.

Zierler:

And, Carlton, it sounds like that might have put you on an early path towards theory, also?

Caves:

Well, it did. At the time, laboratory courses in American physics departments were largely a way to weed out all the people who—well, they were a way to encourage everybody to be a theoretical physicist because they were boring [laugh], the equipment didn’t work, nothing worked. Actually, this is sort of the way experiments actually are in real life. They don’t work. But—

Zierler:

Except for the ones that do.

Caves:

[laugh] And that’s the whole point of experimental physics, to mess around until it does work. And even then, even when you think it’s working, to be incredibly cautious to check that it really is doing the right thing.

Zierler:

Right.

Caves:

This didn’t appeal to me as something I wanted to spend my life doing. The whole experience pushed me, one, to be a theoretical physicist, and the only scientists I could see were professors. So, two, I wanted to be a professor.

Zierler:

Right, right. When it came time for graduate school, where did you apply and what were your motivations?

Caves:

A big influence on me from Rice was that I took, as a senior, the graduate course in quantum mechanics that was taught by Neal Lane, also an Oklahoman. And that was sort of a big deal for me because you don’t meet many Oklahomans who are big-time physicists. And Neal was a great instructor. Even though that’s not what I intended to do when I went to graduate school—and I’ll tell you what I intended to do in a minute—even though it wasn’t what I intended to do, ultimately, it was that real—I mean, when I learn something, I learn it thoroughly, and Neal taught a course where you learned quantum physics thoroughly. That’s where my skill set was when I left Rice, and ultimately, that’s what I ended up doing, quantum mechanics in a way that people didn’t tend to do it.

What I thought I was interested in—and I really was interested in this—was relativity, which seemed to be addressing all those big questions about black holes and the history of the universe. Black holes had just been named at the time I was an undergraduate. This was between ’68 and ’72. What happens in strong gravitational fields? And what are all these puzzles about the way spacetime gets warped and perceptions of time and distance are totally different? How do you describe these objects, mathematically? That’s what I wanted to do. I had more ambitions by then, but I was pretty specific. I wanted to go to a place where you could do relativity. I was advised by Don Clayton, in the Rice Space Science department, “You should try to go either to Princeton to work with John Wheeler, or you should go to Caltech to work with Kip Thorne.”

Zierler:

Right.

Caves:

Ultimately, I got into both. I went to Caltech because Caltech offered me $500 extra on top of a $3,000 NSF graduate fellowship and because Princeton required you to wear academic gowns in the graduate college dining hall. I didn’t want to do that—at all! That’s ridiculous. This was the ‘60s after all, even at Rice, and that had changed me.

Zierler:

[laugh]

Caves:

So, $500 and the absence of gowns got me to Caltech. Caltech is and certainly was a strange place. But for me, it was terrific. [laugh] Kip, who ultimately was my PhD supervisor, is my model, my ideal of a theoretical physicist.

This is the thing. I was scared. In early May of 1972, I had just graduated from Rice. Western schools were on a different academic calendar, so Caltech on its quarter system wasn’t finishing its quarter until the middle of June. So I got to Muskogee in early May, and I certainly didn’t want to spend another summer there. So I called Kip Thorne to see if he’s got anything I can do for the summer. The odds of his actually having anything that I could do were close to zero, but anyway, I called him up. It was the boldest thing I ever did in my life. I can still remember staring at that black, rotary-dial phone in my parents’ house, daring myself to do it. Kip said—I think this is after taking a look at my graduate application and calling back—“Well, yes, I’m writing a textbook on gravitation with Charles Misner and John Wheeler. I need my parts of it proofread to make sure there aren’t any errors in it. And I’ll hire you to do that.” Geez, actual money to get out to California ASAP and start doing something. So I drove to California in June of 1972. I have to say, another reason that I went to Caltech was, at the time, who in God’s name would go to New Jersey for graduate school if they had the chance to go to California?

Zierler:

[laugh]

Caves:

[laugh] So I drove out to California. A huge shock. I can remember driving from San Diego to Los Angeles, and this huge wall of smog looms in front of me as I approach Orange County, and I was almost completely repelled by it. June 19th, 1972, I arrived in Pasadena and went to work immediately proofreading Kip’s part of Misner, Thorne, and Wheeler, and, of course, I couldn’t really proofread Kip’s chapters effectively without reading the whole thing so I understood what I was doing. That’s how I learned relativity, reading MTW as it was being written. That was [laugh] a tremendous experience. I learned such an incredible amount from that. There was one chapter, I think it’s Chapter 18, on linearized gravity, where eventually there were so many mistakes that I couldn’t figure out what Kip wanted to do. So I went to Kip and told him, “I don’t know what you want to do here because I can’t figure out—there’s so many mistakes—what you want to do and what’s right.”

Zierler:

Now would Kip have been impressed that you caught all of these mistakes, on a technical level?

Caves:

I believe he was. Although Kip, I think, later on got sort of skeptical of my abilities, since I didn’t really produce much for quite a long time, that early proofreading probably was a big deal. There’s a footnote to Chapter 26 on stellar systems that says I found and fixed numerous errors in that chapter and others.

Zierler:

This is a pretty big responsibility to give an incoming graduate student sight unseen. Did he have any recommendations from your professors at Rice attesting to your abilities or is that just the kind of person he is?

Caves:

Kip must have gone off and read the letters of recommendation from Rice. I mean, analysis is what I was good at. I wouldn’t say I was very imaginative at that point, but I was really good at analysis. A lot of that proofreading job was drudgery, like Kip’s assistant and I were in charge of copying the 3,000 pages of the manuscript, and I don’t know if you’ve ever worked on a 1972 Xerox machine, probably not, but they broke down roughly every 200 pages.

Zierler:

[laugh]

Caves:

Geez, it took us forever to do that copying! But, most of it was making my way through the thing very carefully and learning relativity, not just in a broad-brush sense, but in detail. I’ve never been any good at learning anything broad-brush. I have to learn things in detail really to know anything at all. Even now there are loads of things I know only because I did that proofreading of MTW.

Zierler:

Now, Carlton, you mentioned how Caltech was an odd place. In what ways? What struck you about its oddness, at the time and since?

Caves:

I don’t think I would ever send a kid there as an undergraduate. The competition is just too intense. But it was a great place for a graduate student, because the faculty was truly inspirational and because of the peer group of graduate students. Its features of probably not being very friendly—well, no, not probably—not being friendly to women and minorities was pretty much lost on me at the time. I wasn’t a woman or a minority, so I didn’t pay much attention to that when I was a graduate student. Now I can look back and recognize some incredible things. But for me Caltech was an ideal place to be. And I was there for a long time. I was a graduate student from ’72 to ’79, and then I was on the research faculty until the end of 1987.

Zierler:

Now, beyond the textbook, what else was Kip working on in those early years?

Caves:

The distinctive thing—it was really Wheeler who had turned general relativity from a subject that was predominantly mathematics into something that was physics. Kip worked very much in that vein and went much further than Wheeler, looking at ways to connect general relativity to physics. Kip did a lot of astrophysics. A lot of the other graduate students were doing astrophysics. He did a lot of black-hole physics. He pioneered—following Ken Nordtvedt and then with Clifford Will—the parameterized post-Newtonian approach to general relativity, which is the way you set up what would a general metric theory of gravity look like in the slow-motion approximation at post-Newtonian order? And then what predictions would it make for what happens in the solar system? That’s where tests were done, mainly, but not entirely in the solar system. But the PPN formalism allowed you to sit and say, “Okay, here is what any metric theory would predict for this experiment or this observation.” It is very hard to test a theory unless you know what the alternatives are.

Zierler:

Right.

Caves:

The PPN formalism was a way to formalize all those alternatives. You give me an alternative theory of gravity, and if it’s a metric theory—and that’s nearly required by the equivalence principle—it will have some values of the PPN parameters, so then you know the range of possible predictions, the range of things you can test. I worked on that for a while, and I also worked, with Kip and Vladimir Braginsky, on potential laboratory tests of general relativity, some of which were outside the slow-motion approximation. Sometime in the mid-70s, Kip stepped back and said, “What’s really the most important thing to do?” And that, he concluded, was to detect gravitational waves. Detecting gravitational waves from astrophysical sources would be the ultimate confirmation of Einstein’s relativity, and it would also be, as Kip came quickly to understand, a new window onto the universe. Every other time physicists had set down a detector of some kind of radiation, typically electromagnetic, no matter how rudimentary it was, “Oh yeah! The cosmos is doing that!” You didn’t have to work very hard, because the cosmos is doing it. Gravitational waves were different because they are so weak that the initial efforts using big metal-bar detectors all failed and never detected anything. We understand why now. They were just woefully too insensitive. The fact that Joe Weber claimed to have detected gravity waves and then that was later shown to be incorrect, well, that put the whole field in a bad light in the minds of many, perhaps most astrophysicists and physicists. Many concluded that gravitational-wave detection was just a huge waste of money and a place where people … well, where you can’t rely on what people reported.

This had two effects. One, funding and interest in gravitational-wave detection sort of dried up, but, two, that meant that there was a lot of time to think about what sources would be and how strong they would be and to design detectors that would have sufficient sensitivity to detect those sources. This was a quite different process than for any of the electromagnetic or even neutrino sources. You had the luxury of being able to do a lot more thinking both about sources and also about detector design and what limits sensitivity. There was, not surprisingly, complete confusion about sources at the time. The discovery of the binary pulsar was a big deal because, provided it was even remotely typical, it allowed one to guess how often per galaxy compact objects collide and coalesce. With the limited observational data and some theory, organized ultimately by Kip and others, it became possible to get a rough idea of how sensitive an Earth-based detector would have to be to detect gravitational-wave events often enough to make it worth the effort. It became clear that Rai Weiss’s (and Bob Forward’s) idea of using an interferometer was going to be a better way to look for gravitational waves than bar detectors would ever be. This led ultimately to the current LIGO and Virgo gravitational-wave interferometers.

