World Science Festival: w/ Edward Witten on String Theory (Transcript)

Use quantum mechanics where it’s meant to be used—small things. Use general relativity for big things, stars and galaxies. Stop trying to put them together.

EDWARD WITTEN: Well, the obvious counter to that is that the big things are ultimately made out of small things. So the sun, for example, is one of these big things. We study its gravitational field using general relativity, but it’s ultimately made out of atoms and molecules that we study quantum mechanically.

So it’s incoherent to expect to have one theory for the small things and one theory for the big things. I think Freeman’s skepticism was a little bit less crude than that, at least when he was talking to physicists. But Freeman in general, I must tell you, was a great scientist, but he also was one who was a contrarian. And if I was a contrarian, I’d be wrong close to 100% of the time. So as a contrarian, Freeman had an excellent batting average.

BRIAN GREENE: So given that it is important—yes, I mean, like you said, you know, big things are many small things. You can also say, I guess, you know, center of a black hole, the moment of the big bang, are realms where we think you’ve got to have quantum mechanics and gravity play well and so forth.

Given that, what would you say is special about string theory, that it makes headway where the previous approaches were running into these non-linearity problems?

The Restrictive Nature of String Theory

EDWARD WITTEN: Well, string theory is much more restrictive than—before we even get to string theory, though, I should explain that physics as it developed in the 20th century was extremely restrictive. So when the standard model of particle physics was discovered, the experimental data was very limited.

And the inventors of the standard model didn’t invent it because they were fitting very, very rich experimental data which told them what to do. They had very limited clues from experiments, but the framework of quantum mechanics and special relativity was sufficiently rigid that just trying to fit very limited clues, they were able to invent what came to be known as the Standard model and has been very successful since then.

So with gravity, well, unfortunately with gravity, our luck ran out. So the framework we had with quantum mechanics and special relativity was very restrictive. But it barely made it possible to incorporate the other particle forces. And that “barely” was why, as I just told you, physicists made so much progress with limited experimental clues.

With gravity, because of Einstein’s highly nonlinear mathematics, our luck ran out. And in that framework, it actually doesn’t work. The other forces barely worked, but Einstein’s theory doesn’t work in that framework.

String theory is even more restrictive than the framework in which the standard model was built. String theory kind of forces the theory upon you, whether you like it or not. The inventors of string theory were not trying to make a theory of gravity. They were trying to make a theory of the nuclear force. And that ultimately, at least in the original form, didn’t work.

But one of the reasons it didn’t work was that although they didn’t like it, they kept finding this massless spin-2 particle which isn’t there in the nuclear force. But eventually some people started to take it seriously that it is there with gravity, according to Einstein.

So in the very rigid framework where the standard model was developed, gravity actually didn’t fit. String theory provides an even more rigid framework that actually forces gravity upon us. That’s probably what I was telling you in 1986.

General Relativity Emerging from String Theory

BRIAN GREENE: Sure. And when you first saw gravity, as you’re saying, forced upon us, general relativity forced upon us—in fact, I’ve heard you say, and I know it’s tongue in cheek a little bit, but the counterfactual on another planet, maybe their scientists found string theory first before their Einstein, and they extracted general relativity from it as opposed to having to put it together with quantum mechanics.

EDWARD WITTEN: I don’t find it completely far-fetched that that could happen.

BRIAN GREENE: Right?

EDWARD WITTEN: Yeah, actually.

BRIAN GREENE: And so when you first realized or learned that general relativity emerges from—was that one of those?

EDWARD WITTEN: Well, to avoid misunderstanding, I want to stress that “learning” was the right word rather than “realizing it,” really. Physicists of a generation, slightly more senior than I am, had understood this in the 70s, although the insight did not catch fire in the physics world. It seemed too outlandish to a lot of people, but still it was there.

And by the early 80s, a very small number of colleagues—John Schwartz, Michael Green, Lars Brink—were reviving this. And so I would say that where I learned this was in the summer of 1982, which I spent reading a review article by John Schwartz. Yes, that was extremely electrifying to understand that there was a framework even more rigid than the standard framework of physics, in which gravity was unavoidable instead of being impossible.

The Question of Extra Dimensions

BRIAN GREENE: Now when you say it’s more rigid—in some ways, it is. In other ways, and perhaps this is part of the resistance that I was feeling as a young researcher in the 1980s, the theory also requires more than three dimensions of space.

EDWARD WITTEN: Well, that’s part of its rigidity, formally speaking. Formally speaking, in pre-string theories, any dimension is possible, right? In detail, in quantum—in the standard framework, you do find that four seems to be the maximum, although we’ve learned some loopholes in that, right? More recently, yes.

But at the time in the 70s and 80s, four was believed to be the maximum. In the standard framework, string theory does force more dimensions upon us, and that’s actually one of the reasons it does have the chance to unify gravity with the other forces, because—well, of course, this was very important input to your own work.

But making a four-dimensional world, starting from higher dimensions, you build it in with potentially a very rich geometry that gives a very simple theory, conceptually, potentially the freedom to be rich enough to describe the whole world of elementary particles and forces. So I think without the extra dimensions, it would be hard for such a simple theory to lead to something as complicated as the real world. Possibly.