Where I finally found my footing is when I started thinking in the mid ‘70s, at Kip’s urging, about gravitational-wave detection, either using bars or using interferometers. It was Vladimir Braginsky who pioneered thinking about the ultimate limits on detection of gravitational waves and who pointed out, both for metal-bar detectors and interferometers, that there are going to be serious quantum limitations on how well you can detect these gravitational waves. That’s what I got interested in.

For bar detectors there was the question of getting past what Braginsky called the standard quantum limit for detecting changes in the amplitude of motion of a bar detector. The bar—more precisely, a mode of oscillation of the bar—is a harmonic oscillator, and you want to look for small changes in the amplitude of its motion produced by a gravitational wave. You need to do that in the presence of thermal noise, but it took quite a long time to get this stuff straight. Eventually it was realized—this was really Gibbons and Hawking—that if you used high-Q detectors, so the damping time is very long, thermal fluctuations can be made irrelevant. It takes a damping time for thermal fluctuations to take full effect, so if you can make strong measurements on a time scale short compared to the damping time, it’s not thermal fluctuations that are important, but rather the noise—ultimately, the quantum noise—associated with measurement. Basically, the size of the ground state of a harmonic oscillator, the scale set by the zero-point noise in momentum and position. In bar detectors that wouldn't be even close to the sensitivity required to detect any waves, so it was finally realized that this zero-point noise was the issue.

That’s what we were thinking about at Caltech. I can remember, partly because I looked it up yesterday in my files, having lunch at Caltech’s Chandler Dining Hall on November 28, 1977, with Kip, Ron Drever, Vern Sandberg, and Mark Zimmermann. We were discussing a paper by Vincent Moncrief that suggested that somehow using coherent states of motion of the oscillator, Roy Glauber’s coherent states, which have only the zero-point noise, was the key to getting around the standard quantum limit. Now, Bill Unruh had already written a paper saying, “No, that’s not going to get around Braginsky’s standard quantum limit. The zero-point noise is Braginsky’s limit.” That was the point. Coherent states were the limit. Their size on phase space is the size of the ground state. But thinking about noise rotating around in phase space, as Moncrief suggested—well, that was a very productive thing, then and now. I realized during that lunch discussion, while we were sitting there, that if you measure the position very accurately at one time, you should just continue to measure the conserved quantity, now called a quadrature component, that continues to have small uncertainty as the noise rotates in phase space. Braginsky had already pointed out that if you measure the position very accurately now, then every half-cycle later the position will have small uncertainty again, and at those times you can measure position again and detect very small effects—these were called stroboscopic measurements. I realized, well, you should just measure the quadrature component, which after one measurement, continues to have small uncertainty at all times as the noise rotates around in phase space.

That lunch discussion is the source of what the group came to call back-action evasion. I went on to do a heck of a lot of work with this group on back-action evasion. Now, the words back-action evasion and quantum non-demolition have been totally bastardized in the intervening years, but the key idea is to measure something where the quantum back-action disturbance produced by the measurement now doesn’t affect future measurements of the same quantity. That was my first introduction to actually making a discovery and following up on its consequences.

[This] turned out not so important in the end, because bar detectors were on their way out, and that was my main interest in the late ‘70s and early ‘80s. But at the same time, Kip could see, encouraged by Braginsky, that bar detectors weren’t going do it. The interferometric detectors were the best way to go. Rai Weiss and Bob Forward had proposed this. Forward had done a very early experiment, and Rai had done a very important, remarkably prescient early noise analysis on how interferometers might be made into working gravitational-wave detectors. Kip had asked around and had determined that the experimenter who was thinking most imaginatively and deeply about experimental design was Ronald Drever, then at the University of Glasgow. Ron had gotten into this because his group basically made a bar detector whose motion was detected by an interferometer. There was also a German group in Garching that took Rai Weiss’s ideas and analysis and ran with it, because they had the money to do it. I knew there were a lot of fundamental questions about the noise in an interferometer. I knew, in particular, that Weiss had, in his famous quarterly progress report, basically said, “Okay, along with all the technical noise sources, there are the fundamental quantum noise sources. There’s shot noise from detecting (particulate) photons, and there’s radiation-pressure noise because when the photons hit the mirrors at the far end of the interferometer, they will bump them around.” Rai had analyzed these two effects and found that together, they gave Braginsky’s standard quantum limit for monitoring the position of a free mass. But Rai had said, “Yeah, but I don’t see how this radiation-pressure effect works, because if you put a laser beam through a beam splitter, the fluctuations will divide equally at the beam splitter, and they’ll hit the two mirrors equally and drive them around in common mode, and the whole point of an interferometer is that it isn’t sensitive to common-mode phase shifts.” I knew that was a problem, so I decided, “Yeah, let’s figure that one out.”

The answer is trivial. The electromagnetic field is particulate. When each photon comes up to the beam splitter, it’s 50:50 to go in one arm or the other arm of the interferometer. So, there will be fluctuations in which too many photons go down one arm and two few down the other arm, and that will produce differential radiation-pressure fluctuations, as opposed to common-mode fluctuations. If you think in terms of photons, which you already did for the shot noise, and you think again in terms of photons for the radiation pressure, it’s just obvious what the effect is.

I was thinking about this in the autumn of 1979 and on into the early winter. For one reason or another, partially because Rai had posed the question in terms of waves and how the waves split at the beam splitter, maybe partly because I knew that there were coherent states and that the laser light is a coherent field and is really like a classical electromagnetic field—well, for one reason or another, I decided to give a wave interpretation of the same phenomena. [laugh] I’d like to think that this is also partly a consequence of my own personal belief that if you can do a problem in theoretical physics more than one way, you are obligated to do it in every way you can think of because you never know which way will be the path forward on future problems. I’d like to think that, but it’s more likely that’s my reading into the past what I learned from this very experience. [laugh] The way the wave explanation works is that beam splitters produce differential radiation-pressure fluctuations automatically from whatever noise enters the unused port. That is, if a fluctuation from the unused port increases the amplitude of the laser beam in one arm, it must decrease the amplitude in the other arm by energy conservation. The result is to push on one mirror and pull on the other, so that’s the source of the differential radiation-pressure fluctuations.

The only puzzling thing was that the source of the fluctuations is the nothing—the vacuum—that comes in the unused port. The surprise was for quantum opticians, who had over-interpreted Glauber’s photodetection theory to mean that the vacuum never contributes to anything. I was fortunate not to know enough about quantum optics to be constrained by that. So I wrote a paper. It was published in the summer of 1980. Even now people quote the abstract because at the time I was full of the Caltech arrogance, and it was brash beyond belief, saying (I looked this up): “The interferometers now being developed to detect gravitational waves work by measuring small changes in the positions of free masses. There has been a controversy whether quantum-mechanical radiation-pressure fluctuations disturb this measurement. This Letter resolves the controversy: They do.” [laugh]

Zierler:

Was it a little more definitive than it should have been? Is that the idea?

Caves:

No, no. It was absolutely definitive. It’s just that people had been taught that the zero-point fluctuations never do anything. Once you’ve let them into the picture, where do you stop? Do they gravitate? But the point here is that the reason I didn’t end up as an astrophysicist is because in astrophysics you have to rummage around in all of physics and try to think up explanations for things. I wanted to work on things where you could take the theory, which you absolutely know is correct and derive a reliable conclusion. And this one was ironclad, even if it required some people to adjust their thinking before accepting it.

Zierler:

Yeah.

Caves:

You can contest it, and experimenters can go out and test it. But if it’s wrong, we’ll have to throw quantum mechanics out the window. So I didn’t mind being definitive. Perhaps it was too definitive for many people. But I think [laugh] if you know something, you should say it.

Zierler:

[laugh]

Caves:

This is where we come to the lucky part of my main discovery, which was the use of squeezed light for interferometry. I was lucky to be working in a place and in a field where you were encouraged to think about fundamental questions. But the implications of those fundamental questions and the impact on real high-precision instruments were so far in the future—10 years, 30 years, a lot of people would have said never, and it turned out to be about 30 years. So most of the world wasn’t paying any attention, giving me the luxury of working on this without having to worry that somebody was gonna scoop me. While giving a wave explanation for the radiation-pressure fluctuations in that first paper, I said that there is a wave explanation for the shot noise, which had always been thought of as due to the particulate nature of light, a Poisson process like rain drops falling on roof. I didn’t spell this out clearly in that first paper. While I did say that the radiation-pressure noise comes from one quadrature of the vacuum entering the unused port, I didn’t say that the shot noise comes from the other. Still, anybody who read that paper could have said, “Well, this is completely obvious. The shot noise is due to the other quadrature of the fluctuations that come in with the vacuum. It is the phase of the vacuum that’s in phase with the laser light that makes the radiation-pressure fluctuations and it is the phase that is out of phase with the laser light that produces the differential phase fluctuations that mimic a differential phase shift and thus are directly responsible for the shot noise.” And once you realize that, it should be obvious that, well, you should replace the vacuum with something else that reduces one noise source or the other.