BRIAN GREENE: But at the same time, I remember in those days, in the 1980s and especially after you wrote a critical paper with Candelas, Horowitz, and Strominger, giving us the mathematical tools for what to do with these extra dimensions, how they could be curled up in a way that was compatible with string theory and so on. Many people, when they encountered the idea of quantum gravity, would be excited, and then you’d say, but it requires extra dimensions. They’d be like—

The Miraculous Escapes of String Theory

EDWARD WITTEN: For many colleagues, it was too big a package. But I might tell you that. So, okay, that was electrifying. In the summer of 1982, learning about the work done in those periods by Green, Schwartz and Braingun, certainly after that, I was following it closely, but I was very reluctant to get seriously involved because I felt that even if it was right, it was potentially too hard to understand. It might take 100 years to understand it, and it seemed like too big a commitment.

And so I actually spent two years watching from the sidelines. I was watching much more closely than most of our colleagues were because I was very interested in it. But I only slightly became involved during those two years. I was mostly watching from the sidelines.

I pointed out to Green and Schwartz a difficulty with their theory, that technically it was hard to see how their theory could capture an important aspect of the weak interactions. The fact that nature is asymmetrical between left and right. It’s not something we see in everyday life, and most of you are probably not too familiar with it, but in the world of elementary particles, one of the most basic, most, deepest, most fundamental facts we know is that there’s a fundamental asymmetry between left and right.

And their theory as it existed in 1982 when I first learned about it, electrifying though it was, appeared incapable of capturing this property. But okay. My only input, I guess, in those two years was to make the few practitioners aware that this was a problem that had to be solved.

And there was an electrifying moment in the summer of 1984 when Green and Schwartz discovered another trick the theory had that hadn’t been appreciated, that completely obliterated this problem. Now, it’s difficult to convey this to you, but string theory is an incredible story of the perils of Pauline. There are all kinds of ways for the theory to prove to be inconsistent, and miraculous escapes. There have been a whole series of miraculous escapes.

And what I was weighing in 1982, when I was not sure if I really wanted to commit myself to the subject, was whether the evidence from the miraculous escapes was convincing enough that the theory was for real, that I should gamble my career in focusing on it. And as I’ve told you, I lacked the courage to make that commitment in 1982 and 1983.

But when yet another of these miraculous escapes happened in the summer of 1984, after that my doubt evaporated, but I felt I’d run a kind of experiment. The question I raised in 1982 was, is this sequence of incredible discoveries all a big mirage, or is it real? If it’s a big mirage, then there shouldn’t be any more such discoveries made. But not only was a big discovery made, but it even involved the what I had seen as the most pressing deficiency of the theory.

So, having gone through that experience, my direction was clear in my own mind after that. But many of our colleagues, to get back to your remark, that many colleagues did not react as I did. They had no inkling of the theory. They never heard of it before the summer of 1984. They didn’t have the experience I have had had of watching and wondering for two years. So sitting on the sidelines, being reluctant to commit myself, waiting to see if there’d be another miracle. And then there was so obviously someone who hadn’t had that experience reacted differently from the way I did.

The Sociology of Scientific Revolution

BRIAN GREENE: And I want to focus this more on science and sociology, but I have to follow up with one question. Do you think that part of the resistance back then was of our own making, an exuberance that felt almost threatening to other colleagues. This idea of the “theory of everything.” And if you’re not working on it, then what are you working on?

EDWARD WITTEN: Well, there could have been some of that, no doubt. And I’m sure that for some colleagues, Greene and Schwartz were respected colleagues, highly regarded, but maybe they weren’t seen as the ones who were supposed to be revolutionizing the world. And some of our colleagues were skeptical, untold that Green and when told that Green and Schwartz had made a discovery that was going to revolutionize the world. I can understand it.

BRIAN GREENE: Yeah, sure.

EDWARD WITTEN: I’m sure. If I had never heard of string theory before 1984, I would have reacted the same way these other people did.

BRIAN GREENE: You really think so?

EDWARD WITTEN: Yes.

BRIAN GREENE: Wow, that’s interesting. Anyway, so back to the science. So 1984, you and the colleagues I mentioned write down this idea of what to do with the extra dimensions. And to someone like me, who was a beginning graduate student, the idea that qualities of particles that had previously been unexplained, like why they come in certain organized groupings, might be explained by just the shape of the extra dimensions. Little space that would be everywhere around us, but so small that we can’t see. But the intricate nature of its shape would explain these qualities. That was astounding.

EDWARD WITTEN: Well, I agree. It’s kind of astounding. Yes, of course. Unfortunately, there’s a question mark about it, because we don’t. We understand that qualitatively it can work, and you can make a reasonable rough draft of the world of elementary particles with relatively simple assumptions about the extra dimensions, but we haven’t gotten past that. We don’t know how to make it work in detail. And that’s a disagreement.

The Landscape Problem

BRIAN GREENE: But can I ask you a question about that? Because one of the critiques. And again, you know, I don’t want to focus solely on the chatter of people who may have other agendas. But one of the critiques, and it’s a real one, is there are many possible shapes for the extra dimension. As time went on, I mean, back in the mid-80s, yes, like your paper, I think, had five possible versions of the extra dimension.

EDWARD WITTEN: But we learned more and more.

BRIAN GREENE: We learned more and more. So in the early days was like, all we have to do is examine these five shapes and maybe we’ll have the answer to everything within there. But five turned into ten hundred thousand million, and the number of possibilities exploded. Now, sometimes people say, well, that means that this is not a real scientific theory because there are so many possible shapes. But then I look at the standard model of quantum mechanics, a quantum field theory. And just so people know, that’s not a single theory. It’s a whole class of theories where you can choose the individual fields, how they talk to each other, their interactions. There are infinitely many quantum field theories. So is that a real critique?

EDWARD WITTEN: Well, it’s been an obstruction to doing some of what we want to do. I think we don’t know the last word. I think I want to describe a little bit of it for our audience, and then we can get back to the question of whether it’s a critique.

So let’s think about the solar system, which we described by Einstein’s theory. But Einstein’s theory doesn’t predict exactly what the solar system has to be. And in fact, we now know that the galaxy we live in is filled with solar systems that also obey Einstein’s equations, but they have different masses and properties of the star and the planets making them up. So Einstein’s theory tells us how solar systems work, but it doesn’t tell us which solar system we’re living in.

A possible interpretation of the fact you’re mentioning is that string theory tells us how gravity interacting with quantum mechanics and elementary particles can work, but it doesn’t tell us which solution of the equations we’re experiencing, somewhat like Einstein’s theory doesn’t tell us which solution of his equations describes the solar system that we actually live in.

The Anthropic Principle

Now, there’s a further angle to this, which I should mention, which perhaps you would have gotten into in a moment if I hadn’t, which is that some colleagues have suggested that this actually has an anthropic interpretation that, okay, there are aspects of the real world that look very strange, like the fact that, first of all, the expansion of the universe is accelerating, but it’s accelerating incredibly slowly.

It’s difficult to explain to people who haven’t studied physics, advanced physics, why it’s so strange that the acceleration is so small, but it is very, very strange from the point of view of fundamental physics. Now, it’s also a fact that we wouldn’t be here if the acceleration was expanding at a decent pace, because we would have been blasted to bits long before we had the chance to evolve.

That’s stating it conservatively. A lot of other things would have gone wrong before then. The solar system would have been blasted to bits before the Earth formed and before the sun formed and before the Milky Way formed, and everything would have gone wrong if the expansion of the universe was accelerating at the rate that physicists would consider easier to understand theoretically.

So then you could ask, well, how fast would the universe be accelerating in string theory? And we don’t completely understand the answer, but the answer seems to depend on which solution of the equations we assume to be right. And so some of my colleagues said, well, no use thinking about the ones where we would have been blasted to bits. We can only live where we can live.

So, inevitably, the solution of the equations that we’re experiencing is one that has the miraculous properties that make life possible. The one I’ve mentioned is the acceleration of the cosmic expansion. But there are a few other things like that, like the scale of the elementary particles compared to the scale of gravity.

Now, I have to tell you that when this argument became popular in the late 90s, it was popularized by people like Steve Weinberg, the pioneer of the standard model, Martin Reese, the distinguished astrophysicist Leonard Susskind and other highly distinguished theoretical physicists and others. And I was very upset. It really got me disturbed.

First of all, well, as a physicist, I wanted to explain the masses and lifetimes of the elementary particles and other properties rather than accepting the fact that they depended upon the choice of a classical solution. Literally, it made me very unhappy for years.

I made my peace with it because I had no alternative. So I made my peace with it by accepting the fact that the universe wasn’t created for our convenience and understanding it. Maybe the universe is harder to understand in some ways than we would have wished. It’s a shame, but our understanding the universe probably wasn’t the criterion based on which it was made.

So I accepted that. I came to accept that I would say by now, almost 20 years ago, roughly 20 years ago, and I’ve had a more peaceful life since then. But I don’t know where the truth of the matter is.

So you see, if the anthropic interpretation, which says to explain why it’s even possible to have this very small rate of acceleration, you have to have a vast plethora of solutions with all kinds of different values, and then we live in a lucky one where we can live. So that’s the anthropic view of the universe.

Well, I can’t say I would prefer to have a more conventional scientific explanation. But I don’t know, maybe the anthropic explanation is correct. It would be silly to throw away the right answer because we don’t like it. But anyway, there’s lots of uncertainty in our lives as theoretical physicists. There are lots of things we don’t understand.

And whether the anthropic interpretation is correct or a better one will be found one day is one of the things we don’t understand. I’ve made peace with it on the grounds that I’ve told you, but I haven’t stopped wishing that there would be a better explanation.

For example, as far as we know, the acceleration of the cosmic expansion is described by Einstein’s cosmological constant, meaning that the rate of the acceleration is constant in time, at least toward the future. I’d be incredibly happy if it was discovered experimentally, though that’s false, because I think that would give a better chance of a more conventional scientific explanation of the world rather than the anthropic one.

In my reading of the data, there’s no significant evidence that the simple interpretation is false. There are experimental claims. Actually, in their present form, they’re, in my opinion, unconvincing. And I’m someone who would like to be convinced.

BRIAN GREENE: And I gather it’d be easier to explain. You know, if its value is, say, decreasing over time. We’re just explaining why we have this value right now as opposed to why it’s that value forever.

EDWARD WITTEN: I can’t promise we’d be able to explain it better, but intuitively, I suspect we’d have a better chance. So if that were found that the rate of the acceleration is going down in time, I’d be encouraged. And there is actually an experiment, the Daisy Collaboration, that has claimed evidence in that direction. Unfortunately, as I talked about you a moment ago, I don’t consider it convincing at the present time, but I’m ready to be convinced if anybody has better data.

What Should We Ask of Our Theories?