In the end it’s not so surprising that nobody else picked up on this, but the question is why didn’t I? The answer is probably one that plagues all of science. Before you can give an answer, you first have to realize that there is a question. I had solved the riddle of where the fundamental noise sources in an interferometer come from, and there they are—it’s the vacuum into the unused port. That is the fundamental noise source, and there’s nothing to be done about it. How did I get out of this trap? Probably because at the time I went to dinner with Ron Drever nearly every night. We would often go to a Mexican restaurant called Ernie’s on West Colorado Boulevard in Pasadena, then a very dingy area, now recovered as Pasadena’s Old Town. Ron is a very unusual person. [laugh] He didn’t live in the world that the rest of us live in—he’s off in some space of his own—and for that reason he could see things that eluded everybody else. An absolute genius, but a genius in a completely different way than what I had been taught was a genius. That is, I looked at Kip as a genius, and that’s the way I should try to do things. But Ron could get answers to questions that Kip couldn’t, by a totally different kind of genius, wholly intuitive. [laugh] He didn’t do any calculations. He’d draw little pictures and tell you the answer, and then he’d say, “Can you do a calculation to show I’m right?” But he knew he was right. So I go to dinner with Ron and every time he’d ask two questions: “Carl, what’s a taco?” “Ron, you’ve had a taco hundreds of times. I don’t need to tell you what a taco is.” [laugh] But he would also always ask, “There must be something we can do with these quantum non-demolition ideas in the interferometer.” And I’d say, “Nah, it’s a different system. Free masses, not an oscillator.” But maybe just Ron’s asking me over and over again got me off the motion of the mechanical system and to think hard enough about the light, to get out of the trap I had set for myself in thinking I had finished the problem with that first paper, to realize that if the whole deal is the vacuum coming in the unused port, if that is the ultimate fundamental source of noise in a balanced interferometer, then if you want to improve the sensitivity against shot noise, which is what needed to be done, you need to squeeze the vacuum, that is, reduce the fluctuations in the quadrature that’s responsible for shot noise.

I realized that at the end of May 1980, and notes in my files show I gave little blackboard talks at Caltech early in June of 1980 explaining this. The paper was published in April of 1981. I knew about squeezed states from work that had been done by David Stoler and then even more by Jim Hollenhorst, who had applied squeezed states to ideas of back-action-evading measurements on oscillators. So I knew about that. There was additional earlier work on squeezed states by Horace Yuen and Jeffrey Shapiro, which I really didn’t know about until I started writing my paper up and researching the literature. But nobody had told the story of how the noise runs around in an interferometer and how to reduce shot noise using squeezed light. And that’s why what I did got a lot of attention. Because I told a story of noise that people can understand, and they could see that if you want to design a better system, squeezing is what you need to do. You have to reduce this noise at the expense of increasing that other noise. People have elaborated this idea enormously over the years, and now it’s understood how to get around both the shot noise and the radiation-pressure noise simultaneously. That’s a thing I did not understand at the time. But what they’ve installed in LIGO and Virgo—they installed squeezed light and started doing observing runs with it in April 2019—is basically what I proposed long ago to reduce shot noise. I was at a talk by Nergis Mavalvala at Harvard last January, and Nergis showed a schematic of the design of the LIGO interferometers and how the squeezing is produced and used, and then she said, “Well, if you want to see how this works, here’s Carl’s diagram from his 1981 paper, and you can see it is pretty much what we’re doing.” [laugh] It’s incredibly gratifying to see—it took 30 years of in-the-trenches experimental effort to get the squeezed-light sources working properly. Maybe a 100 people worldwide working to make the squeezing work in the way that will actually make LIGO and Virgo better, and now it’s made them better.

So there I was in June of 1980. I look now at my notes, and I cannot believe the alacrity with which I produced a paper by August that even described in a rudimentary way how to make the squeezed light. Geez, I must have done nothing but physics back then. Again, the reason this is important is that I told a story of how the fundamental noise runs around in an interferometer and how the interferometer can be modified to make things better. That story wouldn’t have been told with the particle explanation of interferometer noise—with that story of the noise, you would have thought the problem of interferometer noise had been solved, and there was nothing to be done about it. Only with the wave explanation could one see that the problem is the vacuum into the unused port, so there’s a knob to adjust—replace the vacuum with something else—to make things better.

Zierler:

And, Carlton, I want to ask what was, after you defended, what was your administrative status at Caltech? Was this an extended postdoc? What was your work arrangement?

Caves:

After I defended in 1979, I was a Research Fellow, which is a postdoc. In 1982 I was promoted to Senior Research Fellow, which is, I think, the research-faculty-ladder equivalent of an assistant professor. Caltech’s a place that can afford to have research faculty. There are people who spend their whole career as research faculty at Caltech in these research positions. At the time, in 1980, I was a Research Fellow. And that’s basically a postdoc. And it was funded by Kip.

Zierler:

And did you have ambitions to join the faculty? Did you not care about that stuff because you were so involved in your research? What were you thinking in terms of long-range planning for your career?

Caves:

[laugh] Sure, I would have been happy to join the Caltech faculty, but, you know, I’m terrible at thinking about myself long-term. I talk to postdocs and students now, and I think, “Geez, I’m advising you to pay attention to all this stuff that I didn’t pay attention to.” I just went through life hoping. [laugh] Sure enough, it all worked out, although not quite as expected.

Zierler:

I mean, because of course the normal trajectory is you do one postdoc for a couple years. Maybe you do another one. And then you get a faculty job somewhere. But this was not on your radar? You were not thinking along these lines?

Caves:

I was not and I have to say, [pause] compared to the … Kip’s students had a very good track record of getting faculty positions. But those were the people who worked on either astrophysics or relativity. There just wasn’t a market for … Well, the late ‘70s and early ‘80s were a bad time to try to get a job in physics generally, but I worked in an area that really didn’t exist, a sort of quantum measurement theory applied to high-precision measurements that is really, as far as anybody could see, only relevant to this crazy project that’s not going to get built perhaps for decades. You could say this was quantum optics, but at the time, quantum optics really didn’t exist in most American physics departments. So there just was no market. I went and gave talks at lots of places, but finding a position, that would have been hard. Eventually Caltech got tired of me and suggested that I go find a position. This was in about ’86. By that time, I had gotten married, to Karen Kahn in ’84. We bought a house in Glendale. By ’87 we had a first kid, a son, Jeremy Caves Rugenstein. In ’89 we had another, a daughter, Eleanor Caves. Karen had a very good job in Los Angeles, and I didn’t want to leave Los Angeles. Now, when you combine working in a field that most people don’t recognize as important—even if what you did was important, most people don’t recognize this field as important—with the fact that you want to stay in Los Angeles, well you’re ... [laugh]

Zierler:

Yeah, yeah.

Caves:

This is not what I advise my students and postdocs. Go get married, buy a house, and decide you’re going to work in a very specialized area and only accept a job in a place where there’s just three major universities. You know? No, no. Indeed, my postdoctoral experience at Caltech—not mission-oriented, with no funding-driven focus—was so different from what most postdocs encounter that I wonder why I give advice and why anybody listens to it.

So I was very fortunate—luck again—that there were some people at the University of Southern California, actually in their Electrical Engineering Department, who thought, “Yeah, it’d be good to have Carl over here.”

Zierler:

Yeah.

Caves:

The department had a so-called Electrophysics half, more like a third of EE, which was essentially applied physics. I joined them as an Associate Professor at the end of 1987, when Jeremy was nearing a year old.

Zierler:

So, Carlton, I want to ask two questions about that before we get too far afield. I’m struck—the title of your dissertation, Theoretical Investigations of Experimental Gravitation, baked into that is this idea that you have this dual interest in the interplay between theory and experiment, right? I want to ask you about how self-conscious you were about this interplay in terms of how you wanted to design your own research agenda? And then, in your self-characterization of your area of research being so niche that it was hard enough finding a job generally in physics, let alone in this particular area that you worked on, is that to say that had you given a job talk at a place like MIT that what you were doing was just so narrowly focused that nobody would even think, “Well, here’s an area where we need to have representation on the faculty.” Was that the case?

Caves:

Yeah, well, MIT would be a special case because the other wing of LIGO, headed by Rai Weiss, was at MIT. But MIT at the time was not very friendly to gravitational-wave detection.

Zierler:

Right.

Caves:

And they certainly did not see an institutional commitment to LIGO: “Rai, if you can do that, great. But don’t expect the rest of us to help.” Whereas Caltech did. Caltech had an institutional commitment to gravitational-wave detection. And that’s because of Kip, who convinced the rest of the faculty and the administration that this was the future. And that was a tough sell, I’m sure. And the future turned out to be a lot further in the future than anyone … If you had said the future was in 2017, back there in the early ‘80s and the late ‘70s, they would have said, “Well, that’s a future we can wait for.”

Zierler:

Yes. Right.

Caves:

So, nobody said that. If a physicist tells you something’s a year in the future, well, it’s probably three. If it’s five years in the future, it’s probably ten. If it’s ten, really, nobody has a clue. Ten is just the number you pick when you don’t have a clue how long it’s going to take. [laugh] Because if you say anything more than ten, geez, nobody will ever do anything.