BRIAN GREENE: And so what do you think we realistically should ask of our theories? Everyone has been very gratified by the standard model of particle physics, where we tune all these little numbers inside the equations to match the mass of the electron and the mass of the quarks and so on. We don’t have explanations for those two, and we just pick them. So if in string theory, we haven’t yet, but if we found, say, a shape for the extra dimensions that gave all those numbers, we’d still be picking the shape by hand. Would that be any worse?

EDWARD WITTEN: Well, it would be nice to have. Regardless of what you wanted to say about it, you’d still want to, of course, explain why that was the right solution, but. Yeah, of course, yeah.

The Challenge of Progress in String Theory

BRIAN GREENE: I mean, we always want to push forward.

EDWARD WITTEN: Yes. Well, we learned more first, as you said in the late 80s, early 90s, we learned a lot more ways that you could make something somewhat similar to the Standard Model. And then with the advances in what happens for strong coupling, we got more insight about that. But I don’t think our models of the real world based on string theory have gotten much more precise than they were in the early days.

Yes, and sometimes you have to be pragmatic about where you can make progress. For example, I’ll tell you now where I would most like to make progress. And this is the same thing, actually, I would have said in the late 80s and perhaps did say to you in 1986, I don’t remember specifically.

Well, actually, I kind of explained this about 10, 15 minutes ago. String theory was kind of invented in bits and pieces without understanding the principles behind it. And what I’d most like to understand is what are the fundamental principles behind it. But you can’t always choose where you want to make progress. That’s what I wanted to do.

Okay, I could have potentially been completely obsessed by this problem in the late 80s, except I have the sense to realize that I had to do easier things because it was a little too difficult. But one thing I didn’t try to do was to understand how string theory behaved for strong coupling. I assumed that that was too hard.

Strong coupling means when quantum effects are big. As I’ve told you before, if you’re not a physicist, the phrase “strong coupling” is probably gobbledygook. But if I say that when quantum effects are big, that might also be gobbledygook, but maybe at least is a little bit more understandable.

Anyway, I would have assumed that problem was out of reach, but actually that’s where the most progress was made in the 90s. So sometimes you have to be pragmatic and accept the fact that you can’t solve the problem you want to solve. And there’s a different aspect where you can make more progress.

I thought in the 2000s and tens I didn’t manage to find a good way to contribute myself, but I thought in the 2010s, the most interesting progress came from rethinking Hawking’s discovery of quantum radiation from black holes and trying to understand better what are the lessons from that. And that’s another area that physicists have not gotten to the bottom of, in my opinion.

But very interesting progress was made in the 2000s and tens. A little bit slower recently, in my opinion. I wouldn’t be able to tell you. Yeah, I can’t give you any kind of straight line extrapolation from the last few years to where the next advance will be.

Understanding Duality

BRIAN GREENE: But in terms of the advances, if we can now turn to, say, the 90s and beyond, I think one idea that most people would say is vital to the progress is this notion of duality. This idea. Well, in fact, maybe you just want to say a word. What do we mean when we talk about duality?

EDWARD WITTEN: Well, usually what duality means is that the same theory can be described in different languages, but you can’t translate in a simple way from one to the other. The prototype is that quantum effects are small in one region, but they’re big in the other region. If he asks question A, but if he asks question B, it’s the other way around.

So dual descriptions of the same physical system are alternative descriptions which have the property that whatever questions you can answer in one description, you might not be able to answer in the other description. But if you have several different dual descriptions of the same theory, you can pull the knowledge you get from different points of view.

BRIAN GREENE: And why do you think dualities exist? Right. I mean, I’d say, you know, in Einstein’s day, he had this image which I think many of us grew up on. You have a mathematical description of a theory, and that’s it. There isn’t like a second or a third radically distinct mathematical description. Then over the years, we found more and more examples.

EDWARD WITTEN: Yes.

BRIAN GREENE: What does that tell us about mathematics and describing the world?

EDWARD WITTEN: Well, it’s too big a question to answer completely, but one aspect of the answer, I suspect, is that it’s telling us that string theory is inherently quantum mechanical in some way we don’t fully understand. So string theory does have a classical limit, but it has different classical limits.

And each classical limit looks highly quantum from a different point of view. And because there are different classical limits that are so different from each other, no classical limit can do justice to the theory. So string theory is a theory that really only comes into its own fully in the quantum case. But I might have mystified you, but I can assure you that Brian and I are also mystified.

BRIAN GREENE: Without a doubt. But when you actually look at the famous. Well, maybe you can just give a couple of examples. Like in string theory, in the early days, we realized that you could have one mathematical description of a universe that was very small, like a circular universe with a small radius, and a completely different mathematical description where the universe is really big. You think a small universe and a big universe are simply different, and yet we found that they were the same theory just described two different ways.

EDWARD WITTEN: Yes, and at first it looked very quirky, but then, of course, many other examples were discovered. And. Well, that’s where you did some of your most exciting work, actually, of course, which I will mention, just because you might be too bashful to inform our listeners.

BRIAN GREENE: Thank you.

EDWARD WITTEN: Well, it’s a strange fact about the world that these quantum equivalences between theories that are classically different look so very special at first, and yet they keep coming up and they seem to tell you more and more. And they don’t only turn up in physics.

There’s a theory that’s kind of central in number theory and in large areas of mathematics that goes by the name of the Langlands program. And we now understand that that’s a number theorist version of some of these dualities that physicists study. So clearly the message from duality is far reaching and deep, even if we haven’t completely come to the bottom of it.