Zierler:

Yeah. [laugh]

Caves:

But, for me, you ask specifically about—I had been thinking about experiments that could either test relativity in the lab or detect gravitational waves. I learned more than I will ever learn about any experiment by talking to experimenters about how you do those experiments. This would be, first, Braginsky. And then over many years, talking to Ron Drever and his associates of my generation at Caltech. That’s how I learned about experimental physics. I would never have gotten into it if it was going to be the tedium of trying to design and build something. For example, something that is essential for LIGO is to design and build the isolation from ground noise. I thought a little about this very early on. It’s utterly essential for LIGO, but seismic noise is a nuisance, nothing fundamental about it. It’s a technological marvel that LIGO can be sufficiently isolated from ground noise, but the fact that there was no fundamental question there meant that I really wasn’t interested. My whole life, I’ve just done things because I was interested in them. Somebody was willing to pay me, although it wasn’t necessarily easy to arrange that. I was interested in fundamental questions that I encountered, and I worked on them. If other people were interested after I’d done whatever sated my interest, well, that’s icing on the cake. But if they’re not interested, well, that’s life, too. You know? I worked on what I worked on because I was interested in it. I wanted to know the answer to something that is out there in the world, that other people don’t know. There’s a satisfaction to finding the answer and knowing you are the one that has it. If it has a practical application, if it’s important in some way, and people take it and run with it, terrific, but that’s not the point, at least for me. So, you can see I wasn’t really good at … I wouldn’t have been very good at all at trying to find a job because everybody wants to know, “Well, how are you going to work with us?” Well, I don’t really want to work with you. I want to work on what I’m interested in. “How is what you’re doing going to be important?” Well, I don’t know and, more to the point, it’s not what drives me. [laugh]

Zierler:

Nobody ever taught you how to promote yourself, either, it sounds like.

Caves:

You know, I thought I was promoting myself, but I see now that I just wanted somebody to think I was clever and to hire me to be clever. I don’t want to leave the impression that the only reason I had a hard time finding a job was that I worked in an obscure field. I had plenty of job interviews when I was looking for a job in ‘86 and again in ’91, by which time Karen and I had changed our minds and wanted to get out of Los Angeles. The evidence from those job searches, frankly, is that most physicists just didn’t find me very impressive.

I’ll tell you a story from my job search in ‘91. In March I was at a Workshop on Squeezed States and Uncertainty Relations at the University of Maryland, and during the conference, I had a little talk with the Chair of Maryland Physics about job opportunities. Maryland had a relativity and gravity-wave group, so who knows? Maybe there’s a chance. The chair told me that the department was only interested in hiring people who would be elected to the National Academy. What are you supposed to do with that? Perhaps others would have an effective response, but, geez, I’m having a hard time finding a job, and you want me to persuade you that I’m going to be elected to the Academy? When this gets published, my hope is that readers will be able to feel the anger and frustration radiating from the page even from a distance of 30 years.

[Editor’s note: Caves was elected to the Academy in April 2020, not long before this interview.]

On the plus side, at that same Maryland workshop, I discovered Chris Fuchs—or maybe Chris discovered me—and that was a very big plus.

Zierler:

And, so, what clicked for you at USC? What was the connection there that made it work?

Caves:

Well, the Electrophysics part of the department was doing applied physics, mostly device physics in what was then called quantum electronics, what now would be called photonics. And they could see that what I was doing was relevant to that. When I say that the sort of thing I did, quantum limitations on high-precision measurements, didn’t exist as a field, I mean that it didn’t really exist in physics departments. But it was a thing that electrical engineers were interested in. For actually doing stuff that they were doing. And they could see that so they decided to hire me. I’m very grateful to that group of faculty at USC for having, from my perspective, the vision to hire someone who was not doing stuff directly in their main line of research.

Zierler:

How was EE interested in this stuff? Where a classical physics department was not? Can you go into that a little more?

Caves:

Yeah, well, they were making actual devices. Photonics devices, where quantum behavior was important. There was a very long history of EE people working on this kind of stuff, and the people at USC could see that what I was doing was sort of the same thing from a different perspective, so if I was dragged out of the gravity-wave realm it might be important to them. Now what I actually did there … [laugh] I got interested in a different subject … I got interested in information physics, which morphed into quantum information in the early ‘90s. I still worked on very fundamental topics in high-precision measurements, what’s now called quantum metrology. And on both of these I worked with my former Caltech student, Sam Braunstein, in what was an extremely productive, rewarding, and long-lasting collaboration, even after I went to UNM in 1992.

Information physics and quantum metrology, well, these were not directly relevant to what my colleagues at USC did, but I am grateful that they were patient. Still, EE departments were a mixed bag for me. I wasn’t comfortable in an EE department because I didn’t feel like it was … I mean the motivation was different from physics. And I had a hard time plugging into that much more applied motivation. And I didn’t see that I could teach a lot of things comfortably in an EE department. But, on the plus side, EE departments have had … well, let’s step back to a bigger picture … Physics rests on an over-300-year-old tradition and ultimately, after World War II this meant that there were three interest and power centers in physics. There was astrophysics. There was condensed-matter physics. And there was particle physics. The kinds of questions that I was addressing were the things that were increasingly coming into play in quantum optics and atomic physics, where you think about quantum coherence and its implications. And this ultimately came to be quantum information, which is really all about what you can do with quantum coherence. Physics departments had this long tradition, and it wasn’t a time when physics departments were expanding. Ultimately, I had to think of myself at the time as a quantum optician, but most of the prominent, first-rank physics institutions in the U.S., they didn’t do quantum optics, and later they didn’t welcome quantum information unless somebody prominent in the department happened to get interested. Quantum optics was hardly done in the United States. It was done in Europe, Australia, New Zealand. Now, in contrast to all of this, EE departments, not having a 300-year-old tradition, just do whatever they think is important—and can get funding—as it comes along. Not to say they don’t have entrenched constituencies, but they are far less interested in policing the boundaries of their discipline. In the other part of the EE department at USC, called EE Systems, there were loads of people who anybody would say were mathematicians. Coding theorists and so on. Mathematicians would say that they’re doing applied mathematics, applied to the way we communicate information, but it’s a very abstract mathematics. EE departments could do that because, unlike physics departments, they don’t have those 300 years weighing them down and telling them what their field is.

Zierler:

Now you came in on a tenured position or you earned that while you were there?

Caves:

USC wasn’t keen on tenuring people on arrival, especially ones without extensive teaching experience—I had taught a year-long relativity course at Caltech in ‘84‒85, and that’s it—so I came in as an Associate Professor and was tenured, I think, two years after I got there.

Zierler:

And what were their expectations in terms of the kinds of classes you would teach? I mean, were they looking for you to teach EE classes or were you expected to teach physics classes in the physics department?

Caves:

I did a little bit of the latter—taught one introductory physics class while I was at USC, but that was because the dean of engineering thought the physics department was too tough on prospective engineering majors and was failing too many of them in the required, introductory physics classes. So the dean wanted somebody from electrical engineering to teach an introductory physics class, and guess what, I did it, just to show … well, I would say the experiment, from the dean’s point of view, was not a success. Otherwise, the culture of teaching in different departments is quite different. I think physicists are more interested in effective teaching than the other sciences and far more interested than engineering departments. In addition, in physics departments, certainly the one I was in at UNM, you’re not allowed to teach the same course from here to eternity. You might teach a course as much as three times, and then you need to move on and teach something different in order not to go stale and in order to learn new things and how to teach them. In EE departments, certainly at USC, people had courses, and they just owned them and taught them forever. So I took an existing graduate course in statistical optics and made it into a course in quantum optics. That I would teach one semester, and the other semester I would teach the junior-level course in electromagnetism. That was like a physics course, just more recipe-oriented, which sort of drove me nuts. Even the best students wanted a list of formulas that they should memorize.

Zierler:

And, Carlton, did you feel like you had to—or did you want to—brush up on your own EE knowledge or was that sort of already built into what you were doing?

Caves:

Well, certainly, anyone can afford to brush up on anything, but the two courses I generally taught, I knew cold, and they were sort of the only good opportunities I saw for teaching in an EE department. The two things I knew best were electromagnetic theory and quantum mechanics. There really wasn’t a chance in the EE department to teach an undergraduate or graduate quantum mechanics course. And, actually, I never taught an undergraduate quantum course at UNM, either, but I did teach both semesters of graduate quantum mechanics. At USC, I took over a graduate course called statistical optics which, it turned out, somebody else owned, but hadn’t taught forever, and I changed it into a course in quantum optics, which was very relevant to what I did. The point is that I never felt at all like an electrical engineer, and that contributed to my leaving for the Physics and Astronomy Department at UNM in 1992.

Zierler:

And I’m curious for your own research agenda, a bit of a counterfactual but, how much did your research change as a result of going into an EE department versus had you gone into a more traditional physics department? Did you respond to that change or did you just sort of plow ahead with the stuff that you always wanted to do and you continued to do it?

Caves:

It won’t surprise you that I wasn’t very responsive to hints, subtle or otherwise, about what I should be doing, so I just continued along whatever lines I was interested in. I’m not saying this is a virtue, just a fact, and somehow, I managed to get by. At Caltech Bonny Schumaker and I had developed the theory of wideband squeezed light, which serves as the basis for analyses of squeezing in the big gravitational-wave interferometers and which, generalized outside the context of upper and lower sidebands of a carrier, informs all sorts of work where two electromagnetic modes are squeezed. Also at Caltech, I had initiated new work on high-precision measurements, mainly with Sam Braunstein, basically learning quantum-measurement theory in the process. By the time I went to USC, I had become a quantum optician—certainly no longer a relativity theorist—but a quantum optician with no formal training in the field. What I worked on at USC was topics in the emerging field of information physics, which later morphed into quantum information, and fundamental questions in what became known as quantum metrology. Quantum optics is a very slippery field, really defined by whatever self-identified quantum opticians are doing, and by the turn of the millennium, a good fraction of people in quantum optics had migrated, to one degree or another, to quantum information.