The Maldacena Duality

BRIAN GREENE: Now, a famous example of that duality which you made reference to, Juan Maldacena, your colleague at the Institute for Advanced Study, which really took everybody by storm in roughly 1996, 1997, something like that, just to give a feel for it, I just realized we didn’t actually say what string theory is in the conventional formulation. Instead of having little dot particles, you got little vibrating filaments that look string-like.

EDWARD WITTEN: Except they obey quantum laws.

BRIAN GREENE: They obey quantum laws.

EDWARD WITTEN: The dot was a fuzzy dot, and that’s hard to explain. You should study some differential equations maybe. And then the string also obeys quantum laws.

BRIAN GREENE: And so most of us thought that string theory was one way of describing the world. The dot description with the fuzziness, which goes on to the name of quantum field theory, is its other approach. But Juan came along and found a duality in which string theory in a certain realm was equivalent to quantum field theory in a different realm, in fact a different dimensional realm, and very importantly.

EDWARD WITTEN: Quantum field theory without gravity, since, as our able students here know, quantum field theory is not consistent with gravity, so it had to be quantum field theory without gravity.

BRIAN GREENE: And so that’s a stunning realization. That means in some sense that string theory is not, maybe not quite as radically different from the old methodology in some way.

EDWARD WITTEN: Well, it is true that string theory can emerge from more conventional theories, as Juan essentially discovered, but the world in which Juan showed that you could start with a mathematical world that doesn’t have gravity, and there emerges from it a whole different world that can have stars, planets, gravitational fields and civilizations in it.

So you see, there are all these crazy dualities that have been very important in string theory since the 80s and 90s. But in many cases, the different dual descriptions, although they’re radically different, they seem like they’re radically different in detail. But the differences might be hard to explain to you because from an outsider’s point of view, they might seem like somewhat similar theories.

The Maldacena duality is really different because the alternative descriptions are really of a completely different type. On one side there is the type of theory that the Standard Model is, a special relativistic quantum theory without gravity, but that obeys the same general principles, based on the same general principles as the Standard Model of particle physics, the kind of theory we think we understand in principle, although in practice they can sometimes be hard to understand. It’s equivalent to a world like the one you and I actually live in, a world with stars, planets, galaxies and gravitational fields and Einstein’s theory, all that stuff.

Experimental Evidence and Mathematical Truth

BRIAN GREENE: And so does that, for instance, we don’t yet have experimental evidence for string theory. That’s an important part of the reason why we’re uncertain if these ideas are correct. Does this link to quantum field theory, the more traditional methodology that has been, excuse me, tested for other quantum field theories, not the one that Juan invoked? Does that give you any greater confidence?

EDWARD WITTEN: Well, sort of, but I would have said it a little bit differently. What you see, not only here, also other aspects of this point could be made in relation to other things we said earlier tonight, but in many respects, Maldacena’s duality, but also some of these other discoveries in string theory have shed new light on existing theories in physics, in different areas of physics, and even in some cases on mathematical theories, purely mathematical ideas.

So to me it’s implausible that all this thing that seems to know so many secrets about all the theories we know about exists by accident. So when Maldacena’s duality is successfully applied in a new area of physics, to me that reinforces the belief that it must be physical, not just a coincidence. But as I’ve told you since December of 1984, when that last electrifying discovery was made, I was convinced there couldn’t be a coincidence.

BRIAN GREENE: And so how do you view mathematics? Do you view it as a set of ideas that we invent in order to describe the patterns that we encounter? Or do you see mathematics as somehow out there in some Platonic realm, and we’re ultimately just instantiations of these abstract ideas mathematicians develop?

EDWARD WITTEN: Mathematics is largely self-directed by ideas that the mathematicians find elegant and beautiful. But it also has been throughout history, heavily influenced by physics. Newton invented calculus because he needed it to describe the motion of the planets and. Well, we could point to many other examples in the 19th and 20th centuries.

Partial differential equations, functional analysis, even Riemannian geometry got a big boost because Einstein used it for his theory. It existed before Einstein and that helped him invent his theory. But mathematicians took it much more seriously when it was known to be relevant in the real world. It’s surprising. I think it’s often surprising how much mathematics is influenced by things that are relevant in physics.

BRIAN GREENE: And when we have a case of duality where now you’ve got two mathematical frameworks that are describing one and the same universe, a hypothetical universe, say how should we think of. How do you think about it? Is that one would be right and the other secondary? They’re both right? If they’re both right, then they can’t really be the mathematics because they’re two different languages.

EDWARD WITTEN: Well, in most of the examples of duality, where the different dual descriptions are qualitatively more or less similar, I see them as being completely on a par. And the fact that more than one of them exists shows that none of them is the truth about the theory. And the fact that we don’t understand where it comes from is our problem. We don’t understand the theory very well yet.

This duality between gauge theory and gravity discovered by Maldacena is a little different because the two are so. It raises a completely different kind of question. I’ll tell you a version of that question.

So Maldacena showed or argued and subsequent work bore it out very well, that you can have a theory of a conventional type without gravity that sort of sits on a shelf, but it secretly describes in a holographic fashion a world where we could be living. A world of planets, stars, galaxies and gravitational forces and all the things that make life possible.

So one world seems very static and just, well, doesn’t have any of that. It’s a kind of theory that physicists like to study, but it’s not describing stars and people. The other world can have stars and people and they seem to be equivalent. But this one is totally crisply defined mathematically. But this one, where we live, seems a little bit murky by comparison with our present understanding.

I don’t feel we know the truth about that. I feel that we’re missing a breakthrough that would make the bulk description. The bulk description is the one where we live, that would make that seem closer, maybe never as sharp as the sort of abstract one that lives on the shelf, but at least closer to being equally sharp.

The Future of Experimental Physics

BRIAN GREENE: Can you imagine that? We’ll get to a point in the next—again, it’s always difficult to predict these kind of things—but 10, 20, 30 years, will we understand things well enough? And do you have any kind of confidence that we might gain observational or experimental data to bolster this stuff being more than equations?

EDWARD WITTEN: Well, there’s some hope, but it depends on being lucky with both the theory and the experiment. I would say if you go back to the 1980s and you consider what was in prospect in the 20 years after that for how much the energy increased at accelerators, it was more or less universally believed that with that huge jump in energy, we would discover the next layer in structure beyond the Standard Model.

Now, it came as a huge surprise that that didn’t happen. Actually, the fact that it didn’t happen is interpreted by some as another clue in favor of the anthropic universe. I told you before how much I was upset over the anthropic universe and eventually made my peace with it. But the only—I told you there were other clues apart from the acceleration of the cosmic expansion that suggested that the anthropic universe might be correct.

But the second one worth mentioning is precisely the fact that accelerators improved so much in the 20 years after Brian and I first met, 25 to 30 years after Brian and I first met, without giving us physics beyond the Standard Model. I think it was not only reasonable to expect that we would have gotten a lot more input from those experiments than we did, but I’d say it was virtually universally expected in the field, and certainly by me that that would happen. It seemed paradoxical otherwise.

In fact, the paradox has been reinterpreted as a hint of the anthropic universe. As I told you before, I don’t know if that’s correct, but it’s what we have to live with for now. Apparently now, going back to when Brian and I first met, we would have anticipated, with a huge jump in the power of accelerators that followed, that we learned way more about physics beyond the Standard Model.

The Rise of Cosmology

On the other hand, we might have underestimated what would be learned in cosmology. In 1986, if you wanted to summarize the sum total of cosmological knowledge, what was it? Well, there was the temperature of the so-called microwave radiation, the radiation that fills all space that’s left over from the Big Bang: 2.7 degrees Kelvin, 2.7 degrees above absolute zero, and the same in all directions as far as one could measure.

There was the fact that the universe was expanding. There were two numbers in cosmology. There was the temperature, and there was also the expansion rate. And of those, only the temperature was measured with any reasonable precision. The expansion rate was unknown to a factor of two. And that was cosmology in 1986. I’m probably cutting some corners in that summary, but not very much.

Now, cosmology mushroomed in an incredible fashion, and we know so much more today than we knew in 1986. And while I would have been too optimistic about what we’d learned from accelerators, I would have greatly underestimated what we’d learned in cosmology. We were able to observe that the temperature is not the same in all directions, and that gave us an incredible amount of information about the early universe and how galaxies formed.

Then we discovered this crazy acceleration of the expansion of the universe, which—well, I told you all the angst that it caused me. So that was certainly a big discovery that we weren’t counting on in 1986. That was made in approximately 1997, actually, by coincidence, more or less the same time as Maldacena’s discovery that you mentioned. And anyway, cosmology has become a real precision science, so we’ve learned a lot from it.

So it’s difficult to predict what clues we will or won’t get from experiment in the next 20 years. I perhaps best conveyed that by telling you how far off I would have been in 1986 trying to guess the next 20 years.

BRIAN GREENE: Sure, absolutely.

Einstein and Quantum Nonlocality

And so we’ve been focusing a lot on gravity, quantum mechanics. But if we could pivot for just a moment back to quantum mechanics—you know, Einstein famously had his issues with quantum mechanics. “God playing dice,” didn’t like probabilities. But I think if you look at the history, what really troubled him was this idea of nonlocality, that what you do in one location could have some weird quantum correlation with something at another location. Do you, in retrospect, consider that to have been a reasonable, rational issue that he was uncomfortable with?

EDWARD WITTEN: Well, he’s certainly correct that it’s very strange, the nature of quantum correlations. It’s also relatively clear that Einstein’s hope to get rid of them will not bear fruit. The subject was understood much more deeply by John Bell in the 1960s, and experiment has confirmed—experiment has shown that things are much worse than Einstein feared, frankly.

So experiment has shown that quantum correlations do things that are much worse than what had Einstein upset. So, yes, Einstein raised an important question, but his hope is not going to be realized.

BRIAN GREENE: I mean, is that sort of a sign that obviously he was a revolutionary thinker—

EDWARD WITTEN: Yes.

BRIAN GREENE: —had limits on what he was willing to—

EDWARD WITTEN: Well, we all do, I’m sorry to say. We all reach our limits, Brian.

The Quantum Measurement Problem

BRIAN GREENE: And what about quantum mechanics itself? So there’s another issue that many people have shone light on, the so-called quantum measurement problem that, you know, you have a description of a quantum system in terms of, say, a probability wave. It’s spread out. You measure the particle, you find it here. The wave has to somehow accommodate that new information. We often call it the collapse of the wave function to accommodate that. No one knows how or if this happens. Where do you come down on that kind of thinking?

EDWARD WITTEN: Well, I don’t consider the collapse of the wave function to be a useful concept. So I don’t think that way. I think that unfortunately, to seriously think about it, it’s very eerie because I think when you try to think seriously about what quantum mechanics means, the essence of it turns out to be the question of what it means to know something. And that’s very strange.