But let’s back up a bit on my own story. I was a quantum optician as a consequence of the squeezing proposal and being introduced to the worldwide workers in quantum optics at one of the greatest conferences I ever went to, in August of 1981 in Bad Windsheim, West Germany, a small spa town in northern Bavaria where I learned, pretty directly, about the German distaste for putting one’s feet on a coffee table. The conference, a NATO Advanced Study Institute, organized by Marlan Scully and Pierre Meystre, brought together leading researchers thinking about high-precision measurements, open quantum systems, and all sorts of fundamental questions in relativity, including gravity-wave detection, which might benefit from the increasing ability in optics and atomic physics to control and manipulate the behavior of optical and atomic systems. It was a spectacular opportunity to engage with people from very different fields who, it turned out, had common interests. Just the title of the conference (and the proceedings), “Quantum Optics, Experimental Gravitation, and Measurement Theory,” says it all: When had these three topics ever been put together before? When had Measurement Theory even been in the title of a physics conference?

So, I was really a quantum optician without any formal training in the subject by the time I was looking for jobs. And that’s what, I think, USC would have thought they were hiring, a quantum optician. But your question was … and that’s the thing that really wasn’t represented in American physics departments.

Zierler:

Anywhere. You’re saying anywhere. It wasn’t represented anywhere.

Caves:

Not much at the first-rank institutions. Atomics physics and optics in the US had been pretty much exiled to second-rank institutions—this is not to say the people were second-rank—but for most American physicists, atomic and optical physics was, well, been there done that, let’s move on. Harvard has always had atomic physics, and of course, Roy Glauber was at Harvard. Roy was many things over his long career, but what he won the Nobel Prize for was his theory of (quantum) optical coherence, which absolutely nailed it and was a foundation stone of quantum optics. At the first-rank physics departments in the U.S., you weren’t going to find quantum optics and atomic physics, outside of big presences at Rochester and Arizona and JILA at the University of Colorado, Roy at Harvard, and after 1987, Jeff Kimble at Caltech. I didn’t really understand this landscape very well, probably still don’t. The truth is that despite being willing to call myself a quantum optician, which was better than having no field at all, I really thought of myself as a theoretical physicist jack-of-all-trades. What ultimately changed the situation—things are very different today—was the increasing ability to maintain and control quantum coherence in atomic, optical, and condensed systems, and this is exactly what you want to do in quantum information science.

So USC probably thought they were hiring a quantum optician who didn’t actually have any training in quantum optics. That is, me. And all of that is how I ended up at New Mexico in 1992. New Mexico had a commitment to quantum optics begun by Marlan Scully in the early ‘80s. I had known Marlan since about then, and Marlan is the person who is most responsible for bringing me to UNM. Ironically, as I entered the front door, Marlan walked out the back door and left for Texas A&M—although I have to say nobody at UNM seemed to realize that for a while. Oh, well.

Zierler:

What was Marlan’s main body of research at that point? What was he working on?

Caves:

Marlan was always interested—so, what quantum optics is … it’s isolating a few relevant modes of the electromagnetic field that are interacting strongly with matter, either atoms or condensed systems. And also isolating within the matter, within the atoms or condensed system or whatever, the few quantum states that are strongly interacting. Instead of dealing with all the matter and all of the electromagnetic field, quantum optics is about isolating the parts that are important and studying what happens, so that you have a tractable problem, and treating the rest of the world, the rest of the electromagnetic field or other states of atomic systems or whatever, as a bath, if necessary. So quantum optics is one of the places where open-systems theory was developed. Marlan’s history back into the ‘60s was the theory of the laser, precisely where this quantum-optics style of research was pioneered. That’s how he got started. Marlan was always interested in practical applications of all this—his genius was recognizing when this perspective, this quantum-optics toolkit for doing problems, was relevant to research problems throughout physics and going to work on those problems—but he was also interested in what it says about quantum mechanics. That’s really the interest that Marlan and I shared, although I never worked directly with Marlan on anything. Quantum optics and atomic physics became the place to address interesting questions in the foundations of quantum mechanics, because that’s where the quantum coherence was, quantum coherence that could be used to test quantum mechanics and put to use in the emerging field of quantum information science.

Shortly after I moved to UNM in 1992, my USC postdoc, Rüdiger Schack, joined me at UNM, and I persuaded Chris Fuchs, then a PhD student at the University of North Carolina, to transfer to UNM’s PhD program. The three of us were interested in how to think about quantum probabilities. After some initial muddle, we settled our thinking on the notion that all probabilities, including the probabilities associated with pure quantum states, are subjective. Not out there in the world. In our heads as tools for dealing with the world. These are often called Bayesian probabilities, and for my own part, I was heavily influenced by the work of Ed Jaynes, whose book of collected papers, about what would now be called objective Bayesianism, was one of the revelations of my life. Chris and Rüdiger and I worked for many years on importing Bayesianism into quantum mechanics. This all culminated, much later, in the early ‘10s, after many twists and turns, with the formulation by Chris and Rüdiger, along with David Mermin, of QBism, a consistent way of thinking about quantum mechanics and its predictions. By then I had gotten off the bandwagon—too much subjectivity and too little ontology for my taste, particularly after I proposed, near the end of the aughts, an ontology of unitary operations that didn’t appeal to Chris and Rüdiger. I have enormous respect for the way Chris and Rüdiger (and David) pushed this program through to its logical and consistent conclusion. I just didn’t like where it ended up. Though I am no longer on board, I value this collaboration as much as anything in my life, both for the research content and for Chris and Rüdiger as research colleagues and friends. One thing it does show is that I followed my own interests—so do Chris and Rüdiger—without much thought about the consequences, because nobody in his right mind in physics works on foundations if they want to get ahead in life.

Schack and I also initiated a research program on nonlinear dynamics and chaos, both classical and quantum. We thought of this as a natural follow-on to foundational work in information physics on Maxwell demonology, by Charles Bennett and Wojciech Zurek (and me and Rüdiger, a bit later), which clarified, I think completely, the connection between information and entropy. Rüdiger and I called our work hypersensitivity to perturbation. It was based on a trade-off between information and entropy in perturbed classical or quantum dynamics. Though we worked on this for over a decade, it never really went anywhere, like a lot of work in quantum-chaos theory. Even so, it remains a highlight of my life because of the opportunity to work with Rüdiger. Might get revived now in the context of the complexity of quantum dynamics. Maybe I’ve learned enough math by now to make it convincing.

Zierler:

Now that you’re moving to New Mexico, maybe it’s time to fill in a bit about your personal life. What motivated your move to New Mexico?

Caves:

Some not liking having to fake it as an electrical engineer at USC. But mainly just tired of Los Angeles, both me and Karen. Too big, too crowded. Couldn’t take off work at USC to go to a noon program at Jeremy’s preschool, because fighting the traffic there and back didn’t make sense. I often said that a place where any conversation veers within ten minutes into the shortcuts you take to avoid traffic getting to and from work is a place you can do without. Or maybe that I loved living in LA because you could eat al fresco at nearly every meal, but when that got to be the only reason to live there, we left.

I was lucky to find a job at UNM, because I had always been attracted to the Southwest. I have no regrets about moving to Albuquerque. Great place to live, close to the outdoors, culturally diverse, just big enough to have things to do, but small enough to be easy to get around. I often tell people that if your city is big enough to have major-league professional sports—we don’t—it’s time to find a place that doesn’t. Karen and I bought a house at the foot of the Sandia Mountains—still live there. Great place to raise kids, and I think Jeremy and Eleanor would endorse that.

Zierler:

What were the circumstances leading to your decision to convert to Judaism?

Caves:

I converted in 1989 very shortly after Eleanor was born. Why? … hmm … Karen’s a born Jew, so the kids were Jews, by Jewish rules, regardless of what I did. But I thought it was important for everybody in the family to have the same religion, and I did sort of need a religion. Even though my parents were probably not thrilled with my conversion, it was better from their perspective to be a Jew than to have no religion at all. And the belief structure in liberal Judaism was far more congenial than Christianity, even liberal Christianity. Maybe that’s because Judaism is a religion of actions, of how you live your life—and in liberal Judaism, even those actions are pretty much up to you to choose the ones you find meaningful. Certainly I thought it would be a plus for raising kids to be able to tap into a three-thousand-year old tradition, provided the ethics and morality and traditions are under review for updating, as they are in liberal Judaism. And, it did not escape my attention that so many of the great theoretical physicists were Jews, along with a bunch of intimidatingly capable PhD students in Kip’s group when I arrived in 1972. There’s a correlation there that must mean something.

Zierler:

What movement of Judaism did you want to align yourself with?

Caves:

Really the only choice that suited me was Reform Judaism. Getting back to the basics after a long period of trying to imitate liberal Protestantism, but open to what’s happening in the world in a way that Judaism ought to be. We joined Temple Sinai in Glendale about when Jeremy was born—Carole Meyers was my first rabbi, and she was great—and then we joined Congregation Albert when we moved to Albuquerque. Paul Citrin and Joe Black were also great rabbis, learned, inspiring, and wise, and Joe, in particular, was a very positive influence on my kids. Karen was eventually president of Congregation Albert at the time of Jeremy’s bar mitzvah.

Zierler:

In what ways have you delved into the teachings of Judaism, and what have you learned along the way?

Caves:

Geez, not my specialty. Mainly two things. The Jews have been sufficiently oppressed throughout history to have learned the importance of social justice, and that, for me, is what is most attractive about the Jewish tradition. None is free till all are free. Just as important for me is a lesson from the rabbi who taught my conversion class. Torah study extends beyond studying the Torah to anything that adds to human knowledge about the world and its peoples. As far as I could tell, that means a Jewish scientist is being both a scientist and a good Jew. I liked that.

Zierler:

You quipped that you became a “Jewish atheist.” I am curious if your atheism preceded your conversion, and if it only strengthened afterward.