I can try to explain it. I’m not totally sure this will be understandable to all of you, but I’ll try to make it so. Well, first of all, you have to understand that the quantum equation seems to produce a world of possibilities. And it’s not clear how a definite outcome emerges from that.

What Bohr said was that when a measurement is made, the measuring device records the answer, and the physicist who looks at it learns the answer. But Bohr didn’t say in what sense you’re supposed to know what the measuring device said. But whatever the answer is, you’re not supposed to know it by making a measurement of the measuring device. Because if you had to measure the measuring device and then you have to measure the measuring device of that measurement, you’d soon get to an infinite regress.

So there was a key step in Bohr’s explanation, which was undefined. You were supposed to know the state of the measuring device in an undefined sense. Now, there was a student at Princeton named Everett—I’m forgetting his first name.

BRIAN GREENE: Hugh.

Hugh Everett’s Contribution

EDWARD WITTEN: Hugh Everett, who in 1953 or 1954 wrote a paper that’s often cited, but as far as I can see, rarely read. And I’ll tell you what Everett said. Well, first of all, Everett criticized Bohr. Bohr was not applying quantum mechanics to the observer. But Everett said justly, well, we are ourselves made out of atoms and molecules. So quantum mechanics was applied to us.

And then he said, if you look at his paper, it’s very interesting. He begins his paper by talking about quantum gravity. He says that in quantum gravity, especially in a closed universe, nobody can look at the world from outside. He criticizes Bohr because Bohr had a classical observer who was looking at the quantum system from outside. Everett wanted to rethink it with the quantum observer being part of the quantum system, the observer being part of the quantum system.

So what Everett said was, the experiment is done. The measuring device records the answer. The observer looks at the measuring device and records the answer in his memory or her memory. And there’s what—he doesn’t use the word “entanglement” in that paper, but nowadays we call it—it creates a quantum mechanical entanglement between the state of the measuring device and the state of the observer’s memory.

And then, and this is the key point, Everett says the observer knows the state of the observer’s memory in a sense he didn’t define, but not by measuring the state of your memory. That would have led you to the same infinite regress that Bohr was trying to escape.

Well, for a long time, I was one of the people who hadn’t read this paper. But about 15 years ago, I did read it. And when I did, I was really unimpressed. And I couldn’t understand why people thought it was such an advance. Because all that Everett did was to shift the strange step in Bohr’s work one step further from the measuring device to the observer’s memory. But the observer was supposed to know the observer’s memory in some undefined sense, but not by quantum mechanical measurement, not in a sense that was defined.

So for many years, I was very dissatisfied with this. But in recent years, I’ve actually had an afterthought. And my afterthought is that I don’t know what’s going on when we think we know something. So is it conceivable, as a matter of biology and physics, that we are accessing our memory in a way that can’t be described, that isn’t a quantum mechanical measurement, or in what sense? What is happening when we think we know something?

Consciousness and Quantum Mechanics

BRIAN GREENE: Wait, so are you saying that you think the solution to the quantum measurement problem might involve consciousness at some level?

EDWARD WITTEN: I’m saying that I think you can’t discuss it satisfactorily without discussing the conscious observer who’s making the observations.

BRIAN GREENE: And do you have any thoughts on what consciousness or the conscious observer would be?

EDWARD WITTEN: Well, I share the views of most physicists that consciousness is purely emergent. Now, that phrase won’t mean much to most of you, so let me explain it. Electric charge is well defined. You can count the number of electrons, and that’s the electric charge. Or maybe there are protons too, when you count those.

But on the other hand, whether a piece of material is a superconductor is not. If you have a small piece, the question doesn’t make any sense. In the limit of many atoms, it becomes well defined to say that something is a superconductor. That’s what physicists call an emergent phenomenon.

My view of consciousness is that of most physicists. I believe that consciousness is an emergent phenomenon that arises—that can arise for a sufficiently complex system that we call a brain. I, of course, don’t know that that’s true. And some have other views about consciousness. Believing that consciousness is an emergent phenomenon, I also believe that knowledge is an emergent concept.

And since I believe the interpretation of quantum mechanics depends on what it means to know something, I conclude that the interpretation of quantum mechanics should not be expected to have a sharp answer—that it’s fundamentally emergent. I can say this in another way that would make many physicists less uncomfortable. I think what I’ve said—

BRIAN GREENE: You see me squirming in my seat. Is that what it is?

EDWARD WITTEN: Well, I think what I’ve been saying would—many physicists would want to say something. The notion that the interpretation of quantum mechanics is emergent would be expressed in another language by many physicists who would try to keep the conscious observer out of it. What would they say?

Well, the trouble is that since it’s not exactly my viewpoint, I have trouble expressing it with passion. They say roughly that the world becomes classical when things are sufficiently complex. Well, that’s an oversimplification. Sorry. I can tell my view with more passion than I can tell other people’s view, I’m afraid.

Free Will and the Laws of Physics

BRIAN GREENE: And so if you’re willing to go there for just half a second, I’m just wondering—so I think most physicists would agree with you that consciousness is emergent phenomena. We don’t need anything else.

EDWARD WITTEN: Yes.

BRIAN GREENE: Where do you then take that when it comes to human free will?

EDWARD WITTEN: My interpretation of the statement that we have free will is that our actions are largely determined by the state of our brains. Now, that interpretation of free will would not make everybody comfortable. But I think, to me, that’s the meaning of the statement that we have free will.

BRIAN GREENE: Then, of course, the state of your brain, presumably, is described by the equations of fundamental physics.

EDWARD WITTEN: Yes.

BRIAN GREENE: So when you have a great breakthrough, Edward Witten, with that view, do you take credit in your—you know what I mean? So is it the laws of physics that just chose you and through you yields this insight, or is it something that is intrinsically you?

EDWARD WITTEN: Well, I don’t worry about it too much. If I discover something nice, I enjoy it without worrying about it.

Moments of Awe and Discovery

BRIAN GREENE: That’s probably the best way of going about it for sure. One final question and point. You know, when you’re still extraordinarily active and doing all sorts of remarkable things, but as you look back on what you’ve done so far, is there a single moment, were there many moments that you would describe seeing something and a sense of awe emerging from even rather…

EDWARD WITTEN: Little things can give you a sense of awe because they can be very pretty. So, for example, I did some work around 1997 that’s not so well known on why the low energy effective action in M-theory is consistent. It was very pretty. So it looked like magic and it worked. I got satisfaction from that. It didn’t have to be the biggest discovery I ever made.

I answered the question you asked. I thought you were about to ask me what discoveries had given me most satisfaction. But you didn’t quite ask that.

BRIAN GREENE: But that’s a good follow up. So thank you.

EDWARD WITTEN: It would have to be the work in 1994 and 1995, first with Seiberg on the low energy dynamics of certain supersymmetric theories, but roughly speaking, on those dualities that you were telling us about, and the following year on similar dualities in string theory.

And the last insight that I had, the most recent insight I had that made me really happy, unfortunately was two years ago. I regret to say I’d like to have that feeling again. But, you know, who knows?

The Creative Process

BRIAN GREENE: Sure, you know, Einstein, I think, liked to shape his own public profile in interesting ways. You know, we’ve seen the photograph with his tongue out and so forth. He once described his process as he would sometimes describe as “thinking in music.” How would you describe your process?

EDWARD WITTEN: Trying to do research means most of the time it feels like you’re sitting around and doing nothing. So I’d say you spend a lot of time, it feels like nothing at all is happening. And if you’re lucky, you get a good idea eventually.

Okay, you spend a lot of time. Well, I don’t know if I’m describing everyone. I have no idea how closely this matches the way you work. In my case though, there are an awful lot of days I come in in the morning, and at the end of the day, I just haven’t written anything on the sheet of paper that’s been in front of me all day.

BRIAN GREENE: Well, I need to say, though, and I won’t name any names, but there was a young postdoc at the Institute for Advanced Study, not me, who was in an office either with an adjoining wall to yours or within earshot. And apparently there was a period during a summer when only you two were around.

And he said that he was trying to do his calculations while all he could hear was tap, tap, tap, tap on the computer as you were going, like, from brain to paper, brain to paper. So it’s a little hard to imagine this idea of coming in the morning and sort of nothing happening.

EDWARD WITTEN: There are periods, if something does happen, you might go through a period where you know what you want to write down. So I guess our postdoc was in the next office during such a period. But I can assure you that there’s a lot of the time trying to do research in theoretical physics feels like you’re hanging around and doing nothing.

BRIAN GREENE: And when you… if you don’t mind me asking, like, when you’re sitting there trying to figure something out, are you talking to yourself?

EDWARD WITTEN: I’m not even trying to figure something out. I usually don’t even have a good idea about what I want to figure out.

BRIAN GREENE: And so what do you do when you sit there?

EDWARD WITTEN: Stare at the ceiling? Not much. If I get really frustrated, I might go for a long walk.

A Personal Story of Inspiration

BRIAN GREENE: And the beautiful grounds that you have at the Institute for Advanced Study. Let me just finish with one small story that is sort of… I just thought of as I was coming down in the Uber, you know, I mentioned 39 years ago, whatever. We had that. But even a couple years before that, when I was at Oxford, it must have been 1984 or ’85. Your paper on Calabi-Yau manifolds had just come out. I was trying to understand. I had some… I called you up.

EDWARD WITTEN: Yes.

BRIAN GREENE: And you answered. And you spoke to me, a little graduate student that you’d never heard of. And we went through things, and it was really a wonderfully inspiring moment for me and the other kids in the office.

And so that night, we said, we need that kind of inspiration. So what we did was we took all the chairs in the office, we piled them up to the ceiling, and we put your picture at the top so that we would sort of have you with us.

What happened was, story’s not quite done. That following morning when the cleaning staff came in to the Oxford physics department, they opened the door, they saw a face up near the ceiling. They thought that someone had, you know, so they called the police. By the time I got to the office, there were police cars, there were…

EDWARD WITTEN: Ambulances and so on.

BRIAN GREENE: When it was finally determined what the truth was, the chairman of the department called me into his office, and he basically said, one more infraction, I’d be out of the program. But I just want to say it was well worth it to have you as an inspiration then, and you’ve been an inspiration ever since. So thank you so much for this conversation.

Related Posts

• Neil deGrasse Tyson on UFO Files, Trump & Alien Existence (Transcript)
• Professor John Lennox: AI Is Humanity’s Attempt to Make God (Transcript)
• What’s Changing, What Kids Must Learn w/ Sinead Bovell @ SXSW (Transcript)
• How Emotions Are Made: The Secret Life of the Brain – Dr Lisa Feldman Barrett (Transcript)
• The AI-Generated Intimacy Crisis – Bryony Cole (Transcript)