Caves:

In the end I’m a scientist—a bit of environmentalism, a bit of Judaism—but mainly I’m a scientist, particularly a theoretical physicist. Scientists are natural atheists because if you’re holistic about science and its increasingly powerful naturalistic explanations, there isn’t really any space for a God. I didn’t see any reason to trumpet this point of view for most of my life. In my family, my siblings and I and our progeny never really said what we thought about religion till my parents died in ’06 and ’07. Religion was important to them. What would be the point of announcing a rejection of it?

I think a little differently now, motivated partly by an alarming realization that the fundamentalisms in all the major religions are more and more a major threat to humanity—not as existential as climate change, but gettin’ there. I want to be clearly dissociated from that so, yeah, I’m an atheist. And Judaism, as a religion of action, really doesn’t require belief in a God. The point is how you live your life, not your belief structure on divinity. Besides, I can talk the God talk because I’ve become a Spinozan. Spinoza reclaimed all the space occupied by science simply by declaring, well, that space is God. God is the Universe—not less and couldn’t be more. Omniscient? Sure. Omnipotent? Yeah. Omnibenevolent? Well, no, not even a question that makes sense. Indifferent is the word. Sure the Universe can slap us down big time—look what happened to the dinosaurs—but none of that has any intention to it. Good and bad and ethics and morality—these are human concepts that make no sense outside of us. They originate in the tiny piece of the Universe that is each of us. That’s why it is so important, drawing on the collective wisdom of those who came before us, to maintenance ethical and moral precepts forward, with the idea of increasing the freedom and dignity of all people. Secular humanism, really, but there’s a Spinozan take on it.

Now I’m back to what I said before. We scientists are reading the book of God directly, in discovering how the Universe works. The more we discover—and, geez, we’re just getting started—the more we’re in awe of what we find. Here’s a contrast I like to make. A good scientist has some understanding of the four-billion-year history of Earth, some familiarity with the deep-in-time, finite, but large complexity of Earth and the living things that have evolved on it. The Earth is ours. It’s unique. Won’t happen again. So, by the Spinozan God, take care of it. Such a contrast with the religions. They read their own self-regard onto a God they create who, not surprisingly, turns around and authorizes them to do whatever they want. The one-off, anthropocentric world of Biblical literalists can be just thrown away. Its complexity is merely some whim God had, and God could decide at any time to dump it and start all over. And the afterlife will come to the rescue even if Earth tanks.

So much for ranting.

Zierler:

Let’s return to your move to New Mexico. Did all of what you’ve said make New Mexico essentially the place to be for this field? By default, if nothing else?

Caves:

UNM was a genuine opportunity for me. I was hired as a tenured Professor at a time when the Department of Physics and Astronomy was bringing in several people as investments for the future. We hired Ivan Deutsch in 1995, just at the time when the discovery of quantum error correction and the proposal of the Cirac-Zoller gate for trapped ions was turning quantum information science into something a physicist could take seriously. Ivan and I worked together for many years to make UNM a center in this emerging field. And we did that. But we certainly didn’t accomplish as much as we could have had the university been friendlier to allowing us to build what we were trying to build.

Zierler:

And is this friendly on a budgetary perspective or, sort of, across the board?

Caves:

What are universities made of? Money, space, and people. Ivan and I had a reasonable amount of money from federal funding, mainly NSF after the mid-aughts. Indeed, after our first center grant from NSF in 2009 allowed us to establish CQuIC—me, Ivan, and Poul Jessen at the University of Arizona—we were quite well funded. And, beginning about then, we had enough space for our expanding team of PhD students and postdocs. But what we didn’t have enough of was quantum-information faculty, and it took a very long time to convince the department and even longer to convince the UNM administration to start hiring additional faculty in this area. This despite the fact that we had long been well funded, and we really were a leading place in quantum information.

Michael Nielsen, who became my PhD student in the mid-90s after an undergraduate degree from the University of Queensland, was told by his undergraduate adviser, Gerard Milburn, “Michael, if you’re interested in quantum information, you should go to UNM and work with Carl Caves.” That’s how Michael, who is really quite an extraordinary person, ended up at UNM. Michael used his dissertation work as the basis for writing, with Ike Chuang, the first comprehensive textbook on quantum information. Ivan and I got many great PhD students because we were out in front early on in quantum information. Our initial prominence faded over time, not because we changed, but because we didn’t. Other institutions got into the field, and UNM really did next to nothing to build on our early prominence.

ZEIRLER: Let’s talk a bit more directly about your career at UNM. Tell me more about your professional service to the university, the ways that you have shaped and moved forward your department, and specifically, your work as founding Director of the Center for Quantum Information and Control.

Caves:

I wasn’t really much into university service. For example, I didn’t really want to be Chair of the Department of Physics and Astronomy, nor as far I could see, did anyone else think that was a particularly good idea. The main thing I did do for the Department was to chair two Long-Range Planning (LRP) Committees, one in ‘03‒‘04 and the other in ‘08‒’09. These were comprehensive reviews of the Department’s activities, and the first, especially, served to provide a structure for the Department’s research activities and a plan for future faculty hires to staff those activities. Prior to the first ‘03‒‘04 LRP, the Department’s hiring planning was pretty non-existent. Basically, we reinvented the wheel every time the Dean asked for proposals for faculty hiring. Having a hiring plan reduced friction within the Department and made for a more orderly and far more convincing process for responding to the Dean’s requests. For me personally, these two LRPs were genuine learning experiences. I learned, especially from the first, how all aspects of the Department worked, or didn’t, and how one justifies the existence of a medium-sized physics and astronomy department, with a heavy service-course load and a healthy research enterprise, but few majors, to the administration of a public university.

The main thing I did for the University was to chair a Provost-commissioned study of UNM’s research administration in the spring and summer of ‘07. We were called the Research Study Group, and it was a truly great group of people to work with. Despite the enormous amount of time I devoted to this—essentially all of my time in the spring and summer of ’07—it was a highlight of my career at UNM. The RSG worked really hard and worked together so effectively that it almost made you want to volunteer to do university service. We produced a report that was supplemented by a report from a group of professional research administrators, who visited the campus to provide external calibration. The report was quite influential and actually produced changes in the structure and practice of UNM’s research administration. That is what you want—and rarely get—when you do university service, and that, of course, contributed to making this a very rewarding experience, alongside getting to know faculty and staff from across the University.

The last major piece of service I did for the Department was to chair a three-person, ad hoc committee that wrote the Department’s standards for tenure and promotion. This was absolutely essential to do, because the Department was operating under the typical, lazy attitude that says, yeah, we can recognize good people when we see them. But that won’t and absolutely shouldn’t fly in a process that must be fair and transparent and as objective as possible. So when you get to thinking about what it is that makes a good faculty member, you realize that you have to recognize all the ways of being a good faculty member, including those outside your own skill set, because if you don’t put them in the standards, well, you’re going to be trying to justify exceptions pretty quickly. The committee did an excellent job, but after the Department adopted the standards, it proceeded promptly to ignore them. I was more than a little pissed off, and this prompted me to tell the Department Chair that I wouldn’t do any further service. That was also because I thought being Director of CQuIC was enough.

CQuIC was founded in 2009, when Ivan and I and Poul Jessen got our first, sort of ad hoc center grant from NSF. It was a name change from a center called the Center for Advanced Studies, which went all way back to the ‘80s, when it was started by Marlan Scully. I was its Director for a few years after coming to UNM, and Ivan had been Director since about 2004. The name change to Center for Quantum Information and Control was part of the deal of getting the NSF grant. Ivan and I worked really hard to build CQuIC, and by now, it is mid-size center with a lot of good people.

ZEIRLER: Who among the faculty and staff at UNM have been your most important collaborators? Most important supporters?

Caves:

Ivan and I got good support from P&A Chairs Bernd Bassalleck and Wolfgang Rudolph and from Arts & Sciences Dean Mark Peceny, but it was a slog with higher administrators. That’s changing now, with Ivan as Director, perhaps because Ivan is a more persistent presence than I was and maybe also because of the flood of money into quantum information from the National Quantum Initiative.

For my own part as a researcher, what I did was to work with PhD students and postdocs. That’s what I like doing and perhaps what I’m best at. I don’t have time here to discuss everything we did, but working with Chris Fuchs and Rüdiger Schack, Howard Barnum and Michael Nielsen, Andrew Scott and Joe Renes and Bryan Eastin, Anil Shaji and Alexandre Tacla, Sergio Boixo, Steve Flammia, Animesh Datta, Mankei Tsang, Zhang Jiang, Josh Combes, Chris Ferrie, Saleh Rahimi-Keshari, Jonathan Gross, Raf Alexander, and Chris Jackson—geez, I’m going to forget some names—well, working with all of them was what made my life at UNM great. Thanks to all of them for putting up with me and making my life better. Sometimes we struck it rich, mostly we didn’t, but we always had the reward of doing research together.

And I should probably correct a misimpression. You might think from what I said that I blame UNM’s administration for not providing sufficient support for me and Ivan. That’s true, I think, but it’s not the whole story. Throughout my whole career, there were competing internal interests at UNM that worked against us, and I really only understood those forces, even vaguely, fairly recently. Moreover, I think I was pretty clueless about how to persuade people, and really, maybe my heart wasn’t much in it. As I said, I was mainly interested in having good students and postdocs, and we had those. I didn’t see the need to expand aggressively. Things are different now, what with the flood of money from the National Quantum Initiative. Grow or die is now the imperative, and Ivan, who became CQuIC Director when I retired in 2018, is navigating that terrain pretty well. In this environment, if you don’t grow, you’ll die, and you certainly won’t continue to get good students and postdocs.

One last thing I should mention, about my research life outside UNM. I have a lifelong research collaboration with Gerard Milburn of the University of Queensland, extending back to the ‘80s, shortly after Gerard got his PhD. Gerard, who is a great researcher and a very talented administrator, is the person primarily responsible for building the genuinely amazing Aussie presence in quantum information. For my own part, along about 2004, I started spending extended periods at UQ. I spent a sabbatical year ‘08‒‘09 in Brisbane, and eventually, for much of the teens, I had a deal where I spent three months a year in Brisbane. Multiple reasons to do that in addition to the research collaborations with Gerard and others. I like living in Oz, at least as much as I like living in New Mexico, and I tended to go to Brisbane in the northern spring to get away from the spring winds in New Mexico. The birds in Oz are bigger and brighter and louder than ours. And, truthfully, I was also escaping from my responsibilities at home. I learned to use Skype because of my sabbatical year and used that to keep up with my family and to run my research group whenever I was Down Under. For the rest, you wake up in Brisbane, talk to people at home, yesterday, for a couple of hours, and then they all go home, and you have the rest of the day to focus without interruption. Gerard and the rest of the crew at UQ were great hosts, always ready to tell me what they were doing, but also willing to give me the time to focus if that’s what I needed.

Zierler:

Carlton, let’s broaden out a little bit about quantum information. Who coined the term? Are you aware where that came from?

Caves:

Whoa! [laugh]

Zierler:

I mean is it a relatively new term in the ‘90s or early 2000s or do you see this term come up far earlier than the computers would have warranted it?

Caves:

Well, I will tell you that it’s … what’s now quantum information absorbed work from a variety of fields that was going on in the ‘80s and early ‘90s. There was a long tradition of work on quantum limits to classical communications, and there was the sort of thing I did, what became quantum metrology and is now often called quantum sensing. Both of these involved the quantum theory of measurements and are fundamentally about encoding classical information in quantum systems and then extracting it via measurements, and both were practiced some in physics, especially in quantum optics, but also in other disciplines than physics. Beginning in the mid-‘80s there were initial proposals for quantum computers and a dawning realization that by processing quantum information instead of classical information, a quantum computer could do things efficiently that can’t be done by a classical computer processing classical information. And then, there was this whole effort on the equivalence between information and entropy and how this accounts for how Maxwell demons don’t violate the Second Law. That’s really how I got into information physics, which evolved into quantum information. I went to a series of great conferences on information physics. Worked on it. As I mentioned, had a program with Ruediger Schack, which never really took off, for characterizing classical and quantum chaos that was based on ideas from Maxwell demonology. So there were all these different streams coming together. Everybody could see that quantum computing was a big part of it … but at the time quantum computing appeared to be impossible because decoherence from interaction of the quantum computer with the external world would destroy the very quantum coherence that the computer lived off of. That was before error correction was invented by Peter Shor and Andrew Steane. Just knowing that quantum error correction was possible allowed you at least to dream about doing quantum computations.

Zierler:

This is a technological development? Or a theoretical development?

Caves:

That was a theoretical development. It showed that physicists’ worries about decoherence, though serious, were not a showstopper. When one describes a quantum computer, it looks like an analog computer that runs on interference between waves. And analog computers never work because of noise. That’s where the physicists’ worries came from, that decoherence and noise would always destroy the interference that is at the heart of a quantum computation. What Shor and Steane realized is that you could organize things in such a way that even though the quantum computer is sort of working in an analog fashion, the errors are digitized by clever measurements that leave the encoded quantum information unaffected. And then these digitized errors can be corrected. If you couldn’t do that, it was clear that, okay, suppose you design a quantum computer with errors small enough to do problems of some size; then if you wanted to do problems of a bigger size, you would have to tear the whole thing up and build an exponentially better device. That would be a showstopper, and quantum error correction shot that down. With quantum error correction—more precisely, fault-tolerant quantum computation—you can prove that if you reduce errors to a certain level, then you can do quantum computations of arbitrary length. That changed everything. The other really important development of the ‘90s was the proposal of the Cirac-Zoller gate for doing two-qubit gates in an ion trap. That was the first realistic proposal for how to do the elementary entangling operations that are essential for doing quantum computations.

These developments allowed you to dream semi-realistically about building a quantum computer, but it was clear to everyone that it would be a very long time, involving a huge amount of technology development, before you could build a universal, fault-tolerant quantum computer. Even with all the developments of the last 25 years, the emphasis today is on doing quantum simulations that don’t require error correction and fault tolerance. With quantum computers so far in the future, many people in the ‘90s and the early aughts, wanting to get away from being just about building a quantum computer, were searching around for a different way to describe the whole range of related activities and disciplines and expertise that came together to take advantage of quantum coherence. My initial favorite was Zurek’s name, the physics of information, the emphasis being on the representation and manipulation of information in physical systems. But it became clear very quickly that quantum systems and information stored and processed coherently in quantum systems were the really important new thing. So quantum information was made up as the kind of umbrella name for all of the activities that involved control and manipulation of quantum-coherent systems, to emphasize that the field wasn’t just this pretty far-fetched goal of quantum computers. You know, now many people consider quantum information to be too constraining, because their efforts to harness quantum coherence into technologies don’t appear to them to have much to do with information. So they want to call it quantum technologies or quantum science or quantum engineering or even just quantum. There’s loads of people who would just say, “I’m doing quantum.” I’m a little sympathetic to that, but you could say that nearly all physicists are doing quantum, so it doesn’t [laugh] really tell you very much.

Zierler:

Right, right. Carlton, if you can give sort of a broad overview from the early ‘90s to the present, what were some of the major questions that were being raised in the world of quantum information? And how have those fundamental questions changed in the past 30 years?

Caves:

[laugh] In the end, all situations in … the field is ultimately about … controlling and manipulating quantum systems so that they maintain quantum coherence over long periods of time and so that ultimately relatively big systems, either mesoscopic or macroscopic systems, consisting of many constituents, maintain coherence over big length scales and over long times. That’s what’s necessary to do a quantum computation. The big questions are, how do you prepare systems appropriately to be ready to do this? How do you manipulate and control them to do it while shielding them from the decoherence that typically follows from the very same couplings that allow you to control and manipulate them? And then how do you extract information in the end? So you want to describe such a system, preparing it, manipulating it, and extracting information in the end. The system is made up of parts, and the parts are described in a Hilbert space of a certain size, containing all the relevant quantum states and their superpositions. If this sounds like quantum optics done more generally, well, it is, and this Hilbert-space language for quantum systems is now pretty much de rigueur in quantum information science. Once that abstraction is made, everything can be done without reference to a particular physical system. That is to say, you have a Hilbert space, you want to do stuff in that Hilbert space and extract information by making measurements in the end. You don’t need to say what kind of physical system you have. And everything you can do is described by some piece of linear algebra that ultimately can be formulated without reference to any particular physical system. So much, a heck of a lot of the work, in quantum information, and especially in the theory of quantum computation, assumes that you’ve formulated this Hilbert-space model, and then you just ask what you can do with it. You’ve completely forgotten how it’s instantiated in an actual laboratory system. So one of the big questions has been, what can you do with that kind of stuff? And with abstract models of the noise and decoherence that would plague such a model, what can you do? What can you do if you assume that errors are sufficiently controlled to do fault-tolerant computation, and since that looks like a tall order, what can you do if you can’t control errors that well? Are there things you can do, basically treating a quantum computer as an analog system without error correction?

Of course, the flip side of that platform-independent model is … the other big questions are all about, what can you get particular physical systems to do? What happens when you start at the bottom and start from thinking about the parts, the components? What are those components? How do they interact? How can they be integrated together? How can errors and decoherence be suppressed? Ultimately, the whole enterprise is going to be a huge piece of quantum systems engineering, from the components and their integration, to how you control and manipulate and read out the system, to how you characterize and check on its performance on the fly, to the software, from operating systems to the actual applications. And, just like for classical computers, people working on the various levels in this hierarchy of systems engineering aren’t going to talk much to each other or even care much about each other. Maybe initially there will be a lot of cross-hierarchy talk, but that will tend to disappear if the thing ever gets going. It already is silly to ask people thinking about quantum algorithms how they are going to work with people who are doing the actual quantum-component development.

Zierler:

[laugh]

Caves:

The way I always put it is, when I was a kid, the point of quantum mechanics was to explain what physical systems do naturally. What quantum information is about … it asks what can we do if we make quantum systems do what we want instead of what comes naturally? The things we want to accomplish with quantum systems are typically far away from the natural state of things in the world, and that’s why it’s highly nontrivial to do quantum information processing. To do unnatural things, you need to know what you’re gonna get from it and how you’re gonna get it.

Let me back up a step. The reason it was Peter Shor, a computer scientist, who discovered quantum error correction is because, well, he was pretty smart, but also because he knew coding theory. He knew there were so-called linear codes, which were a little bit like the kind of superposition we have in quantum mechanics. He knew, “I don’t need to think about what physical system you’re using. I just need to know about the Hilbert space that it lives in and see what I can do in that Hilbert space. Basically, I just need to know how the rules of quantum mechanics are expressed in the linear algebra of Hilbert space. You physicists know a heck of a lot of other stuff about real physical systems, but that’s not helping you. You aren’t going to see the forest for the trees because you’re looking at all this extraneous detail you know about, which is totally irrelevant to the question of can we, in principle, detect and correct errors?” Much of the work in quantum information, especially in quantum computing, is of that sort. The flip side is, how do you make physical systems do this? And, of course, in the end, that’s certainly the harder part.

Zierler:

Now on the question of what can this do? Is the field still sufficiently abstract and theoretical that it’s premature to think about the societal impacts or benefit that quantum information might have? Or are we already there? Or are we close to getting there?

Caves:

[pause] This is not the first question I would think about, but maybe that’s because I’m not thinking far enough ahead.

Zierler:

I mean, I’m thinking specifically, AI is coming. It’s already here. Deep learning is coming. In some ways, it’s already here. Where is quantum information in these developments?

Caves:

That is a place where one could use the pretty abstract … and there are people who think about this … trying to understand what a quantum computer could do, for example, in machine learning. To put a quantum piece into machine learning. There’s loads of people who think about this and ask what would be different and maybe better. There are very fundamental questions here that I certainly don’t know the answer to and don’t really think about much. For me a big point is, in AI, after a machine has learned something, are you going to be able to ask it, “Well, what did you learn? Tell us what it means.” And, if you think about it, for the machine to do that, one part of the machine will have to contemplate another part to formulate a model of what has been learned. I believe this goes by the name of deep learning. Basically, the question is, it seems to me, will the machine be like a dog that has learned to catch a frisbee? Or will it be able to formulate the physical laws that would allow one to understand how a frisbee flies?

Zierler:

Right, right.

Caves:

Well, we do that, although it’s not generally how we learn things. No outfielder is integrating Newton’s laws for a ball flying through the atmosphere to catch a fly ball. Still, since we were able to abstract laws and models, it seems reasonable to think that some other kind of intelligence could do that. But I think it’s a real question how that happens, probably akin to what consciousness is.

Zierler:

Right.

Caves:

How one part of a machine looks at another part. It’s got to have something to do with that. But, if you want to extend this to a quantum machine, one part can’t be looking at another part because if it does, it screws things up.

Zierler:

Why? Why does it screw it up? What’s the reasoning behind that?

Caves:

If you intervene in a quantum computation in the middle, it will stop. Because you’ve destroyed the coherence that is essential to the processing of the quantum information. This is a big question for quantum machines; what does it even mean to have one part of a quantum machine contemplating another part? Can’t mean it looks at it classically, has to mean looking quantum-coherently. Understanding this might, in the end, be the key to “interpreting” quantum mechanics. Suppose we had a theory that could tell us what consciousness is, in us and in AI, and suppose we could figure out whether or not that theory extended to quantum machines. Might be the point.

Zierler:

So, I asked you an inside looking out question about potential societal benefits. To flip that question, in what ways, at the beginning or even now, does quantum information offer promises to advance physics, itself, generally? What can quantum information do to push the dial on physics, both in the experimental and the theoretical realm?

Caves:

Well, this might go back to the original pre-motivation for quantum computation, which is attributed to Feynman. That if you want to simulate quantum behavior, because of the exponential explosion of the size of Hilbert space with system size, you can’t efficiently simulate that on a classical machine. So, simulating the truly quantum behavior of a system requires simulation on a quantum machine. Or maybe you should say that truly quantum behavior is precisely what cannot be simulated on a classical machine.

Zierler:

So, for example, and again this would require some speculation on your part. But, in theory, would quantum information perhaps be a means that gets string theory to break out into the real world?

Caves:

That could be. People think about that, but, David, I’ve sort of reached my limit on the speculation front. I’m not one of the physicists who’s licensed to do speculation, and that’s partly because I didn’t ask for a license. Because I know that most of my riffing is just that, so why subject people to it?

Zierler:

Well, then, maybe a concrete way of getting towards what you think is going to happen next is to ask you a little bit about your current and recently former graduate students and the kinds of projects they’re working on.

Caves:

Current and recent. Yeah … well, I don’t have any current graduate students. You know that I retired at the end of June 2018. At that time, I decided that my whole life I had put all kinds of things off, on the grounds that I could do them later. And now the later is here. So, I’m gonna focus on those things I’m most interested in. Since I retired, I have finished off a metrology project with my former PhD student, Jonathan Gross, which puts a geometric gloss on a particular problem in multiparameter estimation. Jonathan is one of my favorite people to work with, but he’s a postdoc at Sherbrooke now and very soon is going to be focusing on his new responsibilities at Google. I’ve also finished off a big project on how to characterize a linear optical network remotely using entanglement. This was done in the context of what’s called randomized boson sampling and was done with Sima Baghbanzadeh and Saleh Rahimi-Keshari. I helped to supervise Saleh’s thesis at the University of Queensland, and we have had a very productive collaboration extending over many years. Sima and Saleh are now married and had a son about a few months ago in Teheran. There’s much left to be done on our common interests in boson sampling, but I’m thinking now I won’t have time to fit that in with my other projects.

Right now, I’m working very closely with two UNM postdocs on projects that I think are very interesting. With Raf Alexander, I am working on a metrology protocol called quantum illumination, which is a goofy way to use entanglement to enhance detection of a target against a bright background. The project seems always to balloon beyond control, but that’s because Raf and I share a weakness for getting too ambitious. There are always new questions, and one of us always seems to have the tool set to address each new question as it pops up. Still, there will be a cap on this at some point, and we will write a paper that blows the socks off people who work on quantum illumination.

The other postdoc is Chris Jackson, and this is really where my heart is. As a UNM postdoc, Chris has formulated a program based on the generalized coherent states that live on the curved phase spaces of semi-simple Lie groups. Chris got into this by showing that the phase space and the generalized coherent states result from the most natural kind of measurement on the group, weak, continuous, isotropic measurements of the group generators, and then he realized that everything can be put in a powerful geometric setting. Geometry is sort of a joint passion for me and Chris, and fortunately Chris is better at it than I am. In the end, what I have learned is that physics, both classical and quantum, happens on phase spaces, generally curved, and the generalized coherent states provide the link between classical and quantum. This project has legs: geometric quantization, nonlinear dynamics and chaos—this is how I hope to revive the Schack-Caves hypersensitivity-to-perturbation program as a phase-space way of characterizing the complexity of quantum dynamics—and maybe even a convincing interpretation of quantum mechanics.

Zierler:

I know that you’re involved, when you talk about now is later … I know you’re involved in environmental issues and things like that, but to the extent that you are still actively working in physics, what are the things that you want to do for the rest of your career? What are the things that captivate your imagination for the limited time and resources that you want to put into these things? What are the topics that are most compelling to you looking forward?

Caves:

[pause] This whole phase-space program of Chris Jackson’s. Physics, particularly quantum physics, on curved phase spaces … I can see from what has already been done that this set of ideas is going to be what occupies my time, so I am very much hoping that Chris and I will be able to continue working together after he leaves UNM. This program has the potential to redefine what we think of as classical, or maybe more accurately pre-classical, and to connect the unpredictability of classical Hamiltonian nonlinear dynamics with the unpredictability of quantum mechanics. It’s clear to me that there is a there in this program, and when I start thinking about it, I can hardly sit still for this interview because I want to run off and get to work on it.

Zierler:

So, the idea is the work is still very fundamental at this point and that the work needs to remain in that fundamental arena before those—

Caves:

I’m only interested in fundamental questions.

Zierler:

Yes, right.

Caves:

And now that rudimentary quantum computers are actually being built … you know, I described to you earlier that I was very lucky as a PhD student and fresh postdoc to be working in a place and on a subject where you could work on fundamental questions, with potential ultimate application, without any fear of being scooped. You can’t do that on any of the hot topics in quantum information theory now. If you choose a hot topic, others will be breathing down your neck, and you have to be awfully confident that you’re quicker than they are.

Zierler:

[laugh]

Caves:

I often tell kids this: You should work on a topic that nobody else is working on, chosen so that because of your work, everybody will get interested in it. Well, yeah, that’s a great strategy, if you can make it work, but … [laugh] you see, I can do that now!

Zierler:

Right.

Caves:

Because at this point in my life, there’s no downside to failing, so at least aim high enough that you can fail in a big way.

Zierler:

Right. Well, Carlton, for my last question, the last portion of our conversation has been so forward oriented. I want to ask you, for my last question, a sort of broadly retrospective question. And that is, a theme of our talk today has been … you’ve taken … in many ways, an against-the-grain or an unorthodox approach, both to your research agenda and your career trajectory, in the kinds of topics that you’ve taken on over the years. What do you see as the major through line that connects your overall career and the things that you’ve worked on? I mean, you could have set yourself up to pursue a much more traditional career, I guess. I mean, being Kip Thorne’s student, and things like that. And you either by temperament or intellectually, for whatever reason, you decided not to … I don’t know if the easy path is the right word. But you certainly didn’t take a more traditional path in terms of the things that you did. So, what is the through line that sort of connects all of those points together, in your mind?

Caves:

The thing that connects everything that I’ve done is, what can you do with quantum mechanics? That is, it really is, even before there was quantum information, what can you do with quantum systems? What are the limits on what you can do? And what can you do if you can take advantage of quantum mechanics, instead of just regarding it as a nuisance?

You remember my saying that quantum information asks what we can accomplish if we make quantum systems do what we want instead of what comes naturally. With the benefit of hindsight, one can say that squeezed-state interferometry is a first example of this. Don’t accept something so natural that it seems inevitable, the vacuum, entering the unused port. Do something different. Substitute squeezed vacuum and make things better. Generally, look around in the vast Hilbert space of quantum systems, and find a place where you can do something that otherwise couldn’t be done. Squeezed vacuum, really, you don’t have to look very far, hardly past the end of your nose. There’s a lot of unexplored terrain in Hilbert space, where who knows what you will discover. That’s always been my interest, and it continues to be. And I will work on things where I don’t have any competition, at least initially.

Zierler:

[laugh] Well, on that note, Carlton, it’s been a delight speaking with you today. I really want to thank you for spending the time with me. [End]